Today I learned that the meme of the white woman (?) wearing a green shirt and a black cap screaming originated from a real person who actually did, in fact, scream during the very first inauguration of Donald Trump back in January 2017.
I feel like I finally understand why are Russians so obsessed with the greatness of their literature.
I mean, every nation has its favorite writers, but it is virtually impossible to have a debate with a Russian or a fan of Russia without arriving at the topic of "aren't Pushkin, Tolstoy, Dostoevsky, etc. the most important writers who have ever existed?" It's as if accepting the fact that these were the greatest writers ever is the mark of a civilized and cultured person. If you disagree, or just think that this is not important because everyone prefers different books, you are hopelessly stupid. The people who say this don't even have to be Russian nationalists, or at least they do not feel that way; they are just telling you an obvious fact that everyone knows. It's in the water supply.
But the fact is, this is an important part of Russian imperialistic propaganda. When Russians conquer another nation, they murder the writers and destroy the books. Then they teach their children the Russian language, and make them read and admire Russian writers. And a generation or two later, Russians can paint themselves as those who brought the light of civilization to the illiterate peasants.
It's not about the literary qualities of the War and Peace per se, because no one reads that book anyway, but about the fact that the glorious Russian nation can produce Great Literature TM, while you rednecks are not even capable of writing a limerick in that thing you call language. (Of course, whenever someone actually writes a limerick, it gets burned, and the author gets shot.) Which proves that you are not true nations, and you need your Russian overlords to provide you spiritual support and thereby keep you culturally above the level of mere animals.
Whether you think that Pushkin, Tolstoy, Dostoevsky, etc. were great or not, the mere fact that you are debating the question means that you have been successfully distracted. The question you should be asking instead is: what did Russians do to the writers of the nations under their rule?
Across Eastern Europe, it's pretty common for people to claim that their nation is the greatest in some respect, not sure Russians stand out so much.
More interestingly, it's simply false that " they murder the writers and destroy the books." Most famous Ukrainian authors (Shevchenko, Lesya Ukrainka, Kotliarevsky) created their works when Ukraine was part of Russian Empire. Mickiewicz and Sienkewicz lived and wrote for the most part in Russia. In the Caucasus and Central Asia, national literatures either experience renaissance in the 19th-20th centuries.
This is not to say that there was no oppression and some systematic efforts at Russification, some of the writers that I mentioned were arrested and/or exiled. But then Dostoyevsky himself also only barely escaped a capital punishment.
Tolstoy and Dostoyevsky are very both good and I have in fact read W&P in its entirety twice. Winters are long where i live too. ;) I think Anna Karenina is a much better novel though. Pushkin doesn’t do much for me. Among the 19th century Russians Turgenev would be ahead of Pushkin for me but again that’s completely subjective. This is complicated by the fact that I am mostly reading in translation. Some Russian speakers say that Pushkin loses a lot in translation.
Calling any writer the most important that ever existed is just stating an opinion. Great writers are great in many different ways for many different reasons. William Faulkner is an important American writer but to compare him to another great but more contemporary American writer, say, Cormac McCarthy would not make much sense. They both set out to achieve different things.
this is a bit of an odd strawman. I mean you can acknowledge Russian writers and filmmakers really have made some of the best artistic works out there, but so has everyone lol. Like trying to argue this doesnt need "but russia kills its artists!" over "this is just a silly argument."
you dont read war and peace, they dont read John Dos Passos. everyone is a redneck, lol
Sometimes I think the fundamental mistake of rationalists was to locate their capital in the Bay Area. Yes, close to the Silicon Valley, etc. But also the world capital of wokeness.
When I read these stories... most people who later appear to be a PR problem (to put it mildly) had obvious red flags since the beginning. In retrospective, the right thing to do would be to tell all people with serious mental problems to go away, right? That would prevent them from getting their symptoms dramatically worse after being exposed to the ideas of AI doom and Roko's Basilisk and acausal trade and whatever. But you can't simply send the crazy people away, because that would be *ableist*, of course!
You can't even send away the drug junkies, because hey, this is Bay Area, that would be deeply uncool! (Even if Eliezer himself disapproves of drugs, everyone else is like "yeah, just ignore him, all the cool kids do their own internet research about which drugs are safe".)
And you can't send away Ziz, because Ziz will obviously accuse you of transphobia. Though Ziz will later accuse you of transphobia anyway.
So you let all these people stay, and then someone snaps and someone gets killed, and then all the normies are like: "omg, the Bayes Theorem drove them crazy, I knew that was some dark sorcery!"
...I know this is written with the benefit of hindsight, but it seems that the project of "raising the rationality waterline" was doomed since the moment the key players decided to establish their base in the Bay Area.
And, you know, we didn't even succeed at slowing down the AI apocalypse. Maybe accelerated it.
I wonder what happened in the alternative Everett branch, where people decided to avoid Bay Area like poison, and instead kept the rationalist bases in New York and Berlin. Are they smarter? Or did they fuck up dramatically in a different way?
If I read it correctly, Ziz joined the Bay Area rationalist community in 2016. In 2019 they became increasingly aggressive against CFAR people, which led to the infamous protest. There was no throwing out -- Ziz decided to leave. I am not sure about the other Zizians.
They wouldn't have been relevant regardless. But I don't know why you're acting like this was an issue contained to the Bay Area, as if it wasn't endemic to every city on Earth worth a damn. And now those titans at Silicon Valley have put their lot in with the force that will provide a long-term solution, so what's the problem?
> But I don't know why you're acting like this was an issue contained to the Bay Area, as if it wasn't endemic to every city on Earth worth a damn.
Could you give me examples of bad things happening in the rationalist community in other cities? Maybe there was a thing or two, but probably nothing comparable in size to things that happen in the Bay Area regularly and kinda predictably.
...What rationalist communities in other cities? The whole movement was centered on the Bay Area, as you said yourself. Of course the bad things would happen there, because that's the only place where things are happening. On the other hand, the unwarranted sympathy for liabilities that you're complaining about is not contained to the Bay Area. This obsession with "tolerance" is everywhere. Any city is going to have a good amount of... problematic elements. The same thing would have happened regardless of where they decided to settle.
nothing to do with wokeness, the conservative example in the past was people bombing abortion clinics. it turns out rhetoric can lodge in crazy people's brains and its not a good idea to go incendiary or apocalyptic when you really don't believe it.
like a lot of people say things but never internalize them, because their internal life is full. but empty or damaged people do, and its dangerous. i don't think much can be done except to reflect on whether you are not kind of giving ammo to it; abortion is wrong but if you start saying abortion is murder...
I agree that crazy people are both in the left and the right. But it's the left that considers talking about crazy people and avoiding them "ableist". (I think the word "crazy" itself is banned.)
But the group's beong too woke to call crazy people crazy is only one of the dynamics by which crazy, destructive people can be accepted by groups. Groups can also adopt people who are way blunter about some issue than them, and more willing to take destructive action. Those people give the group a way to participate vicariously in vicious tirades and physical attacks they themselves can only dream of carrying out.
i mean the right let their various crazies take over the party with the various truther movements. but its a human impulse not a partisan one
its tough to explain. Like imagine you are in the 80s and we watch Threads, a very grim movie about nuclear war.
like 90% of people will watch it, shiver, and the next day put their pants on and go to work unfazed. water off a ducks back. Some may do very light disaster prep that wont save anyone.
Maybe 1% will be moved to make it part of their life to stop nuclear war in healthy ways.
9% though will be various types of crazy, from "how can anyone have kids under MAD?" to selling everything to build a bombshelter to get crazy in, to what have you. the extreme cases turn to violence because they are very crazy and cant let the idea roll off their back. it would be best if they never knew it existed.
the ability of people to entertain apocalyptic scenarios and functionally ignore them is something we take for granted. maybe we should chill a bit with them because some can't.
Are there any good theories as to why Elon Musk keeps talking about birthrates being incredibly important, but also supposedly believes in short timelines for AGI?
It seems like a bizarre preoccupation moving in to a technological singularity.
I'd personally say that AI is the most urgent issue we're facing now, but that stuff like the ever-growing deficit, fertility collapse, global warming, and persistent government dysfunction are all also very serious problems. We should care about all of them, and we should be trying to address them all.
Musk believes in short timelines for AGI *and* also that alignment is possible (at least if the greatest genius in history, ie. Elon Musk, is allowed to work to bring the alignment to reality), so if we're about to move to an AI-created universe-spanned human-oriented Paradise, it makes sense that he wants as many people to enjoy it as possible.
The most reasonable explanation I can think of is that Musk is not a perfect utility maximizer, and is prone to shift to side projects that he then focuses on, almost exclusively. He chooses projects roughly in terms of how interesting and solvable they appear to be. The mean time for him to happen upon such problems seems to be about 12 months, give or take 6. An unusually tough problem might take him longer, especially if it's standing in the way of an even longer term project he has, or if it can be broken into subproblems, each as interesting as the first. But the key here is that once he's decided to work on a problem, he ignores other problems almost entirely, like a cheetah that chases an antelope she's locked on even if the chase takes her past other antelope.
So for now, optimizing the US government is his interesting problem. It will probably stay so for the next 6-12 months.
I think simple reasons are enough to explain it. In order of most to least important:
1. Musk is on Twitter constantly and his timeline constantly talks about birthrates.
2. Musk's political faction sometimes uses low birthrates as reasons to support their policies.
3. Musk's uses low birthrates to retroactively rationalize a history of "pregnancy-maximizing" behavior in his personal life. (https://archive.ph/F3Foa)
There's also probably a motivated reasoning factor. Musk has something like 14 children by three or four different mothers, and presumably that predisposes him to arguments that suggest his prolific fecundity is a highly prosocial activity that others should imitate.
Trying to solve all humanity's problems single-handedly seems in-character for Musk, and I guess that includes the population crisis.
I am surprised that *more* billionaires don't do the same thing. I mean, if you have enough money to get as many women as you want, and feed as many children as you can conceive, then what exactly is the argument against doing that?
My guess would be a quality over quantity preference combined with a preference for a relatively conventional family structure. Looking at the rest of the current top ten billionaires in the US (Bezos, Zuckerberg, Ellison, Buffett, Page, Brin, Ballmer, Bloomberg, and Gates), they seem to be pretty tightly clustered around three children apiece (Bezos has four; Ellison, Page, and Bloomberg have two; and the rest have three). Five of them are divorced and Buffett is widowed, and of these Brin, Buffett, and Ellison have remarried. Only Ellison has divorced and remarried more than once (*). None of them seem to have publicly-acknowledged children born out of wedlock.
(*) To paraphrase an Oscar Wilde character, one failed marriage may be regarded as a misfortune, but five seems like carelessness.
On the quality vs. quantity front, spending time with your children and being involved in their lives and upbringing takes time. The time demands of parenting scale sub-linearly with number of kids (this being a major part of Caplan's arguments in favor of larger families), and billionaires can trivially afford to hire as many nannies and other helpers as they need, but there's a certain baseline of time and attention you need to direct towards your kids if you want to actually have a relationship with them. Billionaires also tend to be hard-driven career men (and women, but mostly men) who spend a ton of time on their businesses, which only leaves so much room for family time.
If you value at least the appearance of a relatively standard family structure, and especially if you value the substance, then your spouse's preferences are a limiting factor on how many children you have. Preferences for very large families are outliers in our culture, and both partners would need such a preference in order for the couple to decide to have lots of kids. And I'd guess that women somewhat less likely than men to prefer very large families, since women are generally the ones who get pregnant and give birth. Some women do actively enjoy the physical experience of being pregnant, but I get the impression that most don't.
Overall, the top tier of billionaires (excluding Musk) seem to have more kids than the overall norm (2.78 vs 1.94 children), with "can afford more help" winning out over "hard-driven career men" to the tune of an extra 0.84 kids. Which doesn't strike me as too surprising.
Also, in a traditional marriage where you have one wife and stay with her, you hit some constraints:
a. There is a limited fertility window that closes sooner for women than for men. I could presumably still have more children, but my wife has gone through menopause, so that's not on the agenda unless I get remarried or take up a mistress or something.
b. It's somewhat common for health problems to crop up that close the woman's fertility window early, so even if you always planned to have five kids, you may find that you're only having one child after all.
c. Random crap can happen in your life that makes it very hard to have kids for awhile--losing your job, having a family member very ill or in some kind of serious trouble, some physical or mental health crisis by mom or dad or one of the existing kids, etc.
I suppose if you want to do your best King Solomon impression, you can also get some parallelism going--nine women can actually deliver a child per month, after all....
> On the quality vs. quantity front, spending time with your children and being involved in their lives and upbringing takes time.
> And I'd guess that women somewhat less likely than men to prefer very large families, since women are generally the ones who get pregnant and give birth. Some women do actively enjoy the physical experience of being pregnant, but I get the impression that most don't.
This does not matter, if the billionaire can choose the ones who *want* to have many kids. Because there are women like that, too.
If that is high on his priorities when choosing a partner, and if he is choosing (or re-choosing after divorce) after becoming a billionaire, then yes, a billionaire would presumably be able to select for that as Musk seems to have done.
The other nine of the current top ten US billionaires seem the have revealed preferences for other priorities besides having more then 3-4 kids apiece.
An educated guess is that Musk believes a species with a birthrate below replacement is headed for extinction unless birthrate goes back up, and a birthrate even close to replacement is an extinction risk if something in the environment goes pear-shaped - such as unaligned AGI, but also a comet impact, giant solar flare knocking out the grid, nuclear war, Black Death, mass famine, ice age, and so on. A high birthrate mitigates a lot of problems at the species level in the same way a healthy immune system and diet mitigates a lot of problems at the individual level - at worst, it buys you time.
Musk probably also believes (as a lot of people do) that the sort of AGI plausibly in our near future isn't an existential threat. It's even possible that it's a non-threat in his eyes, and so birthrate is merely a benefit alongside it.
From the oral argument in the SCOTUS yesterday it sounds like most of the justices view the end-birthright-citizenship EO as obviously unconstitutional.
The matter before the Court yesterday wasn't that question; rather it was whether nationwide injunctions from federal district courts should be allowed to remain in place. But any court injunction is rooted in a likelihood of the plaintiffs eventually succeeding on the merits, so the larger question about that EO was inherently present yesterday.
Both conservative and liberal justices posed hypotheticals to the Trump admin lawyer which assumed the EO is eventually ruled by the SCOTUS to be unconstitutional. They each asked versions of "assuming the EO is ultimately found to be invalid how should that judicial review process proceed?" [meaning in particular, with or without the presence of nationwide injunctions?] Alito, Kagan, Barrett and Kavanaugh all got into that with John Sauer who is Trump's chosen Solicitor General.
Sauer responded repeatedly with the same basic argument: that the presidential EO should receive full deliberative due process. His proposed alternative is the class-action lawsuit process, which Kavanaugh seemed to agree with while Alito not. I'm not sure how serious a suggestion or possibility that is. Regardless, Sauer repeatedly argued that whether class action or not, court cases against presidential EOs should work their way through the federal courts in "the normal manner" without being frozen by any district-court injunctions.
He may even be right about that, I'm fairly persuaded myself that district-court national injunctions have become a broader problem. Still though, the irony! The president who bluntly wants to throw all non-white asylum seekers into dungeons first and sort 'em out later if ever, is suddenly verrry interested in the importance of...due process.
I've said a few times to a friend of mine -- a previous government lawyer -- that we both wished that this question was brought before the court in literally any other administration. Nationwide injunctions are a legitimate problem with serious negative effects, but, fuck it, call me a partisan, we *need* that now. I noticed none of the GOP was calling for the end of nationwide injunctions during Biden's admin.
As it is, Sauer was an absolute wreck. The justices AND the court were outright laughing at him several times. My favorite line of questioning was Kagan, who had two bangers:
- "O, so if the government decided to take everyone's guns, that would be fine?"
- "You keep saying that people can appeal to SCOTUS. But for that to happen, some plaintiff will have to lose. And in a case as ridiculous as this, the plaintiff will win *every time*. And you, the government won't ever appeal, because you don't WANT a national ruling!"
But I also appreciated ACB's almost genuine disbelief at Sauer stating that "generally" the government will follow court precedent, and even Kavanaugh had a good question in there about the mechanics of figuring out citizenship for babies.
Meanwhile it felt like Alito and Thomas and even Roberts were trying to give Sauer layups that he was either unable or unwilling to take, which I thought was funny. I'm sure somehow the former two will find a way to justify why there shouldn't be a nationwide stay here, as they always do.
Luckily SCOTUS isn't stupid. They see through Trump's demands for due process for *him* in this case where he is the defendant, and not in any case where he is the plaintiff. The hypocrisy is *staggering*, and really makes it clear that everyone in the Trump admin sees both the judiciary and the law in general as impediments to be removed instead of principles to uphold.
That Barrett moment was LOL for me, and there were a couple of good memes of it going around Reddit.
A couple different practical answers about nationwide injunctions have been proposed in Congress more than once in recent years but never advanced. Maybe one if those finally will now.
That is because it is a great stalling tactic, and allows Trump to go forward with this in the meantime. The idea that this should wind its way through federal courts for the next year, while children are born stateless, is absurd. The executive order is on its face textually unconstitutional. Whatever arguments you care to make about altering that amendment, It’s still a fact on the ground. The second amendment is another good example. At the end of the day, there was no getting around the text of the thing.
I listened to the argument, and I found it interesting that at the very end the Attorney General Sauer argued the case on the merits,after having spent all his time previously trying to focus it just on the legalistic question of universal injunctions.
I think this is a very bad case to resolve that issue with. The text of that amendment is way too big an elephant to hide in a closet.
I personally think the 14th amendment has become outdated. It needs to be revised. But the Supreme Court can’t rewrite it. This administration brought this case as a stalling tactic. That’s my opinion. Of course they want it to percolate through all kinds of courts and take forever. I think it was pretty clear that most of the justices found both class action and percolation to be problematic to say the least. I know there are justices there that would happily rewrite the 14th amendment if they could only get their hands on it, but the Govt doesn’t want to tee that one up.
Does anyone have stats or have recommendations for how to find stats on the % of teachers that abuse children? I have done some research and can only find % of students.
Mostly I've just used AI search and done some searches on pubmed. Please help!
Do you have in mind sexual abuse or other forms of abuse such as inflicting physical or emotional harm? If it’s the former, I found stats maybe a year ago on what
percent of abusers fell into various categories in cases where the abuse came to light. Teachers were definitely one of the categories. Not exactly what you’re asking but gives a basis for estimating: We know what fraction of kids are known to have been abused at a certain age, and can estimate fairly well how many teachers a a kid has had by the time he reaches a given age
. I don’t know how much data there is on what fraction of abuse come to light. I’m a psychologist and have seen quite. few people who were sexually abused as kids, or raped as adults, and all of them either kept the abuse a secret or only disclosed it years later, or only disclosed to people they knew would not inform the authorities.
We could make a guess based on the following numbers: total number of teachers, total number of students, % of abused students, average number of teachers abusing the same student, average number of students abused by the same teacher.
Total numbers of teachers and students should be public info; you already have the % of abused students. I would guess that the average number of teachers abusing the same student is approximately 1; maybe let's use 2 as an upper limit.
So the last missing part of the puzzle is how many students are abused by an average abusive teacher. You could probably find some numbers in some specific cases, the remaining question is whether those numbers are representative... maybe teachers who abuse more students are more likely to be caught.
If I was the superintendent of a school district where such a thing happened, I'd try hard to keep it out of the news. If I worked in the Department of Education, I don't think I'd be eager to publish this data either, for obvious reasons. I don't see any reason to believe that this data would be easily accessible, or even gathered and stored in one place: the organisations running schools don't have an incentive to gather and publish it, and the courts or child protection authorities probably don't track child abuse cases by the profession of the abuser.
In other words, that data might not be publicly available. Depending on the country you are interested in, you might be able to file a Freedom of Information request with the appropriate authorities.
Yeah I think that's what I'm finding. It's frustrating because everyone points to the Catholics and rates of child abuse, but they wre only able to get the data because of major compliance with the Catholic church. I'm very curious how it would compare with teacher rates of abuse...
In Ireland, where I'm from, Catholic orders ran residential schools for much of the twentieth century. There was terrible abuse there, and some of the most notorious pedophiles were Catholic religious working in schools. You can look up Letterfrack industrial school as an example.
There is some evidence that the Catholic Church is no worse than other large organisations running residential programmes for vulnerable children. I don't have facts and figures, but I think that e.g. the Australian Government programme which took indigenous children into state care in the mid-twentieth century was found to have abuse at similar rates to the Catholic schools I described.
But I think your premise is wrong. Other organisations failed also, the Catholic Church might have more prominence because it's larger and richer. But it's unethical to tell people that child abusers are everywhere, and so the Catholic Church is not that bad. Are you trying to undermine trust in public schools and society in general? Perhaps better to focus on ways in which the Church has helped people, but don't trivialise the suffering of people abused as children.
Kinda late in the open thread, but I just read that the Federal Appeals Court has given UC Berkeley a chance to win back the CRISPR patent from MIT.
Emmanuelle Charpentier and Jennifer Doudna, who shared the 2020 Nobel Prize in Chemistry for devising the CRISPR technique, both worked at Berkeley. But MIT got the patent because they showed it could be applied to eukaryotes. Charpentier and Doudna expressed uncertainty about whether it could in their groundbreaking paper. Because of their expressed doubts, the US Patent Office gave the patent to the Broad Institute of MIT, which showed that CRISPR could edit the genes of eukaryotes. Of course, I'm reading the media's distorted reporting of legal arguments, but it would have been ironic if Charpentier and Doudna could have been awarded the patent if they said, "We're pretty darn sure we can apply this technique to eukaryote genes, but we haven't done so, yet." Of course, the US Patent Office, as a general rule, awards patents to *working* implementations. However, C&D did get CRISPR to work on prokaryotes—but no one really cares about prokaryotes. Eukaryotes are where the action is!
Oh boy, this just so reinforces the basic truth that a patent is not a technical document, it’s a legal one. Don’t ever put in writing “I’m not sure my idea works for ‘thing’”, the other guy will use it as a defense, “see, Fibonacci here is explicitly teaching away from using the idea for ‘thing’, while in [0056] we are teaching how to do ‘thing’”.
Ask me how I know :)
Also, exposes the tension between publishing papers and applying for patents at the same time. The two don’t mix well. I have dozens of patents and 0 scientific papers to my name. Got to pick one or the other.
It seems that there are already enough incentives for scientists to be dishonest about their work. Now we add to the mix the University's licensing lawyers (who dream of fatter endowments) and the nitpicky requirements of the USPTO (not that it could be any other way).
Yes, I used that exact framing to defend a patent submission. We (very few patents have a single inventor these days) had our submission rejected based on certain prior art, where the examiner saw a document that mentioned both the 'method' we were patenting and the 'thing' we were trying to accomplish. But on a closer reading the prior art was actually saying "there is this 'method' but you can't really use it to accomplish the 'thing'". So I constructed a response basically saying, look Mr. Examiner, you're saying that anyone skilled in the art will understand that our 'method' can be used to achieve 'thing' based on this prior art, but ACTUALLY the prior art document says "don't do this", so it "teaches away from using the 'method' to achieve the 'thing', therefore you can't argue that anyone skilled in the art will understand from the prior art reference to use the 'method' for 'thing'.
It worked. We got the patent.
I don't think this means scientists have to be dishonest. What it really means that IF they want to patent what they publish, they have to think very carefully about both timing and content of the relevant papers. In simplest terms, "file first, publish later". Then maybe instead of saying "'method' is unlikely to work for 'thing'", say "possible use of 'method' for 'thing' is still under investigation" - no lies here, just a more careful wording.
The result was: PW 5P 4P 7S KC 7C 7P XVIII -4S -XIV
Position 1 (Heart of the Matter) & Position 2 (Opposing Factor) - Page of Wands, Five of Pentacles
The core of the matter is that Humanity rushes forward with a child's enthusiasm into embracing the creative power of AI, but this could lead to hard times and economic inequility (and maybe even extinction if we stretch the ill-health meaning of 5P).
Position 3 (Root Cause) - Four of Pentacles
The source of the whole situation is the greed of the relevant participants. If it generates profits in the next quarter, why not do it?
Position 4 (Past) - Seven of Swords
In the recent past, there was some hidden dishonor in the situation. This probably refers to the various shady AI company business like the OpenAI boardroom drama, Sam Altman maneuvering behind the scenes, Anthropic pushing frontier not caring about the previously and privately made commitments, etc.. The card is hidden dishonour, so we probably don't know most of it.
Position 5 (Goal or Purpose) - King of Cups
The goal/the ideal future of this whole AI situation is that the AI, like a caring and wise father, enables human flourishing forever in the universe. Think Minds in Culture novels.
Position 6 (Future) - Seven of Cups
In the near future we can expect to have lots of options on how to proceed, but also much wishful thinking ("I'm sure it will behave roughly the same way even with more intelligence.").
Position 7 (Self) & Position 8 (Other)- Seven of Pentacles, Moon
The querent thinks of themself as carefully assessing the possible consequences of AI, but from a more external point of view their thinking is confused and dominated by fear. Or alternatively, if we take the self/other position to be about humanity instead of personally me (because the word "I" was not in the question): Humanity sure does believe itself that they think carefully about AI, just look at all these blog posts, congressional hearings, books etc, but is it not actually an illusion? Won't the relevant people just keep on doing what they have short-term interest in doing and rush toward the end irrespective of the content of those deliberations?
Position 9 (Hopes and Fearrs) - reversed Four of Swords
The main fear in this situation is that humanity won't contemplate, won't prepare enough for the possible pivotal point and rushes on without pause.
Position 10 (Outcome) - reversed Temperance
The final outcome is that the harmony and cooperation will only be apparent. This, to me, suggests deceptively misaligned AI. Or alternatively: The going away of balance which makes me think of a paperclip maximizer, ie an agent which pursues a singular goal maximally hard, without caring for anything else.
Disclaimer: Reading Tarot is merely the drawing of purely random images on cheap paper. There is no supernatural Spirit or unconscious directing the process. This is purely for fun, the verbal equivalent of looking at clouds forming an angry old man or a dog. Peace!
Hi Scott. I won't link to the post in case there's a privacy concern, but in one of your defunct blogs you made a post about dystopian fiction, and you mentioned a novel where "all Earth's countries have been renamed things like FRA-113 and JAP-289". But you didn't name the novel in the post. Do you remember what novel it was?
I recognise the question as you've asked it before. It sounded like an interesting premise for a book but I had no luck searching with all my google-fu or any of the AIs out there. I'd also interested in an answer.
Gemini 2.5 outperforms, suggesting Ira Levin's This Perfect Day; on a quick glance inside via Google Books, people are said to live in regions like EUR55131 and MEX10405, so it's at least a plausible candidate.
"A leaked planning document shared with Fast Company shows the ORA pilot launched in 2023 with a handful of agency HR and payroll offices, serving “a few hundred” retirees. The plan under Biden was to roll out the ORA system government-wide in 2025. "
"Now that he’s there, he says he finds himself surrounded by people who “love their jobs,” who came to the government with a sense of mission driving their work.....
....what he’s found is a machine that largely functions, though it doesn’t make decisions as fast as a startup might. “I would say the culture shock is mostly a lot of meetings, not a lot of decisions,” he says. “But honestly, it’s kind of fine—because the government works. It’s not as inefficient as I was expecting, to be honest. I was hoping for more easy wins.”
An interesting turn in the SA refugee story: SA President ranted on them, calling them “cowards”. I still do think it’s hilarious, having the anti-refugee President welcoming African refugees, but! I now fully support admitting them as refugees. When your country’s President calls you “cowards” for wanting to emigrate, get the hell out while you can.
I was first introduced to the story from a usually anti-China vlogger.(Will look up if anyone cares.) He had video claimed to be black South Africans calling for white people to be killed, and it might be legitimate.
However, it seemed to be a problem at this stage of a high crime level. I don't know whether white South Africans are at more risk than black South Africans.
It looked like adequate security in South Africa is very expensive-- both a fortress home and guards, and my conclusion is that if white South Africans are at special risk, it's the white South Africans who don't have much money, and they aren't the ones being let into the US.
There were a few hundred or so (white) South Africans that were recently admitted to the US in a refugee program. It became a media kerfuffle because a church administering the program pulled out over disagreement that these people were worthy of refugee status. And the SA government was quick to dismiss the idea that they were being persecuted. It's funny that the president called them cowards because that directly implies they were fleeing danger by leaving the country, i.e. they were legitimate refugees all along.
I think it implies that they were fleeing *perceived* danger where there was no actual danger; it's not cowardice to flee serious persecution but it is cowardice to flee mild inconvenience while claiming it's persecution.
If you place that much weight on one-off remarks by a country's President, what do you think Trump opponents should think about the remarks that he makes? Do you think a reasonable person would regard a black lives matter protestor leaving America as a "refugee fleeing persecution"?
That isn't really the issue, it doesn't matter what one guy says so much as what the government policy actually is on the ground. I just find it amusing that he both acts like these people are cowards and fake refugees while simultaneously claiming they deserve to receive justice for historical wrongs. Both of these are obviously not going to be true at the same time.
Well, if I were an Afrikaner in the SA I'd not wait around to find out whether the danger I perceive turn out to be actual or not.
Ramaphosa could address this in so many different ways, like earnestly proclaiming his and his government full support for the rights of every SA citizen regardless of skin color or prior history, he could gently egg them in a stile of "come on y'all, really, there's nothing to fear", etc. etc. He chose differently. With the SA's history of gruesome violence I'd not want to wait for a chance to learn what necklacing feels like from the inside of the tire.
Yeah, he definitely shouldn't have said it, but I think it's wrong to interpret it as a threat or menacing; it's like Trump insults: pathetic and beneath the office, but if a black lives matter organizer tried to claim refugee status elsewhere because Trump called them "thugs", that's obviously dumb.
It may very well be wrong. But I'm here an ocean away from the place, so I have a luxury of suffering no consequences of interpreting the signals incorrectly. They, being there, clearly see things differently. I suspect I'd do the same.
I submitted for the not-a-book review contest on Monday. I was expecting to get some kind of confirmation email or something, but I didn't get one, and it's activated my anxiety. My brain keeps telling me that I messed up the form, and all the hard work I put into my review won't even get read.
Can anyone confirm if I was supposed to get a confirmation email? Is there any other way to confirm that my submission was received? My overactive brain thanks you!
There’s a short thread from primrose further down in the comments asking the same thing; the short of it is that there’s no confirmation email, if the google form said thank you for your submission, then it should have worked. I’m in the same boat, also nervous that something is going to have gone wrong with my submission to the contest, but it sounds like the only way to be completely certain would be to email Scott directly, which I don’t want to do.
So he gets a gold plated 747 (whatever the hell that means) stashes it away in a hanger until if and when he leaves office, and that can’t possibly be considered a bribe?
What is a gold plated 747 worth anyway? If you sold one, how much could one reasonably expect to get for it?
If Donald Trump is the registered owner of the airplane as soon as Qatar transfers it, that's a bribe. If the United States Air Force is the registered owner of the airplane, and uses it to fly Donald Trump around while he is POTUS, that's not a bribe, that's just a bargain. If the airplane winds up owned by the Trump Library and Muesum Foundation after Trump leaves office, it gets murky - there is precedent for the Air Force donating retured AF1s to presidential libraries, but if Donald Trump maintains an unusual degree of control over the foundation, if he derives direct benefit from it, and particularly if the foundation decides that they're not going to put the aircraft on display just yet but use it to fly Donald Trump around in his post-Presidential travels, then it looks a lot more like a bribe.
The lack of any discussion of these distinctions in the media coverage of this controversy is disturbing to me. Nobody seems to care whether it's legal or ethical, just whether they can use it to score points.
I mean with scenario 1 while not illegal per se, having a president flown by a luxury plane gifted by another state whose intentions are murky is objectively a bad decision.
1 it could still influence the presidents decisionmaking related to that country (how i love Qatar who has provided the luxury i am in)
2 it's easily a security risk due to potential listening devices of any kind or worse (though doubt Qatar would risk anything that can kill the president)
3 might become a slippery slope for outright corruption.
This is also doing a lot of heavy lifting. Does it end up with the library *immediately* after Trump leaves office? In which case that is also extremely strange. Or is it only after many years of service to many presidents later, in which case less strange (in part because Trump would likely already have passed of old age) All of the reporting I've seen on this seems to imply it ends up with the library immediately, which I would take as either
The odds of any president not named Trump, making use of a plane that used to belong to Qatar and hasn't been completely torn down and rebuilt, seem exceedingly small.
Are we actually worried about the Trump Library using it to fly anyone? It's a 747, and those are not small or cheap to operate. The VC-25 is at least plausibly necessary given the size of the President's entourage, but all of the 747BBJs in air force colors (including this one) are basically because they were sold to oil sheikdoms who wanted the same plane that POTUS flies in. Obviously, donating it in flyable condition and when the US government isn't necessarily done with it (though there's the obvious question of why he have it in the first place) is extremely icky, but I'm not sure that he's going to get much benefit out of it, and I'd really like the sight of the Trump Presidential Library going bankrupt because they tried to keep this thing in the air.
Oh, there's no doubt a privately-owned 747, especially one whose operation has to be laundered through a "charitable foundation" would be a white elephant of the first magnitude. But it's the sort of white elephant I could see Donald Trump's ego compelling him to accept, and he might be able to make it happen.
Certainly having it drive the Trump Presidential Library into bankruptcy would be the most fun outcome here. The more likely but boring outcomes are that the gift never happens, or the gift happens and the plane spends the next three and a half years being torn down and rebuilt before being put up on blocks outside the Trump Library.
Its a fair criticism that the media are not covering the details. But there are two potential questions.
Is it a bribe? Murky
Is it violation of emoluments? Not very murky. Due to pre-arranged transfer of ownership to closely affiliated organization. Emoluments doesnt require quid pro quo.
And emoluments has a straight forward resolution: congress, not ag, reviews and approves.
It's incredibly hard to keep track of things in the following categories:
* Things that Trump has actually done
* Things that Trump has said he's doing to do
* Things that Trump has said he might do
* Things that Trump has randomly mused about possibly doing
* Things that Trump's opponents have accused him of planning to do.
I have no idea, at this point, where the 747 story goes in all these categories. I'm reasonably confident it will never actually happen and everyone will move on to talking about some other thing.
Judge Dugan (the Milwaukee judge who was picked up by the FBI on bullshit charges) and her legal team filed their petition to dismiss the case.
From the first page:
> The problems with this prosecution are legion, but most immediately, the government cannot prosecute Judge Dugan because she is entitled to judicial immunity for her official acts. Immunity is not a defense to the prosecution to be determined later by a jury or court; it is an absolute bar to the prosecution at the outset. See Trump v. United States, 603 U.S. 593, 630 (2024)
Damn. Citing Trump’s own precedent against him. I aspire to be this petty.
You should try reading the Trump immunity ruling. Roberts makes clear that intentions cannot be used to determine whether or not an act is (un)official. That was kind of the whole point of the only action the SC deemed immune - Trump threatening to fire Cabinet members if they wouldn't send out a fradulent letter to confiscate voting machines from the states could not be used as evidence because it is a conclusive and preclusive ability of the President to fire Cabinet members.
Are you this credulous about all prosecutions? Like, say, all of the prosecutions that Trump was under, and the felonies he was convicted of? Or are you engaging in motivated reasoning?
Note that you started off by saying that the charges are bullshit. I could just as well ask you whether you're this incredulous about all prosecutions
Our takes on the merits of the case aside, to my understanding the central finding to be determined in the case is whether Dugan obstructed ICE in carrying out its functions. That's not an official act, whether or not that that is established in court
O, sorry, yea, that's because the charges are bullshit. I'm not this incredulous about all prosecutions, just the ones done in obvious bad faith by this administration. I don't have any patience left for obvious blatant authoritarianism.
> to my understanding the central finding to be determined in the case is whether Dugan obstructed ICE in carrying out its functions. That's not an official act
this is like saying "the central finding to be determine in the Trump case is whether he tried to soft coup the country. It's not an official act". The whole point of Trump's legal argument was that you cannot even prosecute to begin with. That was also, imo, pretty bullshit.
But, in case it matters, I also suspect this play won't work. Which is why, despite your claim, I didn't do any 'fist pumping agreement'. I do think it's a clever argument because it makes bare the obvious bad faith in which this administration and the entire GOP has been acting for, say, the last 9 or so years.
Sort of an unusual issue in the mmo i play, Final Fantasy 14.
in the game you can go into player housing set up as venues to chat, like cafes or bath house.
in ffxiv its possible to mod your character, by replacing the assets on your local computer. you make your catgirl into a fox girl, etc. only you can see it but you coukd screenshot, etc.
but there is a program that lets you sync this, so you can download and see everyone else's mods. people "connect to the wifi" and now you see the petite catgirl you are chatting with is seven feet tall and packing piercings. The two girls dancing are now actually doing a sex animation, because you tag it to emotes in game.
this...creates an unearthly experience if you dont or cant have the mod (console players)
You are blind, but not blind. A girl makes a joke about people being overdressed in the tub, but to your eyes she is fully clothed; she just has replaced that particular outfit with scandalous textures. Everyone else can see it, but you can't.
it creates a layered reality that is something to experience. A friend described to me what is really going on and you'd never know from the room otherwise.
and a majority of venue enjoyers use it. not always for lurid acts, but to show off their vision of their characters for others to see. but not using it, that world is invisible to you.
overlapping realities but being unable to see them is a very sci-fi thing.
John C. Wright has something similar to this in his sci-fi trilogy starting with "The Golden Age". Essentially everyone is always in augmented reality and sees (and hears/feels/etc.) whatever their filter is programmed to show. If you turn off the filters, everything (in reality) is just spam ads.
Overall, I recommend the trilogy - it describes a largely positive way that a post-scarcity society can function without going quite as far as Banks' Culture universe.
-- the gold-plated 747 that the Qatari royal family has Trump so excited about is 13 years old, and has none of the extensive secure-communications and other upgrades that are required for a plane to be used as Air Force One.
-- it would take several years and hundreds of millions of US taxpayer dollars to install those upgrades, a process which amounts to "taking a 747, disassembling it, reassembling it, and then jacking it up to a very high level" according to an aviations-industry expert quoted in the NBC news article I just read.
-- there would be no possibility of that being completed by the end of Trump's current term in office. Whenever he does leave office the plane would be donated to his presidential library (which currently exists only as a placeholder website) after taxpayers had spent many more millions _removing_ the classified-security upgrades.
-- the Air Force is in the midst of a multi-billion dollar contract to upgrade two newer 747s for use as Air Force One/Two (though that work has had delays and those planes may now not be ready before 2029).
"The White House did not respond to NBC's request for comment...."
This is precisely why it's a complete nonstory. Anyone familiar with the mil/demil process and costs knew this couldn't possibly be any sort of bribe, it never even made sense superficially. You can't fly a demil'd plane, it's an inert art installation to be displayed to tourists, like most other demil'd aircraft (some are used for firefighter training or scrapped for parts). To be in service as AF1, it'd be acquired by the wing at JB andrews, militarized, used until Boeing delivers, demilitarized, and towed to the library.
One can't have an informed opinion on the cost of the mil/demil without knowing the cost of the status quo. (747 Mil/Demil Cost + Operating Cost) = (AF1 Operating Cost) is entirely possible.
The diplomatic thing to do when someone gifts you a white elephant, especially when you're doing a "peace in the mideast megatour" is to say "thanks, I'll definitely use this" and then do something else with it.
Mil/demil cost can be whatever they want it to be. The VC-25 is basically an entirely custom plane that happens to look like a 747 (this is why it is so expensive) but you could easily strap a couple of military radios into this one and call it good if Trump really wants to. Demiling basically requires taking the radios out, which is not all that expensive and would leave the plane flyable. There aren't many ex-military modified airliners in commercial service, but that has more to do with the military's tendency to keep stuff in service forever than the inherent problems of making, say, a C-40A ready for airline service. (Also, paperwork, but a low-time 747-8 is a lot more tempting than a 25-year-old 737, even if it's not got a lot of hours by 737 standards.)
You seem to be very certain that the "mil/demil process" will be applied to this airplane. Granted that is the normal procedure when previous Air Force Ones have been retired, but this is a very abnormal administration proposing a very abnormal sourcing of an Air Force One. What is the basis for your apparent certainty that it will be handled in the normal fashion going forward?
If Qatar offers and Trump accepts and swears up and down to every media outlet that he will accept and use the plane for all his official flying for the rest of the term and then keep it for use by his presidential library and that it's an amazing idea for him to accept and it's all very cool, how is it not a bribe?
Even if he can't practically use it in practice during his term, whether he understands that or not (assuming he decides to comply by his security team's advice), wouldn't the proper non-bribe scenario be a very public "Oh you're so generous Qatar, thank you but I couldn't possibly."
No, the standard diplomatic thing would be "thank you, but I can't accept this because as an elected official I am supposed to avoid things that look like bribes." Don't imply that Trump had to accept the bribe to be polite.
Also, while you're right that it's unlikely to be used as Air Force One, I think it's plausible for Trump to say "I can't use it as President? That's fine, it can sit in my presidential library until I leave office and I get it for my personal use."
It would be very unwise to accept it on any terms. I’m sure Trump saw all the gold trim and expressed great pleasure, and seeing as it is a white elephant, as you say, the Qataris were more than happy to offer it to him. It was apparently put on the market in 2020 and nobody bit. Apparently it has only been in the air for 1100 hours so it’s very unused. It’s original cost was 400 million but it seems a regular 747 of its kind would be worth about 50 million in resale. Unclear how much you could get back on the gold plating and trim, etc..
I wish substack had a !remindme equivalent so that in a year if Trump was actually using the plane we could see who was engaging in motivated reasoning and who was being earnest
> You can't fly a demil'd plane, it's an inert art installation to be displayed to tourists
In principle, if we were dealing with a normal administration, I would agree that this is likely a nonstory. But we are not dealing with a normal administration. We are dealing with an admin that has literally run a crypto rug pull, invited a journalist into a signal chat, and turned the white house into a tesla showroom. So unfortunately I don't share your credulity. In the words of a great man, "fool me once, shame on you. Fool me — you can't get fooled again."
I expect that Trump will want to use his pretty new toy basically as soon as he can get his hands on it, past precedent about what you "can" and "cant" do be damned. Who's going to stop him, if he wants it?
(by the way, the goal posts have already massively shifted on this one story. It went from:
- "this is made up" to
- "this is because boeing is behind and the plane isn't coming from the qataris its from a new defense contractor" to
- "its not a gift its just refurbishing a plane that used to belong to the qataris" to
- "its a gift from the qataris but he just wont use it")
This is pretty much what I was thinking as well. (On DSL, I mostly stressed the security risk.) Including the ideal thing to happen: Trump accepts the gift, security quietly wheels it into a hangar, and Trump never gets within a mile of it until maybe after his term is up. Quiet mothballing avoids a small army of snoops getting a look at the USG's mil process, and a small army of actual baddies looking over their shoulders.
Alternately, Trump gets tired of waiting for the official new Air Force One, proclaims he's going to save the taxpayers a yuge amount of money, and orders the United States Air Force to fly him around on the Qatari jet as is - whatever level of communications, security, etc, was enough for a rich Arab prince is good enough for him. As Commander-in-Chief of the US Armed Forces, I'm pretty sure he can do that if he really wants.
As Paul Botts notes, it is entirely up to POTUS how those responsibilities will be exercised. "I can totally give nuclear launch orders by calling up missile wing commanders via Starlink, reading out the codes on an open channel", would be a mind-numbingly stupid thing for anyone to do, but there's no law saying the President has to make smart decisions and this one is his decision.
I made the point elsewhere that there's also nothing stopping one of his advisors from making the case to him how stupid this would be. I could probably make it, if someone tapped me one morning and said I'd been field-promoted to delivering the PDB.
It's not even like it has to be adversarial. I'm convinced Trump would readily believe someone might try to pull a fast on him in order to get the edge on dealmaking. Heck, I could even make the case in a way that makes him think he thought of it (or at least, make him think I'm assuming he's thought of it).
The SCOTUS awarded him unconditional blanket immunity for actions that are or could be viewed as part of his "core constitutional responsibilities". Ordering the Air Force to fly him around certainly fits that criteria.
And then he personally sells it for $50 million or more and pockets the money without ever having to pay any of the maintenance or upkeep? That doesn’t seem like a good idea to me.
The Financial Times published on May 11th an article entitled "Three things we learned about Sam Altman by scoping his kitchen". Altman was a guest for the segment "Lunch with the FT". Analyzing the video made for that segment, Bryce Elder, the author of the article, pounces on the opportunity to satirize the problems that stand out to him and quips that they might be an sign of troubles to come for OpenAI. Looks like a puff piece, feel good article. I was thoroughly amused.
I have some commentary about it in a post I made on LessWrong. If it's OK, I'll publish the link to my LessWrong post. If it's not OK to direct to my post, them I'm happy to mention my thoughts here.
I feel like any journalist who wants to critique someone else's kitchen choices needs to post a full inventory of their own kitchen to let journalists pick over.
This article makes Altman seem a lot more relatable. Picking the wrong olive oil? Yep, that's something I struggle with too, I feel like I never know what oil I should be using. Buying expensive kitchen gadgets that turn out to be disappointing? Guilty.
(Also FT, $2000 is not an expensive coffee machine, sheesh. If I were a billionaire I'd be rocking a La Marzocco or something, not a freaking Breville. Of course I'd also have a full-time barista...)
True, it does make him more relatable - an instance of the Pratfall effect: making a small, harmless mistake can increase the esteem we have for someone, particularly if that person is already perceived as competent or above average.
I'd venture to say that something similar might have happened to Bill Clinton after his affair with Monica Lewinsky became publlic, at least in the eyes of some people.
Oh my, that might be an error, if the article wants to go after Altman (or if anyone reading it wants to go after Altman). Kitchens and UK politicians have not been a happy pairing, just ask Ed "Two Kitchens" Milliband 😁
Now I must try and read that article to find out if Sam only has the one kitchen, or does he only use the smaller of his two/more kitchens? How many ovens? Does he need a cookbook to make toast?
I also remember Elizabeth Warren doing a video in her kitchen where she wanted to share with us all her drinking her OWN beer in her OWN kitchen with her OWN dog and her OWN husband (as distinct from anyone else's husband, but if you're a politician it probably is a good move to emphasise this, given the amount of affairs and sex scandals that hang around them).
EDIT: Ah, that article was well-worth it. Sorry Sam, the FT is not impressed by your culinary prowess, maybe you should have consulted one of your tame AIs for advice beforehand!
Oddly enough, due to me watching Youtube cooking channels, I *had* heard about Graza olive oil and the Sizzle/Drizzle/Frizzle variations. And I'm not even a multi-millionaire!
They sneer at his knife (either really expensive or really cheap, and they intimate it's probably bad quality priced highly, because Altman doesn't know any better except to go by price) and have a *second* go at the bottles of olive oil.
"Maybe it’s useful to know that Altman uses a knife that’s showy but incohesive and wrong for the job; he wastes huge amounts of money on olive oil that he uses recklessly; and he has an automated coffee machine that claims to save labour while doing the exact opposite because it can’t be trusted. His kitchen is a catalogue of inefficiency, incomprehension, and waste. If that’s any indication of how he runs the company, insolvency cannot be considered too unrealistic a threat."
Honestly, Altman would have come out much less scathed had he just gone "Oh, I don't cook, that's what the staff is for" like a proper billionaire and potential emperor of humanity when doing a tour of his kitchen while the cook and scullery maid did the chopping, slicing, frying, etc. before plating up the "pasta-infused garlic" (wait, shouldn't that be the other way round, or are they jeering at the amount of garlic he had on the chopping board?)
It's not often I get to look down my nose at an oligarch steering the fate of humanity onto the rocks of AI, but the olive oil allowed me to do so! 😀
"We couldn’t find an exact match online. Maybe it’s a one-off piece by an artisan steel forger who shuns tradition. Another possibility is that it’s a Chinese mass-produced blade that’s sold under countless names, usually in sets, often in a fancy presentation box or with fake Damascus patterns etched on the side. There are Sino-Niho-Germanic Frankenknives all over Amazon and AliExpress that look a lot like Altman’s."
I now feel a lot better about my distinctly unhip, purchased piecemeal over the years, some of 'em came with the knife block, kitchen knives! Though I am considering finally breaking down and buying a good knife, but there's a lot of different recommendations out there (usually "pay €€€€€€ for this fancy Japanese knife you will be too intimidated to use at all"). But now I know which to avoid - any knives that look like Sam's knife!
Gentlemen! If you ever somehow find yourselves in the position where you have in some unimaginable fashion totally by accident murdered your wife and buried her body beneath the stairs in your family home, I'm here to tell you that six years later at the murder trial explaining how you put flowers in, and expressing a wish to that it had been roses but you couldn't get them, the grave where you are secretly disposing of the corpse does *not*, in fact, make it all better.
"His wife was “a beautiful woman,” he said. But her face was distorted with anger as she lay dead in their home. He wanted to lay roses with his wife’s remains but could not find any so put tulips with her instead, he said."
I'm going to go out on a limb here and say I think he's lying his backside off and that if you accidentally, in self-defence, not at all premeditatedly, killed your spouse you would try calling an ambulance and/or the police instead of going for a stroll around town then back home to dig that grave, then for six years maintain your story that your wife had left you and simply walked out one day and you hoped she would come back. But that's just me, maybe this sounds reasonable to other people.
Scott's back and forth with Curtis Yarvin and the responses to it have made me wonder: What exactly is the existential crisis that many high-profile Trump supporters (especially SV types who all turned from sorta blue to very, very pro Trump almost over night) seem to think would have certainly occurred had Harris won in 2024? I'm leaving aside already die-hard MAGA voters, because it's at least clear to me why Trump losing is an existential crisis to them. I'm talking about people who would have likely been fine had either candidate won, who suddenly became convinced Harris winning would be the end. Not trying to start a political flame war, I'm actually wondering. Many allude to that being some kind of point of no return, existential crisis for the country because *gestures broadly at wokeness*, and I gotta say, I just don't see it to the point that it makes me think I'm missing something significant.
Scott has mentioned this, but woke stuff seemed to be in decline before this Trump term. Joe Biden won 2020 as a straight old white man, and Harris didn't really seem to be a paragon of wokeness herself, probably because the campaign team rightly realized it was going to cost them votes and the best thing to do was moderate. Now, I realize that says nothing of the actual policies her administration would end up implementing, but I haven't seen anyone who thought it would be a crisis moment actually state what policies they were expecting and dreading. From the way I saw some commenters talking about the hypothetical Harris term we would have gotten if we hadn't gotten Trump II, you would think she was going to start rounding up the unwoke and sending them to camps, which (to me) is a completely ridiculous thing to believe.
If I had to guess, her term would probably be much like Biden's. Pretty uneventful, some attempts at progressive "woke" policies that would get shot down by the courts, and some mild lip service for woke causes in a speech every so often. I cannot see the hypothetical Harris term cracking down on speech any more than the median president, implementing martial law, or becoming a dictator. These are all things that, if they happened, would warrant the reaction the people I'm referring to have, but I don't see reason to believe they would.
Personally I'm a centrist that was on the fence before and then ended up voting for trump. What got to me was immigration. From what I see, the left seems dead set on getting as many illegal aliens into our country as possible, and doing everything in their power to keep them here, abusing our welfare system.
I've listened to Musk speak on his idea of the existential crisis of the election, and the idea is something like this. Democrats want to import a ton of illegal aliens into swing states, and push for laws that allow noncitizens to vote (no voter ID). The combination of these two pushes will leave swing states voting democrat forever. Illegal immigrants need the welfare state to live, the welfare state is primarily supported by the democrats, and so these huge numbers of illegals will vote for democrats, locking in swing states. Democrats then get to win every election into perpetuity, which brings with it all the woke stuff, which is catastrophic for the wellbeing of our nation.
Take it or leave it. I find this convincing enough.
If you don't mind me asking, what made this argument click for you in 2024?
I ask because this is not a new argument. This first thing I thought of was Anton's "Flight 93" (1) essay from the first Trump election, where the fundamentals of this argument are already presented:
"Third and most important, the ceaseless importation of Third World foreigners with no tradition of, taste for, or experience in liberty means that the electorate grows more left, more Democratic, less Republican, less republican, and less traditionally American with every cycle. As does, of course, the U.S. population, which only serves to reinforce the two other causes outlined above. This is the core reason why the Left, the Democrats, and the bipartisan junta (categories distinct but very much overlapping) think they are on the cusp of a permanent victory that will forever obviate the need to pretend to respect democratic and constitutional niceties. Because they are."
Were you unaware of this line of argumentation before? Did something happen between 2015 and 2024 to change your mind on this argument? Did Musk give it additional weight?
Before 2016, there wasn't actually that much difference between Republicans and Democrats on immigration law. Rs might have talked more about it, but there was a general bipartisan understanding that enforcing immigration laws was a basic part of the function of government, and largely acted accordingly. After Trump got elected the first time, liberals negatively polarized themselves into taking "no human is illegal" etc. seriously; under Biden, prosecutions for immigration-related offences cratered to the lowest levels seen in decades.
Personally, illegal immigration is something I don't care much about one way or the other, so this doesn't really matter to me. But if someone considers it to be an issue of supreme importance, it makes sense that they'd take an "anything but the Democrats" stance in 2024, even if they didn't in 2016 or 2020.
I think this is a great example of what I mean when I talk about information diet and epistemology here (https://www.astralcodexten.com/p/open-thread-381/comment/117100524). Elon has fried his brain with ketamine, he may not be the best person to listen to about conspiratorial stuff like this.
But just to respond on the merits:
> Democrats want to import a ton of illegal aliens
You've already walked back on framing this as if this is some intentional scheme on the part of mustaciod democrat villains, which I appreciate. But you go on to say that this is an unintended consequence of Dem policy. Except...even that isn't true. The number of illegal immigrants in the country has remained roughly static at around 10M since 2003. In that time, there have been 7 different presidencies with 4 different presidents. At the end of Biden's term, with all the caterwauling and screaming about immigration we had...fewer illegal immigrants in the country than we had in 2005. So, just empirically, this untethered from reality.
> into swing states
From the data we have, most illegal immigrants live in extremely polarized states in deep blue cities! The states with the largest illegal immigrant populations are California, Texas, Florida, New York, NJ, and Illinois. NONE of those are swing states. They don't even live in swing counties.
> push for laws that allow noncitizens to vote (no voter ID)
And, again, even if there *was* some small amount of fraud by illegal immigrants, they don't live in swing states or places where those votes matter at all.
> Illegal immigrants need the welfare state to live
Approximately 40% of illegal immigration is visa overstays. Approximately 33% of illegal immigrants are from relatively well off countries. It's true that something like 58% of illegal immigrant households are on some kind of welfare...but 50% of naturalized immigrant households are on some kind of welfare. That delta is only 8%. Which makes this next part...
> and so these huge numbers of illegals will vote for democrats
hilariously, hilariously, wrong.
People on welfare vote for republicans *all the time*. The states that are most dependent on federal aid are all red. I can't find more recent polling data, but in 2012/2013 57% of conservatives were on some kind of welfare. And from 2020-2023, the entire country got some kind of welfare through COVID subsidies. There are tons of people who are on ACA who will happily vote against "Obamacare".
So just to keep a tally here. Your primary concern is voter fraud from illegal immigrants, but:
- most illegal immigrants don't live in a place where their votes would matter
- even if they did, a sizeable % of illegal immigrants don't use welfare
- even if they did, welfare recipients are never single issue voters about welfare
- and, even if they were, there is no evidence that they commit voter fraud
> I find this convincing enough.
If you're serious about this take -- as in, you really are a centrist and this isn't just motivated reasoning -- you should not be trying to approach this from a 'first principles' perspective. Motivations for immigration and, like, people in general are extremely complicated. First principles thinking is useful in the absence of data, but there are countless sources of hard data here that are easily accessible.
The level of illegal crossings, you’ll notice, was fairly low when Obama was in office. He deported 3 million illegal immigrants and was known in Mexico as the “deporter in chief.” The left being against deporting illegal immigrants is a fairly new thing.
In fact I'll go a step further and say that the left has never been, and even now is not, against deporting illegal immigrants. Given that the context of conversation was explaining that "conspiracy theories are not a good reason to vote for Trump", it seems like we're in agreement. Thanks for helping to make my point!
George Soros (or anyone) paying protestors shouldn't sound like a conspiracy theory.
Heck, I've been a paid protestor myself, in my student days. Strictly speaking I wasn't paid so much as given a free hotel room and dinner along with a bunch of other folks from the same political group, but when you're a student a night in a fancy hotel sounds just as good as actual pay.
I think the conspiracy theory is that every time you turn on TV and see a large protest, the majority of those people are actors, essentially extras bussed in like a Hollywood set, with no affiliation to the cause and no actual concern for the issue, just mercenaries showing up for a (comically large, if you ask these people to guess) check. You might say, "nobody actually things *that*" but there are people out there who think the parents who go on tv crying after their kids get shot in school shootings are just paid actors, so that's where the bar is.
I incorrectly implied that this is a conscious scheme by the democrats. I think a set of actions taken for different reasons may have unintended outcomes that may be positive for the actor. You are correct, the idea that a cabal of democrats in charge have this in mind is a conspiracy theory. While it is bad Bayesian analysis to give any thought that smells like conspiracy theory a zero percent chance of being correct, I do think it is more likely that this outcome is not intentional.
For your clarified position, I've never been sure how to treat the certainty with which conservatives claim that illegal immigrants use more of the welfare state than they contribute. Its not that I'm certain that they don't, it just that conservatives speak so certainly about this when AFAIK its at least an open question about the long-term costs vs benefits.
The idea that illegal immigrants are voting in elections is obviously unsupported by any data, Republicans have mounted what amounts to a pogrom against supposed illegal voters and have only ever found a handful of people. I guess your fear is that somehow Democrats would reverse this trend in the future?
But even if both those things were true, the idea that illegal immigrants would be out stumping for Democrats just because its in their rational self-interest is the biggest unsupported supposition of all. We know this because the states that benefit most from government welfare the most are some of the reddest states in the country. Rural voters use government aid vastly disproportionately to urban voters and go for Republicans in far, far greater numbers. And even among first generation legal immigrants, we see that the trend towards voting Democrat is dubious at best, because immigrants have a strong cultural bias against a lot of democratic social policies. If anything, if Republicans wanted to sow up every election in the next fifty years, they would be the ones pushing to have a huge influx of uneducated, low-skilled, culturally conservative people voting in every election, because by every metric that is the people most likely to vote Republican. They reason they aren't pushing for that is the same reason those people are voting Democrat, nobody actually cares that much about their self-interest compared to their cultural identity, and Republicans are still pretty hung-up on being the party that opposes an influx of Brown people.
Which is all just to say, if you were being honest and that was really your reason for voting for Trump, it was an extremely poorly reasoned one and you should re-evaluate all of your priors before voting again.
Why not more *legal* immigrants? I mean, if they want more immigrants anyway, why filter for the ones who break the law? Why create a large group of people that can be easily blackmailed by criminal groups? If people already pay money to get illegally into USA, why not let them pay the same money legally?
Policies that make sense, although not each of them for everyone:
* closed borders
* open borders
* select for the smartest / richest / otherwise interesting people
I absolutely agree with you. It just seems to me that right now the only side interested in some sort of sensible immigration reform is the right. So we end up with closed borders because thats the only good option on offer.
The US still has plenty of legal immigration programs, including ones that do select for smart/interesting people (e.g. O-1) along with people who are just kinda useful (H1-B) along with random people who are just lucky (Green Card Lottery). It's not a perfect or even a great system but it's certainly not "closed borders".
Well, it wasn’t really bipartisan then was it? Democrats wanted it, Trump and Republicans didn’t, on the grounds that it was less a border security bill and more an amnesty for illegal migrants who had already crossed the border. Anyway, as Trump proved shortly after taking office, America didn’t need a bill to resolve the border issue, it needed a president to enforce the existing rule of law.
I have many friends who were in this position -- they were blue, and turned strongly anti-Kamala. These are smart people who are experts in their fields and who I have a lot of respect for.
A few things have come up repeatedly:
- a belief that the reactions to COVID were unjustified overreaches of government control (mask mandates, lockdowns, etc.)
- a belief that a continuation of Biden era economic and tech policies would be terrible (coupled with the belief that Trump would not implement tariffs). In particular, there is a lot of handwringing about Biden admin attempts to regulate crypto out of existence + their stated intention to regulate AI heavily + they all seem to think that Kamala was going to implement price controls (lol)
- strong disagreement with the incentive structure of wokeness, along with a specific strong aversion to one tenet of wokeness (e.g. one friend is strongly anti-trans, another is strongly anti-affirmative-action, etc.)
I think you are right to be skeptical that together these are not really meaningful signs of the apocalypse. When pushed, my friends will mitigate Trump's actions and play up the hypothetical downsides of a Kamala government; that said, many of those people have quieted significantly after 'liberation day' and find the ICE enforcement pretty bad.
Many of the people in this crew are downstream of independent news sources. They style themselves as independent thinkers. Many are in the startup world, and love people like Marc Andreesen or Tyler Cowen. They are trained to be contrarian. But as a result their epistemic hygiene is...lacking. They are bombarded with information about woke malfeasance (Chinese Robber style) and end up believing that it is a much bigger problem than it really is. LibsOfTikTok and its ilk (which are descended from TumblrInAction and earlier kiwifarms) are extremely damaging to our nation's information landscape. And confirmation bias keeps them in the hole.
Many of these people know nothing about the eastman memos because it never showed up in their feeds. But they know all about Hunter Biden, because that showed up in their feeds constantly. If I had their information diet, I too would think the world would end if Kamala was elected. Unfortunately it is very difficult to spot a problem in your epistemology -- a bit like finding a bug in your bug report form.
As an outside observer, I think it was the covering up around Biden's mental state. Either Harris knew and was complicit in "it's all fine, he's fine, sharp as a tack" until they had to shiv him in the back and suddenly she was the bestest ever possible candidate, or she knew nothing and was pushed aside and had no input at all into the four years of the administration.
Neither possibility sounds good for someone looking to be voted into a globally influential office during a time of tension with Russia over Ukraine and what is going in Gaza and potential trade wars with China/disputes over Taiwan, etc. A nonentity who could have been replaced by a rubber duck during her time in office, or someone willing to go along to get along in order not to rock the boat purely for the sake of personal ambition? Which do you prefer?
That is setting aside any consideration of Trump, qua Trump, if you're looking at the Democratic candidate and being told "you can have any choice that you like - so long as it's Harris".
>From the way I saw some commenters talking about the hypothetical Harris term we would have gotten if we hadn't gotten Trump II, you would think she was going to start rounding up the unwoke and sending them to camps, which (to me) is a completely ridiculous thing to believe.
Is this any different to any other candidate in any other cycle? I remember when Mitt Romney was going to reinstate slavery.
Personally, I couldn't get past point 2 which was laughably the wrong way around, based on party platforms and reasonable economic forecasting at the time. I don't know if Ackman believed what he wrote or just thought it sounded good and covered his real reasons.
As to what those real reasons could have been: based on the US economy going into the election, generic Republican had a strong advantage over generic Democrat. Add in the disarray on the Democratic side and there was a very good chance Trump would win again. Further factor in the possibility that he would reward loyalty/punish disloyalty to a greater degree than Harris and it is obvious that endorsing Trump was the better trade.
Yes, that list really reads as largely rationalization. No 3 is particularly laughable, given that the Afghanistan withdrawal was negotiated by Trump (and by the rather predictable Trump cancellation of resettlement of Afghans who aided US forces, and the even more predictable suspension of all refugee resettlement. And #14 looks even worse in retrospect.
Not to put too fine a point on it: Nobody actually cares about balancing the budget. Literally not a single person gives a shit. People that say they do are pretending.
Democrats don't give a shit but occasionally balance the budget anyways as part of the neoliberal project almost by accident; democratic voters don't care or notice.
Republicans pretend to give a shit, then slash income and massively expand expenditures while failing to cut entitlements every time. People vote for republicans to "balance the budget", then theatrically shake their fists at the sky when it doesn't happen and vote republican again next election, because they don't actually care.
What about libertarians, you might ask? What about them. We as a society can do nothing but stand in silent pity and imagine that Sisyphus is happy.
I actually do care about balancing the budget, at least to the extent of keeping the debt-to-GDP ratio from growing any further. I am not pretending. If that was just hyperbole on your point, that's not helpful to the sort of debate we usually try to have here. If you actually believed it, you were wrong and I think foolishly wrong.
You are a kind of probabilistic cloud of possible identities and beliefs to me, as a random name on a discussion thread of a blog. I can't know anything about you but population data and what you say, which doesn't have to be true.
From these datum, I can only draw tentative conclusions until you reveal information about yourself.
For example, you say that you care about balancing the budget. I believe that this might be true, but can't draw a conclusion about it until you narrow down your identity on the graph.
For example: did you vote Trump once? Twice? Who did you pick down ballot?
If you voted for Trump/Trump, I can state you don't care about balancing the budget based on your actions. If you voted Trump/Biden, I can say that you might care. If you voted Clinton/Trump, I can say that you definitively want the budget to be even less balanced.
In any case: If you truly care about balancing the budget, you represent a segment of the population that has no influence on politics except as a marketing tool for the political tendency that wants to make the budget as unbalanced as possible, ie, the non-neocon conservatives who are currently ascendant in the ranks of the party and the state, and who are so committed to unbalanced budgets they got the STATE WITH THE WORLD RESERVE CURRENCY downgraded by Moody's without a government shutdown immanent.
Thus, the "Nobody" in "Nobody cares" can be more accurately stated as: "Nobody who matters has a balanced budget as their first priority vs. IE ending DEI or wokeness or reducing taxes on top earners or some such".
We'll see if that changed, but I think most R's in government who are against the current bill actually want MORE tax cuts and pork, not less.
"From these datum, I can only draw tentative conclusions until you reveal information about yourself."
Says the guy posting anonymously, to the man who has been posting under his own name for thirty years including more than a decade here and on SSC. If you want to know whether e.g. I am a Trump voter, it's not hard to figure out. If you want to posture about how people on the internet are just fuzzy clouds of probability (about whom you are nonetheless willing to make absolute and unconditional pronouncements), go look in a mirror.
I don't actually look at the names of people I reply to on the internet, and if I did I wouldn't research them because this is all shouting into the void in any case.
I reply to these posts for my own entertainment/to refine my own ideas under pressure/to practice my short form writing now that I no longer have to do a quick 200 word explanation of the effects of "A Silent Spring" on the ecological movement, etc etc.
I still haven't read your handle, and I still won't click on it. If you would like to collapse your wave function with some observations, I will go off of those; otherwise I won't.
If you want to address the meat of the comment, IE that there is no political tendency in the US that will sacrifice their Tendies for a balanced budget unless you are willing to mount a doomed struggle ala the Rands, say that and I will respond to that.
Or don't, and I will treat it as a creative writing challenge to whip out a couple lines without going back and editing it, as I find that muy ability to write things correctly on the first go is getting degraded by dictating most of my short messages to the robot that lives in my phone on the go instead of sitting down and hammering them out caveman style.
FWIW, I contacted a group that publish math books for kids that I thought might be interested. It turns out that they already investigated translating and publishing Levshin's books, but the hurdles to officially publish were insurmountable.
In case you want other books, there are now a lot of great ones for kids, including some that have something of the Dwarfland flavor. I am trying to revive interest in the CSMP storybooks (https://stern.buffalostate.edu/csmpprogram/Storybooks/bygrade.html) which are generally tuned for younger kids, but are also funny and emotionally touching. Natural Math (https://naturalmath.com/goods/) has many good choices. Finally, for now, anything by Raymond Smullyan is bound to be excellent, but probably also quite challenging.
> the hurdles to officially publish were insurmountable.
Please tell more.
Thank you for the links, they seem interesting. I wish we had funny books for kids on every topic, and a library full of such books at every school. (Even better, a torrent of such books, in epub format, translated to every language.) It is such a waste that we don't have that already. Kids are curious about things, they just need it to be accessible for their age.
I didn't really get a lot more detail about the obstacles, but it was bureaucracy in Russia with a lot of steps to push against people who don't want to help and couldn't care less about the project ending successfully.
New post up in which I argue that the left and the right are not all hypocrites about the rule of law, even though they frequently seem to display behavior that suggests they are. What we are is unsystematic about our reasons thinking discretionary violations of the law is good. https://hiphination.substack.com/p/are-we-all-hypocrites-about-the-rule
Honestly, your post doesn't seem to be about the rule of law at all. It is about the exercise of prosecutorial discretion, and, yes, people disagree about how that discretion should be exercised.
And when you say, "What we are is unsystematic about our reasons thinking discretionary violations of the law is good," nothing in your post relates to that, because the exercise of prosecutorial discretion is not illegal unless done for an improper purpose. "The rigors of the penal system are also mitigated by the responsible exercise of prosecutorial discretion." Cheney v. United States Dist. Court for DC, 542 US 367 (2004).
I was thinking about how AI "thinks", and it is noticeably different from the way humans think. For example, I have a plant identification application on my phone, which I use by taking pictures of a plant, identifying the part (bark, leaves, fruit, etc.) and it provides a guess as to what plant it is.
I imagined someone showing another person various plants, and identifying them to the other person, especially to a NON-plant person. Something like this:
Teacher: See this leaf shape? You can tell it's a maple tree from the distinctive shape of the leaf.
Student: OK, yes, it's all pointy.
Teacher: Now this leaf is also from a maple tree, but a different type. See how the leaf is broader than the first one?
Student: I guess it's a bit thicker.
Teacher: This is bark from the first tree, and this is bark from the second tree.
Student: The bark looks the same. Isn't bark just bark?
Teacher: No, this first one is much more craggy. This second one has straighter lines to it.
Student: If you say so. They both look like bark to me.
It seems to me that AI is BETTER at finding differences and connections than the average person, for a given individual is unlikely to be specialized enough to care about the differences in most things. Yet something is still missing, so that any given person is probably better than AI at finding differences in SOMETHING, but not everything.
I'm reminded of looking at the AI generated art. I'm no art expert, so it was no surprise I didn't do that well in differentiating between AI art and human art. But experts WERE significantly better at it.
It may be that AI is going to be generally better than humans, but not better than top humans. At least, not until we have another AI breakthrough.
You are playing a cooperative variant of Rock-Paper-Scissors where the objective is to tie rather than win. You and your partner don't know one another and don't have an opportunity to coordinate in advance nor by side channels during the game.
Consider the following variants:
1. Only one round is played.
2. The game is iterated across, say, nine rounds. You play several rounds with the same partner with the objective of maximizing the number of rounds that end in ties.
3. Also iterated across nine rounds, but your objective is for you and your partner to have won equal numbers of rounds over the course of the series. Does it make a difference if fulfilling the objective is an all-or-nothing thing or if you get partial credit for, say, a 5-4 final score?
4. Instead of playing Rock-Paper-Scissors, you are playing Rock-Paper-Scissors-Lizard-Spock. The rules are:
2. First round Rock. Remaining rounds, if you tied last round repeat. If you didnt tie last round, theres 2 options: a) pick the one neither of you picked b) 50% own last play 50% partners last play. a) is theoretically the best joint strategy conditional on not agreeing on the schelling point, but never adjusts to constant strategies. b) does adjust, and 50% is the number that most cooperates with itself. Which one I pick depends on how smart I think the other guy is and thinks I am.
3. If youre currently balanced, same as 2. If youre not, repeat with 50% and switch to what would decrease the count if the other repeats with 50%. If you decreased the count last round, you both repeat. Partial credit doesnt matter.
4.1 same as 1.
4.2 The extension of 2a) is determined by putting the options into a pentagon. Then the three unpicked options are either adjacent, in which case you pick the middle one, or two are adjacent and one is separate, in which case you pick the separate one. (Or, always the apex of the isosceles triangle) 2b) is unchanged. Intelligence requirement for a) is slightly higher. I wonder how much difference it makes if youve played 2 before.
4.3 If youre currently balanced, same as 4.2. If youre not, then a) again using the pentagon, to each adjacent pair corresponds exactly one non-adjacent pair that is parallel, and vice versa. You both play the element of the pair corresponding to the last round that makes you win/lose as you should (following this strategy, its impossible to be unbalanced and tied last round, or to be unbalanced by more than one round). 4.3b) same as 3, except because you have two picks you can switch to, you pick the one that would be consistent with 4.3.a) if the last round was not tied, and the one thats not adjacent to the last round if you tied. You use either the a) or b) options consistently.
Question 2 is called Rendezvous Problem or also Mozart Cafe Problem in mathematics, introduced by Alpern in 1976. A common formulation is:
You and a friend have agreed to meet in the Mozart Cafe at noon in Vienna. But arriving there, you find out that there is not one, but n cafes called Mozart Cafe. You and your friend know each other well enough that both of you will go to some Mozart Cafe every day at noon and hope that you both choose the same one.
For n=3 this is the same as your question 2, because after the first tie, both players can continue playing this symbol again. So the question is just how many rounds you need in expectation to reach the first tie, and which strategy achieves this optimum. (Except that you truncate after 9 rounds, which will alter the answer very slightly.)
The question is surprisingly difficult. For n=3 it was only solved in 2012. The answer is 8/3 rounds in expectation.
The problem is also surprisingly nuanced. For example, if both players can prepare for the situation, without knowing anything about the cafes (in a fictive world with completely symmetrical situation between the cafes), then they can do better by deciding that one player stays put (always goes to the same cafe), while the other tries them out one after the other. But this only works if they can break symmetry between the players. Likewise, if there are other symmetry-breakers (e.g., a river flows through the city and divides the cafes into two subsets) then one can improve the 8/3.
Your partner’s choice being public information makes it easier. As pointed out by Schneeaffe, picking the option neither of you chose in round 1 always results in a tie in round 2. (And of course the case where you both chose the same option in round 1 is trivial.)
Yeah, I don't see a better strategy. Rock is the Schelling point. Variant 3 would involve both people randomly playing scissors until one person loses, and then you both play rock until the final round where you do the opposite match-up to the round that had the win/loss.
That's actually the first thing I was wondering when I came up with this. Rock was also my instinct for Schelling point and I wanted to know how universal that was.
If your opponent is also throwing rock every time, then yes. But if your opponent throws scissors or paper in round 1, then that would require the opposite result in another round to balance it out.
3. I do not believe he is a "libtard and a coward". If nothing else, he's got the US conservative mindset of what Hanania described as "overwhelmingly motivated by cruelty and hate" when it comes to immigration and immigrants, as part of the broader "suspicion of outsiders" that is the most common attribute of conservatives. It seethes through every point of his reply chain in the click-through link.
In any case, I agree with you on Trump - mostly. Pretty dumb, but I wouldn't underrate just sheer corruption in what the Trump folks do as well (some of them are very likely trading on the impact of his pronouncements and reap rewards every time he walks a tariff back or such).
First, are there top Top Men fleeing from US universities? Second, if I'm a Top Person fleeing the US, I think I'm more likely to head to the UK or to continental Europe than Ireland (unless I'm linking up with an American multi-national already established here).
Third, if the US academics are going to be paid higher salaries, this will kick off unrest and protests from Irish academics about two-tier systems and unequal wages and court cases about discrimination.
Fourth, what is viewed as a 'high salary' by Irish terms may (or may not) be buttons by US standards. If they are really looking for Big Names, those people with established reputations and careers are going to have expectations around pay and conditions. How long will the Exchequer fund the higher salaries? Are we really going to poach Professor Bigname from Harvard?
Now, to be fair, I have no idea what the average salary of a Harvard professor is, or the average salary of an Irish academic, and how one stacks up against the other. Unreliable online searching tells me Harvard ranges from $165-282k which is around €148-254k, and average Irish academic salary is €153k. So depending on how Big a Bigname they are going after, the pay could indeed be competitive. But I think that if you really want to advance your academic career, you're going to Europe or back to the US, not staying in Ireland.
I think there might be some chance of coaxing bright young academics who are research students or junior faculty from non-EU countries, but I do wonder how many Harvard exiles are going to want to work at Munster Technological University (formerly Institutes of Technology, formerly Regional Technical Colleges) for one?
I get what you’re saying here, but Ireland might make some sense if you’re planning on some major industry collaboration with some company that is there for *cough* tax reasons.
I think that is what our government is hoping for (Professor Bigname has a hot new idea, has lined up funding from American Megacorp, and all that is needed is to roll it out in Ireland where the new factory will be located), but how that would shake out I don't know.
Also, if you’re a pessimist about current political situation, you might guess that the Irish government is less likely to put you a death camp, and this overrides the salary concern.
I know Ireland probably seems boring to you, but when I was living in Michigan I spent a week in Ireland and it was f***ing amazing. It's so mind-bogglingly gorgeous and everyone has cool accents and everywhere I went the food was delicious.
If one day the President cracks down on US-based software engineering and I have to flee to Dublin, I will be pretty chill about it.
Imagine an American university president in 1930 saying “who would want to flee the University of Vienna to come to UCLA or university of Minnesota?” Turns out, a lot of people did, and it put these places on the map.
Damn it, now I want another one of those terrible Paddywhackery movies about an American moves to Ireland, only set in Dublin where our Ivy League Professor of Big Numbers moves to Trinity and it's full of spud-in-the-mouth Paddies and horse Protestants.
The cognitive dissonance between "this is not the hell what Dublin in 2025 is like" and the fakery of "Sure aren't the Cliffs of Moher and Skelling Michael just on the doorstep of dear old dirty Dublin?" would be both enraging and highly entertaining.
They would have a lot to live (or down) to in the tradition of "what the hell is wrong with the Yanks?" movie-making about Ireland:
My personal favourite in that genre is "Blown Away" where I saw it in the local cinema and at the flat pint scene every single person, including myself, groaned out loud. It's a fleeting scene where Jeff Bridges and his father, Lloyd Bridges, are playing an Irish-American father and son (edit: sorry, having read the synopsis, they are both Irish-American cops) having a drink together and they both light up with anticipation about the lovely pints they are about to consume. Except the pints are flat, which would not cause a genuine Irish pint drinker to be delighted if those were put in front of them to drink.
My Irish wife always complains about the geography of Ireland being ignored in movies, but that is true of everywhere. People stroll impossible distances in New York, leave London at noon to be in Scotland by the afternoon, go east to go west, and take a one way the wrong direction. This happens all the time and only locals know it v
Yeah - here in the UK perhaps the most famous one is Robin Hood: Prince of Thieves showing our heroes traveling from Dover to Nottingham via Hadrian's Wall. And the (much, much better) American Werewolf in London appears to show a man waking up in a London hospital after being attacked on the Yorkshire Moors (which you could justify by saying they took him down there to get access to specialist equipment, but the movie doesn't).
As someone who moved from a US University to an Irish one, the pay is competitive - most places in the US are not paying as much as Harvard. On the other hand, there is a major funding squeeze on Irish Universities. The 2009 austerity measures seem still to be in place - permanent staff do well, but it's difficult to get new positions approved (I'm currently in a position which will end in the Summer, mostly due to inertia and bureaucracy). It's infuriating that permanent posts for people on renewing one year contracts will likely be displaced by this initiative, which probably isn't going to attract top talent. The talk of sending out talent scouts to the US strongly suggests a closed nomination process where the universities themselves will have limited say in who lands where - it seems unlikely to me to work well.
Yeah that's the sticking point I see here - where are all the hundreds of millions to be poured into research going to come from? I can see from our government's point of view that they're hoping "smart/big name academics come here, already have idea that can be spun off into profitable commercial product, we fund start up, profit!" but they are not really contemplating "smart/big name academics come here, we have to match US standards of pay and conditions, plus match US levels of funding for 10+ years as they explore 'maybe this will result in commercial product'".
The end goal will be an increase in successful funding applications to European programmes. Ireland doesn't perform well on this front - this has been directly linked to longterm underinvestment in research here. As a personal anecdote, I finished my PhD in 2011, and applied for funding in 2012; the application was returned unassisted, because they weren't funding my area that year. Things have improved, but only slightly.
Whether it will work remains to be seen. It will certainly be the case that the majority of those employed will bring in no external funding at all. Success in these schemes typically runs at around 10%, with a strong bias toward funding applicants with a history of success already. I imagine the most likely outcome will be a small number of large grants being awarded to researchers in this scheme, not nearly enough to cover the costs of the scheme.
" I imagine the most likely outcome will be a small number of large grants being awarded to researchers in this scheme, not nearly enough to cover the costs of the scheme."
I think that is the most likely outcome. I imagine the government expects "big name academics and/or whizz kid junior academics with connections who can draw funding from American multinationals with them to spin off commercial applications of their research" so that the scheme will pay for itself/be a job creation scheme, but how that translates into reality is a different matter.
There's already far far more good researchers than research jobs on both sides of the Atlantic. Maybe not "great" researchers, but great researchers these days are characterised more by their ability to write grant applications, not to actually do work.
If you want to recruit top talent from US universities then you not only need to provide them US-level salaries but also US-level funding for their research. Nobody on that side of the Atlantic is going to be willing to do that. All these numbers are routinely two or three or four times higher in the US than in Europe.
Well, you see, the US has decided to cut their funding to the bone. So really, Europe doesn't need to do anything. The US will do all the self-sabotage Europe needs to benefit.
Are academic salaries decreasing threefold or fourfold? Genuine question.
And if they really are three or four times higher than in Europe, then it's not actually clear if it's higher expected to value to stay in the US, get nothing for four years and then return to normal, or to get European normal per year during that time. And that's assuming moving back to your old job is trivial.
They decrease to zero if you get fired because the university no longer has the money to employ you. And most academics getting this kind of salary are the ones that actually run the labs, and if there are no grants available they will not hire the postdocs and PhDs that do the research (and are generally far worse paid).
> There are no reliable figures yet on US researchers interested in moving to Europe. But 75 percent of US researchers who took part in a Nature survey are thinking about leaving the US. Of the 1,600 participants, more than 1,200 scientists are considering moving to Canada or Europe. The trend was particularly pronounced among researchers who are still at the beginning of their careers. Of the 690 postgraduate researchers who responded, 548 were considering leaving; 255 of 340 PhD students said the same.
> Leading German research institutions are receiving more applications from the US, including some top researchers they would love to attract.
> In early February 2025, the president of the Max Planck Society, Patrick Cramer, reported that applications from the US had at least doubled, and in some cases, even tripled.
PhD students and postgraduate researchers are always considering moving, because permanent jobs are very hard to get even with a stellar record and no geographical constraints. Those with geographical constraints are almost guaranteed to not find a job and have to leave the field.
I imagine that if you're a top tier US academic, you'll be looking at the likes of top tier European universities, not Ireland. This ranking site has Trinity at no. 26 which is a very respectable showing, but the other Irish universities are way down the rankings:
That may well be the case, but I was answering your first question - "Top Men" of Science indeed seem to be leaving the US at a greatly increased rate, looking for greener pastures. Also, money (as in US vs. non-US) is important, yes, but academic freedom, let alone personal freedom (as in "I don't want to be deported to a Venezuelan mega prison without recourse") should not be discounted out of hand as a potential contributing factor.
Suppose a time traveler from the year 1900 materialized in the modern-day US. I would expect him to be arrested in fairly short order for innumerable reasons, but what would happen to him after that? He has no documentation and, as far as the government is concerned, doesn't exist. Does he get deported? To where?
Most homeless people I've encountered are smelly, dressed for sleeping outside in clothes that look like they haven't changed, and appear mentally ill or suffering some sort of drug addiction. My central example of a time traveler from 1900 would not have these properties.
Interestingly, I think what we care about here is really "appears homeless upon a two-second inspection", as opposed to "actually homeless". Someone who appears to have showered recently, is wearing clean clothes, lucid, and not casting about frantically for a bottle or needle, is not going to strike me as homeless unless it comes up in conversation, and if it does, they're probably going to strike me as "temporarily embarrassed" and might indeed be treated similarly to an actual time traveler from 1900.
Because of the "to where" question, I think the answer has to be that no, he doesn't get deported. If the Trump admin makes an agreement with a third party country to just take anyone not in the US legally, then that would become a possibility. It seems that is not currently in place and has not been previously. Third party countries have very little reason to want to take on random people being expelled from another country.
If he makes a nuisance of himself, there is a good chance he spends time in jail or similar, but generally there wouldn't be grounds to hold him long and a bare claim of US birth might be enough to preclude attempts at sending him out of the country.
Even assuming they don't accept his claim of time traveling, what do you do with a guy who is clearly spouting nonsense but seems otherwise mentally whole and responsible? You arrest him if/when he commits crimes, but you otherwise let him go free because it's too much of a hassle to deal with him.
If he makes enough noise about this and gets broader attention, there must ways to ascertain that he is indeed a time traveler. Such as measure the concentration of microplastics in his blood. Something. There must be some biomarkers that would unequivocally support or disprove time traveling theory
Maybe, but he's hampered for a lot of reasons by coming from the past instead of the future. Anything that modern man could verify historically is something he also could have looked up. If modern man cannot verify the information, then they also can't tell the guy didn't just make it up. Someone from the future may be able to predict something to verify his knowledge. There's also the question of how time travel was invented in the past but we never heard about it and no one reinvented it since, even with a supposedly better understanding of time since then.
I don't know the science behind it, so maybe there are means to prove certain age bands. Another thread someone mentioned being able to prove post-1955 by radiation in teeth. No idea if that's possible. Obviously that wouldn't help him very much if he was coming from 1960 instead of 1900, but perhaps there are ages that it would be clearer. I fear that many/most forensic approaches would be hard to falsify. Maybe he has no microplastics because he's from the past, or maybe there are dozens of other explanations. I strongly suspect that we would be very much more likely to come up with such an alternate explanation than to believe in time travel.
>There's also the question of how time travel was invented in the past but we never heard about it and no one reinvented it since, even with a supposedly better understanding of time since then.
They tested it on this guy, and he exploded in a ball of fire. They concluded it was a flawed design, and also were arrested for blowing the guy up.
This is totally tangential to your question, but the scenario reminds me of the Norwegian series "Beforeigners", in which some sort of time portal has dumped crowds of Stone Age, Viking, and Victorian time travellers in Oslo. The somewhat clever twist to the setup is that 'beforeigners' get the same treatment as illegal immigrants, with the non-Victorians mostly living in ghettos where their 'beforeign' ways make them subject to xenophobia - despite them not actually being from somewhere else, just 'somewhen' else. The Victorians integrate a bit better, given their similar culture.
It's Nordic, so the main story involves a pair of cops (one modern, one a Viking) trying to solve a series of murders.
And I'm reminded of a satirical novel, _Look Who's Back_. Hitler appears in modern Germany published in 2012. Hitler appears in 2011. He's in a uniform, but it happens to be one without swastikas.
This is not a science fiction novel-- there's no interest in how it happens. Hitler struggles to understand enough of the world he's in, and is picked up by the media as a brilliant Hitler impersonator.
When I read it about 10 years ago, I thought it was a lot of fun. Maybe not so much now.
It was cool that Hitler would say "Jews are not a laughing matter". He meant it one way and his audience heard it differently.
There's a well-made German-language film adaptation of Look Who's Back which came out in 2015. I watched it a few years ago when it was on Netflix with subtitles. I didn't know there was a book; I'll have to check that out.
I remember the "Jews are not a laughing matter" line being in the movie, too. IIRC, it happened when he was prepping for the first episode of his TV show, and a network executive (or maybe the show's producer?) pulls him aside and asks him (awkwardly and somewhat indirectly) not to make jokes about Jews, which Hitler solemnly agrees to.
There are several scenes in the movie where Hitler is traveling around Germany in uniform and striking up conversations with random people he meets who think he's an impersonator. These scenes hit very differently in hindsight when I found out later that they were real Borat-style unscripted street interviews.
The best part of the movie (scripted, however) is when he meets the leadership of the modern heirs to the Nazi Party. They're clearly very uncomfortable with his presence and he rips them a new one for their worthless, imbecilic "management of the National cause".
If he's a smart time traveler, he'd show up with some kind of proof to show that he's actually a 19th century person and not just a crazy guy pretending to be one. Drop a time-capsule somewhere or do the Marty McFly thing on a sealed letter from Back to the Future II.
I don't think he gets deported unless he claims to be from a particular country that still exists in some fashion, but he gets treated like a crazy guy unless he does the above.
The criminal justice system processes plenty of people who won't give their legal names/records. For a minor offense they'll just take down whatever information is given, process him and then release him. This is the procedure for 'nuisance' crimes like trying to sleep on park bench.
For a more serious offense, he'll stay in jail and they'll spend more time trying to establish his legal identity and then shrug and give up when they can't figure it out. His lack of legal identity will make getting bail impossible but not otherwise affect proceedings much.
Legally speaking they can't deport him without figuring out his native country (and getting the country to accept). However ICE is often cavalier about the legal standards they're supposed to meet. Since he will not be able to provide proof of citizenship while he shouldn't be deported, if he is deported he's going to be pretty out of luck trying to get back in.
In the long term the problems he'll face will be pretty much identical to most illegal immigrants, with the same solutions (working in cash or getting a false identity).
There's an Alfred Bester story about people in the future being deported to the past. A lot of homeless people are actually forced time travelers who can't manage.
In the 2013 movie Only Lovers Left Alive, some several hundred year old vampires travel between Tangiers and Detroit by taking nighttime flights under false identification. It struck me that by 2013, the idea that you could take a commercial flight with false identification was already pretty totally dead, but it made me realize that even just a few years before then, it might have still been alive.
Lots of 20th century media involves the idea of running away and starting a new life somewhere else under a false identity. How much is that still possible if you don’t try to do anything like fly or vote?
I'm spacing on the name, but someone in trouble for crimes against humanity(?) went on the run and hid for years (decades?). He ran some sort of new age-related business when he was hiding.
If you don't try to do something that requires someone to ID you - vote, get a DL, buy beer - then how often do people really get asked for ID? At that point the only required uses I can see would be if you get stopped by the police, which is not so infrequent as to ignore, but I haven't been asked for my ID by the police for ~17 years?
That said, getting a job would perhaps be required for many people and is supposed to require some pretty clear ID. That should be very hard to fake and would end the experiment right away if all employers were really asking like they should be. I don't have much experience with it, but it seems obvious that a number of employers aren't even trying to get ID from employees. Less so if you're doing odd jobs or otherwise self employed and get paid in cash. There's plenty of businesses that prefer to get paid in cash for labor so they can decide how much their taxable income is.
Buying or legally renting housing should also require ID, but there's lots of ways around that as well.
He can prove that the person he claims to be existed and that he knows certain biographical details, but proving that he is that same person is a much taller order.
Fingerprints would probably be the best evidence if you could compare a set of pre 1900 records to current prints. I don't know when fingerprints started getting taken and kept for official records, but my guess is 1900 is far too early.
"I would expect him to be arrested in fairly short order for innumerable reasons"
This doesn't seem obvious at all. I'm pretty pessimistic about the authoritarian nature of the U.S. these days, but unless it's more quickly and more thoroughly than I've heard, there still aren't watchful cops lurking behind every bush, looking for people to snatch up. A particularly oblivious, foolish or belligerent time traveler might quickly attract the sort of attention that would lead people calling the cops. But merely being dazed and bewildered isn't likely to get you there. Having no ID or job history and thus no prospects for legitimate work is definitely a potential problem, but that's a medium-term problem not a short-term problem, and one that plenty of real-life people successfully navigate without running afoul of the authorities.
To address the actual question: now that you bring it up, it seems like there has to be a procedure for this, doesn't there? Not the time traveler part, mind you, but the "personal with no official, legal existence" part. In a population of 300 million people, it has to happen from time to time that the police arrest somebody whose records simply don't prove findable. But I have no idea what that procedure is.
Some more fictional evidence for a hypothetic situation: In the 2001 movie Kate & Leopold, Leopold (Hugh Jackman) gets in conflict with the law by not picking up his dog's droppings. Luckily, he is a rather modern-thinking gentleman for his era, so he doesn't disrespect the policewoman (Viola Davis) or call her a racial slur. This scene could have ended very differently.
The case where there's someone with no official legal existence is probably a lot less common than the case where "this guy might have legal existence somewhere but he's not carrying any ID and he's not usefully answering questions that would allow us to prove his identity".
In which case the procedure, if they haven't committed any serious crime, is to deem them too much trouble and let them go.
If he can persuade a responsible official to take his claim seriously as a real possibility, there are well-established ways of testing his claim. If he will consent to having a tooth extracted, then radioisotope testing of the enamel can establish with a very high level of confidence that that tooth grew before c. 1955 when large-scale above ground nuclear testing started making a huge spike in background radiation in the atmosphere.
If he's newly arrived, you can probably also do the tests with a lock of hair or a bone biopsy, but teeth are the most precise and convincing because of established protocols and because tooth enamel grows at a very specific time in childhood and isn't remodeled like bone or continuously growing out like hair.
If the person persistently maintains they were born in the 19th century and seems confused by modern objects, they might be temporarily held for psychiatric evaluation. But supposing the outcome of that evaluation is that the person poses no danger to themselves or anyone else, and the initial cause for their arrest is sufficiently minor to pose no interest to the DA, they'll have to be released.
If there is evidence of foreign origin, such as limited English-language skills, and if the police department is interested in referring the person to ICE custody, deportation might happen, but more likely the person is simply put back onto the street, or possibly connected with advocacy resources who could assist in re-establishing a contemporary identity.
A few other folks have hinted at this, but I want to be direct about it: @Scott there are so many people in the world who have legitimately interesting things to say. And you only have one second per second. As they say, don't feed the trolls -- why do you spend any time at all on Yarvin, who seems to basically be an edgy teen in arrested development?
There are obvious consequentialist impacts of giving airtime and legitimacy to arguments that are both heinous and stupid -- cf your own writing on Crazy Like Us. If you read the comments under his thread, there are countless people who are discovering Yarvin for the first time and are thoroughly enthralled by his bullshit philosophical justifications to destroy society. And it seems equally obvious that Yarvin is either arguing in bad faith or is so deep in an epistemic pit that your response cannot possibly push back on all of the crap at once. Like, he straight up says "[liberals] invented a pandemic and killed 20 million people" and it just slides by, you do not make the time in your response to call that out as obviously false. The result is that all of those people who are reading Yarvin for the first time will see that "fact", see that you do not respond to it, and go "O, I guess COVID was a farce. Yea, fuck democracy."
This is really bad.
I'm purposely leaning heavily on consequentialist framing here because you're an EA. Do you think that the upside of responding to / arguing with Yarvin is worth it?
EDIT: adding a comment I made from deeper in the thread.
As of this edit, Scott's response to Yarvin got just under 215k views. 1k likes, 65 retweets. Yarvin, by contrast, got 380k views, 2.6k likes, 175 retweets. There are 165000 people who only saw Yarvin's tweet thread, and who never even saw Scott's response and maybe never saw the original piece that kicked off the whole thing *since Yarvin doesn't link to it*. Just to put some numbers on this.
You've made so many long comments in this subthread saying almost exactly the same things in slightly different ways, so forgive me if I missed it, but multiple people have pointed out that mockery and oxygen-deprivation are symmetric weapons that can be equally well used against good ideas and bad ones, and I don't think you've addressed this at all. I agree with you that Moldbug/Yarvin/whatever-his-name-is-now is insane. Probably almost everyone in this thread agrees that he's insane. But your attitude is dangerous and terrifying, because your calls to mock or ignore him are *not based on an objective standard proving that he's insane*. You have not said Scott should mock or ignore all people who 80% of the population agrees are insane, or that Scott should mock or ignore people who have spent time in a psychiatric facility. No, you instead say Scott should mock or ignore Yarvin because clearly he's insane. Which is just another way of saying because *you* think he's insane.
Can you honestly not see how symmetric that is? How it is *just as easy* for someone on Yarvin's blog to say "don't engage with Scott or acknowledge his arguments, he's clearly insane"? How it's just as easy for me to say you're insane so everyone should mock or ignore you?
People here aren't necessarily averse to rejecting or deplatforming some arguments. They *are* for the most part absolutely averse to doing so *solely on the basis that the person calling for said rejection/deplatforming is really REALLY sure they're bad arguments*. Because, you know, they've seen this done over and over to arguments that turned out to be true! Ranging from "your rape accusation statistics have not the slightest connection with reality" to "there's some evidence covid may have been a lab leak".
Making a comment saying that Yarvin is clearly "heinous and stupid" (and so should not be engaged with) is utterly unpersuasive as an argument, because you could say that about anyone. It's based entirely on your own claimed authority to make this pronouncement. But there are two ways that you *could* give people strong, asymmetric-with respect-to-truth-and-falsehood reasons to agree that Yarvin is heinous and stupid. The first is to use some totally or near-totally objective standard to demonstrate this, that doesn't at all depend on your or any other person's personal judgement. Examples might be "Yarvin has been successfully sued for libel five times" or "Yarvin has been clinically diagnosed with paranoid schizophrenia". You didn't, as far as I can see, provide anything like such objective evidence in anything you've written for your claims about what he "clearly" is. The second is to actually engage with his arguments and rationally demonstrate that they don't hold logically and/or that they're dishonestly or maliciously motivated.
> You've made so many long comments in this subthread saying almost exactly the same things in slightly different ways
apparently i haven't been clear enough 😂
> Making a comment saying that Yarvin is clearly "heinous and stupid" (and so should not be engaged with) is utterly unpersuasive as an argument
> I agree with you that Moldbug/Yarvin/whatever-his-name-is-now is insane. Probably almost everyone in this thread agrees that he's insane.
I don't understand the confusion! It really seems exactly right that everyone in this thread agrees he is insane! So what is the point of me spending a bunch of time explaining what everyone in the thread seems to already know? Like, not a single person in this thread has said "actually I think Yarvin has some good ideas". I have explicitly asked people to make this argument and no one has.
So I keep coming back to Alex Jones.
Scott doesn't write about Alex Jones -- is this 'depriving of oxygen'? Scott doesn't spend time refuting whether or not the Sandy Hook Shooting happened. Is this "deplatforming"? No! There is no platform! This is a blog. It's not like, facebook, or substack, or twitter. It is absurd to make the case that everyone who Scott doesn't write about is somehow being deplatformed! Like, what?
And even if this was somehow the case, I don't think I need to spend time explaining on the merits why Alex Jones is insane. I can just, like, point to his body of work and say "It is self evident" and everyone here would hopefully understand that, and if someone did not understand that they are welcome to say "I actually think Alex Jones has some good ideas".
If everyone agrees that Yarvin has lost his mind, what are we even talking about? And yes, I do think he is heinous -- the dude is explicitly, openly racist (https://x.com/curtis_yarvin/status/1922234238444711949) in addition to everything else. Do I have to sit down and explain, in detail, why this tweet is racist? Or does it stand on its own as something that is obviously horrible and *the person who sincerely believes this should not be treated seriously*? In my opinion, no, it is not worth my time explaining why this tweet is racist -- the body of work makes this self evident.
> Can you honestly not see how symmetric that is? How it is *just as easy* for someone on Yarvin's blog to say "don't engage with Scott or acknowledge his arguments, he's clearly insane"? How it's just as easy for me to say you're insane so everyone should mock or ignore you?
Yes, I see the symmetry. I do not think it matters. I am not making a platonic argument about good discussion norms, even though several people have tried to put me in that box. I am asking a much more mundane question: "Hey, Scott, are you sure this is worth your time?" If someone, somewhere else in the world, was like 'dont engage with Scott', power to them. That is their right. If someone, somewhere in the world, said 'dont engage with theahura', that is *also* their right. There are billions of people in the world, you cant engage with everyone.
But as I said elsewhere in thread, if a *really famous person* told their audience 'dont engage with theahura', my standing would immediately increase overnight. If Trump said, on air, 'theahura is the worst, no one should read him', I would get hundreds if not thousands of subscribers immediately. So if you think my ideas are bad, and you have more reach/standing than me, it would be better to just not engage at all, at least from a consequentialist framing. This isn't some enlightenment salon where only 10 people are around, after all. This is the public web.
> The second is to actually engage with his arguments and rationally demonstrate that they don't hold logically and/or that they're dishonestly or maliciously motivated.
Rationally engaging with someone who a) is not acting in good faith and b) thrives off bad publicity is a great way to make Yarvin more well known. I think this is bad. You could rationally engage with a 5yo throwing a tantrum, but it won't get you anywhere, and instead will give the 5yo more attention which is exactly what they want. More generally, you are making a category error -- you admit that Yarvin is insane but think that we should still engage with his arguments. Why? How do you rebut an insane person?
I go back to the numbers I posted in OP. As of right now, Yarvin's thread -- full of blatant misinformation like "liberals created COVID to kill 20 million people" -- is sitting at nearly 400k views. Scott's response is sitting at 240k views. 160000 people have *only* seen Yarvin's show boating, irony filled, steaming pile of bullshit. And unfortunately, Yarvin is a showman. So he knows how to write things in a particular kind of tumblresque way that seems convincing if you don't stop to think about it for 5 seconds, which most people won't. If you think he is insane, you have to deal with the fact that Scott treating him seriously has led to more eyeballs on the work of a madman. This isn't a hypothetical, we can see this objectively. So if you are a consequentialist, I ask a simple question: is this good?
Scott may have other reasons for responding to Yarvin in a serious way, but that is why I framed my question the way I did -- as earnest inquiry. Maybe he thinks Yarvin *does* make some good points, or maybe he thinks that it *is* worth responding to Yarvin despite the negative effects of spreading Yarvin's shit around. Maybe he thinks Yarvin can be reformed. Idk. He's been silent on the matter and I suspect we aren't going to get a response, which is also fine.
You're right, I think... or if you're wrong then I'm also wrong, but... and hear me out on this, it's bad that you're right.
Why? Because basically the upshot of all this is that there are some topics that Scott, or anyone with a substantial audience really, is forbidden from discussing in good faith. Instead these topics must, at least in public, be approached with an attitude of ridicule from the get go. Any failure to submit a sufficiently scornful general disclaimer imposes an extra burden of explicitly calling out every individual point of difference and endlessly caveating every point of assent to forestall any impression of general agreement that may be fermenting in the mind of our impressionable reader.
> The thing about an online comment section is that the guy who really likes pedophilia is going to start posting on every thread about sexual minorities “I’m glad those sexual minorities have their rights! Now it’s time to start arguing for pedophile rights!” followed by a ten thousand word manifesto. This person won’t use any racial slurs, won’t be a bot, and can probably reach the same standards of politeness and reasonable-soundingness as anyone else. Any fair moderation policy won’t provide the moderator with any excuse to delete him.
Scott doesn't quite go this far in his post, but it's reasonable to ask: what *should* be done about this kind of person? You can't, like, do a beat by beat takedown of this manifesto every time it gets posted. Even if you could, you basically immediately elevate this person's standing, while possibly tarring your own.
It would be great if we lived in a reality where the public sphere became a place of enlightenment and earnest clash of ideas. In practice, only the worst ideas get spread, everything is taken out of context, the confirmation bias + Chinese Robber Fallacy dooms us, and our own engagement algorithms amplify our worst traits such that more reasonable takes never even see the light of day. I'm actively trying to think about how to solve for this, but right now no one has a solution
This is only a problem on the internet, and can be added to the long list of downsides that, in my opinion, tip the scales to the side of being a net negative. For all the good on here, I'm becoming increasingly convinced that humanity would be better off if the internet ceased to exist.
Eliminate all of the polarizing pseudo-debates in which both sides make claims and their respective supporters fawn over them and deride the other side in the comments. Create a real physical forum for meaningful political debate between public figures. The current state of the information world is just awful.
> I'm becoming increasingly convinced that humanity would be better off if the internet ceased to exist.
Increasingly of the same opinion. I think there is regulation that could make this better -- a very low hanging fruit example is "Make all algorithmic personalization opt-in instead of opt-out". A less low hanging but still valuable approach is something like requiring all public social media above a certain size to have community notes, or to require that they all show some measure of life experience / expertise.
In general, government exists to try to internalize these externalities. I think there is a way forward, but not if everyone continues blindly following their incentives. This is prime moloch territory.
Funny that you bring up Moloch here because the only solution to Moloch is monarchy, and isn't the reason that you brought this up in the first place because you're scared of the monarchist?
I'm sorry, this is an insane gigabrained take. I know Scott said this in his original post, and I thought he was wrong when he wrote it.
How do you think the EPA was created? Or, like, cigarette legislation? Or insider trading laws? Or GDPR? Or any of the other hundreds of coordination problems that have already been solved by democratic governments? Companies often request additional regulation in democracies because they do not want to be caught in competition traps! Saying "well the only solution is monarchy" is exactly the kind of edgy bullshit that Yarvin thinks he's so smart for saying, without actually doing the hard work of cracking open a history book or even the daily news for 30 seconds. He created an extremely shallow and incorrect model of 'human behavior', then reached a nonsense conclusion, and said "welp I'm done, I've solved political philosophy". Starting from the premise that absolute monarchs don't have to deal with incentives is so far off it's not-even-wrong. "Democracies have all these real world practical problems, but if you pretend like absolute monarchs are all idealized and don't have real world practical problems then absolute monarchy is better, and also cows are spheres". If you believe this, congrats, you have gigabrained yourself.
Also, like, for people who supposedly care a lot about science, this is just such a profoundly bullshit take just on the empirical level. There are something like 150 democracies in the world, representing billions of people. Assume theyve been alive for 50 years on average, thats 7500 democracy-years. You think all of that is just decline? What? There are absolute monarchies in the real world that you can compare against and they all suck! Like *clearly* Oman is worse than the Netherlands.
Government *is* a solution to moloch. All governments exist to solve coordination problems, and also have to deal with the incentives of the people that make up that government. Concluding that therefore you can only have monarchy is maybe an interesting take for an 8th grade history essay, a good learning opportunity. But Yarvin -- and you -- should know better, no matter how many layers of clever weaponized irony he tries to hide behind.
It's definitely Moloch territory. Destructive competition and bad incentives become more relevant every day and I can't wait for the pendulum to swing back in the other direction.
There's one regulation that I'd most like to see enacted and I think most people (of voting age) could agree on, and that's a total ban on algorithms for minors. It was bad enough when we were just publicly ranking our friends and comparing ourselves to others. Giving children their very own Skinner box that's available 24/7 is one of the more deranged things widespread in society today.
Well, the administration seems to have a new idea on how to solve this issue. With force. It seems you and they are in agreement that bad ideas need to be silenced. You're only in disagreement about what those ideas are.
When your concern is "social contagion", yes, they are the same thing. When your goal is to prevent a person from having any avenue from expressing their opinions by pressuring everyone into deplatforming them, there is functionally no difference. Of course, liberals did half-ass this effort. Perhaps they should've pulled the trigger first... but it's too late for that, isn't it?
I said nothing about preventing a person from having an avenue from expressing their opinions. I also did not say anything about deplatforming. At no point have I said that Substack or X should remove Yarvin. I'm not really sure what your political persuasion is, but you seem to be bringing in past scar tissue that is not relevant here.
What I *did* do is ask Scott a question: "why do you spend any time at all on Yarvin, who seems to basically be an edgy teen in arrested development?" And then I gave my reasons for why I spend no time on Yarvin, and why it may be harmful generally for Scott to do so.
If you disagree that Scott's actions are harmful on the merits you are welcome to make that argument. For example, you could take any of the following positions:
- Yarvin actually has ideas worth debating
- We should actually debate everyone regardless of how bad faith they are
- Consequentialist framings are incorrect in this setting, we should instead be looking at this through <some other lens>
Maybe Scott does it because he enjoys it, did you think of that? Scott is not a political operative or a lobbyist, so why are you demanding he act like one?
I'm not passive aggressively demanding anything either. I asked a question: "why do you spend any time at all on Yarvin, who seems to basically be an edgy teen in arrested development?"
I assume there is something that I am missing.
To me, Scott is arguing with the intellectual equivalent of Alex Jones. I legitimately do not understand why he would do this. I provided my reasoning for why *I* would not engage with Yarvin, which is that engaging with him seriously has negative consequentialist effects.
Like it or not, Yarvin has influence and is an important person. Rising up to the challenge is a far greater use of time rather than shirking back and hoping his ideas die on their own.
This constant idea you generally see from the left of censoring/ostracizing bad ideas has clearly *not worked*. You need to be able to give people tools on how to counter them.
It's amazing that the exact opposite of what you said is true. Engaging too much with bad ideas is what has given them fuel in the modern information environment. There is no free market of ideas.
Then why engage with it? There’s just no evidence that debating crazy ideas kills them. The hope that a neutral party looks at two sides of a debate and thinks “I’m going to believe the correct side” is not reliable. When you engage with crazy it has the effect of normalising crazy. The best thing to do is suppress bad ideas into very fringe corners of society. We don’t want the world to be run by people who think measles is curable with chicken soup and we have no time to debate them, we need to use every tool available to determine and shun ideas which are dangerous and wrong (to our best imperfect abilities). It worked extraordinarily well up until unregulated media took hold.
Heliocentrism. Four humors. Phlogiston theory. The rain follows the plough. Divine right of kings. We may have the corpses to admire and study, but many bad ideas are very much dead.
Some bad ideas, though, are /deliberately kept alive/. They keep coming back relentlessly, like superbugs in a dirty hospital ward after each round of antibiotics.
The best takeaways from the rationalist movement, the most useful parts of the sequences are the explorations of the edges of human reasoning - the explorations of when, how and why we fail to arrive at truth, and strategies for reasoning through problems in ways less likely to fail than just trusting our instincts. Universities generally attempt to teach similar things, though success varies.
The crowd here is more likely to have either or both sets of tools available than the average internet denizen; and yet every open thread is filled with discussions of bad ideas that just won't die.
We live in a world of gish gallops and similar firehose-of-concepts strategies, where a few ideas that are bad yet particularly attractive to some are repeated faster than each instance can be dismissed. Ostracising may not work, but neither does giving people tools to reason things out for themselves.
While they remain abstract internet arguments, walking away from the darker internet corners remains a reasonable response. But when there is a direct pipeline from those corners to daily life via the highest levels of government, the situation becomes untenable.
In a world that is tired of experts, how do /you/ suggest we deal with bad ideas?
Alex Jones has influence and is an important person, and arguing with him is a total waste of time. The only thing that arguing with Jones does is give him legitimacy. The same is true here.
More generally, if we believe that social contagion is at all real, then the worst thing that we could do with these ideas is *spread them*.
I am *not* saying "censor" nor am I saying "ostracize", because both of those actions take the underlying argument as something that is serious and necessary to respond to. I am saying "do not treat him seriously". He is a clown. A joke. Not even worthy of a footnote. If other people want to talk about him, sure, whatever. People talk about all sorts of things, like celebrity gossip or which superheroes could beat up other superheroes.
But Scott doesn't go and write 5000 word essays explaining why Iron Man actually could take Batman in a fight. He also doesn't go and write 5000 words explaining that, actually, the Sandy Hook shooting really did happen, or that the water does not actually turn the frogs gay -- both things that Alex Jones has continued to push.
Scott owes no one an explanation for how he spends his time, except for his wife and kids. Certainly not to me. But looking at this from the outside, I'm sad that someone of his talent is using his time to do...this.
What do you want him to do, kill him? Obviously debating him won't amount to anything, but surely you see value in studying this new ideology that has seized control of this country. It's also a little bit late to be worrying about "social contagion". The damage has already been done.
I feel like I've been very clear about what I want him to do, which is nothing
> Obviously debating him won't amount to anything
Unfortunately this is part of the problem. In my OP I say "And it seems equally obvious that Yarvin is either arguing in bad faith or is so deep in an epistemic pit that your response cannot possibly push back on all of the crap at once. Like, he straight up says "[liberals] invented a pandemic and killed 20 million people" and it just slides by". Debating him is actively harmful.
> surely you see value in studying this new ideology that has seized control of this country
I do not, because I do not think there is anything here to 'study' -- there is no intellectualism here, only intellectual people losing their dignity. I prefer Hanania's approach on this, which is to mock it relentlessly.
That to me seems to be what you are arguing about, because otherwise 'this is Scott's blog, he can do what he dang well likes and if he wants to write 5,000 words about gay frogs well that's Hidden Thread 999'.
In the OP I ask Scott a question -- "why do you spend any time at all on Yarvin?" This is legitimate curiosity. Scott is a brilliant person, and I have a lot of cognitive dissonance from him spending time taking a troll seriously.
I back that question up with what my motivations are for NOT spending any time at all on Yarvin, which includes 'oxygen of publicity'. Scott hasn't responded, so until he does I'm more or less left to explain the oxygen of publicity position.
(There are other concerns I have too, like "Yarvin is a bad faith actor who will purposely twist your words to score internet points and cannot be convinced or won over" or "I would like to never have to think of Yarvin again and I wish it did not appear on my feed through ACX", but the former is likely less compelling to Scott and the latter is just me being selfish)
Seems to me that there's a middle ground between "earnestly engage with Yarvin as though he didn't manifestly have end-stage brain worms" and "pay him no mind whatsoever even though he still has influence over people whose hands are on the levers of power."
Yea, I think that's what Hanania does -- a good middle ground between mockery and actually pushing back on bad argumentation. Scott is unlikely to be as vicious as Hanania is, though.
I saw a Note the other day where somebody was bragging about his Substack having reached #4 in the History category, and right below him in the rankings, to my surprise, was Yarvin's Grey Mirror. So he has a decent sized readership, it would seem, and those readers (this being Substack) are probably a good deal smarter and more educated than Alex Jones', and thus might be more persuadable, so I think I'd disagree with the idea that engaging with these ideas is a total waste of time.
> And it seems equally obvious that Yarvin is either arguing in bad faith or is so deep in an epistemic pit that your response cannot possibly push back on all of the crap at once. Like, he straight up says "[liberals] invented a pandemic and killed 20 million people" and it just slides by
> To be clear, *my* position is "If you must talk about such things, do so with the disdain and mockery that such positions deserve -- both because that is the correct response and also because it will help dissuade people from taking such positions seriously."
> I prefer Hanania's approach on this, which is to mock [such things] relentlessly.
> But Scott is kinder than I am, and I know that he will not do this. Given that, the next best thing is to not engage at all.
Its nice to pretend that these debates are all happening in an Enlightenment Salon where only those committed to truthseeking as a project will hear the arguments in the proper context.
Unfortunately, these debates are happening on Twitter.
Scott's response to Yarvin got just under 215k views. 1k likes, 65 retweets. Yarvin, by contrast, got 380k views, 2.6k likes, 175 retweets.
There are 165000 people who only saw Yarvin's tweet thread, and who never even saw Scott's response and maybe never saw the original piece that kicked off the whole thing *since Yarvin doesn't link to it*
I don't think it does. He's much, much better as a critic of liberal "democracy" than he is at outlining a superior alternative.
I think his premise is solid but his conclusions don't follow. Still, I view him as one of the leading intellectuals on the right, so he's worth critiquing for that reason alone.
It's actually pretty common for intellectuals to have a very coherent and sensible critique of the current system and not a very well-thought-out proposal for a better system.
The list of things that "more or less hangs together, kinda", includes the pre-implementation vision of communism. And many others of the same ilk. Do we want a policy of never speaking against such things until *after* they have been implemented and then fallen apart?
Yes, talking about them acknowledges that they are worth talking about. The bit where other people are already talking about them, already established that. They *will* be talked about, and if they're at all clever they'll couple all that talking to the real problems of the current system, and if those problems are substantial, who's going to win? The people who are being talked about because of their bold, decisive plan to fix the problems, or the people who are, what, talking about nothing at all? Talking about boring incremental technocratic measures to mitigate the problems?
Someone needs to be talking about all the horribly bad ideas that "more or less hold together, kinda, until they're actually implemented". This needs to be talked about anywhere people ideologically adjacent to the proposed whatever-ism gather, and for Yarvin's particular whatever-ism, that includes here.
To be clear, *my* position is "If you must talk about such things, do so with the disdain and mockery that such positions deserve -- both because that is the correct response and also because it will help dissuade people from taking such positions seriously." **
But Scott is kinder than I am, and I know that he will not do this. Given that, the next best thing is to not engage at all.
**Caveat that there are of course times and places to respond gently as a teaching moment -- for example, when talking to a young student, or someone who we have good reason to believe is making an earnest mistake. Yarvin is neither.
You want Scott to sneer like John Oliver. That isn't his style and his rejection of that crap is why Scott got popular in the first place. Would you have him delete the Anti-Reactionary FAQ from the archives?
No, I don't want Scott to sneer like John Oliver. *I* would probably sneer like John Oliver if I was going to write about Yarvin, but I doubt I will ever do so. Scott is not me. I literally wrote:
> "But Scott is kinder than I am, and I know that he will not do this. Given that, the next best thing is to not engage at all."
I...pretty clearly think it would be better for Scott to not engage at all. I don't know how you got "You want Scott to sneer like John Oliver" when I straight up said "the next best thing is to not engage at all".
As I just responded to your other comment, maybe there is something that I am missing about why Yarvin is worth a passing thought.
> Would you have him delete the Anti-Reactionary FAQ from the archives
The problem here is that disdain and mockery work about as well for dismissing smart or good ideas as they do for dismissing dumb or evil ideas. If Moldbug has some ideas worth discussing, then we should actuall discuss those ideas, honestly engage with them, and maybe then we decide they're crap and don't support them. Of course, Moldbug's literary style of smoke screens and squid ink makes it hard to actually engage with his ideas (that's one reason I'm not overly fond of his work), but the point is to get smarter, not to consign the stinky outgroup members to the nerds' table.
Suppose Alex Jones suddenly appeared and asserted that he did not think you had any ideas worth discussing. Suppose further that he happened to do it first. Should we just as righteously block you from further engagement?
I've noticed you're employing this tactic that ought to be in a 101-level course on What Not to Do in a Discussion, where you assume you possess the authority on what ideas are worth considering and follow through to which sources are worth considering and proceeding next to a total shutdown, without apparently realizing that you don't naturally possess that authority, and that it could therefore be used against you. And then you compound the problem by abusing everyone else's sense of not employing this tactic, by forcing them to spend their time arguing with you about whether this is a good tactic. At that point, one of the few natural heuristics left to people is to consider which person appears to have rhetorically run amok, and is therefore a threat to the consideration process itself, and shut *that* person down.
I thought I'd give a shout out to my son's school, Achievement Unlocked in Brooklyn, NY, which is currently recruiting for the next school year. It is a twice-exceptional school for gifted neurodivergent children and it is run by a rationalist. It has comments on the wall like, "The map is not the territory,' and "Never ascribe to malice that which can be explained by ignorance." I figure there may be others on this list seeking something like this. I'm attaching a flyer for their upcoming openhouse, and of course if you have more specific questions about my experience you can contact me directly.
This is what another parent wrote about it:
"
Achievement Unlocked (AU) is a small private school in Brooklyn that serves students in 1st-8th grade that have struggled in other schools. AU has a neurodiversity positive environment, reflecting that many but not all students are neurodiverse. AU is not a school that turns kids away who have had behavior issues nor is there any specific diagnosis that they are looking for. They really seek to serve students for whom other schools, both public and private and often both, have not worked. It is so worth scheduling a conversation with them if you are interested but unsure your child would fit there.
There are a max of 25 students. Classes are currently 5 students or less with a max of 6 students at any time. It seems to me there are 2 to 3 teachers/specialists present in every class/subject. My son has differentiated groups - he studies with the youngest kids in ELA and with older kids in math and even within these groups the work is individualized. We get daily report cards showing how he did with academics and behavior in each subject. Also, in addition to their regular academic classes every student takes Independent Study and gets to work on projects of interest that help them stretch their skills and experience and has a choice of many enrichment classes.
Social skills and support is scaffolded through the whole experience - for example, each student works on a self-skill plan in which they pick a skill to develop and negotiate with teachers and staff how to have that skill supported. This can be something they want to develop or something they want to be able to do. The school uses the Collaborative Problem Solving model proactively to do this. They help students proactively figure out how to make something happen and solve problems. Another example is they also have a system in which students are always earning chances to make "reasonable," (easy to earn), "less reasonable" (harder to earn) and "unreasonable" (very hard to earn) requests. A reasonable request might be staying indoors for recess or spending 5 minutes in the quiet room. An unreasonable request might be "the whole school doesn't get homework for a day." Students earn these by doing well, helping others, and accomplishing tasks. To earn an "unreasonable request." students have to do something quite challenging like memorizing the periodic table or beating the school director at chess. This way students are constantly learning about advocating for themselves and what is reasonable to request. My son LOVES this system. Homeroom teachers also regularly help support and prime students for social interactions and choices they might make throughout the day. For example, my son's homeroom teacher always thinks with them about what they want to do at recess and how they can accomplish it, which has really helped him.
Finally, the staff is tremendous! The school founder is an amazing role model, an active participant in developing each child's plan and experience, and the leader for developing the systems and atmosphere of the school. He ensures the school uses research-based approaches and collects a lot of data about how the kids are doing, like how many times they ask for a break and their daily behavior and academic scores. The classroom teachers, related service providers, and support staff and the admin team are also dynamic, talented, and caring.
In parent meetings recently, it struck me that as parents talk about the school, the most common phrase I hear is "no matter what" - for example, "AU will support your child no matter what," "AU staff will welcome your child with open arms, no matter what happened the day before," and "AU staff will never be disrespectful or mean to [parents], no matter what." I have never before encountered a school before that actualized the principle of unconditional positive regard as a whole organization - but AU definitely does."
I had to look up what "twice-exceptional" was because it's a term I've never heard of before. Another example of "all American geese are swans":
"A twice-exceptional school is an educational setting that caters to gifted and twice-exceptional students."
Okay, so "What does it mean for a student to be twice-exceptional?"
"The term “twice-exceptional,” also referred to as “2E,” is used to describe gifted children who have the characteristics of gifted students with the potential for high achievement and give evidence of one or more disabilities as defined by federal or state eligibility criteria."
So kids with special/additional needs who do not have an intellectual disability? Good luck to the school and the kids, but we are getting on the euphemism treadmill here!
To be clear, I'm not criticising you, Laura, but there seems to be this push for "you don't have a disability, you have a superpower! you're not just exceptional, you're twice-exceptional!" around disability (both physical and mental) that is well-meant as encouragement particularly for children so as not to consign them to the basket of "disabled and incapable", but which hampers the people suffering these problems. No, my autistic tendencies are not a superpower, they don't make me super-duper-extra-mega-with a cherry on top exceptional, they've held me back and are something I struggle with daily, and if I had had to handle the rah-rahing about "twice as great because of your problems" on top of that, I'd have melted down a hell of a lot more than I already did and do.
> So kids with special/additional needs who do not have an intellectual disability?
No, not that at all. It's kids who have special/additional needs in addition to being exceptionally bright and not merely "not intellectually disabled". The kids need to meet the gifted cut-off in addition to having additional needs.
I'm sorry you have such a big chip on your shoulder. Twice exceptional is an important term used to differentiate schools catering to children with disability and normal or high IQ from those catering to children with disability and low IQ. This is an important distinction when you are visiting schools. What is the curriculum? What is expected of these students? The public schools stick all kids with disability in the same classroom - so you can have a child with mental retardation in the same room with a kid who has some behavioral difficulties but can do calculus, as well as a normally developing child who has a phyiscal impairment. This is not an acceptable situation. The school doesn't blow smoke up the kids asses about them being 'exceptional,' though I know there are parents who cling to this - but they would anyway.
Twice-exceptional just sounds like such a round-the-houses way to try and avoid at all costs acknowledging "the kids have problems that interfere with their normal development".
Maybe I have a chip, but this kind of "no no there is no *problem* there is only *opportunity*" approach can leave people twisting in the wind - if there is not a problem only a second heaping helping of exceptionality, then there is no need to do anything to address or support the problem which does not exist, now is there?
I'm glad the school is very supportive and seems to be doing the proper job, but the euphemism treadmill seems to be in full swing (you're not retarded, you're special - oh darn people are using 'special' to mean 'retarded' - well now you're not special, you're exceptional - wait, what? okay - you're *twice* as exceptional as the ordinary exceptional kids, by which we mean the ones who are exceptional and not, you know, 'exceptional')
Yeah, I'm with Laura on this. I think maybe it's just that Deiseach isn't as immersed in the context. In North American educational circles "exceptional" is not necessarily a compliment or a euphemism. It's a neutral term for kids who do not fit the mold for whatever reason, be it positive or negative. This is the most literal meaning of the word "exception" after all.
Also, as a neurodivergent myself, I recognize that my autism and ADHD ARE both a disability and a superpower. I'm not trying to sugarcoat reality, this is truly the most accurate assessment I can come up with. It's not all good or all bad but it's different, i.e., exceptional. Basically, almost everyone has a car and I have a motorcycle. Is it an advantage of a disadvantage to have a motorcycle rather than a car? There isn't one answer to this, it depends on the situation and your criteria.
I (UKl also took "twice exceptional" to be intended as complimentary. Perhaps it's a US vs UK/EI thing - if I referred to a kid who was struggling or acting up as "exceptional" I would sound like I was being sarcastic ("wow, that kid sure is something"). Thanks for the clarification!
It does seem to be a US English versus British English thing. Maybe they are so used to "This term that seems complimentary is, in fact, whitewash on a dungheap because we all know it's code for 'got problems' but you can't say that out loud" that they recognise "okay, twice-exceptional means twice the problems".
In his thread, Yarvin suggests that liberal elites are worse than Facebook moms at medical policy because they caused COVID, and also says they turned America's cities into desolate and dangerous places, and in your response you say "I agree there are many bad things about the liberal American regime."
Do you actually, materially agree with Yarvin? Or do you just agree with him in the sense that I might agree with somebody who thinks Joe Biden is a lizard-person (we both would have preferred if he hadn't run for re-election)?
He's either a grandstanding bullshitter or he's delusional. Cities are safer now than they've been since the early sixties, and theyre much more livable -- better urban design and less pollution. And how pray tell did the liberal elites cause COVID? Honestly, his untruths are mendacious, and he's practicing the Paranoid political playbook as described by Richard Hofstadter back in the 1960s. Yet somehow people seem to take him seriously around here.
Even acknowledging that cities are much better than they were in the past (especially the fairly recent past of the 1990s), it's still worthwhile to note negative trends as they happen.
I don't read Yarvin's stuff, but I presume he means either the gain of function research or things like the covid lockdowns that caused significantly damage to people and the economy.
Hey, don't knock the lockdowns. NPIs (non-pharmaceutical interventions) made a significant statistical difference in the mortality rates between states that used them and states that eschewed them.
Also, I got a big chuckle out of MTG who claimed the Dems were giving the Republican caucus COVID. The Dems mostly masked, while the Republicans refused to mask during the B.1, Delta, and Omicron waves. The fact that they kept catching COVID was sinister plot by the Dems according to MTG.
Being a data-driven person, at the outset of the pandemic, I started charting cases and deaths in nine counties of the SF Bay Area (where I live), the Seattle area, and NY City. Dr. Sara Cody of Santa Clara County persuaded the health officers of the nine counties in the SF Bay Area to push for work-from-home policies, restrictions on public venues, masking, and social distancing. Likewise, the public health officers in the Seattle area urged the same NPI programs. These were recommended the last week of February 2020. My employer and most of the other tech companies in the Bay Area had workers work from home. Mayor London Breed of San Francisco announced a shelter-in-place policy on March 16th, but the freeways and cities of the Bay Area were already ghost towns by the first week of March.
OTOH, Mayor De Blasio (sp?) of NY City followed the initial CDC and WHO guidance (wash your hands frequently and go about your business normally). From March through April, NYC's cases and deaths climbed quickly in NYC. Hospitals were overwhelmed. Refrigerator morgues were trucked in, and the dead were carted off to mass burials. While in SF Bay Area and Seattle, cases and deaths climbed at a much slower rate. Emergency rooms and Critical Care facilities were stressed, but we didn't require refrigerator morgues and mass burials. Gavin Newsom was persuaded a couple of weeks into March to apply Sara Cody's NPI recommendations to all of California. By that time cases and deaths in LA were starting to, and their healthcare system was overwhelmed. And a few days later, the Governor of NY (whose name escapes me at the moment) ordered lockdowns across NY State (over De Blasio's objections). Too bad I can't post graph I created in substack comments, because it's pretty stark.
However, serious researchers also noticed a difference between counties that used NPIs and those that did not...
IS THE CURE WORSE THAN THE DISEASE? COUNTY-LEVEL EVIDENCE FROM THE COVID-19 PANDEMIC IN THE UNITED STATES, by Catalina Amuedo-Dorantes et al
>Our estimates suggest that advancing the date of NPI adoption by one day lowers the COVID-19 death rate by 2.4 percent. This finding proves robust to alternative measures of NPI adoption speed, model specifications that control for testing and mobility, and across various samples: national, restricted to the Northeast region, excluding New York, and excluding the Northeast region. We also find that the adoption speed of NPIs is associated with lower infections, as well as lower non-COVID mortality, suggesting that these measures slowed contagion and the pace at which the healthcare system might have been overburdened by the pandemic. Finally, NPI adoption speed appears to have been less relevant in Republican counties, suggesting that political ideology might have compromised their efficiency.
And red states that valued personal freedom over public health fared worse that blue states...
Of course, then Omicron came along in December of 2021. Omi was more infectious than any of the previous strains. NPIs were being loosened and people were being vaccinated. After Omicron the death toll became more evenly distributed per capita across the country.
>Cities are safer now than they've been since the early sixties,
I wouldn't be so quick to gloss over the very noticeable increases in street crime and general public incivility since 2020 nor the even more noticeable increase in the number of visibly homeless and unwell people since 2010 or so.
But even so, it's deeply unhinged to jump from there to "American cities are desolate and dangerous places." There was a huge amount of room between the state of the American city in 2009 and "desolate and dangerous," and even though the movement since then has been in the wrong direction, there's...still a lot of room between 2025 cities and the hellscapes that people like Yarvin imagine our cities to be.
How do you measure public incivility? But it doesn't look like cities are the hellholes that Grima Moldbug claims (see below). Of course, if you consider homelessness to be a crime, then indeed the problem has gotten much worse over the past 25 years. And it doesn't seem to be localized to US urban centers. It's becoming a big problem in Canada and the UK. Canadians blame substance abuse and psychological problems is the cause. UK claims the cost of living drives people to homelessness. I think these explanations are incomplete.
</ChatGPT warning>
Over the past 25 years (roughly 2000–2025), non-violent crime in U.S. urban areas has generally declined, though with some fluctuations depending on the type of crime and time period.
1. Long-Term Trends (2000s–2010s)
According to FBI Uniform Crime Reporting (UCR) and Bureau of Justice Statistics (BJS) data:
Property crimes (which are non-violent and include burglary, larceny-theft, and motor vehicle theft) declined steadily in most urban areas from the early 2000s to around 2014–2015.
Examples:
Burglary and larceny saw consistent drops.
Motor vehicle theft dropped significantly through the 2000s but started rising again in the late 2010s and early 2020s.
2. Recent Trends (2020–2024)
There were increases in some types of non-violent crime after the onset of the COVID-19 pandemic.
Motor vehicle theft rose sharply, especially in cities like Chicago, New York, and Washington, D.C.
Some cities saw increases in retail theft/shoplifting, leading to headlines about “crime waves,” though these increases were often localized or exaggerated.
Despite these spikes, overall non-violent crime levels remain significantly lower than in the 1990s or early 2000s.
3. Caveats
Urban vs. national data: Most national statistics are aggregated and may mask local variations.
Data quality: After the FBI switched to the National Incident-Based Reporting System (NIBRS) in 2021, fewer police departments reported data, making recent year-to-year comparisons more difficult.
Bottom Line:
No, non-violent crime in U.S. urban areas has not increased over the past 25 years overall—it has decreased significantly since the late 1990s and early 2000s, though there have been recent upticks in specific crimes like auto theft.
Let me know if you want a chart or links to specific crime databases or reports.
What you are not taking into account is that the violent crime rate for 90% of the population of a major city has stayed the same since the 90s at basically 0. The violent crime is concentrated in very small parts of the city, so unless you live or go there often (which there isn't any reason to if you don't live there) this doesn't really play into your perception of safety.
Most city residents experience what I did living in SF. Violent crime was never a threat to me and still isn't, but I notice a massive increase in homelessness and public drug use. I go out to the suburbs to go shopping because the CVS down the street locks up half the items and makes shopping a pain in the ass. This didn't happen 20 years ago. And what makes it 10x worse is that there is a degenerate city government that refuses to enforce the law.
If your degenerate city government enforces the law, where will your homeless go? They'll be pushed out into the suburbs where I live. But I live in one of those suburbs you're talking about, and my CVS has locked everything up. We've got a serious homeless problem in our town, and our city council just made sleeping on the street a crime. I still see tents along the sidewalks, though. I wonder if the homeless numbers per non-homeless aren't the same as in SF (I don't know). Anyhow, unless we're willing to fund massive internment camps on the scale of Manzanar, the homeless problem is not going to go away whether our government is run by degenerates or by generates.
Please note: I'm a degenerate libtard and I don't advocate the internment camp solution for homelessness. I'd advocate a massive increase in social services. I'd swamp homeless encampments with case workers to cajole or coerce the homeless into housing that we'd either build or confiscate (something like 1/10th of San Francisco housing is empty now). And I'd have case workers on their asses to make them take their meds (if required) and keep them off of drugs and alcohol. Of course, I'd tax the wealthiest 5% at higher rates to subsidize this.
If you're a genarate repturd, you'd probably opt for internment camps. And given how much it costs to incarcerate prisoners, I'd suspect they'd be just as costly or more so than the libtard solution.
But otherwise, we'd just be playing whack-a-mole with the homeless problem.
No idea, and not much interest. What I said was not that there exist pieces of paper in some official repository, extensively documenting and quantifying the change in rate of public incivility. I said that it was *noticeable*, which is to say, I've noticed it, and quite literally everybody I've talked to about the subject has also noticed it.
></ChatGPT warning>
Genuinely, thank you. I stopped reading at that point. (Just not doing it at all would be better, but warnings are better than nothing!)
Yes, it came out of Central Asia, ripped through Europe (well, at Medieval travel speeds), killing between 30-50% of the population (estimates range between 25 and 50 million people died). Northwest India got hit bad, and.I suspect the rest of India did, too. It spread down into Africa (at least in the places we have records — Ethiopia and Muslim West Africa). And it moved along the trade routes to China, killing at least 13 million people in the first wave. Then it sloshed around Eurasia for a while, resurfacing in the following decades until it burned itself out.
So, the semi-globalized pre-neoliberal economy of the 14th Century was connected enough for the Bubonic Plague to spread almost everywhere along its trade routes. The Americas were spared because no Europeans visited in the right time frame to light the fire there. The Vikings predated the plague, and Columbus was a 150 years later. But then the Europeans brought Measles to the Indigenous people of the Americas, which was as deadly or more deadly to their unexposed populations than the Bubonic Plague was to the Europeans.
As an aside, the current Amazonian rainforests may be a result of the 16th-century neoliberal expansion of Europe. Unexpectedly, archaeologists have discovered vast networks of roads and cities beneath the jungle floor, which were intensely farmed. The density of the communities suggests that the Amazon basin supported 10s of millions of people. The Amazonian civilization seems to have collapsed after the Europeans arrived in the Americas. </ChatGPT alert> Gaspar de Carvajal, who sailed up the Amazon (1541-42) reported seeing large, well-organized settlements along the river, with extensive agriculture, including fields of manioc. Carvajal described a nearly continuous series of villages and towns along hundreds of miles of river, with signs of advanced societies, including fortifications and coordinated resistance to the expedition. </end>
Modern historians had discounted or ignored these reports until the archaeological finds of the past decade. But just as the jungles of Guatemala and Yucatan grew up after the collapse of intense agricultural exploitation by the Maya, the jungles of Amazonia evolved after the collapse of Amazonian agricultural civilizations. I don't know what kind of ecology existed before humans started cultivating the Amazon. If anyone has any links to palynological evidence, I'd be interested in seeing them. But the Anthropocene started a lot earlier than most people realize.
During the actual pandemic of the 21st C we all blamed the immigrants, and closed the borders. If anybody wanted open borders then, they were the fascists.
Purely for clarification, he's saying it's a lab leak and therefore caused by the scientific establishment (who he obviously does not credit for anything else or they'd still have a great track record, medically speaking).
Does anyone have a recommendation for an immigration consultant or lawyer? I want to go to a knowledgeable person and ask: “here are my skills and ethnic background, and here are my husband’s skills and ethnic background. Who will have us, and at what price?” And I want to hear good advice back. Thanks!
I feel like there *must* be a Subreddit all about this where they can point you in the right direction. Most immigration lawyers are focused on immigration TO the country where they practice.
Are you asking which of the many countries in the world are likely to give you and your husband a work visa? Or are you asking about your options for a particular country?
It's a book-review-ish piece about 4 mental illness memoirs that combined give a good view of a variety of moderately-debilitating mental health issues. I've read these over the years and quite like them, and I think they will all help people understand mental illnesses that fall between depression/anxiety and crazy homeless person on your bus level of severity/prevalence.
It was an interesting experience to write it because I haven't written anything that long since graduating college and everything I've written before over 750 words has been term paper with MLA format citations. I've written plenty of 250-500 word things on reddit and substack, but no real blog posts. I think I fell into a lot of term paper ish traps and definitely don't have my writing voice figured out yet, but it was nice to try my hand at it. It definitely helped me get thoughts together in a way I haven't in a while.
I think I'll do a couple more 1k-3k word book reviews just to see what the process is like. I'm curious if any of y'all have gone through the process of starting to write longform pieces and whether you have any suggestions on the process.
I really like the SF apocalypse art created by the Bluesky user "RevHowardArson" (formerly Theophite on Twitter), which has a distinctive but colorfully weird aesthetic. It's a pity he's got a self-hating complex about AI art because he moves in liberal circles where you have endless whining capital-A Artists who dogpile anyone who promotes AI art on social media (although they're thankfully finally leaving Ethan Mollick alone over there, mostly).
I thought of one from a few years back. It was labeled as a building in Budapest, but no one could locate it. Art nouveau with big windows reflecting a forest. Eventually word got out about details not making sense, but there were a lot of people who wished it was real.
There was a lot of prompting by a human involved, and when I saw another picture by him, I recognized the style.
I'm a big fan of Bad Apple Played On AI Paintings as an example of something cool that could *only* be done with AI style transfer: https://www.youtube.com/watch?v=E58aMjthQCM
I like how if you pause it on any frame it just looks like a normal painting, you only see Bad Apple when it's in motion.
Maybe not memorable from an aesthetic perspective, but I think if you had to pick an AI image that captured the "vibe" of the early 2020s AI acceleration, it'd be the DALL-E 2 "astronaut riding a horse" image -- not particularly good, not emotionally moving, but historically important in the sense of "AI can compose something novel and unlikely to have ever been seen before." So, I guess you could call it memorable in the same sense that Muybridge's galloping horse images were memorable.
I loved John Walter's fractal lizards series. Some of them show up if you do a google image search of "John Walter fractal lizard." The rest are probably not hard to find online.
I have a series of images Dalle-2 made in response to a prompt that was deliberately confusing. I like them, and Beowulf did. I'm not sure if any quite make it to the level of memorable art, but I'd say they are very enjoyably striking and novel images. There's a batch here on Imgur: https://imgur.com/a/6vaTrHZ
Would be interested to hear how you react, even if you don't like them. If you find them amusing or appealing, I had a couple projects where I actually made things using images of this kind, but I'd feel sort of silly pulling them out if you're not a responder to this sort of thing so try the litmus test first.
At least passable, and they're right that having this somewhat pleasant stuff in quantity does make it less individually impressive. I'm reminded of surrealist images made by combining jigsaw puzzles.
Has anyone tried to model bipolar disorder as a consequence of dopamine receptor upregulation? If depressive states lead to decreased dopamine production then I could imagine homeostatic factors causing the receptors to upregulate. Then if the depression lifts slightly and dopamine levels return to normal then you get this crazy surge of activity, i.e. mania. Is this plausible?
Doesn't sound plausible to me because I think the feedback loops involved in brain cells influencing each other are much more complex than that. Seems like your model just involves dopamine sensors and dopamine producers, with the pair of types working ok to upregulate when there's depression, but doing a bad job of sensitive adjustment when the depression starts to lift. The real loops involve neural nets and many nested feedback loops.
Oh certainly my model is simplistic but don't most models of mood disorders focus on neurochemical dynamics as opposed to the connectome? Isn't that why mood stabilizers are much more effective at treating bipolar than CBT?
Lithium et al are definitely more effective than CBT, and anyone who has dealt with a manic person. comes away with the feeling that what is going on is very powerfully driven by some biochemical process, and is impervious to even quite clever attempts to access and tweak the person’s state of mind, cognitive biases etc. I don’t know a thing about how lithium works, and maybe it does have a whole different mode of action from what i’m picturing. I don’t see how, though. Those neuro chemicals are the stuff the cells in the neural net squirt out and respond to. Also, last I heard nobody knows how antidepressants work. By the time the person begins to feel better, they are no longer doing the thing people thought would have an antidepressant effect (for SSRI’s that would be slowing down the process by which the cells slurp the serotonin back up). MAOI’s, which are actually the most powerful antidepressants, came to be used for depression not because somebody cooked them up with the idea that tweak brain chemicals in an anti-depression direction but because people who gave the stuff for tuberculosis. which it combats pretty effectively, noticed that many people taking it had big mood lifts.
I don't know that that would be a helpful model. Depressed then manic isn't a particularly common pattern of symptoms - the classic pattern is manic then depressed.
There's also the usual wastebasket diagnosis phenomenon where a bunch of different illnesses are being lumped together, particularly in community mental health settings. A shocking percentage of bipolar diagnoses in datasets are labels for "was a teenager" or "has borderline personality disorder" or "saw a psychiatrist who read too much Ghaemi" or "was on meth." And even once you filter it down to people who clearly have the disorder you run into the division between the classic Cade's Disease lithium responsive cohort and the various other phenomena.
My preferred model for bipolar disorder is clock genes and circadian rhythm disregulation. In my family with classic bipolar 1 (we've got 4 of use who score a perfect 100 on https://www.ohsu.edu/sites/default/files/2021-10/Bipolarity%20Index.pdf), the people with bipolar disorder have notably deranged circadian rhythms and sleep cycles, even before phones and late night blue light. Sleeping 16 hours when depressed, sleeping 0-4 when manic, non-24 sleep cycles when unmedicated or in mixed episodes.
Really? I thought the typical pattern, at least for bipolar 2, was long periods of depression punctuated by shorter hypo-manic episodes. Does bipolar 1 typically start with mania?
Yeah I'm sure it's more complicated than what I said. It was just a brainstorm that seemed like it generated the swings from simple homeostatic considerations.
There's a whole mess as to what bipolar 2 is - it's been used as a wastebasket diagnosis for so long that there's some amount of debate in the field as to whether it's a real phenomenon and what that real phenomenon even is.
Bipolar 1 has a specific pattern for most patients, and our manic episodes are extremely easy to distinguish from any number of other psychiatric phenomena, and the longitudinal course is very different from the psychotic disorders mania can be confused with. In the 80s, researchers noticed that there were people with recurrent depression and subthreshold versions of manic symptoms whose depression responded well to medication commonly used for bipolar 1 and did not respond well to antidepressants, and thus the bipolar 2 label was born. https://www.ohsu.edu/sites/default/files/2021-10/Bipolarity%20Index.pdf gives an idea of what the classic phenotypes of the bipolar disorders look like.
However, it turns out that "checks all the boxes for a history of 1+ depressive episodes and 1+ 4 day period of mild manic symptoms" is not particularly specific for being a responder to lithium/lamotrigine and not SSRIs or therapy. Bipolar 2 as the label is actually used includes anything from "was a teenager" to "has borderline personality disorder" to "self reports mood swings" to "didn't tell my therapist about my stimulant habit" to "depressed person who remembers being happy once," and relatively few people with bipolar 2 in their chart cleanly fit the diagnosis.
Scott's tweet says that America today is a weird place for a revolution, unlike historical revolutions. On the contrary, it looks to me like France 1789.
Not even close. France in 1789 was literally two days from royal bankruptcy at one point, and had been through one of the worst years in centuries for the harvest in 1788. People were hungry, and it had become clear even to the ultra-royalist reactionaries that something had to be done.
US in 2024 had a 3% unemployment rate, a booming economy (so strong that it absorbed a couple million extra immigrants quite effectively), rapidly advancing technology, declining crime rates, and even heavily declining inflation.
I've been thinking about US debt lately. It occurs to me that Trump had a very specific and successful strategy as a businessman -- he would talk a great game (endless evidence he is good at convincing people, ie he has a powerful reality distortion field). He would use his talents of persuasion to leverage so heavily, with such unrealistic promises and projections, that his deals all eventually went bust. Now his business partners start thinking rationally, and rather than sue him to oblivion (he's so leveraged they wouldn't get much) - they basically have no choice but to take cents on the dollar. (ref Getty, "If you owe the bank...)
Isn't he basically doing the same thing on a worldwide scale? I suspect the budget will get passed with the tax breaks and all the candy he wants; Social security and Medicaid won't get more than a pro-forma "waste fraud & abuse" reorg, and the difference will be massive deficit spending as far as the eye can see.
At which point, he can run his game: basically take the country into bankruptcy, through a combination of inflation and possibly direct negotiation with our biggest creditors (can you say China?)
The tariffs start to make some sense: they are a stick he can use to beat countries into submission, and pit them against each other to the point where they are begging him to give them a few more cents on dollars owed than the next guy.
Frankly, I think the guy is a sociopathic, evil genius. His evil plans just might work.
In what way is the contemporary United States like France in 1789? I am honestly not at the moment seeing even a single similarity between the two, at least in the ways that are relevant to revolution. (I’m ignoring the trivial things like “both of these are a place at a time” and “both of these had a strong majority of the population speaking an indo-European language”).
I don't agree with the analogy, but for fun, here are some similarities.
- People are complaining about the cost of gas and groceries. It's not at the level of near-starvation which led to rapid increases in the price of bread in 1789, but one could argue that people's awareness of the increases and ability to compare their lot internationally would lead to the average person being unhappy with their lot.
- There's an excessively rich and out of touch political elite in the US, arguably built on an out-of-date and unresponsive political system. Could one imagine Melania saying 'let them eat cake'? Or Donald surrounded by courtiers in Versailles ignoring a developing political crisis?
- It's been too long since I read 'The age of Revolution' by Hobsbawm, but I remember him discussing rapid technological change and the economic imbalances brought on by the Industrial Revolution and the move to urban living. I'll wave my arms here and say one could draw analogies with globalisation, the digital economy, and AI leading to a disenfranchisement of workers and concentration of wealth.
People have complained and there have been rich out of touch elites since roughly 1890. One could even imagine whoever's leading the country saying or doing dumb things that the poorer people would find outrageous. So this by itself doesn't convince me there's anything special about 2025.
I think there's copious difference between perceptions of the Trumps and the Obamas from the perspective of individuals. I also think that if one could compute the integral of such perceptions over the entire population, there will be amounts of "they're one of us" and "they're out of touch elites" for both, and they'll be close enough in magnitude that anything one wants to say about one, ought to apply to the other.
I checked - Trump has 44% approval, in May of his second term, Obama had 47% approval. So I guess your prediction on US perspective of the two presidents is right. International coverage isn't comparable.
I'm trying to figure out how you could describe a time period that includes two world wars as an Atlas Shrugged-esque slow decline. (Or a continuous movement in *any* direction, given that both wars caused dramatic economic and political shifts.)
Like, if your definition of "collapse" covers "winning the war against the Nazis," then I think you need to explain your definition a bit more.
Look, if you want to define "Atlas Shrugged Style Slow Collapse" sufficiently narrowly then nothing will ever be sufficiently slow or sufficiently Atlas Shrugged-ish.
Europe's decline from unquestioned masters of the globe to an economically stagnant but moderately pleasant corner of Eurasia that's good for holidays did go in fits and starts, with World War 2 being one of those fits, a dumb socialists-vs-socialists war that managed to destroy a huge slice of what was still good on the continent.
Look, it's a better analogy than "present day USA is like France in 1789" anyway.
If *I* want to define it? *You* are the person who made the claim, you're the one who needs to define your terms. What do *you* think is the defining feature of an "Atlas shrugged style slow collapse," as opposed to other forms of collapse, or other forms of decline that are not collapse?
>Look, it's a better analogy than "present day USA is like France in 1789" anyway.
That is the lowest possible bar you could set for yourself.
I mean, they just got eclipsed by much greater powers in the US, Russia, and now China and probably India. There was no realistic way around that unless they actually went full "United State of Europe" with a pan-European state and common identity with US-level economic integration, and they couldn't bring themselves to do that.
The western Roman Empire collapsed because it was an over-centralized military dictatorship that couldn't handle what were actually not THAT large of threats because the western Empire was constantly being undermined by civil wars and a weak center in the Imperial Court.
And because of bad luck. If one or two things had gone slightly their way with the Vandals, for example, we might be talking about the "Crisis of the 5th Century" instead of the fall of the Western Roman Empire.
There were civil wars going all the way back to the Republic, that was just how politics operated back then. The reason they became so much more destabilizing in the 3rd century is because of economic strife. This is evidenced by the heavy debasement of the currency prior to the 3rd century crisis. There's a fairly extensive literature on this.
Look, anything as complicated as the decline of the Roman Empire is going to be multi-factorial, but I think the minimal explanatory model is one of disordered price signals. Those have both economic and social dimensions, as moral norms and social mores function as informal price signals for the allocation of social and political capital. When coin content no longer reflects intrinsic value, or when patronage replaces merit as the currency of favor, the system’s feedback loops break down. From an abstract perspective society is just an information processing machine that coheres and directs itself via mostly-decentralized signaling. Even centralized institutions are only as good as the information they have access to. When those signals become systematically detached from the underlying realities then the system will inevitably collapse. In my view this is what Nietzsche was talking about when he complained about slave morality being responsible for the decline of Rome, but of course he thought about it in moral rather than economic terms. (Although I think that's a superficial distinction as I believe that morality is more deeply tied to economics than most people appreciate.)
Interesting; I'm not arguing but I'd be curious to hear why you think economic disorder is the minimally explanatory model.
To me, each of: the successive Antonine and Cyprian plagues; the rise of the Sassanids; and the increased pressure from Germanic tribes seem like more natural candidates for the minimally explanatory factor.
As another point, it feels to me like the crisis starts to end before the economic situation is stabilized: Aurelian is the clear turning point, and but don't Diocletian's price controls indicate that the economy was still fried a quarter century later?
As you say, it's multi factorial and obviously economic collapse is part of it, but it's easier for me to see those other factors as more important, and limiting the ability to restore economic stability because of the breakdown of trade, etc.
Not so much that, but a dictatorship took over can made it too hard for people to effectively pursue cooperative competition to mutually pursue their interests.
The dictatorship happened because a sufficient percentage of high status people didn't want to do useful work.
It wasn't the starving peasants who revolted in 1789.
Certainly the Flour War 15 years earlier was relevant. Certainly Scott is right to say that bread lines were relevant in 1991. But food wasn't definitive in those places, either. Soviets were rich enough to drink themselves to death, with life expectancy peaking in 1960 and declining until Gorbachev restricted alcohol in 1985. Wealth leads to revolution.
Urbanization leads to revolution; the number of successful peasant revolts can be counted on one hand. Once you have cities, material deprivation becomes VERY important.
No, the expectation of gain combined with enough poverty to tolerate the high variances of war leads to revolt. Only people who have nothing to lose are willing to risk death for a political cause.
In France the third estate paid > 90% of the taxes. In the US, the top 10% of earners pay 70% of all Federal Taxes. Who's gonna revolt? The rich are the ones being exploited but the rich have too much to lose. They'll just leave, much like in Atlas Shrugged.
This is a weird comment. You mention Ayn Rand, but doesn't she say this is all wrong? In Atlas Shrugged it wasn't the rich who left, but the productive. The whole point is that they aren't the same. Taxes to the central state weren't a problem in either 1789 or today. The problem was that the state allowed parasites. 1789 wasn't a revolt against the king, but against the nobility. The money that the state steers to parasites was and is much bigger than the money that the state collects. Prevention of production is much bigger than that.
Elon Musk doesn't pay taxes. He didn't leave California because of the taxes. He left because it wouldn't let him work. His complaint about the federal government is that it stops him from being productive, not that it exploits him.
I don't think Musk would be a hero in Atlas. At best he'd be a complicated ambiguity - a cynical con man who exploits the diseconomic cracks in the system. I don't think Musk is productive. I think he represents a shallow kind of performative capitalism of the kind that Rand would've hated. He built his empire on the back of government subsidy, a cult of personality, and meme stock investment fads. Tesla is/was probably the most overvalued company in history, the archetype of an economic Potemkin village. And actually Trump is exactly the same. He isn't a businessperson in any meaningful sense: he inherited a billion dollars and used it to purchase a real-life Monopoly set, then got famous for being an archetype of the thing that he was only performing. Trump and Musk aren't successful capitalists any more than Elizabeth Warren, Kamala Harris, or Barak Obama are oppressed minorities. Both parties now worship flimsy cutouts that symbolize the values they claim to care about without thinking for a second about what those values really mean.
As for your analysis of the French Revolution I just disagree. Extractive taxation and parasitism aren't really separable in that context - one leads to the other. And American parallels to Atlas are tilted slightly: I think wealth and productivity are still reasonably correlated here, the problem is that the political class has become completely decoupled from notions of fundamental value. And yeah you have to squint a little to put all three things in the same box but what do you want, it's just a metaphor. The thread that ties them together is a delusional elite that worships irrational diseconomic values not understanding that they're slowly poisoning the garden that feeds them.
They pay only about 50% of all federal taxes. Include all taxes, state and local, and it's much less progressive. As to how it feels, consider declining marginal utility. The people I know in that bracket seem materially happy.
The top 1% types that I personally know are uniformly resentful of the US tax system and many are planning on moving abroad in the near future. Not that the European tax structure is less progressive but they spend the revenue more intelligently. Here it's all transfer payments to ineffective entitlement programs. There's higher social cohesion over there and therefore less social fraud. At least you get something for your tax dollar there.
The US is perilously close to having a voting majority that pays zero income tax. That's a disaster, particularly in the context of a culture that feels entitled to treat the successful as piggy banks that can be used to underwrite the poor life choices of the hoi polloi. Once that balance really tips over the wealthy will flee like rats from a sinking ship. Given our rising debt levels and changing demographics I think that's a realistic scenario sometime over the next 20 years.
I've learned somewhat not to use my imagining of the future, short or long term, as feed for confirming my takes. As yogi may have said, "predictions are hard, especially about the future".
I remember some previous Open Threads from early in Russia's Ukraine debacle, where some doubted that drones would be able to blunt the massive deployments of Russian armor. Does anyone still doubt that drones have permanently changed the nature of warfare?
1. First off, an interesting thread compiled from the commentaries of Russian bloggers on how the insect screech of drones has become the defining sound of the Russia-Ukraine war. Drone noise has become a form of psychological warfare.
2. Ukraine is laying down a 15-20 km “kill zone” of autonomous drones along their defensive lines to deny Russian forces the ability to move undetected across the front. And it's trying to extend it to 40km. And it seems to be working. I'm having trouble getting Threadreaderapp to unroll this thread — apologies to those who are allergic to X.
And this made me snicker: "Drawing lessons from the Ukraine conflict, Driscoll emphasized the need for tanks to adapt to the increasing threat posed by inexpensive drones. This strategic pivot underscores the Army's commitment to *preserving its armored capabilities* while addressing emerging battlefield challenges" [emphasis mine]. US strategists still don't seem to have internalized the idea that tanks may be going the way of the Bronze Age chariot. But Driscoll still has confidence in his chariots.
4. Ukrainian sea drones have driven the Russian Navy off the Black Sea back into their ports. And now sea drones have used AIM-9 Sidewinder missiles to shoot down two Russian SU-30 fighters.
Counter example 1) Iran twice has launched multi-hundred drone assaults against Israel to no avail. 2) the recent India - Pakistan unpleasantness also saw multiple drone swarm attacks. It’s less clear what happened there because of fog of war, but it seems to me that on both may 8 and 9 Pakistan launched multi hundred drone assaults where none of the drones got through. Most of the Indian drones launched at Pakistan also appear to have been neutralized, although not all.
Maybe drone warfare is mostly just a nuisance against a competent air defense? (Corollary, Russia does not have a competent air defense).
These are different things. Drones used for some kind of stand-off strike with no ground component, drones used to facilitate a ground assault, and drones used to deny an enemy's ground assault.
The first two have been pretty underwhelming so far, the third has been transformative.
Why the asymmetry between (1,2) and 3? Is it because advancing forces in a ground assault don’t have embedded air defenses against drones? Why can’t they? Are we over indexing on data from a single conflict against a single somewhat incompetent adversary? Like, maybe the Russians problem is just that they are unable to do combined arms warfare, where ‘combined arms’ now includes drone defense.
Even in the Iran case, you cannot really be sure what happened. Of course Israel denies any serious losses, whether they happened or not (I mean, they do it all the time, admitting only what is impossible to deny), and the US, which is otherwise happy to publish satellite images etc. on Russian or Iranian losses, remains silent when it is about Israel.
Don't get me wrong, this is all fine and understandable, but it also means that public information is limited. Sarcastosaurus, for example, concluded that the Iranian attack was successful in the sense that it caused enough damage to satisfy the Iranian public, while being moderate enough to avoid unwanted escalation of the conflict.
I was thinking about this. It seems to me that the fact that drone swarms are what is being tried here is already a big part of the change. This is very different from Obama era drones that worked much like a conventional aircraft bomber. Even if it’s mostly not getting through, it’s very different in kind, and it’s cheaper.
Israel has used MBTs to great effect in urban combat in Gaza. By all accounts the tanks were decisive in many battles.
Active Defence systems like Trophy have made a huge difference with regards to AT weapons like the Kornet.
Admittedly, Hamas' drone technology is probably not the best, but if you want to discuss the future of MBTs in the battlefield, you should discuss the way Israel uses them, not the way Russia does.
In early 2000, the IDF cut its Armoured Corps in half. Following the current war, it has admitted it was a mistake and is backtracking and expanding the tank brigades considerably.
(Although note the post-script where he acknowledges drones having more of an impact in infantry operations.)
A persistent factor in the Russia-Ukraine conflict is that Russia's military is surprisingly bad at stuff. They've lost a ton of soldiers and been forced to come up with new (and therefore inexperienced) ones, and they keep relying on meat-grinder assaults to make progress despite heavy losses. This is bad if you want to avoid getting blown up by drones, but also bad if you want to avoid getting blown up by artillery or airplanes.
Today's main battle tanks are (to a significant degree) meant to fight other MBTs. This dictates their size. If you want to prevail in a direct fire pissing contest, you generally do not want to send twice as many tanks with half the armor and gun penetration.
I do not see a lot to be gained by eliminating the crew. In fact, MBTs have a crew of four or so, including a loader who is presumably employed for mechanical rather than cognitive tasks. (That being said, many MBT designs are from the 80s, when "can we build a machine to replace the loader and then replace the crew cabin with a somewhat smaller housing for a computer" was SciFi.) Sure, eliminating the crew will allow you to reduce the armored interior volume, which is expensive, but you will still need to store the ammo, the loading mechanism, the computer, the fuel tank and the engine somewhere, so there are no huge gains to be had, I guess.
If autonomous tracked vehicles gain dominance, I do not think they will be an evolution of traditional tanks. I am not sure they generally have much of an edge over flying drones, though. Sure, they can operate indefinitely instead of an hour (or however long the batteries will hold) and can pack much more firepower at cost parity. But they also have much lower mobility. The problem of flying a thousand drones through the forest seems much more solvable than the problem of driving a thousand drones through rough terrain.
I kind of envisioned a crew-less tank, still a huge beast but with the ammo and robotics taking the crew's space. What is gained is not having to worry about personnel losses.
With modern Western tanks, you already don't have to worry about personnel losses. You can if you want to, but if you're fighting an actual war there are so many bigger things to worry about that this would be silly. There are weapons that can *disable* an M1A2 or a Leopard 2A6 or a Merkava, but weapons that can destroy one outright are cumbersome enough that almost nobody ever bothers, but settles for the mission-kill. And the crew almost always walks away from that.
Russian tanks (and most of Ukraine's tanks are based on old Russian designs), are an entirely different matter. But the *reason* Russian tanks blow up forcefully enough to loft their turrets two hundred feet skyward any time the armor is penetrated, is because Russian designers a generation back were over-enamored of automation, and the automated gun-loading systems were rather less robust than a guy who knows what he's doing and would really rather not die today.
Interesting, I admit my understanding of tanks’ vulnerabilities is largely based on watching too many videos of Russian tanks blowing up in Ukraine. But still wouldn’t it be nice not to have a human crew at all?
In the sense of "wouldn't it be nice to not have soldiers at all, just turn wars entirely over to the robots", perhaps. But there are problems with that approach.
If you are going to have soldiers anywhere in a war zone, the inside of a modern western main battle tank is one of the safest places for them to be. Possibly even safer than a command bunker fifty miles behind the lines, because that bunker is going to be a much higher priority target and not that much harder to find and hit.
The real interesting part is that, if anything, I see drones getting more dominant in the near future. Some of the most popular current anti-drone techniques are jamming the signal from the operator, or jamming the GPS connection, but even with current "dumb" AI tech that can run on a smartphone, it seems trivial to operate based on pre-specified instructions and pre-loaded satellite images ("fly to this area and if you find a tank, blow it up").
Some of the *other* popular anti-drone techniques involve automated kinetic weapons (simplistically and sometimes literally "guns") that shoot drones out of the sky before they can come close enough to do any real damage. Drones are slow, fragile, easy targets, rather like skeet. These systems are heavy enough that you'll really want a large vehicle to carry them, and the potential for collateral damage is enough that you'll want your people to be under armor when that happens.
So, the future of tanks is, drones keep exploding every time they come within fifty meters of one. The future of everything that *isn't* a tank, is that it explodes as soon as a drone operator realizes "to hell with that tank, I can actually *kill* this SOB". A bit of an oversimplification, but I think closer to reality than your version.
I assume they'll just have an anti-drone* vehicle as part of the tank's screen forces, which would cover both the tank itself and any APCs moving along with it.
* Or just generally a vehicle to provide active defense against threats. I think tanks are going to go the way of battleships and lose some armor for greater speed and range this century.
An anti-drone system that can cover an entire mechanized company's front, is going to need to have a much longer effective range than an anti-drone system that just defends a single tank. And the latter can probably do double duty as the tank's defense against anti-tank missiles. Meanwhile, the dedicated anti-drone vehicle will be vulnerable to everything that *isn't* a drone, and it will be constrained in its deployment by the fact that it needs to maintain constant line of sight to all threat axes against all of the vehicles in its charge.
If at all possible, and I'm fairly certain it's possible, a short-ranged anti-drone-and-missile system on every tank (and maybe IFV) is I think preferred.
Tanks have already been written off in 2022 and 2023 when Javelin & Co got all the praise. It was too early to write off tanks then, and it's still too early today. The combination of mobility, firepower, and protection is a potent one and will always have a place on the battlefield; tracked and armored fighting vehicles will be replaced if and when some new technology makes them obsolete, but not before. Right now drones have the upper hand, but that's just the nature of warfare - it's an incentive to invest more into anti-drone capabilities (new tech & tactics), and that in turn is an incentive to invest in anti-anti-drone capabilities, and so on.
How could tanks defend themselves against drones? Heavier armor reduces the range and/or speed of tanks, but current APVs can carry 5 Kg of explosive. According to ChatGPT (AI warning!), 5 Kg of C4 can puncture armor that's 10 cm (4 inches) thick. The thickest armor an M1 Abrams is 700 mm thick. Likewise, reactive armor could resist a 5 Kg C4 explosion, but it can only resist it once.
Drones are low-cost compared to tanks, and the enemy can send in dozens to swarm a tank to overwhelm the tank's defenses. Videos show Abrams tanks being taken out of action by Russian drone attacks in Ukraine, which demonstrates the effectiveness of this tactic. Ukraine has developed an interesting net-ejector system that tangles drones as they approach, but, at least right now, that can only be used once. It wouldn't work against a swarm of drones.
That's why you shouldn't operate a tank unsupported, as both Russia and Ukraine are prone to doing due to their Soviet upbringing and somewhat limited capabilities. If you expect the enemy to use drones or drone swarms and your tank can't handle that on its own, you give your tanks the support to deal with that threat, such as an ECM suite, airburst anti-air [1] for hardkills, or perhaps a more futuristic, energy-based defense system like lasers, should that prove practical on a battlefield.
For Javelins specifically, it mostly comes down to A: use artillery to suppress the people who would be launching Javelins against the tank and B: if that doesn't work, hope that your active protection system (https://en.wikipedia.org/wiki/Active_protection_system) can shoot it down before it hits, and C: if you don't have one of those (they're fairly new) hope you are fast enough to hit the button that fires a bunch of smoke grenades to create an insta-smokescreen that will block the missile's guidance system.
Against pretty much every other sort of antitank missile, there's also really, really tough armor. It is possible to make the *front* of a tank nigh-impenetrable to all but the heaviest antitank weapons, and mitigate the after-armor effects of those to the point where the crew probably still walks away. Javelin specifically can fly a top-attack trajectory, and it's not practical to armor the entire top surface of a tank to that level.
Option A seems like it would be hard for a missile that's man-portable, can be vehicle mounted, and has a range of several kilometers. What use is the tank if you're already destroying threats before they get within a few kilometers of it? Would they be able to reliably spot the Javelin missile launchers and pick them off at that distance? Are the missiles just rare because they're expensive or heavy?
I am not sure that it is inevitable that the anti-drone tech will catch up.
Consider the cavalry charge. For millennia, it was a staple of land war tactics. Then one day, the machine gun was invented, and cavalry seemed obsolete. Some people might have claimed that it would be inevitable that cavalry would catch up -- armoring horses and riders or something -- but that never happened. It turned out that the armor required to withstand machine gun fire is not feasible for a horse, but it is in fact feasible for a motorized vehicle.
(Naturally, history is likewise full of just-so stories about weapon systems which were declared obsolete by someone only to be proven terribly wrong, e.g. "we can win wars through air power alone".)
My guess is that the main advantages of tanks are that they are mobile, impervious to small arms and a very cost-effective way to deliver firepower (compared to missiles).
If the best way to defeat a tank is another tank, then tanks will stay dominant. However, if you can defeat tanks with other weapon systems which cost a fraction of what a MBT costs, then the MBT will eventually die out.
A MBT is, from my understanding, ruinously expensive. The enemy has every incentive to outperform you in gun and armor performance, so you want to go big or go home. (Also, personnel is expensive, and generally will not survive the defeat of their vehicle.)
I will grant you that in principle, defending a tank against drones and missiles is easier than defending it against pure kinetic penetrators (fired from other tanks). Perhaps drone defenses will catch up and tank vs tank will remain a thing. Or perhaps the economics favor overwhelming the drone defenses with thousands of cheap expendable autonomous drones. Or perhaps someone will build a rocket-based kinetic penetrator which can be safely fired out of sight of the tank using real time satellite images.
There is a saying that armies prepare for the last war they were in. Thankfully, for the last 80 years, we have not seen open conflict between superpowers. We certainly did not have any large naval or armored vehicle battles using cutting edge tech. While this is nice for the humans involved, it also means that to a large degree, we rely on extrapolations and models. Both the military-industrial complex and branches of the military tend not to be perfectly aligned with the defense interests of their host nation in peace time. For example, if the future of air superiority belonged to unmanned drones (which can react faster than a human, can handle much higher accelerations and whose brains are cheaply replaced, at the cost of less versatility), I would not expect the US air force to realize that fact in the absence of peer conflicts.
Quite frankly, if we had stopped doing great power wars in 1870, I would half expect today's militaries still to drill cavalry charges.
I recently saw a Ukraine war video of a stealthy drone following a tank bank to its hiding place, then taking out the tank, the truck next to it and a lot of trained tankers and motor pool guys.
If that becomes more common, tanks will be a lot less useful.
It's better to think in roles and capabilities rather than counters. Economics are important, yes, but the first job of a military is to win a fight. To be cost-efficient is second to that. There is a job on the battlefield that is best fulfilled by tanks and that cannot be fulfilled by drones, and as long as that job is militarily relevant, the military will evolve to enable the tank to do that job, or find a replacement that can - if you simply stop using tanks, the job remains unfinished and there is a hole in your capabilities.
Also, drones are a general threat to everything on the battlefield and beyond, so the imperative to defend against them goes well beyond just to protect tanks. There already exist possible solutions to drones and even drone swarms (e.g. air-burst munitions), plus whatever people come up with when facing an existential threat, and it's only a matter of time until they are deployed at a scale that allows tanks and other vulnerable assets to do their job again. Maybe not every country will be able to afford these solutions at the necessary scale, and yes, those countries might be as screwed as if they showed up with chariots. But everyone who can afford to can and will invest in anti-drone capabilities, and the cycle of offense and defense will continue.
As far as I understood, WW2 was mainly won by outproducing Germany. Cost effectiveness is still a thing.
Of course, if your military can afford to outspend the enemy in 100:1 ratio (like Izrael can outspend Hamas), you don't have to worry about it that much.
Thanks! I hadn't thought of airburst munitions. But once they're expended, the tank would be a sitting duck. Swarm it with cheap sacrificial drones until it expends its airburst munitions, and then drop 5 Kg of C4 on the turret and see what happens.
Maybe tanks would be flanked by anti-drone vehicles loaded with sensors and anti-drone weapons, kind of like how carriers are flanked by a battle group.
Or the tank could travel with its own swarm of anti-drone drones that suicide detonate into an oncoming attacker.
But that begs the question: Why not just have swarms of cheaper suicide drones take out enemy positions? I'm still unclear about the unique function of tanks in 21st-century tactical engagements. My understanding is that tanks were used to overwhelm and/or outflank established enemy emplacements, and to make a path through the enemy that mobile infantry could follow. Seems like drones could perform the same function for less money. But maybe I'm missing something obvious.
For much of the Middle Ages, the Catholic Church wasn't hunting down magic users. The official view was that magic didn't work, and anyone who claimed it did was a fraud.
As might be expected with any high-minded, abstract, and contrary to popular opinion command, compliance wasn't happening.
A lot of magical practice was something congregations insisted on priests walking a host around fields to improve harvests, and compliance with the hierarchy just wasn't worth it.
The later, witch-hunting period was real and horrifying, but it didn't happen in the middle ages. The Catholic Church isn't just one thing, and I wouldn't even say its behavior is especially predictable.
From what I remember of my history classes, witch-hunts were more popular in Protestant areas or areas with mixed religions than in Catholic areas. (James VI & I famously wrote a manual on hunting witches, for example - I don't know of any equally famous Catholic who was as obsessed.) But they did happen everywhere.
I don't think that any witch-trial has been as well-studied as Salem, but I wouldn't be surprised if many of the European ones had similar features: that the cases were brought by members of the community rather than by the religious or secular authorities, that they came during or after times of intense hardship and distrust of neighbours, and that they were (within limits) tolerated by the authorities as means by which the communities blew of steam by scapegoating outsiders. I think you could bring René Girard into the discussion at this point...
"A lot of magical practice was something congregations insisted on priests walking a host around fields to improve harvests, and compliance with the hierarchy just wasn't worth it."
Oh, yeah. Folk religion shades into and out of magical practices very easily. Plus we are a material religion, so we do believe in the efficacy of things like prayers for rain. Blessing the fields and combining that with a Eucharistic procession wouldn't have pinged anyone's radar as "this is magic not religion" (until the Protestants got all neurotic).
My own pet eccentric theory is that the witch-hunting really kicked off due to Protestant neuroticism; the world was haunted by devils, your only bulwark was faith, and if your faith was insufficient you were vulnerable. All the supports, the pseudo-magical protections, had been stripped away and what was left was the bare word of Scripture (hence why Scripture itself began to be treated like a magical text). At the same time, you were told the Devil prowls about like a roaring lion seeking whom he may devour, and it was possible to make bargains with evil spirits to cause harm. How could you protect yourself, except to root out the malicious evildoers?
People seem to find it easier to believe in demons than gods. I don't know how many modern horror/dark fantasy stories I've read where the conventional belief system of Christianity is for the birds, but you bet the monsters or spirits of the old pagan faiths are real and ready to rip your lungs out.
>People seem to find it easier to believe in demons than gods. I don't know how many modern horror/dark fantasy stories I've read where the conventional belief system of Christianity is for the birds, but you bet the monsters or spirits of the old pagan faiths are real and ready to rip your lungs out.
Reminds me of a review of Buffy the Vampire Slayer I read once which complained that none of the characters seemed religious. After all, they live in a world where there are literal demons from hell which are defeated by Christian religious paraphernalia (crucifixes, holy water, etc.), and not once do they say, "Hey, if Christian objects are so powerful, maybe Christianity's true and we should all start going to church."
If I can believe the article, the Church did believe that walking a host around a field to improve the harvest *was* an attempt at magic and wouldn't work. They thought it was fraudulent.
I thought the Catholic belief in the power of Satan rose before the Reformation, but I could be wrong about that.
"If I can believe the article, the Church did believe that walking a host around a field to improve the harvest *was* an attempt at magic and wouldn't work. They thought it was fraudulent."
Not so much that (because processions and 'beatring the bounds' and the likes were embedded into parish life) but as mentioned in the linked Tumblr post:
"In some cases he also willingly shared a bit of the communion wine to the local cunning woman (who was also probably the midwife), since that was the best thing to mix into ointments and potions for everyday ailments. But only a little, of course, as it was precious stuff."
Absolutely you do *not* do that if it's consecrated wine, as that is the Blood of Christ and there are *very* strict rubrics about how to dispose of any leftover wine (the priest is supposed to drink it) and what to do if you wash out wine stains from altar cloths. Giving any dilution of the consecrated wine *is* magic and will get you into big trouble. If you're giving out the *unconsecrated* wine, that probably will get you into trouble as well, but not as much. I think maybe what they are talkigna bout here is not so much "a bit of the communion wine" as "water used to rinse out the chalice" and that too is a grey area.
For instance, when rinsing out communion vessels the water is supposed to go directly into the ground, not into the normal drains/sewerage (see "sacrarium"):
"The purpose of the piscina or sacrarium is to dispose of water used sacramentally, by returning these particles directly to the earth. For this reason, it is connected by a pipe directly to the ground; otherwise presumably, a basin was used. At times the piscina has been used for disposal of other items, such as old baptismal water, holy oils, and leftover ashes from Ash Wednesday.
In the Catholic Church, pouring the consecrated wine, the Blood of Christ, or the Host down a sacrarium is forbidden. Extremely rarely, the Eucharistic species spoils or becomes contaminated such that it cannot be consumed. The host is then dissolved in water until it disappears, and then the water is poured down into the sacrarium. In accordance with what is laid down by the canons, "one who throws away the consecrated species or takes them away or keeps them for a sacrilegious purpose, incurs a latae sententiae excommunication reserved to the Apostolic See; a cleric, moreover, may be punished by another penalty, not excluding dismissal from the clerical state."
So if in the first instance we mean not "having the full bells and whistles procession of the Host in the monstrance" but rather "giving crumbs or pieces of the Host to be scattered on or buried in the fields" hoo boy yeah, magic, big big deal.
This is correct. Until the Renaissance, witch trials were rare and were almost exclusively conducted by civil authorities. In general, the medieval witch trials were otherwise-ordinary murder trial where the killing was alleged to have been done by means of magic.
Even in the Renaissance, it wasn't just Catholics doing witch hunting. Protestants did a lot of witch hunts, too. And at least one Orthodox country (Russia) did thematically-similar stuff in the same time period, but they were hanging sorcerers instead of burning witches and didn't generally consider their sorcerers to be devil-worshippers the way Catholic and Protestants thought of their witches. And in this time period, it was still usually civil authorities doing the burnings: the Dominicans in Catholic countries and various religious groups in Protestant countries did search for witches and do torture to force confessions and compel testimony against other witches, but they almost always turned unrepentant witches over to civil authorities for trial and execution, and a lot of civil authorities did witch hunts on their own.
Popular conceptions about witch hunts, especially in English-speaking countries, seem to be rooted in four sources:
- The Murray Thesis. This is a theory proposed in 1921 by Margaret Murray, an archeologist specializing in ancient Egypt. The theory was that only elites converted to Christianity in Roman times while commoners mainly remained pagan through the Renaissance and well into the Early Modern period and specifically practiced a common Europe-wide pagan tradition centered on a "horned god" which Christians identified with Satan. Under this thesis, the witch hunts were attempts by elites to destroy this pagan tradition. The Murray Thesis got taken seriously by mainstream scholars when it was first proposed, but it has since been thoroughly discredited. In the meantime, though, Gerald Gardner relied heavily on Murrays books when formulated Gardnerian Wicca in the 1950s, which he based on the assumption that modern English folk magic, early 20th century English mysticism, and classical Druidism were all parts of the "horned god" tradition which he aimed to reconstruct and codify. Thus, Wiccans and Wicca-adjacent neopagans (many of whom are terminally online) often take the Murray Thesis as an article of faith. The idea of the "Burning Times" and of witchcraft being "The Old Religion" are references to the Murray Thesis by way of Gardner
- The exploits of Matthew Hopkins, the self-styled Witch-Finder General of England. He was a Puritan con artist who operated during the English Civil War. His technique was to turn up in a village, announce that he'd been appointed by Parliament to root out witches throughout the country, conduct a highly dramatic investigation, and hang whatever "witches" he found that hadn't been killed by his trials-by-ordeal. Along the way, he'd also collect a "tax" to pay for his services. In the lawless chaos of the Civil War, he was able to keep this for several years before dying of natural causes. About half the tropes (at least in Anglo-American culture) of how witch trials were conducted, stuff like "swimming the witch" and searching an accused witch's skin for distinguishing marks, comes mostly from Hopkins. The Witchfinder Sergeant character from "Good Omens" is a pun on Hopkins's title.
- The Black Legend, a body of anti-Spanish and anti-Catholic propaganda originated by Queen Elizabeth I during the 1585-1604 Anglo-Spanish War, and which took on a life of its own afterwards. The Black Legend featured highly graphic exaggerations of the Spanish Inquisition techniques and practices, painting Spain as a totalitarian theocracy. A lot of elaborate torture techniques that have little or no historical basis (the iron maiden, the pear of anguish, etc) come from Black Legend propaganda. The truth of Early Modern Spain and the Inquisition is plenty awful, but the Black Legend exaggerated it enormously. The Inquisition primarily targeted Jews, Muslims, and Christian heretics (especially Protestants), but it also hunted witches and was conducted by the same religious order (the Dominicans) who also did witch hunts elsewhere in Europe in the 17th century, so the Black Legend spilled over into popular conceptions of witch hunts. Most of the other half of the Anglo-American pop culture image of witch trials come from the Black Legend.
- Late 20th century popular edutainment tropes about the history of science in the Renaissance, which tend to heavily emphasize the trial and imprisonment of Galileo for heresy by the Roman Inquisition (also conducted by the Dominicans).
Primarily but far from exclusively. There are a lot of proposed factors, but the big one I remember is that witchcraft was associated with folk magic. Folk magic isn't the same think as witchcraft (this being one of the major errors of the Murray Thesis), as there was a fundamental distinction between beneficial or neutral magic on one hand and harmful of malicious "black magic" on the other. But even though they're separate, people did make the obvious connection that if it seems someone has been doing black magic in the village, then the local wise woman whom everyone goes to for herbal poultices and protective amulets is going to be near the top of the suspect list. And that type of folk magic that got culturally associated with witchcraft was, in most of the relevant cultures, primarily a thing women did.
This is highly culture dependent, ranging from about 60% of accused witches in our records being men (Russia and Estonia) to around 90% women (Southern England). It looks like the norm in most places we have data from is about 70-80% of accused witches were women. The gender ratio for people executed for witchcraft, compared to that for people accused including those acquitted or otherwise not executed, ranges from about the same to somewhat more likely for women to be executed, in the relatively few areas that we have detailed data for both categories.
Other proposed reasons include: societal antipathy for norm-breaking women, anxiety by male decision makers about female sexuality (a lot of the lore around witches being highly sexualized), and the peculiar position of widows and never-married woman without surviving relatives in the Early Modern social order that made them both suspect and vulnerable.
If you want to read whole (relatively short) book on the subject, one of the texts for my witch hunts class focuses pretty heavily on gender dynamics. It's "Witchcraze: A New History of the European Witch Hunts", by Anne Llewellyn Barstow, published 1994.
The other two texts were:
- "The Witch-Hunt in Early Modern Europe", by Brian Levack, 1987, which focuses more on how the witch hunts were conducted and the legal, religious, and political factors influencing the witch hunters.
- "Witchcraft in Europe, 400-1700, A Documentary History", edited by Alan Kors and Edward Peters, 2001. This is a compilation of selected primary source excerpts from the period in question.
The book we were assigned in college history was _The Devil in the Shape of a Woman_ by Carol F. Karlsen. Amazon reports copyright as 1998, but that must be a later edition - I still have mine, and read it almost a decade earlier. It also focused on the gender dynamics, as well as economic (IIRC, most of the accusations were of roughly middle-class women by poorer women, implying that these were the richest rich they could eat).
A thing I've read (it may be hard to find the source) claimed that county by county research found that witch trials corelated with weak government authority rather than with a region being definitively Catholic or Protestant. Is that looking solid?
Governments actually don't like witch trials. It's too easy for anyone to be accused.
There's definitely a correlation, yes, but my professor's take was that the actual trigger of the witch hunts was large-scale social disruption. Witch hunts tended to occur following or concurrent with wars, plagues, and famines and to be concentrated in the worst-affected areas.
Whether or not governments liked witch trials depended on the government and the trials. Some rulers sincerely believed in the narrative that witches were a vast conspiracy that posed an existential threat to Christian civilization, and these rulers tended to push witch trials regardless of other factors. James VI and I of Scotland and England, for example, was conspicuously pro-witch-trial and one of the first things he did after taking the English throne was to vastly expand England's relatively narrow laws criminalizing certain specific forms of black magic.
Another interesting bit I remember from my Witch Hunt class is that Protestants and Catholics had a notable tendency to burn their own witches. That is, the defendants at witch trials conducted by Catholic authorities were overwhelmingly other Catholics and similarly for witch trials conducted by Protestant authorities. Catholics in Protestant countries and Protestants in Catholic countries did face persecution, but they were persecuted as heretics or recusants, not as witches.
I'd clarify a small point, that when the Dominicans tortured and forced confessions, it wasn't the *Dominicans* who were doing the torturing, oh no sir. Besides the bad look it would be for priests to engage in torture, the church didn't have the local resources to detain or imprison people (outside of church strongholds like Castello Sant'Angelo in Rome). That was the domain of the local, secular ruler, and a strong inquisition was only possible with the collaboration of secular authorities. That's why the Inquisition got so bad in Spain (very enthusiastic financial support and authority from Isabel and Ferdinand) but inquisitions elsewhere in Europe at the same time had no teeth -- the local rulers refused to lend their dungeons and torturers and police forces.
That's how you get Pico della Mirandola living it up it Florence even after his excommunication and a clan of barely-disguised pagan sun worshipers operating with abandon in Rimini for decades.
I'd forgotten that aspect of it, thank you. I remembered the Dominicans turning over unrepentant heretics and witches to be tried and burnt by secular authorities, but I didn't remember that the also delegated the torture.
The other important thing about the Inquisition's use of torture is that the Medieval Inquisition knew the failure modes of torture victims trying to guess their interrogator's passwords instead of telling the truth. They'd developed a big set of rules around the use of torture designed to mitigate this. To the best of my recollection:
- Torture was a last resort. Use persuasion, psychological manipulation, threats, and bribes first.
- Getting permission to torture someone required hearings and a fairly high standard of evidence against them, something like the modern standard of "clear and convincing evidence".
- There's a formal escalation sequence before torturing someone, including "showing them the instruments", where you give them a tour of the torture chamber and formally warn them that they will be tortured if they don't cooperate.
- Only one ride per customer. If a suspect is tortured and you don't get what you want out of them, too bad. You can never torture them again.
- Statements made under torture only count as evidence if the subject repeats them later on, under oath, in a formal court setting. Combined with the "one ride per customer" rule, this allows them a chance to recant what they said under torture without fear of being tortured again.
- Only specific techniques were allowed.
- Never lead the witnesses. Ask them questions in ways that don't give hints at the "correct" answers.
- Mix the real questions with questions that you already know the answers to. Ideally, the torture victims shouldn't know which are the real questions and which the controls.
These worked pretty well (by medieval justice standards) when applied to large-scale heresies like the Cathars, Lollards, and Waldensians. But when a vast underground network of Satan-worshipping practitioners of black magic turned up on the Inquisition's radar in the Early Modern period, they turned their sights on it and found, much to their consternation, that the safeguards were stopping them from getting the evidence they needed to dismantle the network of witches that they "knew" existed. So, given the dire threat the witches posed, they rationalized ways to bend practically all of their rules until enough accused witches started confessing and giving up the names of other witches.
Suspiciously similar, actually, to the Iraq-war anti-terror torture (actual) practices of the CIA. If you go back and read the official documentation when they approved "enhanced interrogation", the fascinating part of it all was that on its face almost everything written in the white paper doesn't seem all that absurd, or at least there was a seeming attempt to set some rules -- at least in the original source white paper I read, the vibe is very much about how these "enhanced" interrogation techniques are really only designed to be done once or twice at most, and there wasn't much point to doing more. Whether you believe this to be honest or not is a separate question, but the psychologists in question at least had no experience with torture to start with. The context for this was actually like, one specific guy that the CIA was convinced wasn't talking (since, some have argued he wasn't even an AQ member, but how much of this is on a legal-proof standard vs common-sense probability is still up for debate), it wasn't even a general policy goal.
What IS clear, is that despite the official wording and justification calling for its use sparingly and in extreme situations only, they went on to waterboard the guy something like 80 times in a single month, and of course to rely more and more and more on black sites, and to have "known" terrorists give up the networks they "knew" existed. One might say, "given the dire threat terrorists posed, the rationalized ways to bend practically all of their rules until enough accused terrorists started confessing." Got 2-3 people who know each other blaming each other, occasionally, under extreme torture? Baby, you got yourself a network going! Add some cash bounties to the mix?
There is a very interesting paper collected by Gwern that compares the use and effectiveness torture by the Spanish Inquisition vs. American intelligence:
"Inquisitors tortured for different reasons, with different goals, based on different assumptions, and in a social, political, and religious setting entirely alien to that of modern interrogators…
The Inquisition put in place a vast bureaucratic apparatus designed to collect and assess information about prohibited practices. It tortured comprehensively, inflicting suffering on large swaths of the population. It tortured systematically, willing to torment all whom it deemed to be withholding evidence, regardless of how severe their heresy was or how significant the evidence was that they were withholding. The Inquisition did not torture because it wanted to fill gaps in its records by tormenting a new witness. On the contrary: it tortured because its records were comprehensive enough to indicate that a witness was withholding evidence.
This torture yielded information that was often reliable and falsifiable: names, locations, events, and practices witnesses provided in the torture chamber matched information provided by those not tortured. But despite the tremendous investment in time, money, and labor that the Inquisition invested in institutionalizing torture, its officials treated the results of interrogations in the torture chamber with skepticism. Tribunals tortured witnesses at the very end of a series of investigations, and they did not rely on the resulting testimony as a primary source of evidence.
This systematic, dispassionate, and meticulous torture stands in stark contrast to the “ticking bomb” philosophy that has motivated US torture policy in the aftermath of 9/11… US interrogators expected to uncover groundbreaking information from detainees: novel, crucial, yet somehow trustworthy. That is an unverifiable standard of intelligence that the Inquisition, despite its vast bureaucratic apparatus and centuries of institutional learning, would not have trusted.
The Inquisition functioned in an extraordinary environment. Its target population was confined within the realms of an authoritarian state in which the Inquisition wielded absolute authority and could draw on near-unlimited resources. The most important of these resources was time… It could afford to spend decades and centuries perfecting its methods and dedicate years to gathering evidence against its prisoners… Should US interrogators aspire to match the confession rate of the Inquisition’s torture campaign, they would have to emulate the Inquisition’s brutal scope and vast resources… one cannot improvise quick, amateurish, and half-hearted torture sessions, motivated by anger and fear, and hope to extract reliable intelligence. Torture that yields reliable intelligence requires a massive social, political, and financial enterprise founded on deep ideological and political commitments."
Ironically, the "ticking time bomb" scenario that is usually given as justification for torture is also the one in which torture is least likely to be useful.
I think I took the Witch Hunts class in 2003 and a class on Renaissance and Reformation Europe from the same professor (Paul Hiltpold at Cal Poly in San Luis Obispo) in 2004. I remember him remarking that the then-public information about enhanced interrogation was following patterns he found very familiar.
That is really interesting. Thank you. You say early modern period. Would that be under what we call the enlightenment for instance? In other words was this a reactionary backlash against the age of reason?
What Erica says in her comment - the heyday of magic and mysticism was with the scholars, not the common folk engaging in superstitions and traditional practices. This was always there, you had them drawing up horoscopes and trying to straddle the line between permissible and impermissible (not for fortune-telling or trying to discover the future) usage of same, as well as sailing damn close to the wind where evoking angels (to help you get rich and get women) gradually over time starting evoking devils/necromancy (to help you get rich and get women).
It really kicked into high gear during the Renaissance, when all the old Classical texts were being rediscovered (including Classical era magic and religious systems which were then reworked into philosophy and mysticism - see the revival of Neoplatonism) and the Early Modern Period - the tale of Doctor Faustus comes to us via Christopher Marlowe in 1594:
That leads us into the reign of James I who had a lively belief in the supernatural and who, in the wake of stormy weather which prevented his new bride from sailing to Scotland and the subsequent Danish witch trials, set up his own tribunal of investigation into the alleged witchcraft plot:
He was interested in witchcraft because of his interest in theology, which probably arose from his upbringing by those determined to make him a good, godly, Protestant (Presbyterian) king who would respect the clergy and the limits of the power of the monarchy. So he had it dinned into him about Scripture and the Devil from a very early age.
"King James wrote a dissertation titled Daemonologie that was first sold in 1597, several years prior to the first publication of the King James Authorized Version of the Bible. Within three short books James wrote a philosophical dissertation in the form of a Socratic dialogue for the purpose of making arguments and comparisons between magic, sorcery and witchcraft, but wrote also his classification of demons.
In writing the book, King James was heavily influenced by his personal involvement in the North Berwick witch trials from 1590. Following the execution of an alleged sorcerer in the year 1591, the news of the trials was narrated in a news pamphlet titled Newes from Scotland and was included as the final chapter of the text. The book endorses the practice of witch hunting in a Christian society. King James may also have been prompted to write the Daemonologie by an anonymous critical pamphlet John Upaland which circulated in 1597 and alleged the Scottish court was bewitched."
At this point we are in the late 16th century, well into the Renaissance and Early Modern Europe, and far from the mediaeval period. The Enlightenment is peeking over the edge of the horizon, though it is still a long way away. The foundations of what will become the scientific method are being laid, as is the ferment about society and government that will spill over into civil wars at a later period.
"Early Modern" means the whole period after the Late Middle Ages and before the Industrial Revolution. It includes the Renaissance, the Age of Discovery, the Protestant Reformation, the European Wars of Religion, and the Age of Enlightenment. There are different schools of thought on its endpoints, but generally from some time in the 15th century until within a decade or two on either side of 1800. 1453 (Ottoman conquest of Constantinople), 1485 (Battle of Bosworth Field), and 1492 (Columbus's first expedition to the Americas) are popular start points. The American Revolution (1776) the French Revolution (1789), and the Congress of Vienna (1815) are popular end dates.
The witch hunts weren't a reactionary backlash as such. Nearly the opposite in some respects: the ideological underpinning of the witch hunts was downstream of Renaissance-era changes in how Europeans conceptualized how the world works. As Nancy alluded to in her original comment, the medieval Catholic view was that Satan was essentially powerless in the world except as a tempter and deceiver and that society is relatively static and exists as God ordered it. In the Renaissance, this shifted towards attributing to humans more agency in shaping the world, and at the same time Satan came to be seen as a more active and dangerous entity. This is essential to the fundamental idea behind the witch hunts, that witches represented an existential threat to European Christian civilization rather than (as they were thought of in the middle ages) small-scale local problems whose activities at worst endangered their own souls and their neighbors' health and safety.
The other significances of the early modern period are:
1. Massive social disruption. As I said in the other branch of this thread, witch hunts tended to happen in times of suffering, disruptions, and uncertainty. The European Wars of Religion were a really enormous factor here, as there was a period of about a century there were a bunch of overlapping international and civil wars that were massively destructive and wrecked lives, property, and social order in affected areas. More countries than not spent at least a decade (sometimes several decades) being fought over:
a. The 80-years war (1568-1648) in the Low Countries, also involving England, France, and the Habsburg realms. The Anglo-Spanish war (1585-1603) was an offshoot of this.
b. The French Wars of Religion (1562-1598), which also included England, Scotland, Spain, and several Italian states.
c. The Thirty Years War (1618-1648) in Germany, which also included the Habsburgs, France, Sweden, and Denmark. This also had several notable foreshocks of lesser scope and scale.
d. The Wars of Three Kingdoms (1639-1653) in the British Isles, including the English Civil Wars, the Scottish Covenanter War, the Irish Confederation Wars, and the Cromwellian Conquest of Ireland.
There were probably others I'm overlooking.
2. Big increase in state capacity. Secular governments got more powerful and more centralized, allowing large-scale witch hunts to be organized.
I thought your response to Yarvin's response was very good.
Like you I was slightly cheered to learn that the Dems have started an Abundance Caucus and that as of last Friday it already had 30+ members. An equally positive sign is that the knee-jerks at places like The Nation absolutely haaaate it.
Meanwhile since Thursday afternoon,
-- another Trump-appointed federal judge ruled that the administration must immediately bring back another specific deportee from the El Salvador dungeon, and this judge added that she will start placing administration officials under oath for updates on their work to comply with her order;
-- Trump when asked live on-air whether he is obliged to defend the Constitution replied "I don't know";
-- the FBI Director denied under oath that the Constitution's guarantee of due process applies to noncitizens despite having Antonin Scalia's ruling to that effect read to him;
-- Trump appointed yet another Fox News personality to a high federal office;
-- Stephen Miller is convening White House meetings to explore a national suspension of habeas corpus;
-- Trump defended accepting a personal gift from the ruling family of Qatar of a half-billion dollar glammed-up 747, which he would graciously lend to the government for use as Air Force One until he leaves office at which time the plane goes with him. [Never mind that the Constitution's Emoluments clause _specifically_ bars the acceptance of gifts from any “King, Prince, or foreign State” without prior Congressional approval.]
I don't know the details of this case, but "due process" means different things in different contexts. Someone in the country illegally may be subject to due process, but that may only mean in practice that the government follow a standard practice in immediately deporting them with no trial or legal counsel. In fact, that happens a lot at the border. Illegal crossings where the people are rounded up and sent back, with no courts involved at all.
Do you have more details on the court case of the guy in El Salvador? My understanding of the SCOTUS ruling in the other case was that a court can't order the Executive to engage in any particular type of foreign affairs, negotiations, etc. At least your summary of the ruling seems to directly contravene that SCOTUS decision.
Yes of course "due process" means different things in different contexts, and every federal judge ruling on this topic has noted that. The courtroom arguments about whether due process was honored are always facts-specific.
Your understanding of the various SCOTUS actions on this topic is muddled, which is entirely understandable and I've given up trying to read their tea leaves. The important point is that the SCOTUS has thus far issued only temporary orders related to this topic, allowing full judicial review to play out in the lower courts. That's a totally normal SCOTUS approach given that this isn't part of SCOTUS's constitutionally-specified "original jurisdiction". But the mediots mostly fail to understand any of that and persist in writing stuff up as if the Court has settled a big question when it hasn't at all.
Presumably at some point SCOTUS will have to consider and make a substantive ruling on the AEA and/or what exactly a district court can and cannot order the executive to do. That though is a ways off yet at best.
In the meantime this specific court case is not related to any of that. This case is about the current administration refusing to either honor or try to undo an existing federal court settlement. [Trump wants to be able to just do whatever without bothering about any legal or constitutional considerations, he thinks being POTUS is supposed to be like being a king.] This judge wrote, "this is just a breach of contract case. It is not a habeas case, or a case assessing the propriety of the government’s recent invocation of the Alien Enemies Act (“AEA”)."
From a first reading of this situation, it does sound similar. A quiet fight between the Executive and Judiciary on exactly what a court can require a president to do. I don't think anyone is quite sure where the limits are. As with the previous case, there's an order to "facilitate" the return of someone the court says was illegally removed. No one seems to be arguing about whether the removal was actually illegal, but taking that as given. A more detailed definition of what that would require could easily run afoul of what SCOTUS did rule, while a less detailed one leaves a lot to the imagination.
Would you agree that a court order for the Executive to enter into negotiations with a foreign government to procure the release of a prisoner would be foreign affairs? Would you agree that it would therefore overstep what SCOTUS was willing to lay out? Is there an alternative order a court could implement that actually leads to the release of Daniel without requiring something along those lines? Maybe someone can be really creative, but I'm not seeing anything obvious and haven't heard any suggestions.
The only material difference I see in this case is that the DOJ tried to go through the process to make the case moot, which the court essentially said would not be sufficient. There's still the open question about what remedy the court can order.
US contract law, like the older English contract law that it's rooted in, is very clear that the party violating a contract must cure that violation; and, that courts are supposed to prioritize very highly the enforcing of that requirement.
And then quoting from this judge's most recent order:
"Of course, [her previous ruling] puts this case squarely into the procedural morass that has been playing out very publicly, across many levels of the federal judiciary, in Abrego Garcia v. Noem, No. 8:25-cv-00951 (D. Md.). As of the date of this order, discovery is underway regarding the government’s efforts to comply with court orders (including from the United States Supreme Court) to “facilitate” Mr. Abrego Garcia’s return to the United States. See id. at ECF 79. This Court is mindful of the Supreme Court’s reminder to afford the “deference owed to the Executive Branch in the conduct of foreign affairs.” [Given that this is a Trump-appointed judge quoting an order issued by Trump-appointed justices, I'm sure that she is genuinely "mindful" of that SCOTUS order.]
This judge then reiterates that " 'Facilitate’ is an active verb. It requires steps to be taken as the Supreme Court has made perfectly clear….Standing by and taking no action is not facilitation. In prior cases involving wrongfully removed individuals, courts have ordered, and the government has taken, affirmative steps toward facilitating return." [citations listed]
Therefore, "this Court will [now] order Defendants to facilitate Cristian’s return to the United States so that he can receive the process he was entitled to under the parties’ binding Settlement Agreement. This Court further orders that facilitating Cristian’s return includes, but is not limited to, Defendants making a good faith request to the government of El Salvador to release Cristian to U.S. custody for transport back to the United States to await the adjudication of his asylum application on the merits by USCIS."
Does that new order represent a court trying to direct foreign policy? I have no idea and neither yet does anyone else, pending the administration appealing this new order up to the SCOTUS.
Yeah, sounds like we're in agreement then. We have no idea if this order, or any other, actually meets the requirement to not interfere with foreign policy.
I suspect, when push comes to shove, that the courts will ultimately be unable to require the return. Perhaps if SCOTUS were 5-4 the other direction they'd find a way to make that order, but even then that's a tricky thing to require. It would mean that the courts have some identified right to overrule the executive in foreign policy, which is extremely ripe for abuse. An easy comparison would be a conservative Texas or Florida judge ordering the Biden administration to enter into negotiations with Mexico or another foreign source of illegal immigrants in order to "facilitate" the deportation of illegal immigrants. I would imagine that would not be palatable to many of the people who want the courts to overrule Trump now.
Sorry, then I misunderstood your point. Don't agree at all with that legal/constitutional analysis.
(Nor your "easy comparison", I mean come on now...if the administration tries to pose that equivalency in SCOTUS oral argument it will be a gift to the other side.)
>Trump defended accepting a personal gift from the ruling family of Qatar of a half-billion dollar glammed-up 747, which he would graciously lend to the government for use as Air Force One until he leaves office at which time the plane goes with him. [Never mind that the Constitution's Emoluments clause specifically bars the acceptance of gifts from any “King, Prince, or foreign State” without prior Congressional approval.]
In other news
>May 1 (Reuters) - A stablecoin launched by Donald Trump's World Liberty Financial crypto venture is being used by an Abu Dhabi investment firm for its $2 billion investment in crypto exchange Binance, one of World Liberty's co-founders said on Thursday.
Thanks for the update, I was going to come in and post roughly the same thing. The only things I would add include: DHS has threatened to arrest democratic house reps for visiting an ICE facility that they have every authority to visit, claiming the house reps "assaulted" someone.
> the FBI Director denied under oath that the Constitution's guarantee of due process applies to noncitizens despite having Antonin Scalia's ruling to that effect read to him;
Does that ruling only apply to non citizens in the US, does being in the US apply special status? Because there’s been a lot of extraordinary renditions prior to this.
And then there’s Maher Arar from 2002. Deported from JFK to Syria, back when Syria and Assad were not on the naughty step.
Yes that SCOTUS ruling in particular, and the Constitution generally, applies to people within the US. (With the caveat that "within the US" is not legally limited to the 50 states...."the US proper plus territories currently controlled by it" would be the fuller answer.)
Well, the way Yarvin is expressing himself might be overly poetic and flowery for my tastes, not to mention profane, but it's not that hard for me to mentally translate it into what I would consider plain english, at which point I could engage with it in the same way I would engage with anything else.
Assuming it's Yarvin you're accusing of being postmodernism, I'm not sure.
I, also, can translate the words of Yarvin into something plain and reasonable. But he chose to translate his thoughts into words hateful and profane, so why would I want to do such a thing?
After seeing Scott recommend the Center for AI Safety free online course on AI Safety, Ethics, and Society, I took the course, and can report that it was great! Good weekly discussions, the textbook has good coverage of the relevant topics, and the project phase at the end was a great opportunity to dive deep into an area of interest. For my project I wrote a post clarifying the conclusion I have come to after reading much of Scott and Zvi's writing and taking the course - namely that we absolutely are not ready for AGI and it's kind of crazy that building it is still the stated goal of many of the top AI companies, and we should thus work to build consensus on (1) that this is stupid and dangerous and (2) what to do about that fact given the systemic forces pushing us forward. See the post here, I welcome any feedback or corrections! https://natezsharpe.substack.com/p/what-if-we-justdidnt-build-agi
It seems one of the most popular arguments for imminent AGI goes "skeptics said AI would never do x and then AI did x! They said it would never do y and y happend a few weeks later! Then they said those things didn't really count! When will you all wake up?????"
I think this is a bad argument. Bear with me.
ALICE: AI may solve equations, but it will never write a story.
BOB: Actually, AI has just written a story. "A man planted a tree. The tree grew. The end."
ALICE: I meant a real story. Like, with description and a multi-part plot.
BOB: Lol, moving the goalposts. But as it happens, AI has just produced such story.
ALICE (*reading it*): That's...just a copy of an existing story, with some paraphrasing and word-swaps. That's not what I meant.
BOB: Ha ha, you just keep shifting your claim. Be precise.
ALICE: AI will never produce a story that will win an award.
(A month later)
BOB: An AI story has won a prestigious award.
ALICE: Only because the judges wanted to "take a stand against ableism and human supremacy". That doesn't count.
BOB: I can't believe it. You'll never ever admit you were wrong.
The point is that Alice does, in fact, have something she means when she says "story", even if she can't describe it in advance. Even if she gives ten conditions, and gets presented with something that satisfies all of them, she can still say "that's not quite right, that's not what I meant" and be telling the truth. "I know it when I see it, and this is not it".
Yes, there is some Carol who really is satisfied by the stories AIs produce, and some David who, even when presented with exactly what he was thinking of, will lie and say that wasn't what he meant to avoid admitting he was wrong. But Alice is not Carol or David. There *is* some story that were an AI to produce it, she would be completely satisfied, and happily admit she was wrong. But that simply hasn't happened yet.
And I think there are a lot of Alices in the AI debate, even though there are a lot of Davids and Carols as well. And so there's a fundamental circularity in the pro-AI argument. The other side say that intelligence is really subtle and not something we understand or can model or describe. The AI side say that if they think that, they need to model and describe exactly what things AI will not be able to do, in advance. Which is just assuming that the nature of these things ("art", "emotion" etc) are things that can be modelled and described. Which is precisely what's at issue.
I feel "People have been wrong in the past, therefore you're wrong now" is maybe too strong a statement, but I think there's _something_ to that logic.
In some instances it seems like a legit argument. If tell you "I'll buy a lottery ticket because my horoscope said it'll be my lucky week" I feel it would be fair to respond with "good luck with that, but have you seen how it turned out for everyone else who tried that". I could answer with "what have the failures of past people got to do with me?" but I'd be wrong.
In other cases it's clearly an error. If I told the Wright brothers that "people have tried building flying machines, it never works, give up, man is just not meant to fly" then I'd be wrong.
I haven't yet found the good rule when the argument applies and when it doesn't. At least not in a way I could articulate. I tend to think that "people have been wrong in the past" is evidence but not proof, though it depends on the details ofc.
It's a bad general heuristic, but there are much better heuristics that can be used instead. "Horoscopes are fake" is the better heuristic. "The odds of winning the lottery are incredibly rare" is a good general heuristic.
Betting that nothing ever changes just seems like a worse argument than individualized arguments against the specific claims.
I mean isn't the horoscopes are fake heuristic basically just the "people have been wrong in the past" heuristic?
The two things I know about horoscopes is that a) we don't know of any mechanism for why they would work and b) so far they've never worked (better than luck)
But it seems that (a) is not load bearing, in the world in which horoscopes reliably predict the future we would say "they work, we just don't know why", argument (a) notwithstanding.
And if you'd ask me for a reason for why I don't buy horoscopes I'd answer something like "It's been tried, systematically, and it's never worked. Why would it work now?" which is basically exactly the general heuristic.
Also the general heuristic has the upside that you can use it on things you don't understand. If you offer me a game where I pay $2000 and may win $2million and I observe that none of the first 300 people who try ever seem to win anything, then I won't play your game and I believe that's reasonable. And that works without any knowledge about what your game is.
$2 million divided by $2,000 is 1,000. You would naively expect at least greater than 1,000 people to play the game before reaching a winner. Otherwise the game clearly loses money and in my mind would be even more suspect. I still wouldn't play it, because $2,000 is a reasonably significant amount of money and 1/1,000 odds is bad odds (the lottery would be something like 1/2,000 odds for the same payout).
You don't need to use "people have failed to win before, so it must be rigged" to conclude that this game is a bad idea. You can do the math on that one, as well as the lottery.
In terms of a blank slate review of horoscopes, I would agree with you. With that said, we don't actually have a blank slate. We have a rich and recognized history of scams and shady behavior related to such things. You could actually argue that we do live in the world where your "a" is true about horoscopes - they are reliably correct. Millions of people can read the same horoscopes from the newspaper and conclude that they were correct. They do this by being so vague as to be unfalsifiable. "You're going to meet a stranger" or whatever is so broad as to apply to almost everyone, and actually ends up conveying no information at all. I look at them and conclude that even if they were correct, it would offer no meaningful information or change in my life.
I should've added "(better than luck)" to both conditions. I'm with you that horoscopes are so broad that in some sense they work. But not better than luck. Or, perhaps more precisely, that to the extend horoscopes predict things, these predictions come true about as often as statistics would predict anyway.
And you don't have to go to "people failed it must be rigged", I agree. My point is that even absent all other knowledge it's reasonable not to pay 2k to play this game, because what makes you think you're the special one that'll win it, when the other 300 weren't?
I believe that stance is reasonable without requiring any further assumptions about how the game works, how likely it is this is a scam or anything of the like.
Doing some back of the envelop math mostly projects your own assumptions unto the scene, they might well be accurate but they might also be wrong. If you can find a reasonable way of action without having to make further assumptions, then all the better.
Which is exactly the usefulness of general heuristics. Yes, when you have accurate info then I agree, better to math it out. But for enough scenarios irl it's just not that realistic to gather all the facts, and I'd argue "has been wrong before, often, so is probably wrong now" is quite useful then.
Horoscope.com for today - This isn't a good time to make any big relationship decisions. If you're thinking about starting a relationship, moving in together, or getting married, you may feel a bit troubled by your decision.
Even outside of the "If" in the second sentence, that's insanely vague and of obvious applicability in just about any relationship where people are considering a major life step. And even then it's "may feel" a bit troubled. It's plausibly true for everyone, and therefore worthless.
It's much simpler, actually, it's a bad argument because it fundamentally misrepresents not only what actual "skeptics" (as in, people interested in the technology, not some rando human exceptionalists on the internet) say, but [literally everything we actually care about].
The key trick is "be precise". The hypesters insist on easily demarcated goals so that they can either a) brute-force something that meets them out of the AI via infinite monkey theorem b) teach it to the test. Then they just won't accept "uh, this obviously does not generalize" for a rejoinder, and at this point I'm inclined to believe many genuinely do not understand why it matters.
Agreed. Like the r's in strawberry thing - of course no one is going to be impressed that hundreds or thousands of man-hours and millions of dollars will allow some of the most powerful companies in the world to get a computer to spit out the number "3" when asked that specific question. Maybe along the way they also figure out how to get it to count letters in other words. That's far from impressive to anyone. It was embarrassing that the systems would consistently get it wrong, but not praiseworthy to get it right.
I'm much more interested in generalized ability and reliability. A system that counts basic information wrong is not generalized or reliable, even if the specific examples can be fixed.
Does generalized ability have to mean ability at every task? Why can't computers, like humans, have their own peculiar strengths are weaknesses? Maybe AIs will be able to run Fortune 500 companies and come up with Nobel Prize winning theories of physics without being able to count letters in words. It's just as unfair to call a machine unintelligent because it can't count the number of r's in "strawberry", as it is to call a human unintelligent because he can't multiply 100-digit numbers in 2 milliseconds.
The question is whether the AI can in fact "run" a Fortune 500 company, or whatever other task is set before it. There's an amazing level of detail required to work in such an important position. The impression that people get from a system that cannot do a simple task is that there are one or more significant flaws in that system. Maybe the system would do well running the company most of the time, and then once a year might do something really bad like sell off all the stock for a completely unknown reason. That would be really bad, bad enough to never use the system even if it was 100% perfect the rest of the year.
If we tried to get an LLM to run a company right now, it would be terrible. It would fail in very obvious ways, but worse, also fail in many non-obvious ways. We don't actually know what it would take for current LLMs or a new approach to AI to be able to run such a system, so it's difficult to know how close or far we are away from being able to do it. Small details like miscounting letters (a task many humans do easily) is potentially an important clue. Even knowing that the reason LLMs struggle with this has to do with tokenization (instead of stupidity) doesn't help much, because it's impossible to determine how many things the LLM does might also be related to tokenization. Brute forcing a system to know how many letters are in a word doesn't give us any assurance that the actual generalized problem is or can be fixed.
Yes, by definition. AIs that do a specific thing well, including much better than humans, are just, well AI, G not included. They can be and are immensely useful, but singularity they won't bring.
Humans' intelligence is general not because we can multiply in 2 miliseconds, but because we can (teach ourselves to) do it at all.
(Though I have to say, contra the other commenter, the reasoning models' ability to count letters, from what I've seen, does generalize to performing more complex tasks on character strings described in natural language. I'm not sure how impressed I should be with that, since I'm pretty sure they've been specifically trained to perform that particular task in response to generative models being a running joke in this area, but it does look like a genuine step in the right direction.)
Then neither humans nor computers count as General Intelligence. Humans cannot learn to multiply 100-digit numbers in 2 milliseconds no matter how much they train, nor can they beat Stockfish at chess or AlphaZero at Go. On the other hand, as Mr. Doolittle said, AI *can* be trained to count the number of characters in words, but undoubtedly they'll still be unable to do things a human can do easily. Insisting on judging AI by the skill they lack the most while not doing the same for humans makes no sense.
2 miliseconds is an arbitrary restraint that has nothing to do with the concept of generality.
Jack of all trades needs not be a master of any of them, again, it just needs to be capable of figuring each one out. Simple tasks like counting letters in a single word are particularly telling because if you can't get even them out, then, well, you just can't do it at all.
It has a lot to do with trust. I would trust an AI chess program to give good advice on moves, but not infinitely. I heard someone came up with a really silly way to beat AlphaZero that would beat approximately zero human opponents. I don't know Go, but I understand it had to do with initial board setup and doing something that would never have happened in any real game for the training run. AlphaZero would repeatedly fall for the same setup and lose due to the weird technicality. This could be fixed for future versions, but it's an example of how a "superhuman" system can be incredibly limited - not just to the fact that it only plays Go, but also to not being able to work with many simple variations or unexpected changes.
Real life is far more complicated, so even a system the repeatedly did well at a common function (like self driving cars in SF) isn't trustworthy outside of where it's strong. In a city that isn't mapped out precisely (or if something big changed like an earthquake or a fire) the cars would not be able to function like they do in a fully mapped city. I don't expect to see many self-driving cars in any cities that experience winter, at least during the cold months.
I think it's important to remember that in 1980, people said computers would never be able to play Chess, and they were phenomenally wrong by any metric you want to use. In the long run, there's no reason to think there's anything a human can do that they can't also automate.
I'm very skeptical of AIs being some kind of fundamental shift. I happen to agree that computers can be programmed to automate almost, maybe all, human tasks. But that's far less interesting than it sounds. If you had the best and brightest team on the planet trying to automate a regular function you could have done amazing things in the 1980s as well. We in fact did a lot of that! But none of it was even considered AI. A CNC machine can be programmed to cut very precise tools regularly at superhuman levels, but no one argues that makes them AI.
Can a chess bot help you run a factory or fold my laundry? It can't even try. It's nonsense to think about. Billions of dollars can definitely allow the brightest people on the planet to automate a factory. That's neither new nor particularly useful. It's certainly possible without AI, and at this point trying to get any of the AIs we have to do that kind of work would be a disaster.
The goal of AI companies is to produce a general intelligence, not individually automate thousands of functions through a non-ideal interface. I happen to think that LLMs do a great job with natural language. Spending lots of money and time on getting them to do okay at math, okay at programming, etc. is probably bad ROI. The fact that they are better at programming is very likely an artifact of the people who work on them being programmers and making that a high priority. We've seen much less progress on much simpler functions (still waiting on that laundry!).
...great, now I'm disappointed that my dryer doesn't fold my laundry.
A chess bot and a factory bot can both be put under a supervisor bot, who would be trained to select whether the problem is related to chess or factories. Put all three in a box and it looks like a general AI from the outside.
But more to the point, everyone's endgoal is automating the scientific method, from hypothesis to conclusion, and everyone's fanciful futures stem from their vision of what happens after computers can pose novel questions to themselves and then answer them. And every time someone cajoles a computer into producing a better poem, it's an indication they're getting better at the overall cajoling process, that might at some point lead to automating research.
Quite possibly! That's the aggravating thing about these conversations. As you say, on one level an AGI *might* be as simple as putting many different bots together with some kind of supervising system that identifies the kind of problem in front of it and then choosing the correct superhuman response.
But in your example system with three bots, what happens when instead of running a factory or playing chess, the problem involves playing checkers? If we keep stapling on new bots to solve every individual problem, we're not talking about AGI, we're talking about the brightest people in our society individually automating thousands of discreet functions.
And this is important to your stated goal. If AIs are just referencing an existing automated bot, then they'll probably never be creating a new hypothesis (unless humans already automated science, in which case "AI" wasn't needed) and not answering novel questions.
There's three other issues I see.
One is related to the novel questions. If an AI is referencing hundreds or thousands of other bots for specific tasks, it's very likely to be bad at any new task. If I invent a brand new game, does the system need millions of examples of playing the game to figure it out? A whole new sub-bot? That might be useful in limited senses, but terrible for regular use.
Another issue with that approach is that the supervisor program might misidentify which problem it's dealing with and use the wrong subsystem. In a simple sense, that might be the system thinking there's a biology problem in front of it, but it turns out it's actually chemistry or whatever. This gets way way more common and way worse when the system is doing something truly complex (like running a national economy or taking over the world, but even something much smaller like running a factory).
Thirdly, what happens when something changes? Chess doesn't change, so the chess bot is probably going to continue being fine for most uses. But people like to vary things, so maybe someone comes out with a chess variant they want to play. Is that AI bot going to play the game at a superhuman level anymore? Is it even going to be able to play at all? Something more complex like a factory involves hundreds or thousands of individually automatable points, but also regularly involves significant changes. Even getting a new order for more/less of a widget will result in massive changes throughout the factory. Is the bot running the factory also automated to change itself to a new bot that runs the new setup of the factory? Each step is fraught with difficulties.
In the long run, I completely agree. However, that's the long run. Perhaps it should've been possible to predict Stockfish in the 1980s, perhaps not; but if you went back to 1820s and predicted that no mechanical difference engine (built out of gears) would ever beat a human at chess, you would've been *correct*. It took several radical paradigm shifts to go from arithmometers to Stockfish, and if you were to predict that such paradigm shifts are expected sometime in our future... then again, I might provisionally agree with you. But that's not the same as saying that present day LLMs would ever become superintelligent quasi-omniscient human-squashing oracles.
I'm not sure this demonstrates what you think it does. We didn't build Turing machines out of gears (or pencil and paper, for that matter)...because we found something *better* than gears. I'm really not sure what your point is supposed to be.
My point is that the following two predictions are perfectly compatible and can both be true:
1). Present-day LLMs will likely never become "superintelligent" in any non-trivial sense of the word (just as mechanical calculators could never beat humans at chess).
2). Perhaps one day we will discover some completely new and hitherto unheard-of technology that will lead to the creation of a superintelligent AI (just as electronic computers backed by novel algorithms, both of which were unimaginable in 1820s, managed to beat humans at chess).
But note that statement #2 is not compatible with the following:
3). ...And I can predict exactly when this completely new and hitherto unheard-of technology will arise and how it will function.
> just as mechanical calculators could never beat humans at chess
But this simply isn't true. You *could*, in principle, run Stockfish on a mechanical Turing machine. There's just no reason to, because electronic ones are better!
Alice: "AI will never produce a story that will win an award"
Bob: "An AI story has won a prestigious award"
Also Bob, but quietly, so Alice doesn't get the full story (sorry, intentional bad pun): "An AI story selected out of N attempts, after a painstakingly engineered series of prompts, by an actual human".
That ought to count. We don't penalize humans for asking their friends for feedback on their stories, or for making changes an editor requests. Why should we penalize AI?
If the prompt is multiple pages long and walks out all the important points, that doesn't feel like the AI is doing the important parts. It also seems silly to say an AI wrote the story when you ask it for thousands of examples and then pick the best. That's not too far from the monkey typists "writing" Shakespeare.
"An AI story has won a prestigious award" implies that AIs can generally write good stories, and at least occasionally write very good stories. Similarly, no one would be impressed if the award was "best AI written story" which could only be won by AI submissions.
For reference, this is Scott actually taking the view of Bob against Gary Marcus's Alice. I think it's illustrative to see how close to Bob and Alice they end up being.
My view is that this isn't even close to what has happened. That Alice does not get to wriggle out of their claims just because there's a cosmic sense in which they're still right. In addition, in what sense does Alice have actually good predictive ability if they *can't* articulate what they're looking for? Do we also implicitly trust Alice to keep track of all the reasoning tasks that she can't come up with concrete examples for but AI has implicitly accomplished? I don't think so, because we also don't trust Bob when he says "I've been using AI and there's an ineffable sense in which it's qualitatively better"!
I agree with the debate between Scott and Gary Marcus. But when you look at all the various benchmarks that supposedly measure intelligence, it’s clear that they don’t measure what we want them to measure. GPT-4 (not even 4o) was already maxing out various LSATs and bar exams and such, but you definitely don’t want even the modern LLMs practicing unsupervised as a lawyer!
This isn’t moving goalposts - this is the difficulty of coming up with good goalposts to begin with. Precisifications usually miss some significant features of what we actually cared about when we came up with the vague idea.
I think tests designed to evaluate whether a human student learned what he was supposed to learn in a course of study are just the wrong tool for the job of evaluating the performance of a computer. Ultimately, the test comes down to how well it performs at an actual job.
Consider IQ tests. The reason we care about them is that they predict performance in school and at work. That correlation is why they matter--a test that was similarly heritable and consistent and all but didn't help us predict anything would be pretty pointless. But that's only for humans. If Claude can get a perfect score on an IQ test, it won't tell us much--we don't know the correlation between LLM IQ scores and performance.
That works in both directions. It's quite possible to have many things AIs can't do nearly as well as humans, but also to have AIs be more capable than humans in the right set of domains to let an AGI take over the world or something. There's nothing that says that an AGI that paperclips us all has to also be able to write consistently funny jokes, or make beautiful art, or whatever.
We have a little experience with nonhuman "intelligence" in things like evolutionary processes, markets, bureaucracies, and eusocial colonies. It's entirely possible that you can write poetry and develop philosophy, but your species gets wiped out by unthinking evolutionary processes that lead to the rise of some virus that kills everyone off. In the same way, it's possible for AI to have a very different kind of intelligence than humans, but still be very dangerous. In some sense, the potential danger from LLMs is easier to see because they can talk to us like humans
I agree with all the points in the last paragraph!
But I think the people who believe in the concept of AGI really *do* think that it requires *general* intelligence, which they think is what is behind all
human cognitive abilities, and thus would allow these things.
I don’t believe in general intelligence - I think it’s important to note that things can be dangerously capable in many ways, even while lacking many major capabilities average humans have, because there is no one thing that is general intelligence.
"...in what sense does Alice have actually good predictive ability if they *can't* articulate what they're looking for?"
This may sound backwards, but Alice would actually have *better* predictive ability the less likely she is able to articulate an advanced concept. It suggests that she is able to think of things at a higher level of abstraction by integrating information that is not easily available to Bob. The more pieces of information you work with, the harder it becomes to condense them into a sentence.
For me at least, Bob appears much dumber here because he should know perfectly well what Alice means when she talks about a story. He should have read enough books to know, too.
> This may sound backwards, but Alice would actually have *better* predictive ability the less likely she is able to articulate an advanced concept. It suggests that she is able to think of things at a higher level of abstraction by integrating information that is not easily available to Bob. The more pieces of information you work with, the harder it becomes to condense them into a sentence.
The operational word here is predictive and not smart. What is Alice predicting? Do you think Alice would even grant that the AI has crossed any invisible threshold? Do you think Alice would be any good at predicting capabilities the AI does end up with? Because your response would indicate no, she just has an undefinable and unaccountable ocean of invisible knowledge you can attribute to her.
Reverse the situation. If Bob confidentially says that he sees AI getting smarter in some ineffable way, Alice makes a precise prediction that that capability doesn't emerge, so we then say that no, actually Bob is more predictive than we thought?
> For me at least, Bob appears much dumber here because he should know perfectly well what Alice means when she talks about a story.
Yes, because ascend wrote it so that Bob looks dumb, and then doesn't mention that quietly, 10 months later the AI actually does end up writing a passable short story. But of course, this in fact doesn't come up on any news story, *BECAUSE* it is subjective and because Alice is exactly the genre of person to not earnestly check their claims against reality (oh yeah, says Alice, I used the free model 10 months out of date without any prompting and it failed. I'm definitely still right.) In what world do you see *any* Alices have a mea culpa? Please, if you believe this is true, list them.
The thing is, once the AI is general that's past the point anyone cares to contest the issue, so of course, Alice gets a free pass to say whatever she wants while not actually thinking, only vibing, and everyone attributes deep 10 dimensional chess cosmically and philosophically correct insight to her.
"If Bob confidentially says that he sees AI getting smarter in some ineffable way, Alice makes a precise prediction that that capability doesn't emerge, so we then say that no, actually Bob is more predictive than we thought?"
I think we try to understand what is the true claim being made. Bob doesn't have to lay out every step for how he thinks AI will manufacture nanobots, but he does have to rescind his claim when a bottleneck is suggested and he is unable to pull a single credible counter from his ineffable pool of knowledge. The current method being employed in this space is to trust whoever has the more ostensibly "precise" claim as a type of meta-credibility, without actually evaluating properly the facts of the claim. Whoever can make their claim *signal* more precision by using specific cognitive techniques and figures ends up being trusted more regardless of actual truth, whereas I would urge for true understanding of what is said. This might mean you cannot make or trust some predictions at all until more information is consolidated, and that's almost certainly a good thing.
"Alice is exactly the genre of person to not earnestly check their claims against reality (oh yeah, says Alice, I used the free model 10 months out of date without any prompting and it failed. I'm definitely still right.) In what world do you see *any* Alices have a mea culpa? Please, if you believe this is true, list them."
I don't think you want to go band for band over who rejects reality more, and I don't think it's productive, but at least some Alices *can* have a mea culpa. It is literally impossible for Bob to have a mea culpa because he is predicting fiction and can endlessly shift goalposts (Alice can shift them too, but she can also actually be proven wrong with empirical events).
> I would urge for true understanding of what is said. This might mean you cannot make or trust some predictions at all until more information is consolidated, and that's almost certainly a good thing.
You have also deflected and not answered the question, in the reversed situation, would we say that Bob is more predictive? Or is it really some other generically good sounding word?
Because, my impression is that the type of person who claims the existence of "higher order unspoken concepts" having validity are the type of person who are most eager to excuse the word and status association going on in their minds. Unfortunately, the margins of a deeply nested ACX thread are TOO SMALL to express the complicated social model in my head, and I can only ask forbearance that you do all the intellectual heavy lifting to make what I'm saying consistent. Please and thank you.
I agree that concepts can be too complicated to express, but what I heavily disagree on is that we should consider this a merit instead of a demerit. Adding degrees of freedom in how you express statements goes *exactly* against what good predictors do, as expressed directly by Philip Tetlock and indirectly by Kahneman and co. And I think it's illuminating that the instant we reverse the hypothetical, epicycles get added to make your point hold anyway.
> The current method being employed in this space is to trust whoever has the more ostensibly "precise" claim as a type of meta-credibility, without actually evaluating properly the facts of the claim.
And anyway, if you believe this is what is being done, please link to actual examples of this that have been falsified, in your view. Several times I've asked this of people, and it turns out they were doing word association and not reading the actual statements made. (See: claiming that data centers need to be unilaterally nuked, instead of as a clarifying statement on what constitutes a real vs fake treaty signed between two countries)
> I don't think you want to go band for band over who rejects reality more, and I don't think it's productive, but at least some Alices *can* have a mea culpa.
Between Gary Marcus and Scott Alexander, only one has a mistakes page and has confessed to changing their minds substantially. If you have concrete examples, please link to them instead of hinting at them. I think both in terms of "fake" metrics of things like having mistake pages and "real" metrics like actually changing their mind, Scott is the superior here, and so are people like e.g. Daniel Kokotajlo, who do scrupulously keep track of things. Is there a skeptic who also keeps track of what they're saying and compare it to reality *at all*? I think the closest is someone like nintil or Matthew Barnett, but they hardly strike me as Alices.
I would change my mind if you find such an example that I can agree on, and I'm registering that right now. I will also change my mind if you find an example that I don't agree is reasonable but I can at least see an attempt was made.
So, since you seem confident that mea culpas can be had, where does your confidence come from?
"You have also deflected and not answered the question, in the reversed situation, would we say that Bob is more predictive?"
I mistakenly thought that the answer was implied. Bob *is* more predictive, just as anyone who is capable of higher level reasoning and knowledge will naturally be more predictive. But you asked what should we do to resolve the dispute if both parties have complex ideas in their head which signal a high degree of complex knowledge, and I said that a better framework is to work through the true object-level of the claims than to just believe whoever has written down more information.
"Unfortunately, the margins of a deeply nested ACX thread are TOO SMALL to express the complicated social model in my head, and I can only ask forbearance that you do all the intellectual heavy lifting to make what I'm saying consistent. Please and thank you."
I think this person is not who I am talking about. I am talking about people who:
- claim to know something you don't (ie: Hypothetical AI is not possible, trust me bro im an engineer)
- put down 1 or 2 points of reason which explain a part of this knowledge (ie: Hypothetical AI will suffer bottlenecks 1 and 2 which I know of because of engineering experience)
- claim that the rest of their knowledge is too complex to express, but invite you to dialogue so you can scrutinise or understand it better. (ie: These bottlenecks might not look like enough information to you, and it certainly seems unfair that I have countered your 50000 word blog post with a 1 sentence bottleneck, but I have all the information in my head so we can go over it in detail as best we can)
"I think it's illuminating that the instant we reverse the hypothetical, epicycles get added to make your point hold anyway."
I addressed this one already, but I applied the same logic in the reverse case.
"And anyway, if you believe this is what is being done, please link to actual examples of this that have been falsified, in your view."
You can just look at Scott's AI 2027 predictions a couple of weeks ago. Several people in the comments noted physical bottlenecks or information-theoretic limits associated with the arguments around AI-based algorithm improvement. Multiple times now Scott has dismissed or ignored them as basically just "We expect AI will solve this". That's not reasoning, but if Scott was willing to put forward a small counter-claim and then invite readers to a greater well of knowledge then sure. He can't really though because he is not an expert in anything related to AI and his lack of technical knowledge shows.
"Between Gary Marcus and Scott Alexander, only one has a mistakes page and has confessed to changing their minds substantially. If you have concrete examples, please link to them instead of hinting at them. I think both in terms of "fake" metrics of things like having mistake pages and "real" metrics like actually changing their mind, Scott is the superior here, and so are people like e.g. Daniel Kokotajlo..."
Big problem with these discussions is that the only true/correct way to evaluate which side accepts reality more is to take a large sample of people in a controlled way and determine it from there (it would be rather difficult). You have only listed 3 guys which is obviously not a fair comparison. That's why I said it's not productive. People should be analysing claims independently of who made them anyway.
"I would change my mind if you find such an example that I can agree on, and I'm registering that right now. I will also change my mind if you find an example that I don't agree is reasonable but I can at least see an attempt was made."
"So, since you seem confident that mea culpas can be had, where does your confidence come from?"
Notice I never said I had examples of Alice having a mea culpa, I said it was theoretically possible because their claims may resolve as either true or false when an example is clearly presented in reality. With Bob, the claim is that one day, AI will kill everyone or something outlandish. When it doesn't, Bob can argue it might still be true. I have seen this happen multiple times with people like Yudkowsky. It's very similar to the classic "You can't prove god isn't real" problem.
I was and still am a moderate AI skeptic. I confess to having moved my mental goalposts by a nontrivial amount. LLMs and image-generative AIs have gotten a lot more capable much faster than I expected early on, and I've had to update based on this.
I am a more radical AI skeptic, and yes, I do admit that AI had developed faster than I thought it would -- but in a different direction from the one doomers are constantly predicting. In terms of "intelligence" (whatever this means) AI is still comparable to a powerful search engine; but its usability had skyrocketed, to the point where you can reliably use it for repetitive low-impact tasks (as you would use a search engine).
Indeed, I see AI proponents moving their goalposts. The term "superintelligence" used to refer to some kind of a quasi-omniscient oracle who can answer any question with perfect accuracy even as it converts the Earth to computronium using the power of molecular nanotechnology. Now, it refers to e.g. an engine that can outperform most (though not all) human geoguessers (some of the time) at figuring out where a photo was taken.
My worry is not that AI will awaken and kill us all, but that we will downgrade our notion of "intelligence" to the point where even modern LLMs fit the bill. They would indeed become "superintelligent", but only compared to our new level of natural stupidity.
- a quasi-omniscient oracle who can answer any question with perfect accuracy: yes, sure, why not.
- even as it converts the Earth to computronium using the power of molecular nanotechnology - nope.
The fact that so many smart people think the first one inevitably and "almost instantly" leads to the second one tells us a lot about the ability of smart people to create fantastic tales.
I guess it really hinges on our definition of the word "any": if "any" refers to the "currently-available knowledge", i.e., the thing has read every published paper and reasoned which ones are actually "true"... as I'm typing this I begin to see how even that is not anywhere near within the capabilities of the current LLMs...
If it means answering "how to cure a patient with a stage 4 malignant melanoma" with a perfectly designed treatment proposal, no, that will not be happening anytime within this century.
I would argue that "reading every published paper and reasoning which ones are actually true" is, for the most part, beyound the capabilities of *any* LLM, be it present or future. The reason people write scientific papers to begin with is because they actually went out into the world and ran experiments to determine whether their hypothesis is true; the "replication crisis" in science refers to the lack of papers written by people who -- again -- go out into the world and attempt to replicate another team's findings. The key step in this entire process is going out into the world and running experiments: growing plants, launching stratospheric balloons, building supercolliders, running brain scans, whatever. You can't discover very many truths about the physical world merely by thinking about it really hard in your head -- not even if you can think 1000x faster than any human.
Thanks for writing this, I hadn't really fully appreciated how Alices feel. That said, I think that one of the tenets of rationality is that shifting our claims *should* make us anxious, even if it's not "our" fault. We need to try our absolute best to precommit to criteria that make sense, and if for whatever reason we have to abandon those precommitments, we should "feel guilty" about doing so. In other words, because we can't tell Alices and Davids apart, Alices should be extra careful. Actually, we can't even trust our *own* internal feelings to be consistent over time. Maybe someone who feels like an Alice is actually being a David and their values are shifting over time. Sometimes you need to abandon the inside view of yourself and act according to the outside view.
There are some particular failure modes that are prevented by precommitting to criteria, but in am not convinced that these failure modes are so frequent or important that it makes sense to say “We need to try our absolute best to precommit to criteria that make sense, and if for whatever reason we have to abandon those precommitments, we should "feel guilty" about doing so.”
Precommitment involves your less informed self promising something without the benefit of all the distinctions you’ve learned to make after seeing things in reality.
Early consumers of canned vegetables said they thought they tasted like they were fresh from the garden - and they are in fact much more like fresh vegetables than they are like pickled or dried or smoked or salted vegetables, but once you’ve tasted them a couple times it’s easy to tell the difference. Thomas Edison ran a series of “tone tests” (worth googling) where he sent a singer and a phonograph, and wowed audiences when he put down the curtain while the singer was singing, and lifted the curtain but it was just the phonograph, and the audiences hadn’t noticed the shift. (Modern audiences can easily tell.)
Precommitment is good for something you are going to test statistically, and good for certain kinds of adversarial environments, but not good for getting real understanding.
Good point! I agree it doesn't make sense to *never* reassess things, but we should be painfully aware of our ability to deceive ourselves without realizing, and feel duly incentivized to try to anticipate how things might change. This situation is a little different in that the people guessing at the future have at least some sense of the technology they'll be up against, so they are better equipped to carefully craft criteria. I agree it shouldn't be disqualifying to have to readjust our criteria, but it'd be nice to make that as unnecessary as possible since it damages a really good defense we have.
One thing going on in the dialogue above is that Alice's metrics are corrupted after she issues them by motivated opponents. When Alice says "AI can't write a story that wins an award", and then a group forms to give an award to a story written by AI for the purpose of invalidating Alice's claim, it is obvious that this "doesn't count" - we may assume that what Alice has in mind is for her claim to be true whether or not other people have heard her make it.
Compare "Goodhart's Law".
Dismissing Alice as "moving the goalposts" in this scenario looks very bad.
I agree bad actors complicate this, but I think this is mostly a reason to be more careful when choosing criteria, imagining how others might try to Goodhart them. In your example, though, it's overt enough that I think it's fair to say "clearly that doesn't address the thrust of her claims".
Maybe a nitpick, but you're describing things that don't happen in the story: it doesn't say that the award committee knows anything about Alice, or that the award was founded after Alice made the prediction. It's not clear anyone is Goodharting.
A better way to describe the dynamic is that Alice had correct beliefs about AI (it wouldn't produce a story to her satisfaction) and incorrect beliefs about the judges of literary awards (they would have a similar judgment to hers). She placed the goalposts based on those two beliefs, but only meant to be making a strong claim about AI.
She's definitely moving the goalposts now: she chose a clear criterion that was unambiguously met, and only after that wants to revise it. The question is what we should make of her doing so. In this case, I think she has a very good argument that the details of the case didn't prove her wrong about AI, and if she'd thought about this from the start she would have placed the goalposts differently.
But if we didn't have the ability to read Alice's mind, or it were more ambiguous, it would be reasonable to see "Alice's prediction was falsified" as stronger evidence against her beliefs.
No, you've overlooked the part of the dialogue where the committee makes its award "to take a stand against human supremacy". That is a reason unrelated to the story; Alice was not mistaken about whether the judges would have judgment similar to hers - she was mistaken about whether the judges would consider the quality of the story when giving the award.
I didn't overlook that, I took it to be included in "different judgment than Alice" -- they judged based on very different criteria than what Alice was predicting, and which she doesn't see as legitimate. That's why we know she's justified in not changing her mind. I was pointing out that you framed this as an adversarial play against Alice's prediction, which is not how it's described above.
Quick edit: by "judgment" I meant "judgment about what should win an award." Alice's mistake was to think this meant "judgment about literary quality," but that's not the criterion she specified.
I think it's not unreasonable for Alice to think this criterion is implied. Otherwise the issue is reduced to a triviality: make the only criterion the fact that the story must be written by AI, AI writes "Once upon a time, the end", award goes to AI, boom, done.
Generally, I think "skeptics said AI wouldn't be able to do X, and then it did X" is still a pretty good argument against someone saying "AI will never do X." At least, if Alice doesn't have a decent response to this, I feel justified in not giving her position much credence.
AI systems keep improving. If you claim to have some insight into the kinds of tasks AI systems fundamentally can and can't do, you should have been able to apply this insight at any point in the past and been correct. You can't personally do this. The best available proxy is the smart people from the past who confidently and incorrectly said "AI will never do X."
You need to be able to say why you have better insight into the fundamental limits of AI than they did. There are some good reasons you might: you have access to more evidence than they did! But if you're not somewhat humbled by the failures of people to make this exact same prediction in the past, and don't have an explanation for why you're not making the same mistake, then I'd think you're the appropriate target for this argument.
While I agree that hypothetical "AI systems" in general can do anything you can imagine, present-day LLMs seem to be quite limited. For example, they are incapable of learning on the fly and synthesizing their discoveries into a continuously changing worldview, the way humans can (even, or perhaps especially, in childhood). They are also significantly more prone to hallucinations than humans (even humans who have imbibed a lot of alcohol, weirdly enough). Also, all of the most spectacular feats of AI intelligence that I've ever seen require a huge amount of cherry-picking (by humans).
In addition, most of those smart people from the past who said "AI will never do X" turned out to be... correct ! You are significantly underestimating the sheer number of AI claims that have been made over the years, and focusing on the few that panned out; and even those have not yet been fully realized. For example, people claimed that AI had achieved intelligence since ELIZA, but while modern LLMs can hold a conversation much better than ELIZA could, talking to them is still kind of like... well... talking to a really sophisticated ELIZA. So, thus far, I remain unconvinced.
anecdata, part of a conversation with Claude yesterday:
Claude:
For the metals in the list (Ga, In, Sn, Zn, Al, Ag, Au, Cu), the freezing points are used because these materials have very high triple point pressures that would be impractical for calibration purposes. These freezing points are realized at standard atmospheric pressure.
Is there a particular fixed point you're interested in learning more about, or would you like to explore another aspect of thermodynamic fixed points?
me:
"For the metals in the list (Ga, In, Sn, Zn, Al, Ag, Au, Cu), the freezing points are used because these materials have very high triple point pressures that would be impractical for calibration purposes." This sounds strange. What is the vapor pressure of Ga at its melting point?
Claude:
You raise an excellent point, and I need to correct my previous statement. You're absolutely right to question this.
The vapor pressure of gallium at its melting point (29.7646°C) is extremely low, on the order of 10^-35 Pa, which is essentially negligible. This is true for most metals at their melting points
( I don't have data on the _rate_ of hallucinations, nor of the analogous rate for humans, sober or drunk. )
I asked the Tencent LLM about Steeleye Span a while ago, attempting to lead up to the question about The Victory that I've mentioned in the past.
The conversation went something like this (the LLM responses are quite long in reality, but I'm providing only the parts that I consider of interest. All of this takes place in Chinese except that English names appear in English):
[me] Do you know England's Steeleye Span?
[LLM] Yes!
[me] I like their song The Victory.
[LLM] There is no record of Steeleye Span performing a song by that name; maybe you're thinking of another band, or a song that doesn't appear on their studio albums?
[me] What? What songs appear on Steeleye Span's "Storm Force Ten"?
[LLM] [A tracklist, with blurbs for each song. None of the songs listed appear on Storm Force Ten. None of them are by Steeleye Span. Some of them are actual songs by bands similar to Steeleye Span. One of them is a title structurally similar to the title of an album (but not a song) by Steeleye Span. One of them is "The Storm Force Ten", noted in the description of the track as the song that gave its name to the album (no such song exists). Mostly they're just made up with no obvious characteristics of interest.]
-----
I followed this up by asking Google Gemini, which provided a correct tracklist that was noted as coming from Google search results rather than text generation. I can't fault the choice of approach.
I kind-of wonder whether any of the current LLMs that have been put in customer service roles have advised any customer to do something fatal... Of the "helpfulness / harmlessness / honesty" triple, I really wish the AI labs would put _way_ more emphasis on honesty...
I called my credit card company a few days ago, and in the time between the automated system telling me my balance and me trying to pay my balance, the system forgot how much I owed and changed the number on me. I had to jump over to an actual person to resolve it. Which they did instantly, because that's the dumbest kind of mistake to have.
No, and I also admit that drunk humans can see a lot of weird shit. However, when humans get to that point they usually just shut down; in addition, the weird shit they see is still reasonably comprehensible, e.g. pink elephants and such. LLMs usually hallucinate things like perfectly detailed reports of court cases (complete with references) which do not exist; and they do it all the time. Although I also do admit that there exist many drugs other than alcohol, some of which can trigger much more potent hallucinations (in humans, that is).
I think the issue is that LLMs hallucinate things that are *much* more comprehensible than the things people actually hallucinate. Though some amount of human confabulation is very much like LLM hallucination. Ask a person why they did something, or ask them to fill in some explanation of why some steps work in something they do all the time but have never studied academically, and they’ll come up with something that sounds plausible but is actually totally made up.
I agree that Alice is right not to change her mind here. She has held a consistent position that she's just been unable to precisely articulate. I also think Bob is right to be a bit annoyed with her: she espoused some positions that turned out to be false, and now she's claiming she was secretly right in some inarticulable way the whole time. Unless Alice is able to specify her position more clearly, she shouldn't expect to convince anyone except herself. (Which is unfortunate, since she's in possession of a true belief that others would benefit from!)
So Alice shouldn't be too comforted by being technically correct here. It's easy to shift your own goalposts without noticing, and it's basically impossible to check after the fact whether they were "really the same" all along. One of the best ways to prevent this drift is by making clear predictions -- and making predictions that are later falsified is *evidence* that you had a false belief. (In this case, Alice has a good argument that the false belief was about the award committee rather than about AI, but in many cases it won't be as clear.)
"Unless Alice is able to specify her position more clearly, she shouldn't expect to convince anyone except herself."
But Alice has already convinced me. I think she can convince any reasonable person, because obviously "in the beginning there was a tree. Then it died. The end" isn't what any reasonable person means by "story", and obviously an AI winning a literary award just because the judges wanted to fight human supremacy isn't what Alice meant by winning a literary award. Yes, Alice could have been more clear about her criteria, but there are obvious things in the world that are hard to define.
Sure -- like I said, Alice isn't doing anything wrong with respect to her own beliefs.
But if you can't communicate your beliefs, you shouldn't expect to convince anyone else or be recognized for having been right. If you try telling Bob that you were right all along but can't explain why, he will have no reason to believe you.
I don't know, what I am trying to say here is that it is not possible for Alice to convey her point because it is too advanced, so asking her to be precise is unreasonable. The onus would be on Bob to reach her level of inference, rather than expecting too much out of human language.
From a personal epistemic perspective, Alice is correct and has done nothing wrong. This is what I take you to be saying and I agree.
From a social perspective: she's talking to Bob and trying to convince him of something. Once she decides to do that, she needs to be able to communicate! If she's not able to communicate her ideas to Bob, she will not be able to convince him of anything.
The issue might be with Alice, with Bob, or with the content itself. My point is that if Alice accidentally miscommunicates, Bob isn't wrong to point that out and say she's moving the goalposts. She is, and he has no way to distinguish her from a David who's doing so maliciously.
(I don't quite understand your last point: It seems like you're both trying to say the content is not transmissible, but also this is Bob's fault for not being on her level? But if he were at a high enough level to make the right inferences based on Alice's words, that's another way of saying the content *is* conveyable through language. I'm not sure assigning blame for the communication breakdown matters though.)
"She is, and he has no way to distinguish her from a David who's doing so maliciously."
Good point, I think. It's one I wrestle with a lot in online discussions. I've seen people who -might- be right, but are doing such a terrible, abuseable job of making their point that they end up doing it a disservice.
And it takes two to tango here as well. An audience can be so bad at absorbing an idea that they might claim Alice is David, not realizing their problem.
I think it can be very hard for people to differentiate the internal experience of "I hold a consistent position that I can't articulate" from a bundle of aesthetic(/moral/whatever) preferences that don't actually cohere to a falsifiable test. As non-mind-readers, we should be suspicious of Alice' claim, whether it's about AI or anything else.
I think it’s problematic to assimilate “aesthetic/moral/whatever” to “don’t actually cohere” rather than to “consistent position that I can’t articulate”. Usually it’s probably going to be some mix of the two.
I didn't mean to imply that because subjective preferences inform/support a position that we should disregard that position as incoherent.
I'm not really sure what you mean by "Usually it's probably going to be some mix of the two" - surely, for deciding if any given thing belongs to a category (in this instance, whether or not the output of the AI counts as a story), either there is a consistent position for what the category encompasses (there is some explicit definition of story which Alice would agree to, even if she couldn't formulate that definition herself), or there is no consistent position, and we can't reason about the category (there is no string of text that Alice would consider a story, if it happened to be assembled by an AI, even though she would not endorse that position).
If we're making arguments about what a technology can/can't do, those have to be falsifiable in order to have predictive value. Arguments of the form "AI can never do X because things are only X if I personally determine them to be so" don't get us anywhere.
If it were up to me, I would refine this to, "If Alice can conceive of a claim that is worthy of our serious consideration, then she ought to be able to phrase it coherently."
What do you mean by "ought" here? If Alice doesn't phrase her claim coherently, is it because the claim isn't valid, or is it because Alice is just deficient in articulation?
David's comment seemed to suggest that we should be skeptical of Alice's claim for various ambiguous reasons. My point was that I am okay with rejecting claims purely on the basis of inability (or unwillingness) to phrase coherently. Part of this is that intentionally ambiguous phrasing is a way to keep moving the goalposts but really, if one has a claim that is worthy of our consideration, then it should be well thought out. And if they were capable of that, then they should either write it properly or get help from somebody who can.
Partly, I guess, it feels like we are doing engineering here and, as an engineer, nothing annoys me quite like being handed a document which shows a whole lot of engineering work was done, but completely failing to communicate what the result of that work was.
I don't disagree with you, but for the sake of sparing us all more ambiguity I'll spell out my argument a little more clearly.
Claims like Alice', that "I can't define it but I know it when I see it" can resolve in one of two ways: Either Alice eventually arrives at (or sees and endorses) an explicit definition for the thing she currently can't describe, or alternatively there is no explicit definition which Alice would ever endorse, and the thing we're talking about isn't a coherent category.
We can't come to a coherent conclusion arguing from an incoherent premise, so until we know which way Alice' implicit definition resolves, we should avoid engaging on those terms. We can have fruitful discussions to come up with an explicit/falsifiable definition of "story", but we can't argue about if AI will every write a story if there's no way to know ahead of time what counts as a story. We should take seriously the argument that "there's something meaningful here that bears on {AI}/{Other important topic}, but we can't describe that right now" but we should disregard the argument that "This thing I can't describe right now bears on {AI}/{Other important topic}, failing to address this is a hole in your argument.".
As a software developer, I can sympathize with ambiguity annoyance.
My angle here is that our goal is to get the right answer, whatever it is. We normally go about it by temporarily playing the role of an adversary, but we and Alice ought to share this goal. It might be that Alice is being deliberately ambiguous; it might be that Alice is trying her best, but is bad at communicating; it might be that Alice is trying her best, but her claim can't be defended in the limit - a master communicator couldn't defend it, either.
Part of our task is to play the adversary role, but unless there's a designated arbitrator, we should probably also have a side process trying to understand what Alice is saying, including asking her questions to try to clarify her claim. We're permitted to consider the possibility that she's ambiguous on purpose (or stalling for time, etc.), and disengage and perhaps even decline to engage again in the future.
Matching your engineer's annoyance (and mine) with ambiguity is mine with cases where I've seen people abuse the principle of concluding the other side is ambiguous, by insisting on hyper-specificity, pummeling an Alice who might be trying her best, and if she stumbles even once, they invoke the ambiguity clause in order to win. In other words, I've seen abuse in both directions. (I've even seen abuse of this observation of abuse. Some people can be very clever, and a lot of people have less incentive to get a right answer than they do to just win arguments.)
I wrote about why we should not moralize procreation, and why there should be different measures of epistemic position depending on what criminal justice goals you have. https://hiphination.substack.com/archive
The only meaningful response to that idea was best expressed by Razib:
"Should Religion Play a Role in Politics? [link to a blog post featured on the New York TImes] This is like asking 'should men have sex with women?' It's always going to happen."
Following on from AI’s success at GeoGuessr, I decided to test the free platforms on radiology. Neither was a big success. ChatGPT undercalled an acute radial head fracture, saying it was a normal elbow X-ray, and Grok overcalled it saying that there was complete dislocation in addition to fracture. Neither was able to correct with prompting by query (Chat doubled down on normal when I asked it specifically if a radial head fracture was present, Grok insisted the elbow was dislocated when I asked if it was sure.) I only got correct diagnosis from Chat when I insisted “this image shows a radial head fracture.” Oh well. Radiologists to stay employed for now.
Chatbots were not (generally) trained on medical data or images, and it likely would have been low-reliability data. There are quite a few radiology companies that are using machine learning specifically with high quality data of MRI, CT, and PET scan images. So far, it has seemed quite promising. However, a chatbot would not be the relevant usage of ai in this case. Take a look at companies like Therapanacea, deepc, Aidoc, or Siemens (partnering with Britain's NHS). For example, if you can take a look at the white paper for Rad AI (Siemans), so far results seem to be very good, particularly for expediting radiologists' workflow. There are white papers available with relevant studies and sourcing.
Sure, but what does that say about some "ChatGPT ver. X" becoming literally superhuman in every respect and taking over? Having a bunch of narrowly specialized expert systems sounds useful, but is a far cry from your typical doomsday scenario where that radiology AI becomes sentient, finds an Internet-connected device, and takes over the planet.
Sure, if an LLM would suddenly know, how to anaylyze Xray images out of the knowledge, how the human body is composed and how Xray works, that would be a step towards superhuman intelligence.
However, I am pretty much sure, an LLM canot be better than the training data. Without superhuman training data, no superhuman output.
If the creators of ChatGPT ver. X want it to be good at analyzing CT scans, they could train it on a high quality dataset and it would be good at radiology and still have its general knowledge. The specialized radiology ML isn't going to become general purpose; the general purpose AI will incorporate the specializations.
Yes, this is why my views on "AI" are a perfect mix of wild-eyed optimism for specialized systems and an "I can't believe smart people expecting this to happen in two years" pessimism for "AGI" or "ASI" or whatever acronym is supposed to do Magick.
Would you expect to see a NN-based specialized radiology software designed to do one thing only, read x-rays, with modules for, e.g., "Fractures", "Lung disease", "Brain Tumors", to be much better than a generic ChatGPT?
That's kind of where I see the path forward, huge open field for specialized AI applications as we already see in image processing, for example, while "AGI" remains mostly hot air and fearmongering.
This dialogue with Moldbug baffles me, mostly because there's a dialogue at all.
From the outside, it's like watching a twitching, seething mass of concentrated hatred and contempt, scream in a voice of a thousand howling voices "Coward! Tear down society! Build the God King a throne of skulls if he will tear down our rivals for us! REOPEN AUSCHWTIZ!"
Granted, 82 of the eyes if his twisted frame are winking in an exaggerated fashion when he says the last line, but other half-formed hands are sharpening knives and drafting legal memos justifying the active under unitary executive theory.
And then Scott nods sympathetically and says "hmm yes, you make some good points and nobody can doubt your lucidity, but on balance I think we should perhaps not drown the streets in blood so that our biological lessors are swallowed up before we are. I will of course update my priors if conditions change. Thank you for giving me much to consider."
This is such a confusing position to me. None of what Moldbug says sounds like it's in a uncivil/violent tone at all. His and Scott's levels of courtesy are comparable. The points being made are real, so they receive counterarguments. Why wouldn't they?
Isn't the point not so much about tone as much as coherence? Scott did a great job making as much sense of Curtis's 'argument' as he could in his response. But, it did seem like he was contorting himself to make a normal counterargument. In a lot of places it was really hard for me to understand what Yarvin was even saying. For example he refers to a 'regime' twice:
"You forget that this regime has spent the last century managing public opinion with every carrot and stick it can find, up to and including asking professors to compose their own inventive and detailed loyalty oaths to gay race communism."
"You forget that during the lifespan of this regime, whose crimes have no statute of oblivion, many once civilized places became desolate and dangerous, forcing whole populations to flee—including most of most of our cities, even the hometown of the new Pope lol. Oh yeah that"
What is his definition of regime here? There has been a regime for at least a century, but during this 'regime' people have fled 'most of most cities?' It's committed "crimes with no statute of oblivion". Is the regime the US's current governmental system, is it the political left, is it democracy itself, or classical liberalism? What existed before the regime? Has it been here forever? Presumably someone built the 'civilised places' to begin with. Who sets the regime's positions on managing public opinion, how has orthodox opinion changed so much over time if it's managed so tightly, surely the 1962 version of the 'regime' wouldn't approve of 'oaths to gay race communism'?
I realise things like this might be more evident to people more familiar with his writing. But I honestly don't get it, just comes across as rambling stream of thought and it's kind of funny comparing it to Scott's response.
To be clear, I don't give a single shit about whether Yarvin is courteous to the people he speaks with, although I think throwing around the term "libtard" disqualifies him from that, as does his threat to re-open Auschwtiz to show the libtards how much worse Trump could be. Curtis Yarvin should be cast in the Outer Darkness because his stated political goals are to overthrow democracy (not narrowly, but any form of government where the consent of the governed is a thing that exists and rulers need to worry about) in favor of some mishmash of Nazism, Imperial Russian tsardom, and worship of corporate profits.
I do not believe we as a society should legally declare Yarvin to be a wolfs-head or an enemy of all mankind who sits outside of any legal protection. That would not be a fair or just act. But Yarvin has said that he would do exactly that to all of us and every single other political opponent he has if the boot were on the other foot. Just something to keep in mind when deciding how we all voluntarily choose to interact with Curtis.
Indeed, the stakes are high. Seems like all the more reason to engage with the arguments properly. Proper debate leads to better accuracy. Or, if you want to dispose of investigation entirely and go into the "arguments as soldiers" mindset, then at least ensure that the battlefield is the one where correctness has an inherent advantage, rather than the coin-toss of conflict. https://slatestarcodex.com/2017/03/24/guided-by-the-beauty-of-our-weapons/
My guy, you're lecturing me on being insufficiently polite and charitable while Curtis is saying we all deserve to be thrown in Auschwitz as a reminder to know our place and not complain.
It's in Yarvin's reply to Scott. It's phrased as a joke of course, some sort of elaborate demonstration that liberals need to stop comparing Trump to Hitler. Not that Yarvin really condemns Hitler anymore, except in so far as Hitler lost.
I find the debate frustrating as well. In my opinion Yarvin is little more than a comparatively eloquent charlatan. But I also think it’s likely useful for someone, at least, to engage with charlatans matter of factly and politely.
If a Nobel prize winning virologist were to publicly debate RFK jr on vaccines, it’s not at least self evident to me that the optimal approach would be to treat the latter with the degree of contempt he deserves. Once someone’s inside the Overton window, that’s not really an option anymore.
I would fully expect this debate resulting in even more people concluding that the virologist was one of those contemptuous elites who want to chip everybody, and RFK was just the Man to resist the Cathedral.
One of the core tenets of rationalism is that you treat the argument itself, not the person making it, or what it would imply if true. Scott is the true Caliph of Rationalism.
A lot of people disagree that this is a good thing. You're also allowed to hold that position.
But you coming to his blog and complaining that he's taking Moldbug's arguments at face value is like telling the Pope that Jesus wasn't the son of god.
This is a bit problematic coming right on the heels of “the purpose of a system is what it does”, where he didn’t really make any attempt to come up with a steel man of what the position might mean, and just criticized the literal claim.
I'd be more sympathetic to your point if _anyone_ could agree to what the position might mean.
Quoting Scott:
> In this comment thread, people have claimed that the real meaning of POSIWID is:
Chesterton’s Fence
Moloch
Alienation of labor
Pournelle’s Iron Law of Bureaucracy
People follow incentive gradients
If nothing is changed, things will stay the same
If a system keeps going despite side effects, it’s okay with those side effects
If a system has side effects, those side effects are secretly the whole point
It’s about machines and was never intended to apply to social systems
These are pretty different things! So I continue to think that, if you like one of them, you should consider the possibility that this phrase isn’t a clear way to communicate the thing you like.
> ENDQUOTE
If after 2 articles and a couple of flamewars people can't agree on the definition, I'd say using the empirical way people use the phrase is valid.
I don't like the phrase myself! But there are several valuable and meaningful ideas that are adjacent to it in obvious ways, and Scott didn't even bother to identify those.
(I'm personally partial to "the purpose of a system is those aspects of what it does that explain why it persists and works the way it does".)
One of the core tenets of any way of thought worthy of the name "rationalism", is that a rational being must have a way of quickly and efficiently discarding many, many arguments *without* engaging with them. And frequently the behavior and/or reputation of the person making the argument, will be decisive in that. If it's a good argument, some less obnoxious person will make it sooner or later. Until then, we at least have a defense against the DDOS attack.
Unfortunately, enough "rationalists" have engaged with the bundle of obnoxiousness that is Curtis Yarvin, that someone like Scott probably can't just dismiss them out of hand. But I agree with the OP that it is sad this is so.
A pervasive problem with the Rationalist Movement™ and with people (like me) with personalities drawn to it, is that we don't have robust defenses against Gish galloping. We feel instinctively that anything that looks like a syllogism or a factual claim is worthy of engagement, so we spend hours contemplating it, while the galloper has distracted throngs of us with pseudofactual or pseudological claims and has since moved on to make unchallenged pronouncements to vulnerable people. You're right that any movement that prizes logic needs to reckon with this failure mode, but ACX readers have largely not figured out how to do so.
Mehti Hasan (Win Every Argument) recommends dealing with Gish gallopers by picking out one weak argument and hammering on it. I believe he's talking about interviews.
It's worse than that. Firstly, plenty of people hang out on rat forums who are themselves not rationalists. Secondly, some of them (and even some of the rationalists) *have* figured out a defense against Gish galloping - namely, declare it's Gish galloping and use that to justify not engaging.
The catch is that they'll also use it against arguments that aren't Gish galloping, but resemble it. Hence comes the strategy: Step 1 - make a succinct claim that smuggles a large number of mistaken or controversial premises. Step 2 - wait for the inevitable response that points out all the smuggled premises. Step 3 - declare "that's a Gish gallop! I must be be right after all!".
To a third party observer, it's not enough to side with the party that called out a Gish gallop, because there's often no substitute for inspecting the actual argument, especially if that argument is in a domain the third party isn't familiar with (which happens a lot).
Many such third parties are rationalists, who are well aware of Gish gallops and other debate anti-patterns, who are willing to use their presence as a heuristic to decide who's right, but who are also aware of the anti-pattern of calling out other anti-patterns in order to appeal to the rationalist!
Yarvin does not make an argument. Or if he does, he purposely uses obscurity to hide what his actual argument is, likely so that he can shift goalposts later. The most generous thing I can say about Yarvin's thread is that I legitimately don't know what he's saying, and would invite him to actually make an argument grounded in statistics instead of vibes. I suspect he would never do that because his personal identity would be quickly be annihilated by the contradiction in his views and the actual facts on the ground.
This may be a tenet of "Rationalism" but it is not a tenet of "rationalism." It is clear that in developing a rationalist philosophy, it is necessary to tame the reflex to judge everything without listening but clearly this can be taken too far. One also needs exercise rationalism in reevaluating the consequences of blind adherence to tenets.
To live in the world, you need to filter out the crazy people, bored trolls, conmen, spammers, and the like. But IMO rationalism is about the tools you use to understand the world and engage with ideas once you've decided to pay them some attention.
You don't need Ed Jaynes' version of Bayesian probability theory as the more useful and complete version of logic to evaluate the claims of the guy ranting on the subway about how Donald Trump is the antichrist or the widow of the Nigerian oil minister who has just emailed you an amazingly generous offer. But at some point, you've got to evaluate some idea or argument or claim. Sometimes, that idea will be weird, will violate everything you grew up believing, will be covered in the stench of the outgroup, will be offensive and upsetting and low-status, etc. And that's where you need the tools of rationality most.
But is Moldbug really adhering to tenets? Or using whatever tenets and 'rationalist sounding' arguments to whitewash a personal, emotional response?
I tend to pretend that I am more objective and rational than most, but have learned that we all need to be more aware that emotion has a much stronger influence than we'd like to believe (not that we shouldn't try to control it at times).
Silverax appeared to be excusing a behavior that Scott has, on occasion, been accused of here. That is, "taking at face value the argument of individuals whose arguments may not warrant quite so much deference." He was excusing it based on, "...core tenets of rationalism is that you treat the argument itself, not the person making it." My comments were that real rationalism requires less tenets and more (continuous) self examination. A poor performing tenet should be discarded or modified.
I really agree with this, and to be more direct, Scott:
I absolutely love all of your posts on medicine, psychology, pharmacology, art, and rhetoric. I don’t agree with everything you say but it’s all fantastic writing and has contributed a huge amount to my view of the world. I’ve shared many of the ideas in these posts with my family and friends.
But despite this, I’m extremely hesitant to recommend Astral Codex Ten to anyone who isn’t super close to me (when they ask where I saw an idea, I usually say “a blog”), because of this deliberate courting of and deference to open bigots. Yes, they are free to express whatever beliefs they choose and it’s fine to engage with them seriously rather than with ridicule, but the level of focus and steelmanning they get can seem to evince an actual preference for them.
Would you treat an “extreme woke” viewpoint with the same level of consideration? If not, why not?
I think that if you don't recommend ACX to the sort of people who would get mad at this sort of thing then this is best for you, for them, and for ACX.
Open bigots run the world's most powerful country now, a country in which Scott happens to live.
The only realistic alternative at this point to engaging with and pushing back against their ideas is sealing yourself away in smaller and smaller bubbles, until the power of the outside world inevitably pops them.
If you can't clearly defeat open bigotry in free debate then perhaps you should reconsider your prejudices about what, exactly, you consider bigotry. I would point out the danger of conflating positive and normative views as an obvious failure mode there - HBD isn't bigoty by any reasonable use of the term, for example. From an Outside View perspective the attitude of "well we can't talk to them because they believe in <taboo topic>" generally doesn't come from people who have the truth on their side. Consider if creationists or evolutionists are more likely to require bans of the other perspective in an open forum. If your beliefs require special dispensations to protect them then they're likely not good beliefs.
I doubt anyone has defeated bigotry by open debate before. Therefore, bigotry doesn't exist at all, everything is permitted, nothing is taboo, it sounds fun to be sure. But the taboo exists to prevent you from getting sick, you're not a brave rebel for eating roadkill.
You can "defeat bigotry" in the sense of persuading outside observers. Bigotry is by definition a bad heuristic so it should be easy to prove it wrong if it is indeed bigotry.
You can't "defeat bigotry" in the sense of persuading those who are themselves bigots, again pretty much by definition of the way that bigotry works.
I don't believe so. I'm having a hard time imagining an outside observer who is genuinely unsure on whether we should open an Auschwitz to imprison political opponents, as Moldbug colorfully puts it, and who will be swayed one way or another by anonymous commentators bickering about statistics. This sort of stuff is probably already solidified to some extent by the time you leave childhood.
It's not so much that the observer is on the fence about torture camps. It's more often that both positions don't say "therefore, we should start torture camps" and also say "but the other side will totally lead to torture camps" and the observer can't trivially tell, especially given that each side has an enormous incentive to do that.
The worst thing about Hitler are all the "literally Hitler" arguments. (Cue Norm McDonald spinning in his grave.)
Meanwhile, Melvin's right. Especially on online forums, your audience isn't just the person* directly responding to you, but whoever else happens to be reading but not responding. If they see both sides screaming "no, *he's* the real fascist! Shoot *him*!", then one of the few heuristics they have left is noticing which side seems most willing to give up that type of rhetoric and at least attempt to argue rationally.
Did you read his post about “the purpose of a system is what it does”? Here he just stuck with a few very surface level attempts at something and didn’t attempt to charitably read what truth (or at least, plausible point) it might be pointing at.
Not exactly, but I associate it with the claim that police violence isn’t a malfunction of the system but just evidence that the point of policing is to oppress minorities, or the claim that the number of uninsured people in the United States is not a failing of the healthcare system but the intention. It goes along with revolutionary ideology rather than reformist ideology.
Moldbug has written many thousands of words on the topic.
Point me to an equally interesting post about the truth behind 'the purpose of a system is what it doesn't that a minority of the people using the phrase at least somewhat endorse and there will be something to talk about.
Or, would careful treatment of some other, different, woke point be sufficient evidence?
Uh, I hope it was obvious I didn't mean 'the majority of people agree with moldbug, and crowds are wise' or 'smart people are smart, and smart people are more likely to agree with moldbug'. If unclear, I was saying 'a portion of the people who agree with moldbug are persuadable by honest, logical arguments'
Couldn't agree more. It's like his depictions in Unsong of Satan being at UN meetings! If i had to steelman it, he needs to appear reasonable *to the target audience* aka Moldbug sympathizers, rather than his audience. In politics you need converts, and he is confident any bridge will see net positive immigration... even if his current base don't want them around...
It really does all come down to mass immigration. If one believes in a strong version of HBD, and observes that the normal political process has completely failed to stop the problem, then building the God King a throne of skulls gets a lot more reasonable.
You’re right that these bravery debates are counterproductive, but no one has figured out how to have the real debate yet.
> then building the God King a throne of skulls gets a lot more reasonable
No, it doesn't. There is never a scenario where it is reasonable to build a throne of skulls. If you are advocating for building a throne of skulls, you have lost the plot.
Like, just step back and think about what you wrote. "Because I believe in IQ differences, it is correct to murder a bunch of people in service of a dictator."
Except modern immigration is far from the problem it's portrayed to be. It has not increased Crime, it has not brought the US or most countries for that matter to poverty and it has not systematically changed the culture and law of a single country.
Anyone who wants a dictator forgets 2 of this 3 points, and depending if you count repression and crimes against humanity even the 1, would strongly increase in most democracies where this is applied.
Yes if it's a benevolent king/CEO whatever that won't happen, and yet the cases of an illuminated dictator in modern times can be counted on your hands.
> It has not increased Crime, it has not brought the US or most countries for that matter to poverty and it has not systematically changed the culture and law of a single country.
Crime has increased in Europe.
The claim about poverty is true but neither has immigration helped gdp per capita or wage growth. Poland is catching the U.K. in GDP per capita (at least in PPP) and it’s doing that with no immigration and falling population. (So much for that population decline doomsterism).
And the best we can say is that laws haven’t changed much yet, although sharia courts do exist in civil law.
I was thinking about the US but interesting information, is there any proof that this is related to immigration?
Poland is not a good example since it's an emerging economy whose GDP increase has nothing mysterious, you would need two mature economies of similar nature and significantly different immigration flows.
The Sharia courts do you mean in Europe or US? Because i found no mention of that, would you have an example?
There are no general Sharia courts in civil law in Europe or the US.
The vague claim that they exist in Europe rests on one misunderstood anecdote plus a couple of tightly-limited examples that are outside of civil law.
The misunderstood anecdote is that in the UK in 2018 a court ruling recognized the application of Sharia law when a couple had married only in an Islamic religious ceremony not under civil law. The court said that in that circumstance they could not pursue divorce under UK civil law (which considered them to be simply cohabitating); they could however pursue divorce from a religious marriage under the particular religious law that applied. UK courts remain clear that if couples have legal (civil) marriages then they are bound by UK divorce laws without regard to religious status.
In the far-NE corner of Greece, in a single small province bordering Turkey that has long had a sizeable Muslim minority, there is a hundred-year-old treaty allowing practicing Muslims to use Islamic courts on matters of marriage, divorce, adoption, etc. That treaty currently applies to around 100,000 people.
In Germany marital matters between Muslims with no German citizens involved are handled by German courts attempting to follow Sharia law, with the somewhat-vague stipulation that the outcomes can't violate the German legal system's "broad principles". That has thus far resulted in muddled rulings making, it seems from this distance anyway, roughly nobody happy.
"Prompted by fears that they were discriminatory, Theresa May launched a government inquiry in May last year. Just a month later, the Home Affairs Committee announced its own. In December, the Casey Review by Dame Louise Casey into integration included claims that sharia councils “supported the values of extremists, condoned wife-beating, ignored marital rape and allowed forced marriages”."
A Sharia council doesn't need to have formal legal power to have de-facto power over Muslim women. They are dangerous for the same reason that cults are dangerous even though any member is legally free to leave at any time, and for the same reason that abused spouses often stay with the abuser despite being legally able to file for divorce.
Non-sex crime has been on a long-term trend of decline except for a very modest rise starting in 2021. Sex crime has been significantly rising, but that is probably higher reporting rates.
> "A refugee is someone who has to leave their country out of fear of political persecution, religious persecution, or economic persecution," Ramaphosa said. "And they don't fit that bill."
Considering "Kill the Boer" is a popular political slogan there, I'm not sure how Ramaphosa is saying that with a straight face.
I think saying it's a "popular slogan" is misleading: one very provocative politician is famous for singing it; he was tried though acquitted for hate speech for it (which I think the general pro-free speech vibe of this forum should at least have some sympathy for); in the most recent election his party got just under 10%--not trivial, but "popular" seems pretty strong.
I think Afrikaners face the *prospect* of persecution if the EFF were ever to come to power, which isn't out of the question, but I think currently their grievances are crime, corruption, and mismanagement that don't target Afrikaners specifically. If Afrikaners are refugees for having a government that doesn't deal with crime, then so are all other South Africans, regardless of ethnic background.
> he was tried though acquitted for hate speech for it (which I think the general pro-free speech vibe of this forum should at least have some sympathy for)
OK, but the problem there isn't hate speech. It's incitement.
The one good thing about Trump is how idiotic he makes all his "smart" supporters look. The first go-round it got most of the IDW types. Now he's come for the neo-reactionaries.
Qatar (trying to manage the obviously bad look of a bribe): a Qatari spokesman described the claim that the plane would be a gift as "inaccurate".
Trump: "They're giving us a gift," Trump told reporters at the White House on Monday, adding that he would be "a stupid person" if he did not accept it.
Actually, Trump does that to everyone. His enemies make fools of themselves overreacting to some random outrageous tweet or comment, and then his friends make fools of themselves by jumping in to agree with the outrageous tweet or comment.
I'm not saying Trump is all bluster, I'm saying he gets his opponents to make fools of themselves overreacting to things he says and does. That doesn't mean he's not also doing bad things (and indeed, often enough he is), just that a lot of the loudest outrage is about dumb stuff instead of the really bad stuff.
I disagree: the majority of his actions are being undone by judges, and it looks like his second term will ultimately have no more lasting consequence than his first.
If AGI does everything better than humans can, then it will improve the charity sector. You'll be able to know exactly which charity creates the most benefit per dollar donated, and exactly how much you'll need to give and to whom to save a human life.
Can you say more on why you think this? I can see disadvantages of charity, but large parts of the world don't seem to be able to pull themselves up by their bootstraps. Is your view that the externalities of charities are directly worse than any good they do, or is it mianly about opportunity cost ("the money would do more good here")?
Both, actually. Yes I think charities do long-term harm because they establish poor incentives and those inevitably lead to suboptimal outcomes. But I think the stronger argument is that a utilitarian economic analysis shows that it's better maximize your return on investment regardless of how you value human life. I've made this argument many times here and never gotten a robust response. Here's a more general argument:
I've actually considered starting my own Substack just to collect my anti-EA arguments into a single resource. I feel that I have a fairly comprehensive philosophical and economic refutation of the movement and no one has ever provided a good counter either here or on reddit. By all means feel free to cross-examine me if you think you have a good reply.
If you're still using Google for internet search, here's a One Weird Trick to get rid of all the AI junk and other various and sundry "innovations" of the last decade or so: add ?udm=14 after the url, like this:
After doing a very cursory search, it seems like literally all it does is remove the AI summarization. That is so far from the only problem with modern Google and not even close to "Ye Olde Goode Google". Google has been (relatively speaking), garbage for years before it implemented the AI summaries.
The closest I have actually gotten to Ye Olde Goode Google (and still not quite there) is the paid search service Kagi.
Not everyone will think that paying for search is worthwhile, but I'm personally definitely not going back.
I stopped using Google for search years ago, DuckDuckGo is my default go to. But this Tedium piece came across this morning, so I posted this after doing a quick test. Looks like the rot is more systemic and this hack only removes the top layer of decayed matter.
I tried DDG a while back, and what I found was that I kept having to add the !g because searches weren't finding what I needed, and even in it's fallen state, google was better. Kagi also has the !g bang option, but I don't find myself using it.
This is not to argue against your experience at all. I believe you when you say that you find DDG better.
In discussions about these searches, I have found a lot of people that have your experience, and (on hacker news where a lot more kagi users can be found), people who share mine. My best guess is that perception of search experience across the various options must depend a lot on the kinds of things one is searching for and the way in which different people search.
For whatever reason, my combination of search targets and search technique gets better results on Kagi than on either google or DDG.
This is exactly why it's nice to have other multiple options, and I would encourage people who are using google by default to try other things. It is no longer the head and shoulders obvious best option (and, to bring it back around to the original point: AI search is not the reason for it's lack of Pareto-fronteir superiority)
I'm kind of not even claiming DDG to be "better", it's just that it doesn't track me (at least not to the extent Google does), and it gives consistent results for a given query. It was more of "given that I don't want to use Google, what's an easy alternative that works well enough?" question. DDG fit the bill.
Re #3, in the comments to the response Scott says that housing prices are a stronger case for the claim that the median person is worse off than in the past, but housing does not actually seem to be all that strong a case
You really want to look at rent. I was in the rental market until a decade ago and even then I was getting poorer in discretionary income every year. With recent price rises everywhere it’s probably worse now, even controlling for rent.
Consider the context. We know that in the past it was the norm for people to afford kids and their own home while in their early twenties, even on a lower middle class income. But now most people can't afford that. It's likely that something is missing in the data or the analysis. Otherwise we have to explain why everyone in society seems to have noticed a change that isn't real.
People in their early twenties don't choose to get married and have kids any more (because we've decoupled sex from marriage and kids).
If you do choose to get married in your early twenties, and you don't choose to live in a super expensive area, and you make an okay income, then home ownership in the US doesn't seem that difficult to me.
I think we need to stop listening to people who live in San Francisco or New York whe they talk about property prices. We all have a tendency to assume that our local problems are more universal than they are. I want to hear what people in Milwaukee think about house prices.
Which people can’t afford that? People who want smaller houses in poorer cities on 1950s salaries can still have them. But modern houses are bigger and people don’t want the smaller older ones any more, so they hold off to save up for the big modern ones.
>We know that in the past it was the norm for people to afford kids and their own home while in their early twenties, even on a lower middle class income. But now most people can't afford that.
The homeownership rate is Owner-Occupied Homes/Total Homes. If I rent out a room in a my home to someone who can't find any other housing, the rate is unchanged. If I'm forced to live with my parents because there isn't any available housing, the rate is unchanged. That's why I'm hesitant to jump to conclusions. I've seen historical data showing that the ratio of median home price to median household income has increased dramatically. If I take those numbers and run with them then the housing situation looks terrible. So clearly nuance is needed to figure out what's actually going on.
>If I rent out a room in a my home to someone who can't find any other housing, the rate is unchanged. If I'm forced to live with my parents because there isn't any available housing, the rate is unchanged
>historical data showing that the ratio of median home price to median household income has increased dramatically
Not so much, if interest rates are taken into account. And much of the increase has been very recent (last 5 years), whereas the claim is about long-term trends.
Well, it isn't my graph, but it is mortgage payments as pct of disposable personal income. Wasn't the whole subprime mortgage thing all about people with relatively low income getting mortgages they couldn't get previously?
I think the problem with the first one is that it is total quarterly required mortgage payments divided by total quarterly disposable personal income. So it isn't median, its mean, and as you say, lots of new people on the market, and lots of investment properties. From memory houses were actually really cheap at that time for the median person. Lots of zero deposit loans.
So you'd want to find something like mortgage payments for the median house, then divide by disposable personal income of the median person (or household?). Still need some way to factor in the deposit.
>From memory houses were actually really cheap at that time for the median person. Lots of zero deposit loans.
But they weren't cheap, because the mortgage payments were high. Take a look at an amortization calculator. Someone who buys a $200k house with 20 pct down and a 5% 30-yr mortgage pays $308k over the life of the loan. With zero down he pays $386k (actually more, bc he won't get the same interest rate).
Moreover, if homes were so affordable, why did so many borrowers default?
Finally, while it would be nice to have median, since we are comparing over time, what is important is that we are comparing like with like, unless we have reason to think that the gap between median and mean has changed over time (if anything I think the gap has grown over time, so the graph probably overstates current costs).
Curious Scott's or any psych researcher's thoughts on Cheng, Y. C., Huang, W. L., Chen, W. Y., Huang, Y. C., Kuo, P. H., & Tu, Y. K. (2025). Comparative efficacy and tolerability of nutraceuticals for depressive disorder: A systematic review and network meta-analysis. Psychological Medicine, 55, e134. I read it (and comments elsewhere about it) as saying a series of nutraceutical may match or beat SSRIs for depression.
I've been considering whether it would be possible and interesting to run prediction markets in a LARP setting. Does anyone know of a system I could use to run this which I could execute with nothing more advanced than a slide rule and ledger? (Ideally I could pre-calcuate some ratios and have a reference tome to look them up in.)
(The system in question is Empire in the UK, which has ~4000 players trading such that there is actual economic activity happening, like a recent liquidity crisis caused by spending too much coin with foreign merchants.)
Further specifications that might be relevant:
* I'd want to denominate things ideally in the smallest coin, a "ring" - players get an income worth somewhere around 120 of these, though not all of that in pure coinage. Expecting people to have ~10 rings in their pocket is reasonable; basically I'd like to keep the stake sizes a) integer and b) small enough they're accessible by as many people as possible.
The simplest method to implement would be a parimutual betting pool. The Wikipedia article is pretty good. There are reasons why prediction markets use other methods - the expected return on you bet depends on the bets placed *after* your bet - but it should be simple, transparent and straightforward to implement. Note that payouts might result in fractional rings.
To remove the risk of fractional rings entirely you could have some sort of open outcry system. If there is a contract that pays out ten rings when the blue team wins and zero when the red ream wins, then people could meet in the "pit" and call there buy/sell offers: "I am willing to buy 1 blue-team contract for 4 rings", " I am willing to sell 2 blue-team contracts for 10 rings", "2 blue-team contracts for 10 rings? I accept your offer; here are ten rings, thank you for the two contracts", etc. If you've ever played the cardgame Pit, you should get the vibe. This method solves the fractional rings problem but can be very very chaotic.
The constant product automated market maker is actually fairly simple to implement with pen and paper - whatever liquidity the market starts with, that will be the liquidity throughout; you only need one reference table with only two columns - but there will be a huge fractional rings problem.
My recommendation: Avoid constant product market makers. Pick from Parimutuel pools or open outcry or maybe an auction based model.
a collaborative Streetview roadtrip where users online at any time get to vote on which direction to go. You can also vote to honk the horn or to change the radio station.
I've been checking in every now and then. The trip started a few days ago in Boston, and right now it's covered 376 miles (it's much slower than a real road trip). The trip started heading southwest as if it was planning to head to California, but upon reaching Providence it twisted around towards the north and now there seems to be a consensus in favour of heading for Canada. At the time of writing (while everyone in the US is asleep anyway) we are heading up a coastal road in Maine.
What's interesting is the social dynamics of the voting. There's clearly people who want to stick with the plan -- we're heading north, let's just keep doing that as efficiently as possible. But then there's people who want to mess things up, taking every turn they possibly can because heading in a straight line is boring. These two teams fight it out every time an intersection or turnoff comes along. Often the chaos party will win, but then we all discover that the side road is frustrating and collaborate to get back on track, leading to a path which is overall progress marked with occasional fits of wandering. There's a comment section which will sometimes turn into name calling between these two parties as well. Oh, and occasionally there's a fit of interest in honking, which dissipates the moment we've honked. The only thing I haven't seen happen yet is a change in the radio station, which has been set to a jazz station in New Hampshire for a long while now.
Implications for real world politics are unclear but definitely present.
Without looking I can tell you the jazz station is WJSK out of Jackson, and that IRL it doesn't reach more than 15 miles into Maine, let alone the coast, but I'm happy that you are all enjoying it!
Ah well it looks like we've now switched to WBOR 91.1 FM - Bowdoin College Polar Bear Radio which is playing an eclectic mix of... I'm not even sure what but right now it's Talking Heads. It's broadcasting out of Brunswick and I have no idea whether you can hear it down here on the far side of Portland but they're at least making some effort to give us state-appropriate radio at least.
Right now it seems like the host has figured out he's currently internet famous (not _that_ famous, there's only 550 people online right now but that's a lot more than usually listen to Bowdoin College Polar Bear Radio at 10pm on a Monday) and he's having a great time and playing all his favourite songs.
The divide between the "chaos party" and the "keep going party" sounds a lot like the dynamics of the original Twitch Plays Pokemon. The original mode was "anarchy mode" where the game just took the first input it saw in the chat, and then they added a "democracy mode" where people voted for the moves, which made it a lot easier to actually progress in the game, and the game shifted between the modes based on voting.
A lot of the dynamic there was people trying to actually win the game trying to get it into democracy mode (required a supermajority), while the other group tried to keep the game in anarchy (requiring just a majority) and would vote for "start9" (press the start button nine times - did nothing but waste time) while in democracy mode. The latter group was a mix of people who just wanted to make trouble for its own sake, and people who just felt like anarchy was the "truer" way to do TPP.
Maybe a similar dynamic here where some of the people voting for turns are just making chaos for its own sake, but others just generally prefer the original goal and don't want to go to Canada.
What's interesting is that I can't quite decide on a team myself. Obviously if I'm going to be watching this darn thing then I want the car to go somewhere, but if we just head inexorably north by the most efficient possible route then that's no fun at all, the game is only fun if there's a reasonable balance between collaboration and conflict.
In the thirteen hours since my original post the car has gone another 36 miles, so we're doing a brisk walk. Team Progress has managed to get the car onto I-95 heading north and Team Zen Wandering is annoyed at the lack of exits. People in the chat are naming themselves things like "#1 interstate hater" and "#12 interstate lover" and occasionally the haters will stop the car to honk the horn in protest for a few turns before getting bored and deciding that they'll never get off the interstate unless they keep going to the next exit.
Other social dynamics are interesting. The big question of where to go seems to have been resolved, and even the "just dick around" party seems to be less active, the big question is whether to take the interstate or the side roads. Given that you don't actually travel faster on the interstate and it's much less interesting, the back roads seem like the best option, but they provide more opportunities for team chaos to dick around.
I think there was original enthusiasm for heading across the country (which was probably Neal's intention when he started off in Boston) but once people realised how long that would take they chose Canada as a more achievable goal.
How can one find good editions of public-domain books? A popular public-domain book will usually have a lot of different versions available on Amazon, but no way to tell which ones have decent typesetting and printing. I am not looking for expensive collector’s versions, just good-quality trade paperbacks. For example, I recently ordered a copy of The Wind in the Willows that I figured would be nice because it had the first-edition cover design, but it was actually low-resolution facsimile and poor type. Is there a good brand in this market, or good heuristics?
I happen to have a Barnes & Noble Classics sitting on my desk right now. It's pretty nice. I've also been reading some Penguin and Dover books recently. I think the Penguins are probably a bit nicer than the Dovers.
If people had noticed, it seems clear that there wouldn't be a problem, because noticing the phenomenon by definition requires being aware of both people.
If people haven't noticed, there could potentially be a problem in that people might treat one person as if he were the other one, but I don't think that's what you have in mind.
Funny, you prompted me to check and the AI researcher really does come first for me. Stable across different search engines, too. I just took OP at their word that the director comes up first.
Perhaps Daniel read OPs post and decided to have this Problem corrected in the meantime :D
There's great Russian literature but even the authors who were dissenters have been used by Moscow.
“Why Russian literature must lose its innocence. Germany's "Russia-complex" is linked to an uncritical adulation of Russian literature. The colonial mindset expressed in many Russian classics is too often overlooked.”
Imperial Russia was a state in the business of colonising Eurasia. Novorossiya was perhaps the flagship of this enterprise, with Kherson being the first major city founded as part of this colonisation effort. It is no different in kind to saying "The colonial mindset expressed in many American classics is too often overlooked."
American colonialism isn’t on the past either. The U.K. poodles along with the US and is supplying Israel with arms. I’m no fan of the British empire (although I’m a fan of Britain) but I’m going to read British literature from the past and present.
I'm in favor of this, but not because I think any of these lunatics have a point. However, these books and writers really are painfully overrated and often uncritically lauded by what one suspects are pseuds who didn't read any of them. I read quite a few, enough to satisfy myself on the subject, and can honestly say that they were all meritless dogshit. If they could get canceled out of social existence and nobody would ever again have to listen to some imbecile braying on about "Dosto" that would be great.
I love Dostoyevski. I can't deny, however, that I started liking him less and less every time an idiot started going on and about his work.
Dostoyevski attracts the kind of people that the author would despise. Most love to ignore the fact that our russian author was a raging orthodox christian.
Similar to how people love to ignore that Lord of the Rings is fundamentally a christian story. Really goes to show how even the most surface characteristics go completely unnoticed by the pseudo intellectual.
Yeah, being dead white colonialist guys is bad enough, but Russian as well? Used by Moscow? An utter outrage! To the (long overdue) scrapheap of history with the lot!
Quotes like "Kherson is a Russian city" don't justify a genocidal war, which should be obvious to anybody not entirely mindkilled. So what's actually proposed here is suppressing any perspective contrary to what the western "polite society" deems acceptable. If this endeavor seems worthwhile to you, godspeed, I guess.
Could you summarize what's being proposed then? I read the first link and it includes this: "In Western debates, the calls from Ukraine to stop promoting Russian culture as long as this war is ongoing are often framed as a problematic call to ‘cancel’ certain authors who are still held in high regard. But, as Viktoria Amelina wrote before her death in 2023, should we in the West really be debating the supposed ‘canceling’ of dead Russian authors while Ukrainian authors are being killed right now? " which sounds a lot like proposing to chuck it all out.
> In Western debates, the calls from Ukraine to stop promoting Russian culture as long as this war is ongoing are often framed as a problematic call to ‘cancel’ certain authors who are still held in high regard.
> The latest utterly pointless sanction is the Cardiff Philharmonic Orchestra's announcement that it would remove music by Pyotr Ilyich Tchaikovsky, the Russian composer, from its all-Tchaikovsky concert, calling it "inappropriate at this time."
> the objection hardly necessitates scrubbing Tchaikovsky in his entirety solely because of where he was born. That's especially true in light of what was supposed to be the program's main course: his Symphony No. 2, which, in a sort of cosmic irony, is built around…three Ukrainian folk songs.
> For those familiar with Tchaikovsky, that likely won't come as a shock. The composer spent several months a year in Ukraine and had close family ties to the region; his paternal grandfather was born there. "I found the peace of mind here that I had unsuccessfully sought in Moscow and Petersburg," he once wrote
But "Tchaikovsky was explicit that he preferred Ukraine to Russia, it's just that Ukraine was called 'Little Russia' at the time" is probably even more offensive to the movement to censor Russian culture than Russian culture is.
I feel like focusing on the historical question of whether Ukraine was "part of" Russia or not is to give into Putin's framing of the question.
It's utterly irrelevant whether or not Ukraine was part of Russia a hundred or two hundred or five hundred years ago, all that matters is whether the Ukranians want to be part of Russia now.
Similarly it's irrelevant whether Taiwan was traditionally part of China, or Southern Ireland was part of the UK, or the United States part of the British Empire; it just matters what the people who are actually alive today think.
Sure. But given that premise, how do you censor Russian culture without simultaneously censoring all of Ukrainian culture? We know that one is good and the other is bad, but how do we apply that?
There's a big difference between actively suppressing a culture like what Russians are doing in occupied Ukraine, and in ceasing to actively promote a culture, which is what I've largely seen proposed for Russian culture while Russians pursue war.
What’s your feeling on Israeli literature (or perhaps Jewish literature) in the aftermath of Gaza. I suppose German literature is already out of the question. We are deep into the decolonisation of Shakespeare here in Blighty and I suppose that makes sense since though not a colonialist himself he probably brushed shoulders with a few people whose descendants might go on to colonise the falklands or somewhere.
The Spanish and Portuguese are already banned, one expects, and as to why any American literature survives at all is a mystery for the ages. Maybe because most of this cancellation malarkey is driven by the American Academy.
I am less familiar with the details of Israeli literature and culture. That said, to the best of my knowledge, there hasn't been a concerted effort across multiple iterations of an Israeli empire to wipe out someone else's culture.
Yeah, so I discovered that it's my wifi! When I turn off my wifi and use mobile data (different network) it works! I can't access it when connected to Optus (Australia), but can on the Vodafone network. Weird.
I've had random websites block my home IP for unclear reasons (maybe my ISP is giving me IPs that have been used for abusive purposes? I don't know). Try using tor or a VPN if you can, and see if the problem persists.
nxdomain is unlikely to be an IP block by the website, and is more likely to be related to the user's ISP either having a misconfiguration, or not having picked up a recent change, or blocking the motte.
Abstract: This paper explores the phenomenon of musicians being moved to tears during performance, interpreting such moments as portals into emotional, physiological, and spiritual transformation. Drawing on personal narrative, cognitive neuroscience, and Tantric philosophy, it argues that these experiences reflect a confluence of subcortical emotional release and the disciplined control of artistic expression. Within a Tantric framework, they represent karmic purification and chakra activation—moments where the performer becomes a conduit for Shakti, the dynamic pulse of the cosmos. Music, in this view, functions not merely as art or communication, but as sacred ritual capable of dissolving ego, transmuting karma, and restoring the self through sound.
Contents: Contents
Introduction: How Do We Understand Tears in Music? 3
From a Lump in the Throat to Nothing 6
Tears for Johnny 8
Acting in the Inner and Outer Worlds 10
Where Music Breaks the Shell: Karma, Tantra, and the Tears of the Musician 15
Introduction: How Do We Understand Tears in Music?
As I have said in my preface to Beethoven’s Anvil: Music in Mind and Culture (Basic Books 2001), my oldest memory is of a piece music: Burl Ives singing about a fly marrying a bumble bee. I played a record of that song over and over, driving visiting uncle to distraction when I did it at five in the morning. That little ditty is thus my anchor to the world, my omphalos.
I started trumpet lessons when I was 10. I was a reluctant student for two or three years until my music teacher, David Dysert, read me the riot act and got me to take music seriously. Since then music has always been with me, sometimes more so than other, but always. It has given me great joy and pleasure, and it has soothed my wounded heart.
It has also given me experiences that we do not talk about. Perhaps we don’t talk about them because they are difficult to talk about. But then all sensory experience is difficult to talk about. No, I suspect we don’t talk about them because we are apprehensive about where such talk might lead us.
In the next section of this document, “From a Lump in the Throat to Nothing,” I talk about two such experiences. One of them – a lump in the throat – is about being moved nearly to tears while playing my instrument. It has happened to other musicians – I know, because I’ve read stories – but we don’t talk about them. Most of this document is about that kind of experience. The next section, “Tears for Johnny,” is about a performance Bette Midler gave to and for Johnny Carson in his last week of television. Both were moved nearly to tears and I rather suspect that many in the audience that night were weeping as well.
The last two sections are about understanding how such things happen, first from a Western scientific point of view, “Acting in Inner and Outer Worlds,” and then from an Eastern philosophical point of view, “Where Music Breaks the Shell: Karma, Tantra, and the Tears of the Musician.” How do we reconcile those two ways of looking at the world? I don’t know. But I’m working on it.
I’m working on it, in part, because the other experience I offer to you – nothing – shook me to the core. It was a mystical one, one where the world dissolved in a blaze of light and sound. What are we to make of such experiences?
For we must make something of them, must we not? The psychologist William James wrote a famous book about them, The Varieties of Religious Experience, countless philosophers and theologians, both East and West, have pondered them, sometimes to dismiss them (all too common in the West), but often to understand them, to understand what the world must be if such experiences are possible.
It is easy to psychologize such experiences, to explain them as something that the brain does. It’s highly unusual to be sure, but it’s just some kind of brain event that we can’t yet explain. And yet those experiences are so very compelling. How do we honor that? Or even: Should we honor that? Perhaps we explain it as entering higher realms of being. That is, in effect, what Hinduism and Buddhism do. Is there a way to assert that and acknowledge scientific psychology at the same time? How do we treat those experiences as an invitation, an invitation to a larger experience of the world? That’s what I’m interested in.
I don’t pretend to provide an answer in this short document. My aim is simply to put the issue before you. What you make of it, that’s your responsibility.
Ancient Hindus believed that certain types of art jolted you out of your thinking. Like the discombulated patters I recently saw in the very very old Marundeeshwarar temple in Chennai in the Tripurasubdari section.
And they believed devotional music could be a trapdoor to experience oneness with the universe...i.e. a way to experience the dissolving of your own self and experiencing your being part of a larger universal self ("Brahman").
If Moldbug genuinely wants “rule by CEO” then why support Trump, who has been a disaster in every real management role he’s ever held? Good CEOs in the real world are probably too consensus driven and results oriented for Moldbug’s taste.
Most founders are not great at being CEO. It’s a different skill set. Founders take a lot of risk and generally can’t delegate well. CEOs are usually paid to mitigate risk, grow the organization and guarantee stability and predictable growth to investors. It may be we already live in a society where traditional CEO skills are less valuable. Someone like Musk would have been forced out of management a long time ago if this were still the 1990s.
That’s a) not universally true - Microsoft is doing just fine, plenty of examples of companies that have thrived after the founder was kicked out, and obviously some founders, like Jobs, do have good CEO skills as well and b) kind of irrelevant - you may well simply be making the case that traditional CEO skills, which is what I understand Moldbug to be referring to, are not as important as founder skills. In which case Moldbug‘s argument also falls apart. The U.S. cannot be led by a „founder“ unless you really want to tear the whole country down and rebuild. You can argue Mao was more of a „founder“ than a CEO, and that’s why he needed to incite a Cultural Revolution to keep himself relevant but is that what most Americans want?
Strong agree, especially about Musk. In rare instances (like him) being willing to continue doubling down on bigger and newer risks pays off. In most cases it means a company makes a few big mistakes and dies. A competent manager is, on average, better than a maverick. For newer companies a competent manager can't get it off the ground, as they are too risk-averse to take the necessary jumps that make the company take off in the first place.
I think that in most fields this hasn't really changed since the 90s. It's just that tech has been a huge part of the new company growth and several high profile founders have made it big, so it feels like there are new options. Coca-Cola, Exxon, or even IBM aren't hiring a young CEO to "move fast and break things" because there's more to lose than to gain from even above-average performance from a CEO trying that strategy.
Which is why people try to bet against Musk succeeding. Whether he's unusually good or unusually lucky (my guess is strongly both), he's got a really good streak going by any metric.
Agreed. Beats me. I guess Trump *acts* like he's one of those guys even though he's far less competent and maybe Moldbug can't tell the difference. Or maybe Moldbug just takes what he can get. I dunno. I'm not trying to defend either.
I'm not an expert on Trump's business by any means, but I understood he did get a number of buildings constructed and managed around the world. I suspect that's top 1% ability level at least - meaning 99/100 would fail to accomplish that, probably closer to 9,999/10,000 at least.
Of the people who manage multi-million dollar real estate companies, maybe he's a loser and bad at the job. But that's still a pretty exclusive club with a high bar to entry. I'm certain that if I was given control of his company at any stage from 1970 to present, I would have failed to run it. Even if the only thing he did was hire effective people, that's still a pretty significant skill. Maybe a rarer skill than being a good CEO alone.
He had a couple good projects in the eighties but most of his projects since then have been licensing deals. The Trump Tower project was run by a woman named Barbara Res. She worked for him for 18 years and doesn't have a lot of good things to say about him.
Yes, Trump is a genius at self-promotion and brand management. He was a man ahead of his time for most of his life - his skill set is perfectly suited to the social media/influencer 21st century. He was mostly a failure back in the 20th when organizational and management skills were still in vogue. Seems to me „President“ is still a job that requires the latter skills but maybe I‘m old fashioned.
I did not deeply invest time into researching this, but I don't believe the trump organization was actually much involved in the building process for (any?) of the buildings. My brief internet research implies that they were mostly involved in securing financing and name licensing deals.
Also, most of them seem to be permanently closed or no longer be Trump affiliated. Not a great omen.
Wait, really? Can you point to an enterprise he has made significant money on that isn’t primarily about branding and entertainment? His real estate portfolio in the 80s and 90s had several famous bankruptcies, which were his main claim to fame before his reality show, and everything since then that has been profitable has been based on this bending.
He had several businesses, including real-estate businesses, *some* of them filed for bankruptcy. Moreover, I don't understand how leveraging his public persona to make his profitable businesses even more successful is somehow an indictment to his business acumen, rather than the opposite.
It’s not indicting his business skill - it’s just noting that his major successes are all based on converting attention into money, and not based on creating a product or building that does something of its own, separate from the attention of the public.
Hard to be sure because he's taken such care to suppress his tax returns. He inherited a total of around $400 million from his father. Guesstimates of his net worth have bounced around in the low-billions range, though those took at face value stated valuations of real estate holdings which have been exposed in a couple different courtrooms as hilariously fictitious.
Trump was always very aggressive with financial leverage (debt) which is why business bankruptcies have been such a central part of his operations. (Stiffing the lenders as a strategy rather than a last-ditch option.) Some unknown amount of personally-guaranteed debt still exists on paper, and the guy who ghost-wrote "Art of the Deal" has been quoted as predicting that Trump's probate will reveal his actual all-in net worth to be negative.
The new wrinkle which may wash away all that though is the crypto scam -- early indications are that it is proving _extremely_ rewarding for Trump and whichever family members are in with him on the back end of it.
People believe they're buying into a Trump-endorsed investment vehicle, but the reality is more like a donation to his bank account in exchange for near-worthless scrip
> Guesstimates of his net worth have bounced around in the low-billions range, though those took at face value stated valuations of real estate holdings which have been exposed in a couple different courtrooms as hilariously fictitious.
While this is accurate, it should be noted that all other estimates of anyone's net worth have exactly the same problem.
The valuation of publicly traded stock is public, but it's also obviously fictitious. If a major stockholder attempted to cash his stock out, its value would crash, and everyone knows this. Net worth calculations do not even pretend to take it into account. They are not attempts to calculate the value of someone's net worth; they are only attempts to produce a number that is tenuously related. Valuations of other property meet the same standard.
Though not to equal degrees. Depends on how much of a person's net worth is in real estate.
Forbes just a few weeks ago published their latest estimate of Trump's net worth. They arrived at a total of $5.7B minus $600 million to date in legal liabilities. Nearly half of the total is the highly-inflated Truth Social stock. ("From a financial standpoint, Trump’s social-media venture is one of the most absurd businesses in America, generating sales of just $3.6 million in 2024 and recording a net loss of $401 million. Revenues dropped 12 percent in a year, even while Trump dominated most news cycles. Yet Trump-loving traders still bid up shares to head-scratching prices....") If Trump were to start cashing out of that stock it would plummet in value given its underlying business fundamentals, so that's kind of a platonic example of the concept of paper wealth.
Of the remaining $3.1B of Trump net worth, Forbes puts real estate at $2.2B. They don't say where they are sourcing the valuation estimates from, so it's hard to say how much those figures depend on Trump Org's own highly-aspirational numbers.
Your cookie policy banner doesn't seem to work. At least whenever I click any of the options it says "saved" and then refreshes and the banner appears again. There doesn't seem to be any way to get rid of it.
Interestingly, the Washington Post did a piece on Curtis Yarvin last week making a rather different case on his Trump pivot, titled “Curtis Yarvin helped inspire DOGE. Now he scorns it.”
Curious what’s driving the seemingly different points of view here.
Their leader, imprisoned since 1999, had called for the group to disband this February. I assume this is great news, not just for people in Turkey, but for those in Syria and Iraq too (there are Kurdish militias affiliated with the PKK controlling territory in those countries), but I want to know what people from the area think.
These news may also have an effect on the US role in the Middle East, given that they're allied with the YPG (Kurdish militia in Syria), and possibly the Iraqi Kurds.
I think you greatly simplified the situation. PKK is a terrorist group just as any rebel force which has armed disputes with a somewhat hostile occupying power (which Turkey, without any doubts, is). Also, the territories the Kurdish militias control in Iraq and Syria happen to have a Kurdish majority in most cases (while Turkey frequently makes alliances with islamist militias).
I don't expect this to affect Syria very much. YPG has quite different focus and goal than PKK. It's not clear that they can survive if the US should withdraw from that region, because Turkey would probably conquer the YPG territory pretty quickly if no one else intervenes. But that has little to do with their ties to PKK.
Curtis Yarvin. As a long time SSC/ACX reader, I kinda get it when Scott writes on X: "I agree ... there's still a outside chance Trump manages to come through and do great stuff. Certainly I was happy to see him take steps against affirmative action, NEPA, etc. " - But why give the "others" (NYT and armies of trolls) such a fine quote to "proof" to non ACX-readers: "See: S.A. Siskind really is a racist hating nature - we told you!" ... From a more obviously Straussian way-to-look at the whole thing: What is the aim of all that Curtis-critique - the one effect it can have is giving him more readers. (Oops: the purpose is, what it does.) I did too much history to care zilch about Curtis and his fantasy-monarchies - or his opinion about Trump. At SSC Scott did some great writing steelmanning and refuting those wet dreams on monarchists/reactionaries. Chapter closed.
Nothing dishonest about phrasing it: "Certainly I was happy to see him take steps towards color-blind college-admission and YIMBY." - A tweet is not a personal e-mail, but a public statement. Scott is enough of a brand to attract the haters, mind-police - well, the NYT even. And those do have influence over the collective consciousness. - Most people will never be ACX readers. - I will always be, and paid. But I see nothing worthwhile in former or present Curtis to pay attention to.
Makes sense. I am not confident enough of my media savvy to give detailed advice like that to Scott; he seems to know what he's doing, he's got ...how many subscribers now?
He is number 2 in the substack-category "science". Number one is a mainstream team of medics, very liberal with a capital D - one really wants to comment - Ay, there's the rub: to comment, you need to pay-subscribe! Free to read. https://yourlocalepidemiologist.substack.com/p/birth-rates-are-falling-but-solutions Ok, I did it: a shout-out for substack worse than ACX. All are. But much better than Curtis and his ... in German, I'd call it: Dreck.
I actually have a paid account that never comments. I seriously expect the Democrats to get back into power and start throwing people who have said politically incorrect things in jail.
He probably feels like we're way past the age where the NYT and similarly aligned people had any real power over the collective consciousness, so he feels like he can speak freely. I would agree and I think it's a great thing.
Scott, calling yourself a libtard and a coward seems like poor rhetorical innovation. Just giving the other side ammo. “Scott Alexander admits that people who oppose Trump are cowards.”
It's not giving them ammo if they're going to call him that anyways; it's leaning into being Shylock. The self-awareness is refreshing, if he means it.
Signed,
Someone who has admitted to authorial cowardice in other comments elsewhere
it's also puerile and unclear, what does "libtard" even mean, never heard it in a single context where it made a minimum of sense. At the same time the association of not wanting to risk a monarchy and/or dictatorship = being a coward that's a strange association at best
Well, actual monarchies have occasionally been decent governments. (I think the rule is 1 time out of 5. Which is perhaps better than the record for presidents.) The problem is when they aren't. It has certainly been asserted "For forms of governments let fools contest, what ere is best administered is best.", though I thing is rather depends on the goals of the administration.
The thing is, democracies, and form adjacent governments like the US, are historically less reliable at keeping deals than other forms of government. Both internally and externally. They tend to feel that they don't want to be bound by the prior administration, and the populace forgets why the deal was made in the first place, if they ever agreed it should be made.
I think a modification of the old Anglo-Saxon idea is perhaps the best. The new king must be elected from a group of people, but not a small group. Say those who are either descendants, nephews, or nieces of the current king. (Perhaps expand the group if the list of candidates is less than 7. But expand by consanguinity. If there aren't enough cousins, pull in the second-cousins.) This gives a small enough group that they can be raised responsibly, and also have their characteristics known, and a large enough group that SOMEONE in it is likely to be decent. It also gives the current ruler a stake in the succession. As to who should do the electing...that's a tough one. Most people are clearly incompetent at that. I know that given a field of 12 candidates I wouldn't have paid enough attention to them to be a reasonable decision maker. Saying "just let everyone vote" is a recipe for disaster, but every filter I've thought of is subject to corruption. "Let the governors vote on the king", for example, is a guarantee that the current power structure will be supported.
> Well, actual monarchies have occasionally been decent governments. (I think the rule is 1 time out of 5. Which is perhaps better than the record for presidents.) The problem is when they aren't.
I read somewhere that for a while Turkey's succession system was to have a civil war whenever the sultan died, with two obvious consequences:
1. Turkey was ruled by something like eleven highly competent and effective monarchs in a row, noted by the essay as a record that puts every other known state to shame.
2. Turkey spent much of its time enduring brutal civil wars, which wasn't worth it.
It isn't actually true that the problem with monarchy is when the king is bad at governing. That problem is easy to solve. We prefer not to solve it. The point of monarchy is to assign rulership regardless of what people might think of the particular ruler.
1 time out of 5 for modern monarchies seems pretty optimistic and yes as you point out the consequencces when they go badly are particularly severe.
Apart from the fact that those records were written under that king without any real possibility of criticizing the king even if that were true having one example hundreds of years ago is not remotely a proof that a monarch would be even halfway successfull today.
If we look at modern dictators or kings the ones we see are not particularly inspiring. Russia is under a dictator who brought it into the bloodiest war since ww2. Sure it's marshaling it's resources quite well for the war, except the mistake was starting it in the first place.
Since you mentioned Turkey we can look at Erdogan who brought Turkey an inflation of almost 38%. Similar stories of economic missmanagement you will find in Hungary, Belarus, most African military juntas and Venezuela and in all this the population also bears the human cost of censorship and dissent being eliminated.
Even the cases where dictators have been a lot more economically successful such as China (technically the gulf monarchies too, but their economic success is purely based on oil extraction not on wise leadership) the population has to suffer a human cost.
Among the very few cases of illuminated dictatorship i can count Singapore, an essentially tiny country, a bad example to institure a dictator in any populous country.
I mean, they're worse at keeping deals, but are they better at other things? Why would you go for a form of government that hasn't been used in a millennium and existed in a much smaller polity with a lower tech level when some form of parliamentary democracy (unlike the US) seems to work OK in, say, Europe or the non-PRC parts of East Asia?
In Charles Williams' _Descent into Hell_ (1937), a professor creates a succubus. No overt magical methods, just obsession about a young woman who isn't interested in him. He doesn't like that she's not interested. As I recall, once this is clear to him, he avoids her, and invents a false version with her distaste for him edited out.
He spends more and more time with the false Amelia until she can even be seen by someone else. His lack of interest in truth leads to his mind disintegrating.
I was surprised to find in this discussion of Paracelsus, a major Renaissance writer about magic who put much emphasis on the power of imagination (at about 25:00), a description of making a succubus by imagination, and I'm willing to bet that Williams, who had a considerable interest in magic, had picked the idea up from there.
I thought I had just found a really cool reference, but this does rather look like concerns about AI companions.
The rather hectic story of a manuscript copying Archimedes' letter about his "method", a socially unacceptable way of using infinitesimals to calculate areas.
The ancient Greeks didn't like them, the counter-Reformation Church didn't like them. (Let me know if that's true.) Fortunately, Newton didn't have to please the Jesuits. I feel like there's a whole conversation about gatekeeping and Damned Things* in the topic.
The text barely survived. There's one known copy, and it was bleached out for a prayer, but some of it was barely visible in the margins. A scholar copied what he could see-- recognizably lost Archimedes-- but a lot of it wasn't visible, and then the manuscript was lost and getting moldy, what with being hidden from the holocaust.
Fortunately, it was found, and modern scanning was able to recover the text. Watch the video for details of the method and animated diagrams.
The most interesting thing on the abstract level is the possibility that there is revulsion against actually good ideas, and probably not just ideas some current rebels are defending. Is there any way to get the actually good ideas to surface when there are so many bad ideas competing with them?
*Damned Things-- Robert Anton Wilson's term for things people seriously don't want to think about
"The Book of the Damned was the first published nonfiction work by American author Charles Fort (first edition 1919). Concerning various types of anomalous phenomena including UFOs, strange falls of both organic and inorganic materials from the sky, odd weather patterns, the possible existence of creatures generally believed to be mythological, disappearances of people, and many other phenomena, the book is considered to be the first of the specific topic of anomalistics.
The title of the book referred to what he termed the "damned" data – data that had been damned, or excluded, by modern science because of their not conforming to accepted belief. Fort charged that mainstream scientists are conformists who believe in what is accepted and popular, and never really search for truth that may be contrary to what they believe. He also compared the close-mindedness of many scientists to that of religious fundamentalists, implying that the supposed "battle" between science and religion is just a distraction for the fact that, science, in his opinion is in essence simply a de facto religion. This is a theme that Fort developed more in his later works, New Lands and Lo!, particularly."
It’s years and years since I had to study history of mathematics as part of math undergrad, but…
- you have the exhaustion lemma in Euclid’s elements
- techniques used by Archimedes, considered even at the time not to be respectable mathematics
- Newton and Lebnitz get a lot of flak from, e.g. Bishop Berkeley, for not being adequately rigourous.
- Eventually (nineteenth century or so) we hav an account of calculus that looks respectable
Now, when ‘m teaching ungraduate digital signal processing, we do point out that if the student is paying attention they ought to notice that what we’re doing is not mathematically respectable, and if you care you can go read a book on measure theory.
I maybe misremembering my undergrad history l course here, but possibly there is a step circa Newton and Leibnitz, where people know Archimedes had a method but don’t know what it was, try to rediscover it, and end up inventig something different.
Would appear analogous to the bit where the US circulated a fake story that lasers were used in isotype refinement for nuclear weapons, the soviets believed this nonsense story, and actually got it to work.
You say Archimedes' method was socially unacceptable - I don't think that's quite accurate. The mathematicians of his time had a geometric & Platonic view of mathematics. There wasn't a proper understanding of functions, or of limits in mathematics. The 'paradox' that an arrow could never strike a rabbit because in the time the arrow moves 4 feet the rabbit moves 1, and so they never meet was regarded seriously.
Aristotle's fairly sophisticated methods for working out areas (essentially by integration) didn't have the appropriate foundations, and mathematicians of his time thus couldn't either understand these ideas or work with them.
Newton & Leibnitz developed better descriptions of calculus (though Newton, as I understand it, spent several years developing 'geometric' justifications for the rules of calculus, which were probably not at all related to how he discovered the rules originally). These took mathematicians some substantial time to appreciate, and in fact analysis was placed on proper foundations only in the second half of the nineteenth century.
It's worth pointing out that the Church of England was hostile to Newton, for mathematical reasons as well as non-mathematical (his views were likely heretical). Traditionalists everywhere dislike new ideas, I guess.
Okay, interesting link but I had to stop because I was rolling my eyes so hard when we got to THE JESUIT WHO ROSE TO POWER and TOLD THE CHURCH IT COULD USE MATHEMATICS TO ACHIEVE ITS GOALS.
Said goals being of course "to re-establish its social, cultural and political dominance". Because the one thing every peasant in the middle of Europe was fighting about was which particular mathematical theory one supported, naturally.
If Clavius - who has a crater on the Moon named after him - 'rose to power' via mathematics, it was on the grounds of calendar reform. I guess the Protestants were right and the Gregorian Calendar should be junked because it's a sinister Papist plot to - what was it again? oh yeah - "re-establish social, cultural and political dominance".
Though I have to admit, the idea of a Renaissance figure pondering "how do I achieve power, status, wealth and success? I know, I'll become a mathematician!" does tickle my fancy.
I'd have to read this guy's book to figure out if what he is describing is so, and how the theory of infinitesimals was received outside of Italy in Protestant nations (was there any opposition from Protestant mathematicians? if so, that puts a hole in his proposition that it was purely Catholic power-grabbing) and to be frank, nothing will make me read a book about mathematics, so some braver soul must do this for me:
Infinitesimal: How a Dangerous Mathematical Theory Shaped the Modern World by Amir Alexander
So far, it's sounding like good old Whig history and the pernicious Papists, but since I'm only going by a video produced by some whippersnapper, that may be unfair.
Opposition to infinitesimals wasn't just a Catholic thing. From the book:
"Yet, useful as it was, and successful as it was, the concept of the infinitely small was challenged at every turn. The Jesuits opposed it; Hobbes and his admirers opposed it; Anglican churchmen opposed it, as did many others."
"The struggle over the infinitely small in the early modern world took different forms in different places, but nowhere was it waged with more determination, or with higher stakes, than in the two poles of Western Europe: Italy in the south and England in the north."
I think what Andrews is trying to do (and this is only an impression from the reviews) is set up the sexy opposition model of "side A versus side B, A wins because goodness, goodness = our modern stuff" and thus wants to have Italy - central! Catholic! - versus Britain - isolated! Protestant! as the boxers in their relevant corners.
But that leaves out Leibniz, for instance, who was working away in Germany. Was that not also an isolate Protestant periphery? But that doesn't suit the narrative as well, though Hobbes versus Wallis is not the first thing that comes to mind if I'm thinking of Big Maths Fights from the past.
I think the guy was trying to write the mathematical version of The Swerve:
Big Concept books that pitched Science versus Religion in a sweeping historical context where the Classical past by way of the Renaissance came to save us all from the Dark Ages were very big in the mid teens of this century; Catherine Nixey and "The Darkening Age" was another such:
The Catholic Church has an interest in philosophy and theology, and I can imagine it wanting to spend a small fraction of its resources on what it thought was pinning down the corners, especially if one Jesuit thought it was important and no one else had a strong opinion.
The Church has been both fairly maligned and unfairly maligned, so what's needed is actual history.
I nearly made a fool of myself. I've only ever learned calculus through limits, so I just assumed Newton and Liebnitz used limits. Apparently that was a later formulation.
Newton and Leibniz used limits - they just thought of limits as defined by infinitesimals (inputs infinitely close to x have outputs infinitely close to x’s output) rather than by triply nested quantifiers (for every x and every epsilon there exists a delta such that inputs within delta of x have outputs within epsilon of x’s output).
There are roughly two poles with regard to mathematical rigor. One camp I might call pure math, and it generally cares very much about solid foundations. (Not always enough, mind you, naive set theory was a thing, after all, and Goedel has put a hard limit on how good your foundations can be.)
The other pole I might roughly call "physicists". There, anything goes, if you can write it on a blackboard and don't get struck by lightening, that means you are allowed to do it. Using mechanical (a la Archimedes) or numerical methods is fine. Playing fast and loose with infinitesimals is fine, e.g. cancelling dx in (df/dx)*dx. If a student comes up with a counterexample where these tricks don't work, you just call it a pathological example and move on, irrespective of the fact that almost all functions might qualify as pathological. What do you mean, a plane wave in an unbounded space can not be normalized? In practice, you can take the integral of a lot of sets and never once encounter a Banach Tarski decomposition.
The difference between these poles is exemplified by Dirac's delta "function". It can be defined by a physicist in ten minutes. However, if you want a sound definition which tells you what you can actually do with it, you require distribution theory, which takes a semester or so.
Both approaches have their uses, however it is important to be aware if you are cutting corners or not.
Also, early mathematicians were a bit of a cult, see for example WP on irrational numbers:
> Hippasus, however, was not lauded for his efforts: according to one legend, he made his discovery while out at sea, and was subsequently thrown overboard by his fellow Pythagoreans 'for having produced an element in the universe which denied the... doctrine that all phenomena in the universe can be reduced to whole numbers and their ratios.'[7] Another legend states that Hippasus was merely exiled for this revelation. Whatever the consequence to Hippasus himself, his discovery posed a very serious problem to Pythagorean mathematics, since it shattered the assumption that numbers and geometry were inseparable; a foundation of their theory.
> [Dirac's delta function] can be defined by a physicist in ten minutes.
Only if you want the formality of setting it up as a limit of Gaussians. If you'll accept the piecewise function (δ(0) = +∞ and δ(x) = 0 if x ≠ 0), the integral, and a plot thrown for good measure, it shouldn't take more than one.
Differentiating this function is when you might need to watch out for lightning.
Personally, I would take the ten minutes to go through the limit of gaussians. It will still leave students uneasy (can you treat something which implies "when evaluating this, take the limit for epsilon-->0 over the expression it is in" as a function in good conscience?), as it should, but at least they will kind of know where you are coming from.
The piece-wise "definition" is a function which is zero outside of a null set (which happens to be {0}), so its Lebesgue integral should be zero. And even if you do not have students who have heard of measure theory, someone will notice that δ(x) = 42 δ(x) in the piece-wise definition (unless you want to argue that 42 ∞ !=∞, which is definitely not a hill I would want to die on).
So the "and the integral shall be 1" is very much part of the definition.
Now I wonder if it would be worthwhile to use infinitesimals (1/(2eps) for |x|<eps, 0 elsewhere -- or gaussian if you like) for a more robust (but likely still unsound) definition. Not that infinitesimals are typically rigorously defined in lectures where delta will pop up.
> Now I wonder if it would be worthwhile to use infinitesimals (1/(2eps) for |x|<eps, 0 elsewhere -- or gaussian if you like) for a more robust (but likely still unsound) definition.
Why would that be "likely still unsound"? Any proof involving limits can be translated _mechanically_ into one involving infinitesimals and vice versa.
From a perspective based on infinitesimals, the delta function is nonzero over an infinitesimal halo around the real value 0, and its value over that region is one of the infinitely large values such that the real part of the area under it is 1. This is related to the fact that the limit of 0 (the real part of the width of the halo around 0) times infinity (the value of the function) is indeterminate. It is possible to choose the two values, the one with 0 real part and the one with infinite magnitude, so that their product is 1.
My mistake. I was under the impression that the theory of distributions was younger than hyperreal numbers / non-standard analysis (which offer a rigorous definition of infinitesimals), and hence concluded that if a rigorous definition of the Dirac function was trivial using infinitesimals, Schwartz would have used that instead of inventing distributions.
However, the actual timeline is:
1930: Dirac introduces delta (with some prior art)
1951: Schwartz published his book on distributions
1966: Robinson published Nonstandard Analysis
From my PoV, one of the following three must be the case:
(1) Distribution theory is contained in NSA (and Schwartz was just too early to use them)
(2) A rigorous definition of Dirac delta is trivial with NSA, but other results from distribution theory elude NSA (and Schwartz was interested in these results, and not providing a rigorous definition of physicist's syntax)
(3) NSA can not trivially provide a sound definition of delta (e.g. feeding infinitesimals to the delta-function might be problematic)
Based on what I know about these topics (which is sadly not very much), I will happily plead the fifth on which one is the case. If I had to wager a guess, I would pick (3), but don't quote me on it.
Is the Hippasus story true? In a lovely example of serendipity, I never heard of it or him till this comment, and when I went looking for something about Clavius and infinitesimals, there he pops up again:
"When there are good true stories to tell, everyone suffers when we tell the false ones. This is especially important when the stories directly impact the matters students are studying. If one tells the story of a man named Hippasus being tossed from a ship with regard to incommensurables, better to talk about what it says about worldviews that such a story seemed credible when first written down centuries later."
Anyway, the paper about infinitesimals and why they were so controversial was more educational for me. First off, unlike the framing in the video, it wasn't a tidy "religious bigots on this side, Enlightenment scientists on that side". Cavalieri, Galileo's pupil and defender of infinitesimals, was a member of a religious order himself. And the ban on teaching infinitesimals only held in *Jesuit* schools, since Clavius was creating the mathematics curriculum for these schools. If you weren't attending a Jesuit institution, you could (or could not, depending) learn about these.
"A typical analogy, most often associated with the Italian Jesuat brother [As always has to be explained regarding Cavalieri, he was a member of the Jesuat/Gesuat, not Jesuit, order. It flourished from 1361 until suppressed by papal decree in 1668] and mathematician Bonaventura Cavalieri, is of infinitely thin pages of a book.
1. The pages could be so thin that, no matter how (finitely) many you stacked, all of them together would not be as thick as any actual book; you would need infinitely many. Such pages are infinitesimal.
2. The pages could be so thin that, no matter how you tried, you could not slice them any thinner; they are indivisible. They are parts of the book, but if infinite in number, whether they comprise the whole book was open to question."
So, as ever, the story is slightly more complex than the easily-digestible Youtube video version with (bad old) Church on this side, (good new) mathematicians on the other side.
"Now, there is plenty of truth in this story. First, there is no doubt as to the remarkable success of the Jesuit order in (among other things) establishing rigorous, desirable, ‘safe’ schools for the minor nobility and nascent bourgeoisie – all in the name of the Catholic/Counter-Reformation. Their most prominent mathematician, Christopher Clavius, was a solid proponent of Euclidean geometry in these schools; under his aegis the church (and hence much of Europe) achieved the long-sought goal of calendrical
reform in 1582, where October 15th followed October 4th to make up for the solar year not being evenly divisible by standard days.
...But the overall story, alluring as it may be in our society of absolute freedom and intellectual inquiry, and as exciting as it might be to think of calculus (!) as having political importance so early on, is selling a bill of goods. The suppression of
indivisibles was surely part of a long fight over new ideas and to what extent the Church could direct or restrict many activities, but not the only one, and much of the hyperbolic verbiage used was par for the course at that time.
Despite Alexander’s often deft handling of the distance between (post) modern secular readers and the worldview landscape fifty years on either side of Galileo, much of the political discussion is speculation, or simply wrong. As just one example, while in such a political age it was possible that the Jesuits got the Pope to suppress the Jesuat order, and (barely) conceivable they would have been motivated because some of the most prominent advocates of indivisibles (notably Cavalieri) were from it, among his copious endnotes there is not one to be found about this topic.
We should teach about incommensurables, and that the Pythagoreans may have had strong feelings about this; but we should also make it clear that the death of Hippasus is probably just a story. Likewise, the example of the triangles is a good warning against setting up integrals without caution, and the very real fights waged over infinitesimals/indivisibles for some of these reasons (Galileo was an early advocate)
is a wonderful topic in a Calculus II course. But let’s not suggest it is really about preparing the West for a modern secularist worldview (as even his subtitle How a Dangerous Mathematical Theory Shaped the Modern World implies) that Alexander acknowledges no one in question was actually fighting for."
As for the second half of the book and the Hobbes-Wallis hairpulling:
"Now, Hobbes reasoned if he could achieve the long-desired (Euclidean) construction of a square with the same area as a circle using his techniques, then surely his entire philosophy would be accepted – including otherwise-distasteful-to-all outcomes such as the Leviathan totalitarian-yet-not-monarchist regime. (Here is where Alexander draws a direct comparison to the top-down regime promoted by the Jesuits, adherents of papal authority as their raison d’etre.)
...For the man primarily responsible for Hobbes’ downfall was another Oxford don, erstwhile Presbyterian-party preacher and Parliamentarian-party cryptoanalyst John Wallis.
Through all the regime change of the English Civil War era, Wallis retained favor by adroit maneuvering; unlike many similar men, he seems to have been addicted to producing (and publishing) correspondence aimed at defeating anyone he disagreed with about anything. Since Hobbes was just as stubborn, and since Wallis disliked Hobbes’ theology, views on the university, and (to him) inadequate mathematics,
their dispute lasted through over twenty years’ worth of letters, publication, and ‘transactions’. They argued primarily over mathematics, but interspersed accusations of plagiarism, bad Latin, and proper authority of a minister of the Gospel.
...Still, this was not somehow an English version of the Italian controversy. In fact, Wallis was even more cavalier about his use of infinitesimals than Hobbes was about his indivisibles (see the next section for an example). In Alexander’s telling, Wallis, not Hobbes, plays the role of Cavalieri’s school, with repeated direct references to the
Baconian ideals of experimental induction and free inquiry espoused by the Royal Society of London. All Wallis wanted, in the best experimentalist tradition, were “theorems that were sufficiently ‘true’ for the business at hand.”
Hence the Society’s (with Wallis) repudiation of Hobbes (and so his philosophy) was an example of the key to English (and, by extension, later American?) pre-eminence in science being tolerance of differing opinion, where “a land of many voices. . . discover[ed] its path to wealth and power”. This seems to be an awful lot to lay at the feet of the infinitesimals, or even the Royal Society, alone.
This is especially so since the technicalities of these questions looked so different in the dawning eighteenth century, where people abandoned Cavalieri or Wallis as dead ends to grapple with utilizing Leibniz’ tools (and Newton’s in England), with different controversies. Similarly, it is very hard to imagine the political development of Italy or England going very differently with or without infinitesimals. Even if the Royal Society’s openness to work with the sort of deficiencies Wallis’ had might perhaps be a token of a more general openness in English society to once-heretical ideas, it is not a main reason we should care about this dispute."
There is no reason to believe that it is. Note that "Pythagoras" is known to history only as the name behind a religious cult, making him parallel to Jesus, Mohammed, etc.
It is true that infinitesimal methods were disliked by Archimedes, probably for reasons of perceived rigor. He used a "method of exhaustion" to derive his results, but once he had a result, he would prove it by other means that were more acceptable to him.
So I recently found out how extremely difficult it is to get reasonably priced ADHD meds in the US and then I got linked to this pretty moving video: https://www.youtube.com/watch?v=eRPeU1DYOWA
Now im not an American, this could all just be entirely wrong for all I know but I remember hearing American private pharmacies were all on the the verge of bankruptcy a year ago - is this their response ? Is there something else going on with the actual manufacturers ?
This definitely feels like the sort of thing our resident Adderall Merchant and True Caliph should have done a deep dive on.
A useful workaround I know of for people having trouble getting the extended release version: I've known people who just get the regular version and take a fraction of a pill every couple hours. It's not much trouble, it's cheaper, and I wouldn't be surprised to learn that it actually is a better way to keep a steady lowish level of the stuff in your system. Whatever system an extended release version of a drug uses to keep up a slow release inside the person's body can't possibly work equally well for everyone.
Now I am wondering if someone could build empty diffusion-based extended release capsules for DIY pharmaceutical upgrading.
I don't think that is easy, though. The release rate depends on the solubility of the active ingredient and how well it diffuses through the membrane. And some drugs might rely on stomach acid, which will stop affecting the drug once the XR capsule moves to the intestines. (Also, one would require a two part capsule which seals on assembly, and any commercial manufacturer would likely get sued by the pharma industry.)
Re the contest: if you start organising things only tomorrow, does that mean that changes to the docs after submission but before today are automatically taken into account?
“ But you're the one who likes talking about curate's eggs”
That’s the egg that has good and bad in it which is I suppose how Moldbug uses it. I used to listen to some creaky intellectual Sunday morning review on BBC 4 where every movie or book considered as mediocre was compared to the curates egg.
Well they were wrong! The curates egg is a totally bad egg.
A lot of people think there was no humour in the 19C and when you realise that the “curates egg” has survived for more than a century (however misremembered) as their best joke, then we realise that’s probably true.
The joke is in comic book form, if I recall, the curate is visiting the bishop who has offered him an egg. The bishop asks if the egg is good, the curate replies it’s good in parts.
That means the poor curate is eating an egg that smells of sulphur and he has to eat around the bad parts, which would be greenish tinted if I recall. ( we don’t really have bad eggs any more as the hens are better educated).
Anyway that’s not a good egg, or even a mediocre eggs, it’s a rotten egg. That’s the joke!
You now have more reason to dislike Moldbug and fusty BBC 4 review shows, if you haven’t already. You’re welcome.
The fecal transplant was pioneered by mixing material into a chocolate milkshake; I tend to think that everyone involved would take issue with Moldbug's characterization.
Part of the joke is that the curate (pretty much the most humble clerical rank in the Anglican church) is having breakfast with a bishop, and when the latter asks him how his egg is his ridiculous reply that it is good in places is a desperate attempt not to offend his exalted host.
The modern usage is, in my experience, to say with wry irony that something has good and bad parts (while implying that the bad parts entirely ruin the good part).
That's the same as the non-modern usage.
I'm prepared to believe that someone out there is misusing the phrase, I just haven't seen any examples provided.
Hmm. My experience is that the modern usage is always “good in parts” unironically, and never “rotten entirely” which ignores the original context. Otherwise why not say bad.
Anyway language usage moves on as chatGPT reminded me when I complained about this.
Edit:
It’s possible that i was misreading these BBC 4 types and they were totally condemning whatever they were reviewing as rotten while pretending to find some merit in it.
Apparently, I haven’t much noticed this phrase being used. But this modern usage is illogical and also pretty pointless (what’s the purpose of making a reference to something like this if one just wants to unironically say something is partly bad and partly good?).
Has anyone personally noticed end-user effects from the tariffs yet? I was hearing that by early to mid May the economy would in massive turmoil, and I've heard stories about ports being empty, but so far no complaints about products being unavailable or overpriced. Should I be surprised?
There are definitely business impacts. A bunch of small board game companies have wound up operations in various ways - apparently it’s uneconomic to print custom cards for games or make plastic meeples in the US.
My own employer buys components from a Chinese manufacturer that is basically the only possible supplier. We have stocks, but eventually it will be a problem. Also half our bespoke manufacturers capacity is in Mexico. So our margins will be impacted. But we can still make stuff.
Stocks of the board games that have gone out of print are starting to run out. The impact of my own employers cost issues won’t be apparent to the public for a year or so. Supply chains are long and very complex
If I try to buy a suit from Spier and Mackay (Canadian store), a $270.30 tariff charge is added to the order when choosing the US as the shipping destination. I wouldn't call that massive turmoil, of course.
Most things (that I've been buying) that were initially getting the China tariff markups were simply rerouted and taking an additional week or 2 to deliver.
Hannafords has avocados at $1.50 now. It's possible I'm misremembering, but I think the price has been from $1 to $1.19 for as long as I've gone to the store - I don't think I've ever seen it that high.
I'm in some SF groupchats, and the people doing electronic engineering or computer hardware stuff have been fairly regularly posting about the overnight shortage of parts, some parts skyrocketing, orders being canceled by suppliers or changed in a 'pray I do not alter the deal further' way, for both the hobbyists and the hard-tech startups. People are definitely postponing projects or not starting them at all, in the usual Unseen way. (Just another example of how life is much easier in the software world than hardware world - you don't have to worry about your next `git pull` from Github suddenly costing you $10k extra because Trump 'truthed' something last night.)
And now the God King of Skulls has folded and slunked away into his corner with the tail between his legs as is his custom when met with a push back. Tariff pause with China for another 90 days! A "reset" (how on Earth did he use Obama's term for it?!) with China!
My strategy has been to buy whenever he announces something outrageous, under the assumption that he will walk it back to some degree and stocks will rebound
I closed my eyes and increased my stock allocations a bit, knowing that when I was terrified in the past it turned out to be a good time to invest. I wish I did more.
My Italian Wine Club offered to release me from my club obligations because they can't guarantee that I won't be hit with tariffs when my shipments arrive in the US. But they admit they don't know how they'll be enforced at our border. I decided to forego my club shipment for the immediate future because I couldn't find any clear guidance from the CBP on the current tariffs on alcohol from the EU, if any, and how they'd be collected. 10% I think, but those may only be for major importers, and not people who have alcohol shipped for personal use.
The job market in my field finally picked up, due as far as I can tell to foreign product equivalents being a bit more uncertain long-term, and I was finally able to get a good position. Can't say I've noticed anything on the consumer end.
The board game space is getting hit hard. I know of at least three companies that have already been shuttered or significantly reduced, and a few others on the larger side are dramatically changing the way they do things to not have non-American customers bear the costs of tariffs while charging Americans a surcharge.
I've heard (unofficially) on social media and message groups and such that many other smaller companies are going to just try to sell current stock and hope the situation changes. I'd be shocked if anyone's actually ordering new Chinese game manufacturing right now in the hobbyist space, and I'd guess 95%+ of hobbyist games are manufactured in China.
I'm not aware of what's happening with the actually big companies (Hasbro, Mattel, etc).
I suspect the story is the same for other industries that rely on heavily tariffed sources: they're big enough to start doing something of a pivot or they're selling out current stock and crossing their fingers. Or they were able to lobby for an exception, I suppose.
Another thing to consider is that while the de minimus exception has been removed (last I checked), I don't think anyone has any clue how it's supposed to be enforced, so many places might be getting away with direct to consumer shipments at the moment.
Yes, GMT games sent out a moving message to subscribers about how their operations are going through major changes and risk company finances going deep, deep red.
Obviously, anyone working in finance has had a hard month. Thus my spouse is more stressed.
Our resident substation engineer doesn't know yet how to get new transformers of the right size within budget. But we'll see...
Yep, I've got a kickstarter board game that's on indefinite hold from shipping to the US. Rest of the Anglosphere should have gotten theirs a week or two ago, I think?
More broadly, we're likely in a recession right now so that's pretty bad. I believe I also read a few weeks ago that there are indications that people are burning through their savings, the effects of which we'll see play out in the coming months.
More narrowly, I remembered that a few weeks ago we bought a trampoline. A couple of days after we purchased it the item nearly doubled in price.
The lab where I work is already seeing "tariff" line items on orders. I work with electronics (repairing old equipment and designing/building new tools) and I've seen prices go up. Mouser Electronics is one of the main distributors and they've got notices on their website to the effect that they're trying to minimize the impacts of tariffs on prices.
At least one board game company has closed. And another company for a game I preordered had to 1) raise prices and 2) collect payment in advance (which they normally don’t do)
They're not all made in China. In fact, this is the first I've heard of that level of manufacturing in China - based on the Kickstarters I've watched and occasionally funded, a lot of the printing happens in the Netherlands or Belgium.
I'm now curious what the relative market shares are, broken down by board game material type (card printing, box making, misc. cardboard art and punching, plastic molding, woodcutting, metal casting), but as beleester said, it's not exactly a universal core industry.
To be fair, custom-printed bits of cardboard and plastic are not exactly what you call "core manufacturing capabilities."
Like, if you say "America needs a robust auto industry," you'll get applause. If you say "America needs a robust capability to cut out giant sheets of printed cardboard," not so much.
What is striking (and to an extent, frightening) me is how much we've outsourced to China about things we don't even think about or consider.
"Cutting out sheets of cardboard" is something you'd imagine a medium-sized factory in the Midwest could trundle along doing, neither making "line go up" profits nor in danger of closing down in the morning, but no. Everything comes from someplace else. Just exactly how vulnerable are any one of our domestic markets if something like Covid comes along again, not to mention tariffs/trade wars/hot wars?
As someone who was looking at the car market specifically during this:
25% Canada/Mexico tariffs went into effect March 4th, then March 5th there was a one month exemption for automakers, then March 6th further delays on general Canada/Mexico tariffs.
April 3rd the 25% auto tariffs actually started, and then April 5th the overall import 10% tariff started.
April 29th there is a change to give 3.75% of MSRP of an auto as a tariff rebate.
That's all to say: supply chains are long, we're not yet at the predicted price increase date of June, and the only increase I've seen confirmed so far is Ford in July. Whether others follow or not will depend on if tariffs stay in place or get more exemptions/changes. TBD, stay tuned.
For the Germans here or anyone interested in salaries in Berlin, I recently published the Salary Trends report 2025, a comprehensive overview of tech salaries in town:
I don't have the time to read it right now, just skimmed the intro. The two things that stood out to me is the that it's around 1.8k respondents and the gender pay cap numbers. At 20%+ and 15% they're a lot higher than the German average at around 16% and 6%[1]
I can think of a couple of explanations, but I worry it's selection bias in who answers your survey. If so then the data might just not be representative
Could also have different reasons and it's still interesting and I'll read it. But it'll probably not change my mind much, whatever the results, due to uncertainty about the issues above
Edit: Now that I've had the time to read a bit further into the article I see you've addressed those points in the report :)
Edit 2: I'm actually really impressed by how clear the report is at every level on what's measured exactly, what the sample sizes are for every group and so on. Great read :)
I mean, sure, but is it representative of the Berlin tech sector? That is, should we expect the gender pay gap to be much wider in Berlin tech than in society at large. Or should we expect that OP's getting a skewed sample from their respondents.
Thanks for your comment. As Peter mentioned, yes, def a selection bias. Plus, it's anonymous on top. I cannot be sure it's representative, but I did a bit of checking on particular job categories, and the gap is narrower in some cases.
The data def isn't perfect, but it's still better than Glassdoor or Kununu.
This leaves me a bit confused about how to update. In some sense he's selling out, but he's selling out from a position that he never should have had. So it's a bit tricky to tell how much it reflects a sensible retreat from untenable views, versus him becoming less principled.
One question that would help me thinking about this: lately he has been using FDR, Lincoln, and Washington as examples of the type of "CEO" that Trump should be. But these are non-central examples of dictators. Should I interpret this as him backing away from his position of supporting dictatorship?
"Yarvin's main intellectual contribution was diagnosing the problem of the cathedral"
You... seriously write this? As a serious thing?
These "diagnoses" are worthless, they are not even dime a dozen, every ranting boomer on facebook has "diagnosed" the cathedral/whatever with a great deal of detail and memes, who cares.
I disagree with this. Of all the references Scott and others have made to Moldbug over the years, the Cathedral is the only concept that I found insightful. Everything else just seemed like park bench ramblings.
Like, literally, the only time I can imagine ever referencing him is to reference that concept. He's far from the first to identify a nebulous set of elite institutions that shape public orthodoxy and constantly reinforce each other to make contrary ideas almost unthinkable, but he gave it a catchy name and a somewhat-clear description (I think; I'm going by quotes because I can't stand tryingv to read a single sentence of his).
I find the general prejudice that you're not allowed to say a person had one good idea and everything else was crap extremely annoying.
eric's metaphor emphasizes the distinction between centralization (cathedral) vs decentralization (bazaar).
moldbug's metaphor emphasizes the role of the Priest Caste (Mass Media and Academia) in shaping public opinion. In this framing, checks & balances hardly matter because The 4th Estate is the defacto head of state.
I agree that we should allow that a person can have a good idea in a sea of crappy ones. But is the whole Cathedral thing really that insightful? Or, quoting beowulf888,
"Yarvin Cathedral premises were always based on his mistaken (bordering on delusional) belief that elite institutions, such as universities and the mainstream press, were *colluding* to pull the strings of public opinion. This has been a mainstay of rightwing nuttery long before Yarvin came along."
Meditations on Moloch was insightful. "Shadowy Elites conspire against you and me, dear reader" is... like I said, every Boomer on Facebook has one of those.
For the record, I too, believe that Moldbug's main contribution was identifying The Cathedral.
> "Yarvin Cathedral premises were always based on his mistaken (bordering on delusional) belief that elite institutions, such as universities and the mainstream press, were *colluding* to pull the strings of public opinion. .
Beowulf is mistaken. I'm pretty sure there's a UR post which unambiguously says that The Cathedral is an emergent phenomenon, rather than an explicit cabal. Though I can't remember the exact terms, which makes it hard to search for, but I'll search upon request.
> "Shadowy Elites conspire against you and me, dear reader" is... like I said, every Boomer on Facebook has one of those.
Imma describe to you my own evolution in understanding Moldbug.
1. Chomksy's Manufactured Consent is absorbed by osmosis. I.e. the media lies, vietnam, watergate, yadda yadda. Like you said, boomer stuff. zzz.
2. I read moldbug's UR a bit. I don't really grok what he's rambling about.
3. Scott reviews Albion's Seed. "U.S. Politics is descended from Protestant religious fundies." Huh. Interesting.
4. The Pull Request [0] (not any specific article): "The Printing Press caused not just Protestantism, but also Nationalism. Nationalism sired Democracy, Communism, and Fascism (in that order)."
5. eureka! moldbug read a bunch of old, musty books and decided that Nationalism (and its progeny) is mind-killing, people-killing, toxoplasmic brainrot. Also, Democracy doesn't answer to *We The People*, it answers to Walter Cronkite and Harvard. In this sense, acknowledgement of The Cathedral goes back even further than the boomers. It goes back at least as far as Thomas Carlyle [1]:
> "Burke said there were Three Estates in Parliament; but, in the Reporters' Gallery yonder, there sat a Fourth Estate more important far than they all."
Understanding the nature of The Cathedral comes packaged with other implications. Such as "oh, Wokism comes from Harvard". And "oh, Protestants are oughtistic [2]". And "actually, violent revolutions are kinda degenerate". And "wait, America is actually 4 nations (at least)". And "actually, voting feeds the egregore". And "maybe literacy was a mistake". And "maybe HBD has a point". And "American Civic Religion isn't really a metaphor". And "Democracy has a lot in common with Fascism/Communism". And "wait, maybe governance by kafka-esque bureaucracy is overrated".
Boomers are still stuck on the Chomsky stage. Like, they'll share memes about that one Simpsons episode. Or maybe ask questions about the Rothschilds and the Moon Landing, if they're feeling extra spicy. But generally, they don't look at the history of the Modern Age and conclude "politics is crack, and we're all microdosing addicts".
P.S. Although gunpowder was probably important to the birth of Nationalism, too. As Carlyle once said: "the real use of gunpowder is to make all men tall." I.e. it's the great military equalizer, since it's cheap & easy compared to knighthood. And since the distribution of political power tends to mirror the distribution of military power... therefore universal suffrage.
But still... It's almost trivial? For example, that Democracy answers to Walter Cronkite. It's the next question, where does Walter get his power that the Democracy feels the pressure to answer to him, that gets to some interesting observations., that is, whoever speaks to masses in the language they understand has significant power. We can really dig here, but I think my point stands - even if Yarvin is an interesting thing to study to understand "Alt-Right" or some such, there's 0 reason to engage with him and to promote him (yes, engaging is promoting, especially when you're Scott Alexander).
> It's the next question, where does Walter get his power that the Democracy feels the pressure to answer to him, that gets to some interesting observations., that is, whoever speaks to masses in the language they understand has significant power.
And "The Cathedral" gestures at a fuller answer. Again, it's much more than just "the media is dishonest". It's a window into a heterodox reading of modern history and modern political institutions. Let's dig into some examples that are harder to dismiss.
Remember our debate over monetary/fiscal policy? You concluded with something like "fine, debt is kinda like drugs. But drugs can be very useful if used responsibly." LOL. Please read this post by Dmitry [0] and tell me with a straight face that the U.S. national debt is going to be managed "responsibly". I'm 99% certain that Dmitry was influenced by a UR post that goes something like "hmmm... is it weird that there's no Worker's Cooperatives in the Fortune 500? Maybe because Democratic Agonism inflicts agony on the bottom line? What could this mean for the USG?" However ethically or morally upstanding you think our politicians/bean-counters are, there's a *structural incentive* for the debt to keep increasing. Because the leviathan is myopic. This is one of the reasons why nrx types say things like "cthlulu always swims left" and "conservatives are just progressives going the speed limit" and "conservatives love being beautiful losers".
Here's another example. Recently, you made a negative comment about trump's tariffs and also referenced some sort of obscure aluminum alloy. (I think it was you, correct me if I'm wrong.) I don't agree with the spirit of the alloy clause. If the Cathedral is obvious to you, surely you can articulate why your aluminum alloy clause tilts me to the moon.
Hey, remember Scott's post about various theories on the origin of woke? If The Cathedral is so obvious, then why did such a bright, erudite, elite human-asset like Hanania point to Civil Rights Law instead of our priest caste (viz. the Boston Brahmins)?
I don't think any of this is obvious. And it fills an *indispensable* niche in my epistemology, because it's foundational for understanding a wide array of tangential interests. So I'm gonna keep promoting moldbug, thank you very much. Just as everyone continues promoting Isaac Newton, despite believing in alchemy and secret bible codes.
> Beowulf is mistaken. I'm pretty sure there's a UR post which unambiguously says that The Cathedral is an emergent phenomenon
You're probably right. I can't dig too deeply into his twaddle without shouting at my laptop screen. But I'd agree with you that...
> Boomers are still stuck on the Chomsky stage. Like, they'll share memes about that one Simpsons episode. Or maybe ask questions about the Rothschilds and the Moon Landing, if they're feeling extra spicy. But generally, they don't look at the history of the Modern Age and conclude "politics is crack, and we're all microdosing addicts".
I don't know that it's a *good* idea, but I think it's a comprehensive (unsurprising given Moldbug's style) articulation of a *common* idea, which can still be valuable.
Like, I think the main thing I got from Moldbug was a better understanding of the grievances that conservatives and reactionaries have; it's not that I agree with his presentation, but it made me able to understand the point of view.
What I mean is, if someone described the Cathedral as "how the world works", I wouldn't find it that insightful. But if someone told me it as a description of the world view that animates reactionaries, I *would* find that insightful (and did).
My interpretation is similar, Moldbug was an idealist writer about the best possible solution. Yarvin is a realist that recognizes his old solutions were never feasible, and has found a realistic way to accomplish the same objective. You could call him a sellout, but honestly I think most people do the same over their life. People in their 20s are idealistic and shoot for the moon, but now he’s in his 50s, and wants to accomplish something, even if it’s not perfect.
Scott's point is that *Moldbug's own stated views on this* are that the non-ideal version of his program literally leads to Hitler.
Most idealists soften into accepting second or third best over an unachievable perfect; if Moldbug is doing that then it implies either he thinks "literal Hitler" is a second-best solution (which strikes me, as I suspect it strikes most people, as completely insane), or it means he no longer believes his old ideas in full--which is fine, but undermines the rest of his program. If the democracy in the Trump movement is worth contaminating oneself with, does that mean democracy isn't actually so bad overall?
For me, part of what makes moldbug confusing is that "democracy" is weasel-word that can mean lots of different things, depending on the context. E.g. "the abstract notion of political enfranchisement", "local decision-making among friends at Red Lobster", "formal voting schemes", "implicit voting by audience cheering", "neoliberal managerialism", etc. In the context of moldbug, i think what he really means is "modern liberal democracy as descended from *Nationalism*" (see my reply to Mr Fibonnaci [0]). I think Moldbug understands what's going on conceptually, but colloquial English isn't really precise enough to easily disentangle all the stuff going on under the hood of the concept of "democracy".
And as for moldbug's pivot, it's hard to say for sure. But my own hypothesis is also that he's decided to compromise on his ideals. But if you look at his twitter exchange with Scott, the "Miller Light" and "Jan 6" references make it clear that he thinks there's not really enough energy/edginess to actually reboot the NSDAP in America. On the contrary, he's more worried that Trump isn't powerful/edgy *enough* to overthrow the Catheral.
And for further context as to my mindset lately, I've taken a bit more of a Straussian reading of history lately. E.g. I think HBDchick [1] once hypothesized that Communism was popular in Russia/China/etc because the former Mongol Empire follows a certain kinship structure (viz Exogenous Communitarianism) which engenders a lot of resentment. And Marxism was just a convenient justification to Fuck Shit Up. Likewise, I think Scholar's Stage [2] once discussed how Jihadism comes not from Islamic Theology, but from the underlying Arab Culture. So I get this feeling that an ideology's popularity is less a function of Absolute Truth, and more a function of how well it rationalizes adherents' behavior post-hoc. And this is how I view the Vance and the Dissident Right deciding to wear UR's ideas as a skinsuit. So moldbug is just like "welp, according to my Asimovian Psychohistorical analysis [3], I can either settle for Trump, or endure 1,000 more years of Gay Race Communism".
----
{edit: I didn't explain the "democracy" thing well. there's "democracy qua nationalistic hysteria". there's "democracy qua concentration of political power" (cf Aristotle). there's "democracy qua FDR's bureaucratic leviathan".
To clarify, when moldbug says "democracy is bad", he's usually referring to nationalism and the bureaucracy. But when he says "Trumpism is democratically powered", he's referring to "diluted concentration of political energy" (neutral connotations) (for reference, Aristotle said that the 3 good types of government each has an evil twin. Democracy's evil twin is Ochlocracy, aka rule by mob).}
Au contraire, Yarvin Cathedral premises were always based on his mistaken (bordering on delusional) belief that elite institutions, such as universities and the mainstream press, were *colluding* to pull the strings of public opinion. This has been a mainstay of rightwing nuttery long before Yarvin came along. The Cathedral used to be the Trilateral Commission, and before that, it was Freemasons, and before that, it was the Illuminati. In their ignorance, people like Yarvin revive the bogeymen of the past and repackage them in an attractive garment of contemporary memes and discontent. The Paranoid Style in American Politics by Richard Hofstadter is still relevant sixty years after he wrote it, and nothing has changed except that rightwing nuttery has finally gained power in the US. There's something to Marx's quip that "History repeats itself, first as tragedy, second as farce."
What puzzles me is that the Substack ecosystem treats Yarvin as a Very Serious Person when his ideas are and always have been manifestly screwball.
I think he's said he wants Trump to reduce Congress and the Supreme Court to "advisory" roles only, which I interpet as a polite way of saying completely ignore them if they disagree. I think this is beyond FDR/Lincoln, although of course FDR's appointed court packing was a step in this direction. I really do think he means "rule by decree".
I think all of Moldbug's critique of the Cathedral survives this pivot, but I think the objection to him was always "doesn't dictatorship have some very obvious failure modes?", and that most of his answer to that question involved the weird stuff which is now obsoleted. I don't know what his new answer to that question would be.
Hopefully little. Last time I disqualified one entry for obviously being like this (unusually high score with most voters from the same unusual country as the author). I do spot checks to make sure most voters are ACX subscribers (obviously not all, and it's not cheating if a nonsubscriber votes, but if one entry has way more nonsubscribers than others, I'll be suspicious).
I have to say, I think the rules on these things should be more clearly spelt out. Last year I wasn't sure if it was acceptable to rate your own review. I didn't, because it seemed untoward, but I definitely didn't see anything suggesting it was frowned on.
Getting friends to vote triggers a much stronger negative reaction in me, but I imagine norms on this could be very culture-dependent.
I had something of an idea for the everything-but... contest, but was not able to implement it in time with sufficient quality due to other projects.
Perhaps there's room for two contests per year? One for books and one for everything-but... or is the intention to just return to the normal contest schedule next year and hope there's more everything-but reviews than previously?
You can do double the contests with the same fatigue if you only allow people with even birthdates to participate and vote in one and odd birthdates to participate and vote in the other
(replace birthdates with some other random oracle you can afford to assign to people)
‘Skill issue’ is a useful meme - on agency, learned helplessness, useful beliefs and systems
wrote a short essay on the usefulness of the meme “skill issue” that some of you might enjoy. I wrote it as a way to reconcile my own belief in personal agency with the reality of supra-individual forces that constrain it. The point isn’t that everything is a skill issue, but that more things might be than we assume and that believing something is learnable can expand what’s possible.
It’s part cultural critique, part personal essay, weaving through tattoos, Peter Pan, and The Prestige to ask: what happens when belief does shape reality? And how do we keep choosing, even when the choice feels like it’s left us?
Agreed. Take this as sincere criticism, Maja — but you lost me about halfway down your essay. At that point, I wasn't sure if you believed personal agency could overcome supra-individual constraints—or not—or whether you were sorta in-between on the issue. Instead of opening with a paragraph about Dawkins and the origin of the word meme, state your opinion succinctly and resolutely, then work in the meme theme further into your arguments.
Today I learned that the meme of the white woman (?) wearing a green shirt and a black cap screaming originated from a real person who actually did, in fact, scream during the very first inauguration of Donald Trump back in January 2017.
c.f. https://www.wbur.org/endlessthread/2021/11/12/memes-the-scream for an interview and https://www.youtube.com/watch?v=wDYNVH0U3cs if you want to watch the original footage
What did you think it was before that?
No I just didn't know where it was from. I was not terribly surprised of the date.
I feel like I finally understand why are Russians so obsessed with the greatness of their literature.
I mean, every nation has its favorite writers, but it is virtually impossible to have a debate with a Russian or a fan of Russia without arriving at the topic of "aren't Pushkin, Tolstoy, Dostoevsky, etc. the most important writers who have ever existed?" It's as if accepting the fact that these were the greatest writers ever is the mark of a civilized and cultured person. If you disagree, or just think that this is not important because everyone prefers different books, you are hopelessly stupid. The people who say this don't even have to be Russian nationalists, or at least they do not feel that way; they are just telling you an obvious fact that everyone knows. It's in the water supply.
But the fact is, this is an important part of Russian imperialistic propaganda. When Russians conquer another nation, they murder the writers and destroy the books. Then they teach their children the Russian language, and make them read and admire Russian writers. And a generation or two later, Russians can paint themselves as those who brought the light of civilization to the illiterate peasants.
It's not about the literary qualities of the War and Peace per se, because no one reads that book anyway, but about the fact that the glorious Russian nation can produce Great Literature TM, while you rednecks are not even capable of writing a limerick in that thing you call language. (Of course, whenever someone actually writes a limerick, it gets burned, and the author gets shot.) Which proves that you are not true nations, and you need your Russian overlords to provide you spiritual support and thereby keep you culturally above the level of mere animals.
Whether you think that Pushkin, Tolstoy, Dostoevsky, etc. were great or not, the mere fact that you are debating the question means that you have been successfully distracted. The question you should be asking instead is: what did Russians do to the writers of the nations under their rule?
See: https://en.wikipedia.org/wiki/Executed_Renaissance
Across Eastern Europe, it's pretty common for people to claim that their nation is the greatest in some respect, not sure Russians stand out so much.
More interestingly, it's simply false that " they murder the writers and destroy the books." Most famous Ukrainian authors (Shevchenko, Lesya Ukrainka, Kotliarevsky) created their works when Ukraine was part of Russian Empire. Mickiewicz and Sienkewicz lived and wrote for the most part in Russia. In the Caucasus and Central Asia, national literatures either experience renaissance in the 19th-20th centuries.
This is not to say that there was no oppression and some systematic efforts at Russification, some of the writers that I mentioned were arrested and/or exiled. But then Dostoyevsky himself also only barely escaped a capital punishment.
Tolstoy and Dostoyevsky are very both good and I have in fact read W&P in its entirety twice. Winters are long where i live too. ;) I think Anna Karenina is a much better novel though. Pushkin doesn’t do much for me. Among the 19th century Russians Turgenev would be ahead of Pushkin for me but again that’s completely subjective. This is complicated by the fact that I am mostly reading in translation. Some Russian speakers say that Pushkin loses a lot in translation.
Calling any writer the most important that ever existed is just stating an opinion. Great writers are great in many different ways for many different reasons. William Faulkner is an important American writer but to compare him to another great but more contemporary American writer, say, Cormac McCarthy would not make much sense. They both set out to achieve different things.
this is a bit of an odd strawman. I mean you can acknowledge Russian writers and filmmakers really have made some of the best artistic works out there, but so has everyone lol. Like trying to argue this doesnt need "but russia kills its artists!" over "this is just a silly argument."
you dont read war and peace, they dont read John Dos Passos. everyone is a redneck, lol
https://www.youtube.com/watch?v=BLzxuIfD9rU&ab_channel=TomCardy
A great silly dance song. Rationalist influenced.
Yet another article about Zizians:
https://www.sfchronicle.com/crime/article/ziz-rationalism-lasota-pasek-20323332.php
Sometimes I think the fundamental mistake of rationalists was to locate their capital in the Bay Area. Yes, close to the Silicon Valley, etc. But also the world capital of wokeness.
When I read these stories... most people who later appear to be a PR problem (to put it mildly) had obvious red flags since the beginning. In retrospective, the right thing to do would be to tell all people with serious mental problems to go away, right? That would prevent them from getting their symptoms dramatically worse after being exposed to the ideas of AI doom and Roko's Basilisk and acausal trade and whatever. But you can't simply send the crazy people away, because that would be *ableist*, of course!
You can't even send away the drug junkies, because hey, this is Bay Area, that would be deeply uncool! (Even if Eliezer himself disapproves of drugs, everyone else is like "yeah, just ignore him, all the cool kids do their own internet research about which drugs are safe".)
And you can't send away Ziz, because Ziz will obviously accuse you of transphobia. Though Ziz will later accuse you of transphobia anyway.
So you let all these people stay, and then someone snaps and someone gets killed, and then all the normies are like: "omg, the Bayes Theorem drove them crazy, I knew that was some dark sorcery!"
...I know this is written with the benefit of hindsight, but it seems that the project of "raising the rationality waterline" was doomed since the moment the key players decided to establish their base in the Bay Area.
And, you know, we didn't even succeed at slowing down the AI apocalypse. Maybe accelerated it.
I wonder what happened in the alternative Everett branch, where people decided to avoid Bay Area like poison, and instead kept the rationalist bases in New York and Berlin. Are they smarter? Or did they fuck up dramatically in a different way?
If you eliminate all the crazy people (this may be harder than you think), will you get sufficient weirdness?
Weren't the Zizians thrown out of the rationalist community?
> If you eliminate all the crazy people (this may be harder than you think), will you get sufficient weirdness?
We should not optimize for weirdness per se.
> Weren't the Zizians thrown out of the rationalist community?
https://www.lesswrong.com/w/j-lasota-ziz
If I read it correctly, Ziz joined the Bay Area rationalist community in 2016. In 2019 they became increasingly aggressive against CFAR people, which led to the infamous protest. There was no throwing out -- Ziz decided to leave. I am not sure about the other Zizians.
They wouldn't have been relevant regardless. But I don't know why you're acting like this was an issue contained to the Bay Area, as if it wasn't endemic to every city on Earth worth a damn. And now those titans at Silicon Valley have put their lot in with the force that will provide a long-term solution, so what's the problem?
> But I don't know why you're acting like this was an issue contained to the Bay Area, as if it wasn't endemic to every city on Earth worth a damn.
Could you give me examples of bad things happening in the rationalist community in other cities? Maybe there was a thing or two, but probably nothing comparable in size to things that happen in the Bay Area regularly and kinda predictably.
...What rationalist communities in other cities? The whole movement was centered on the Bay Area, as you said yourself. Of course the bad things would happen there, because that's the only place where things are happening. On the other hand, the unwarranted sympathy for liabilities that you're complaining about is not contained to the Bay Area. This obsession with "tolerance" is everywhere. Any city is going to have a good amount of... problematic elements. The same thing would have happened regardless of where they decided to settle.
nothing to do with wokeness, the conservative example in the past was people bombing abortion clinics. it turns out rhetoric can lodge in crazy people's brains and its not a good idea to go incendiary or apocalyptic when you really don't believe it.
like a lot of people say things but never internalize them, because their internal life is full. but empty or damaged people do, and its dangerous. i don't think much can be done except to reflect on whether you are not kind of giving ammo to it; abortion is wrong but if you start saying abortion is murder...
I agree that crazy people are both in the left and the right. But it's the left that considers talking about crazy people and avoiding them "ableist". (I think the word "crazy" itself is banned.)
But the group's beong too woke to call crazy people crazy is only one of the dynamics by which crazy, destructive people can be accepted by groups. Groups can also adopt people who are way blunter about some issue than them, and more willing to take destructive action. Those people give the group a way to participate vicariously in vicious tirades and physical attacks they themselves can only dream of carrying out.
i mean the right let their various crazies take over the party with the various truther movements. but its a human impulse not a partisan one
its tough to explain. Like imagine you are in the 80s and we watch Threads, a very grim movie about nuclear war.
like 90% of people will watch it, shiver, and the next day put their pants on and go to work unfazed. water off a ducks back. Some may do very light disaster prep that wont save anyone.
Maybe 1% will be moved to make it part of their life to stop nuclear war in healthy ways.
9% though will be various types of crazy, from "how can anyone have kids under MAD?" to selling everything to build a bombshelter to get crazy in, to what have you. the extreme cases turn to violence because they are very crazy and cant let the idea roll off their back. it would be best if they never knew it existed.
the ability of people to entertain apocalyptic scenarios and functionally ignore them is something we take for granted. maybe we should chill a bit with them because some can't.
Are there any good theories as to why Elon Musk keeps talking about birthrates being incredibly important, but also supposedly believes in short timelines for AGI?
It seems like a bizarre preoccupation moving in to a technological singularity.
He is manic, and his views are much less coherent than those of someone who's not flooded with senseless energy, confidence and optimism.
+1
I'd personally say that AI is the most urgent issue we're facing now, but that stuff like the ever-growing deficit, fertility collapse, global warming, and persistent government dysfunction are all also very serious problems. We should care about all of them, and we should be trying to address them all.
Musk believes in short timelines for AGI *and* also that alignment is possible (at least if the greatest genius in history, ie. Elon Musk, is allowed to work to bring the alignment to reality), so if we're about to move to an AI-created universe-spanned human-oriented Paradise, it makes sense that he wants as many people to enjoy it as possible.
If he believed that, wouldn't he be working 24/7 on AI alignment instead of on DOGE showmanship?
The most reasonable explanation I can think of is that Musk is not a perfect utility maximizer, and is prone to shift to side projects that he then focuses on, almost exclusively. He chooses projects roughly in terms of how interesting and solvable they appear to be. The mean time for him to happen upon such problems seems to be about 12 months, give or take 6. An unusually tough problem might take him longer, especially if it's standing in the way of an even longer term project he has, or if it can be broken into subproblems, each as interesting as the first. But the key here is that once he's decided to work on a problem, he ignores other problems almost entirely, like a cheetah that chases an antelope she's locked on even if the chase takes her past other antelope.
So for now, optimizing the US government is his interesting problem. It will probably stay so for the next 6-12 months.
I think simple reasons are enough to explain it. In order of most to least important:
1. Musk is on Twitter constantly and his timeline constantly talks about birthrates.
2. Musk's political faction sometimes uses low birthrates as reasons to support their policies.
3. Musk's uses low birthrates to retroactively rationalize a history of "pregnancy-maximizing" behavior in his personal life. (https://archive.ph/F3Foa)
There's also probably a motivated reasoning factor. Musk has something like 14 children by three or four different mothers, and presumably that predisposes him to arguments that suggest his prolific fecundity is a highly prosocial activity that others should imitate.
Trying to solve all humanity's problems single-handedly seems in-character for Musk, and I guess that includes the population crisis.
I am surprised that *more* billionaires don't do the same thing. I mean, if you have enough money to get as many women as you want, and feed as many children as you can conceive, then what exactly is the argument against doing that?
My guess would be a quality over quantity preference combined with a preference for a relatively conventional family structure. Looking at the rest of the current top ten billionaires in the US (Bezos, Zuckerberg, Ellison, Buffett, Page, Brin, Ballmer, Bloomberg, and Gates), they seem to be pretty tightly clustered around three children apiece (Bezos has four; Ellison, Page, and Bloomberg have two; and the rest have three). Five of them are divorced and Buffett is widowed, and of these Brin, Buffett, and Ellison have remarried. Only Ellison has divorced and remarried more than once (*). None of them seem to have publicly-acknowledged children born out of wedlock.
(*) To paraphrase an Oscar Wilde character, one failed marriage may be regarded as a misfortune, but five seems like carelessness.
On the quality vs. quantity front, spending time with your children and being involved in their lives and upbringing takes time. The time demands of parenting scale sub-linearly with number of kids (this being a major part of Caplan's arguments in favor of larger families), and billionaires can trivially afford to hire as many nannies and other helpers as they need, but there's a certain baseline of time and attention you need to direct towards your kids if you want to actually have a relationship with them. Billionaires also tend to be hard-driven career men (and women, but mostly men) who spend a ton of time on their businesses, which only leaves so much room for family time.
If you value at least the appearance of a relatively standard family structure, and especially if you value the substance, then your spouse's preferences are a limiting factor on how many children you have. Preferences for very large families are outliers in our culture, and both partners would need such a preference in order for the couple to decide to have lots of kids. And I'd guess that women somewhat less likely than men to prefer very large families, since women are generally the ones who get pregnant and give birth. Some women do actively enjoy the physical experience of being pregnant, but I get the impression that most don't.
Overall, the top tier of billionaires (excluding Musk) seem to have more kids than the overall norm (2.78 vs 1.94 children), with "can afford more help" winning out over "hard-driven career men" to the tune of an extra 0.84 kids. Which doesn't strike me as too surprising.
Also, in a traditional marriage where you have one wife and stay with her, you hit some constraints:
a. There is a limited fertility window that closes sooner for women than for men. I could presumably still have more children, but my wife has gone through menopause, so that's not on the agenda unless I get remarried or take up a mistress or something.
b. It's somewhat common for health problems to crop up that close the woman's fertility window early, so even if you always planned to have five kids, you may find that you're only having one child after all.
c. Random crap can happen in your life that makes it very hard to have kids for awhile--losing your job, having a family member very ill or in some kind of serious trouble, some physical or mental health crisis by mom or dad or one of the existing kids, etc.
I suppose if you want to do your best King Solomon impression, you can also get some parallelism going--nine women can actually deliver a child per month, after all....
> On the quality vs. quantity front, spending time with your children and being involved in their lives and upbringing takes time.
> And I'd guess that women somewhat less likely than men to prefer very large families, since women are generally the ones who get pregnant and give birth. Some women do actively enjoy the physical experience of being pregnant, but I get the impression that most don't.
This does not matter, if the billionaire can choose the ones who *want* to have many kids. Because there are women like that, too.
If that is high on his priorities when choosing a partner, and if he is choosing (or re-choosing after divorce) after becoming a billionaire, then yes, a billionaire would presumably be able to select for that as Musk seems to have done.
The other nine of the current top ten US billionaires seem the have revealed preferences for other priorities besides having more then 3-4 kids apiece.
Now I'm wondering how many billionaires came from stable homes.
An educated guess is that Musk believes a species with a birthrate below replacement is headed for extinction unless birthrate goes back up, and a birthrate even close to replacement is an extinction risk if something in the environment goes pear-shaped - such as unaligned AGI, but also a comet impact, giant solar flare knocking out the grid, nuclear war, Black Death, mass famine, ice age, and so on. A high birthrate mitigates a lot of problems at the species level in the same way a healthy immune system and diet mitigates a lot of problems at the individual level - at worst, it buys you time.
Musk probably also believes (as a lot of people do) that the sort of AGI plausibly in our near future isn't an existential threat. It's even possible that it's a non-threat in his eyes, and so birthrate is merely a benefit alongside it.
From the oral argument in the SCOTUS yesterday it sounds like most of the justices view the end-birthright-citizenship EO as obviously unconstitutional.
The matter before the Court yesterday wasn't that question; rather it was whether nationwide injunctions from federal district courts should be allowed to remain in place. But any court injunction is rooted in a likelihood of the plaintiffs eventually succeeding on the merits, so the larger question about that EO was inherently present yesterday.
Both conservative and liberal justices posed hypotheticals to the Trump admin lawyer which assumed the EO is eventually ruled by the SCOTUS to be unconstitutional. They each asked versions of "assuming the EO is ultimately found to be invalid how should that judicial review process proceed?" [meaning in particular, with or without the presence of nationwide injunctions?] Alito, Kagan, Barrett and Kavanaugh all got into that with John Sauer who is Trump's chosen Solicitor General.
Sauer responded repeatedly with the same basic argument: that the presidential EO should receive full deliberative due process. His proposed alternative is the class-action lawsuit process, which Kavanaugh seemed to agree with while Alito not. I'm not sure how serious a suggestion or possibility that is. Regardless, Sauer repeatedly argued that whether class action or not, court cases against presidential EOs should work their way through the federal courts in "the normal manner" without being frozen by any district-court injunctions.
He may even be right about that, I'm fairly persuaded myself that district-court national injunctions have become a broader problem. Still though, the irony! The president who bluntly wants to throw all non-white asylum seekers into dungeons first and sort 'em out later if ever, is suddenly verrry interested in the importance of...due process.
I've said a few times to a friend of mine -- a previous government lawyer -- that we both wished that this question was brought before the court in literally any other administration. Nationwide injunctions are a legitimate problem with serious negative effects, but, fuck it, call me a partisan, we *need* that now. I noticed none of the GOP was calling for the end of nationwide injunctions during Biden's admin.
As it is, Sauer was an absolute wreck. The justices AND the court were outright laughing at him several times. My favorite line of questioning was Kagan, who had two bangers:
- "O, so if the government decided to take everyone's guns, that would be fine?"
- "You keep saying that people can appeal to SCOTUS. But for that to happen, some plaintiff will have to lose. And in a case as ridiculous as this, the plaintiff will win *every time*. And you, the government won't ever appeal, because you don't WANT a national ruling!"
But I also appreciated ACB's almost genuine disbelief at Sauer stating that "generally" the government will follow court precedent, and even Kavanaugh had a good question in there about the mechanics of figuring out citizenship for babies.
Meanwhile it felt like Alito and Thomas and even Roberts were trying to give Sauer layups that he was either unable or unwilling to take, which I thought was funny. I'm sure somehow the former two will find a way to justify why there shouldn't be a nationwide stay here, as they always do.
Luckily SCOTUS isn't stupid. They see through Trump's demands for due process for *him* in this case where he is the defendant, and not in any case where he is the plaintiff. The hypocrisy is *staggering*, and really makes it clear that everyone in the Trump admin sees both the judiciary and the law in general as impediments to be removed instead of principles to uphold.
Well said, all around.
That Barrett moment was LOL for me, and there were a couple of good memes of it going around Reddit.
A couple different practical answers about nationwide injunctions have been proposed in Congress more than once in recent years but never advanced. Maybe one if those finally will now.
That is because it is a great stalling tactic, and allows Trump to go forward with this in the meantime. The idea that this should wind its way through federal courts for the next year, while children are born stateless, is absurd. The executive order is on its face textually unconstitutional. Whatever arguments you care to make about altering that amendment, It’s still a fact on the ground. The second amendment is another good example. At the end of the day, there was no getting around the text of the thing.
I listened to the argument, and I found it interesting that at the very end the Attorney General Sauer argued the case on the merits,after having spent all his time previously trying to focus it just on the legalistic question of universal injunctions.
I think this is a very bad case to resolve that issue with. The text of that amendment is way too big an elephant to hide in a closet.
I personally think the 14th amendment has become outdated. It needs to be revised. But the Supreme Court can’t rewrite it. This administration brought this case as a stalling tactic. That’s my opinion. Of course they want it to percolate through all kinds of courts and take forever. I think it was pretty clear that most of the justices found both class action and percolation to be problematic to say the least. I know there are justices there that would happily rewrite the 14th amendment if they could only get their hands on it, but the Govt doesn’t want to tee that one up.
Does anyone have stats or have recommendations for how to find stats on the % of teachers that abuse children? I have done some research and can only find % of students.
Mostly I've just used AI search and done some searches on pubmed. Please help!
Do you have in mind sexual abuse or other forms of abuse such as inflicting physical or emotional harm? If it’s the former, I found stats maybe a year ago on what
percent of abusers fell into various categories in cases where the abuse came to light. Teachers were definitely one of the categories. Not exactly what you’re asking but gives a basis for estimating: We know what fraction of kids are known to have been abused at a certain age, and can estimate fairly well how many teachers a a kid has had by the time he reaches a given age
. I don’t know how much data there is on what fraction of abuse come to light. I’m a psychologist and have seen quite. few people who were sexually abused as kids, or raped as adults, and all of them either kept the abuse a secret or only disclosed it years later, or only disclosed to people they knew would not inform the authorities.
sexual particularly
We could make a guess based on the following numbers: total number of teachers, total number of students, % of abused students, average number of teachers abusing the same student, average number of students abused by the same teacher.
Total numbers of teachers and students should be public info; you already have the % of abused students. I would guess that the average number of teachers abusing the same student is approximately 1; maybe let's use 2 as an upper limit.
So the last missing part of the puzzle is how many students are abused by an average abusive teacher. You could probably find some numbers in some specific cases, the remaining question is whether those numbers are representative... maybe teachers who abuse more students are more likely to be caught.
Or you could draw straws.
Yeah that's kind of my response lol
Eh, seems like quite a lot of guesswork.
If I was the superintendent of a school district where such a thing happened, I'd try hard to keep it out of the news. If I worked in the Department of Education, I don't think I'd be eager to publish this data either, for obvious reasons. I don't see any reason to believe that this data would be easily accessible, or even gathered and stored in one place: the organisations running schools don't have an incentive to gather and publish it, and the courts or child protection authorities probably don't track child abuse cases by the profession of the abuser.
In other words, that data might not be publicly available. Depending on the country you are interested in, you might be able to file a Freedom of Information request with the appropriate authorities.
Yeah I think that's what I'm finding. It's frustrating because everyone points to the Catholics and rates of child abuse, but they wre only able to get the data because of major compliance with the Catholic church. I'm very curious how it would compare with teacher rates of abuse...
In Ireland, where I'm from, Catholic orders ran residential schools for much of the twentieth century. There was terrible abuse there, and some of the most notorious pedophiles were Catholic religious working in schools. You can look up Letterfrack industrial school as an example.
There is some evidence that the Catholic Church is no worse than other large organisations running residential programmes for vulnerable children. I don't have facts and figures, but I think that e.g. the Australian Government programme which took indigenous children into state care in the mid-twentieth century was found to have abuse at similar rates to the Catholic schools I described.
But I think your premise is wrong. Other organisations failed also, the Catholic Church might have more prominence because it's larger and richer. But it's unethical to tell people that child abusers are everywhere, and so the Catholic Church is not that bad. Are you trying to undermine trust in public schools and society in general? Perhaps better to focus on ways in which the Church has helped people, but don't trivialise the suffering of people abused as children.
I'm not trying to trivialize the suffering, just trying to get a good understanding of the topic!
I'm definitely not going to tell people that since as I said I don't have the data to back it up.
Kinda late in the open thread, but I just read that the Federal Appeals Court has given UC Berkeley a chance to win back the CRISPR patent from MIT.
Emmanuelle Charpentier and Jennifer Doudna, who shared the 2020 Nobel Prize in Chemistry for devising the CRISPR technique, both worked at Berkeley. But MIT got the patent because they showed it could be applied to eukaryotes. Charpentier and Doudna expressed uncertainty about whether it could in their groundbreaking paper. Because of their expressed doubts, the US Patent Office gave the patent to the Broad Institute of MIT, which showed that CRISPR could edit the genes of eukaryotes. Of course, I'm reading the media's distorted reporting of legal arguments, but it would have been ironic if Charpentier and Doudna could have been awarded the patent if they said, "We're pretty darn sure we can apply this technique to eukaryote genes, but we haven't done so, yet." Of course, the US Patent Office, as a general rule, awards patents to *working* implementations. However, C&D did get CRISPR to work on prokaryotes—but no one really cares about prokaryotes. Eukaryotes are where the action is!
https://www.msn.com/en-us/news/technology/uc-berkeley-wins-chance-to-reclaim-lucrative-gene-editing-patent/ar-AA1EEjXv
https://www.technologyreview.com/2025/05/13/1116344/a-us-court-just-put-ownership-of-crispr-back-in-play/
Oh boy, this just so reinforces the basic truth that a patent is not a technical document, it’s a legal one. Don’t ever put in writing “I’m not sure my idea works for ‘thing’”, the other guy will use it as a defense, “see, Fibonacci here is explicitly teaching away from using the idea for ‘thing’, while in [0056] we are teaching how to do ‘thing’”.
Ask me how I know :)
Also, exposes the tension between publishing papers and applying for patents at the same time. The two don’t mix well. I have dozens of patents and 0 scientific papers to my name. Got to pick one or the other.
I gather you went through something similar?
It seems that there are already enough incentives for scientists to be dishonest about their work. Now we add to the mix the University's licensing lawyers (who dream of fatter endowments) and the nitpicky requirements of the USPTO (not that it could be any other way).
Yes, I used that exact framing to defend a patent submission. We (very few patents have a single inventor these days) had our submission rejected based on certain prior art, where the examiner saw a document that mentioned both the 'method' we were patenting and the 'thing' we were trying to accomplish. But on a closer reading the prior art was actually saying "there is this 'method' but you can't really use it to accomplish the 'thing'". So I constructed a response basically saying, look Mr. Examiner, you're saying that anyone skilled in the art will understand that our 'method' can be used to achieve 'thing' based on this prior art, but ACTUALLY the prior art document says "don't do this", so it "teaches away from using the 'method' to achieve the 'thing', therefore you can't argue that anyone skilled in the art will understand from the prior art reference to use the 'method' for 'thing'.
It worked. We got the patent.
I don't think this means scientists have to be dishonest. What it really means that IF they want to patent what they publish, they have to think very carefully about both timing and content of the relevant papers. In simplest terms, "file first, publish later". Then maybe instead of saying "'method' is unlikely to work for 'thing'", say "possible use of 'method' for 'thing' is still under investigation" - no lies here, just a more careful wording.
I conducted a Tarot reading about the question "Will AI serve humankind?".
I use a Rider-Waite-Smith deck, Joan Bunning's tradition (https://www.learntarot.com/cards.htm), and a Celtic Cross Spread (https://www.learntarot.com/chcc.htm).
The result was: PW 5P 4P 7S KC 7C 7P XVIII -4S -XIV
Position 1 (Heart of the Matter) & Position 2 (Opposing Factor) - Page of Wands, Five of Pentacles
The core of the matter is that Humanity rushes forward with a child's enthusiasm into embracing the creative power of AI, but this could lead to hard times and economic inequility (and maybe even extinction if we stretch the ill-health meaning of 5P).
Position 3 (Root Cause) - Four of Pentacles
The source of the whole situation is the greed of the relevant participants. If it generates profits in the next quarter, why not do it?
Position 4 (Past) - Seven of Swords
In the recent past, there was some hidden dishonor in the situation. This probably refers to the various shady AI company business like the OpenAI boardroom drama, Sam Altman maneuvering behind the scenes, Anthropic pushing frontier not caring about the previously and privately made commitments, etc.. The card is hidden dishonour, so we probably don't know most of it.
Position 5 (Goal or Purpose) - King of Cups
The goal/the ideal future of this whole AI situation is that the AI, like a caring and wise father, enables human flourishing forever in the universe. Think Minds in Culture novels.
Position 6 (Future) - Seven of Cups
In the near future we can expect to have lots of options on how to proceed, but also much wishful thinking ("I'm sure it will behave roughly the same way even with more intelligence.").
Position 7 (Self) & Position 8 (Other)- Seven of Pentacles, Moon
The querent thinks of themself as carefully assessing the possible consequences of AI, but from a more external point of view their thinking is confused and dominated by fear. Or alternatively, if we take the self/other position to be about humanity instead of personally me (because the word "I" was not in the question): Humanity sure does believe itself that they think carefully about AI, just look at all these blog posts, congressional hearings, books etc, but is it not actually an illusion? Won't the relevant people just keep on doing what they have short-term interest in doing and rush toward the end irrespective of the content of those deliberations?
Position 9 (Hopes and Fearrs) - reversed Four of Swords
The main fear in this situation is that humanity won't contemplate, won't prepare enough for the possible pivotal point and rushes on without pause.
Position 10 (Outcome) - reversed Temperance
The final outcome is that the harmony and cooperation will only be apparent. This, to me, suggests deceptively misaligned AI. Or alternatively: The going away of balance which makes me think of a paperclip maximizer, ie an agent which pursues a singular goal maximally hard, without caring for anything else.
Disclaimer: Reading Tarot is merely the drawing of purely random images on cheap paper. There is no supernatural Spirit or unconscious directing the process. This is purely for fun, the verbal equivalent of looking at clouds forming an angry old man or a dog. Peace!
Hi Scott. I won't link to the post in case there's a privacy concern, but in one of your defunct blogs you made a post about dystopian fiction, and you mentioned a novel where "all Earth's countries have been renamed things like FRA-113 and JAP-289". But you didn't name the novel in the post. Do you remember what novel it was?
I recognise the question as you've asked it before. It sounded like an interesting premise for a book but I had no luck searching with all my google-fu or any of the AIs out there. I'd also interested in an answer.
Gemini 2.5 outperforms, suggesting Ira Levin's This Perfect Day; on a quick glance inside via Google Books, people are said to live in regions like EUR55131 and MEX10405, so it's at least a plausible candidate.
I'd be surprised if that wasn't it! Thank you very much!
Two recent articles in Fast Company (the tech-world business magazine) about DOGE:
https://www.fastcompany.com/91332785/doge-ai-retirement-system-built-under-biden
"A leaked planning document shared with Fast Company shows the ORA pilot launched in 2023 with a handful of agency HR and payroll offices, serving “a few hundred” retirees. The plan under Biden was to roll out the ORA system government-wide in 2025. "
https://www.fastcompany.com/91330297/doge-sahil-lavignia-gumroad
"Now that he’s there, he says he finds himself surrounded by people who “love their jobs,” who came to the government with a sense of mission driving their work.....
....what he’s found is a machine that largely functions, though it doesn’t make decisions as fast as a startup might. “I would say the culture shock is mostly a lot of meetings, not a lot of decisions,” he says. “But honestly, it’s kind of fine—because the government works. It’s not as inefficient as I was expecting, to be honest. I was hoping for more easy wins.”
An interesting turn in the SA refugee story: SA President ranted on them, calling them “cowards”. I still do think it’s hilarious, having the anti-refugee President welcoming African refugees, but! I now fully support admitting them as refugees. When your country’s President calls you “cowards” for wanting to emigrate, get the hell out while you can.
Edit: source link https://www.bbc.com/news/articles/ce82e67p8p2o
As an American, I am *not* personally against South Africans as a group. They seem to be pretty fine people as far as I can tell.
I was first introduced to the story from a usually anti-China vlogger.(Will look up if anyone cares.) He had video claimed to be black South Africans calling for white people to be killed, and it might be legitimate.
However, it seemed to be a problem at this stage of a high crime level. I don't know whether white South Africans are at more risk than black South Africans.
It looked like adequate security in South Africa is very expensive-- both a fortress home and guards, and my conclusion is that if white South Africans are at special risk, it's the white South Africans who don't have much money, and they aren't the ones being let into the US.
I don't understand how it's hilarious, can you explain the joke?
I meant that Trump is anti-refugee, but is inviting these Africans to seek refuge in the US. I think it’s amusing.
The rest is exactly what Rothwed said - it seemed all kind of overwrought until SA president opened his mouth.
There were a few hundred or so (white) South Africans that were recently admitted to the US in a refugee program. It became a media kerfuffle because a church administering the program pulled out over disagreement that these people were worthy of refugee status. And the SA government was quick to dismiss the idea that they were being persecuted. It's funny that the president called them cowards because that directly implies they were fleeing danger by leaving the country, i.e. they were legitimate refugees all along.
I think it implies that they were fleeing *perceived* danger where there was no actual danger; it's not cowardice to flee serious persecution but it is cowardice to flee mild inconvenience while claiming it's persecution.
He also said they were fleeing "social justice for historical privilege". I don't know about you but I wouldn't be waiting around to receive that.
If you place that much weight on one-off remarks by a country's President, what do you think Trump opponents should think about the remarks that he makes? Do you think a reasonable person would regard a black lives matter protestor leaving America as a "refugee fleeing persecution"?
That isn't really the issue, it doesn't matter what one guy says so much as what the government policy actually is on the ground. I just find it amusing that he both acts like these people are cowards and fake refugees while simultaneously claiming they deserve to receive justice for historical wrongs. Both of these are obviously not going to be true at the same time.
Well, if I were an Afrikaner in the SA I'd not wait around to find out whether the danger I perceive turn out to be actual or not.
Ramaphosa could address this in so many different ways, like earnestly proclaiming his and his government full support for the rights of every SA citizen regardless of skin color or prior history, he could gently egg them in a stile of "come on y'all, really, there's nothing to fear", etc. etc. He chose differently. With the SA's history of gruesome violence I'd not want to wait for a chance to learn what necklacing feels like from the inside of the tire.
Yeah, he definitely shouldn't have said it, but I think it's wrong to interpret it as a threat or menacing; it's like Trump insults: pathetic and beneath the office, but if a black lives matter organizer tried to claim refugee status elsewhere because Trump called them "thugs", that's obviously dumb.
It may very well be wrong. But I'm here an ocean away from the place, so I have a luxury of suffering no consequences of interpreting the signals incorrectly. They, being there, clearly see things differently. I suspect I'd do the same.
I submitted for the not-a-book review contest on Monday. I was expecting to get some kind of confirmation email or something, but I didn't get one, and it's activated my anxiety. My brain keeps telling me that I messed up the form, and all the hard work I put into my review won't even get read.
Can anyone confirm if I was supposed to get a confirmation email? Is there any other way to confirm that my submission was received? My overactive brain thanks you!
There’s a short thread from primrose further down in the comments asking the same thing; the short of it is that there’s no confirmation email, if the google form said thank you for your submission, then it should have worked. I’m in the same boat, also nervous that something is going to have gone wrong with my submission to the contest, but it sounds like the only way to be completely certain would be to email Scott directly, which I don’t want to do.
Okay, thank you for your reply. It's good at least to know that others are in the same boat. Have a nice day.
So he gets a gold plated 747 (whatever the hell that means) stashes it away in a hanger until if and when he leaves office, and that can’t possibly be considered a bribe?
What is a gold plated 747 worth anyway? If you sold one, how much could one reasonably expect to get for it?
If Donald Trump is the registered owner of the airplane as soon as Qatar transfers it, that's a bribe. If the United States Air Force is the registered owner of the airplane, and uses it to fly Donald Trump around while he is POTUS, that's not a bribe, that's just a bargain. If the airplane winds up owned by the Trump Library and Muesum Foundation after Trump leaves office, it gets murky - there is precedent for the Air Force donating retured AF1s to presidential libraries, but if Donald Trump maintains an unusual degree of control over the foundation, if he derives direct benefit from it, and particularly if the foundation decides that they're not going to put the aircraft on display just yet but use it to fly Donald Trump around in his post-Presidential travels, then it looks a lot more like a bribe.
The lack of any discussion of these distinctions in the media coverage of this controversy is disturbing to me. Nobody seems to care whether it's legal or ethical, just whether they can use it to score points.
I mean with scenario 1 while not illegal per se, having a president flown by a luxury plane gifted by another state whose intentions are murky is objectively a bad decision.
1 it could still influence the presidents decisionmaking related to that country (how i love Qatar who has provided the luxury i am in)
2 it's easily a security risk due to potential listening devices of any kind or worse (though doubt Qatar would risk anything that can kill the president)
3 might become a slippery slope for outright corruption.
> after Trump leaves office
This is also doing a lot of heavy lifting. Does it end up with the library *immediately* after Trump leaves office? In which case that is also extremely strange. Or is it only after many years of service to many presidents later, in which case less strange (in part because Trump would likely already have passed of old age) All of the reporting I've seen on this seems to imply it ends up with the library immediately, which I would take as either
a) a bribe
b) an extremely bad use of funds
The odds of any president not named Trump, making use of a plane that used to belong to Qatar and hasn't been completely torn down and rebuilt, seem exceedingly small.
Are we actually worried about the Trump Library using it to fly anyone? It's a 747, and those are not small or cheap to operate. The VC-25 is at least plausibly necessary given the size of the President's entourage, but all of the 747BBJs in air force colors (including this one) are basically because they were sold to oil sheikdoms who wanted the same plane that POTUS flies in. Obviously, donating it in flyable condition and when the US government isn't necessarily done with it (though there's the obvious question of why he have it in the first place) is extremely icky, but I'm not sure that he's going to get much benefit out of it, and I'd really like the sight of the Trump Presidential Library going bankrupt because they tried to keep this thing in the air.
Oh, there's no doubt a privately-owned 747, especially one whose operation has to be laundered through a "charitable foundation" would be a white elephant of the first magnitude. But it's the sort of white elephant I could see Donald Trump's ego compelling him to accept, and he might be able to make it happen.
Certainly having it drive the Trump Presidential Library into bankruptcy would be the most fun outcome here. The more likely but boring outcomes are that the gift never happens, or the gift happens and the plane spends the next three and a half years being torn down and rebuilt before being put up on blocks outside the Trump Library.
Its a fair criticism that the media are not covering the details. But there are two potential questions.
Is it a bribe? Murky
Is it violation of emoluments? Not very murky. Due to pre-arranged transfer of ownership to closely affiliated organization. Emoluments doesnt require quid pro quo.
And emoluments has a straight forward resolution: congress, not ag, reviews and approves.
Murky is a good word for it.
It's incredibly hard to keep track of things in the following categories:
* Things that Trump has actually done
* Things that Trump has said he's doing to do
* Things that Trump has said he might do
* Things that Trump has randomly mused about possibly doing
* Things that Trump's opponents have accused him of planning to do.
I have no idea, at this point, where the 747 story goes in all these categories. I'm reasonably confident it will never actually happen and everyone will move on to talking about some other thing.
Judge Dugan (the Milwaukee judge who was picked up by the FBI on bullshit charges) and her legal team filed their petition to dismiss the case.
From the first page:
> The problems with this prosecution are legion, but most immediately, the government cannot prosecute Judge Dugan because she is entitled to judicial immunity for her official acts. Immunity is not a defense to the prosecution to be determined later by a jury or court; it is an absolute bar to the prosecution at the outset. See Trump v. United States, 603 U.S. 593, 630 (2024)
Damn. Citing Trump’s own precedent against him. I aspire to be this petty.
Full petition here: https://storage.courtlistener.com/recap/gov.uscourts.wied.111897/gov.uscourts.wied.111897.15.0.pdf
Helping an illegal evade deportation is not an official act, as much as progessives might wish it to be
You should try reading the Trump immunity ruling. Roberts makes clear that intentions cannot be used to determine whether or not an act is (un)official. That was kind of the whole point of the only action the SC deemed immune - Trump threatening to fire Cabinet members if they wouldn't send out a fradulent letter to confiscate voting machines from the states could not be used as evidence because it is a conclusive and preclusive ability of the President to fire Cabinet members.
Sorry, what are you hoping to add to this conversation? Are you just signaling?
I was pointing out the immediately obvious flaw in your fist pumping agreement with Dugan's submissions to the court
Are you this credulous about all prosecutions? Like, say, all of the prosecutions that Trump was under, and the felonies he was convicted of? Or are you engaging in motivated reasoning?
Note that you started off by saying that the charges are bullshit. I could just as well ask you whether you're this incredulous about all prosecutions
Our takes on the merits of the case aside, to my understanding the central finding to be determined in the case is whether Dugan obstructed ICE in carrying out its functions. That's not an official act, whether or not that that is established in court
O, sorry, yea, that's because the charges are bullshit. I'm not this incredulous about all prosecutions, just the ones done in obvious bad faith by this administration. I don't have any patience left for obvious blatant authoritarianism.
> to my understanding the central finding to be determined in the case is whether Dugan obstructed ICE in carrying out its functions. That's not an official act
this is like saying "the central finding to be determine in the Trump case is whether he tried to soft coup the country. It's not an official act". The whole point of Trump's legal argument was that you cannot even prosecute to begin with. That was also, imo, pretty bullshit.
But, in case it matters, I also suspect this play won't work. Which is why, despite your claim, I didn't do any 'fist pumping agreement'. I do think it's a clever argument because it makes bare the obvious bad faith in which this administration and the entire GOP has been acting for, say, the last 9 or so years.
Interesting. I wonder if it will succeed. I kind of doubt it. I think this case is going to be around for a while.
Sort of an unusual issue in the mmo i play, Final Fantasy 14.
in the game you can go into player housing set up as venues to chat, like cafes or bath house.
in ffxiv its possible to mod your character, by replacing the assets on your local computer. you make your catgirl into a fox girl, etc. only you can see it but you coukd screenshot, etc.
but there is a program that lets you sync this, so you can download and see everyone else's mods. people "connect to the wifi" and now you see the petite catgirl you are chatting with is seven feet tall and packing piercings. The two girls dancing are now actually doing a sex animation, because you tag it to emotes in game.
this...creates an unearthly experience if you dont or cant have the mod (console players)
You are blind, but not blind. A girl makes a joke about people being overdressed in the tub, but to your eyes she is fully clothed; she just has replaced that particular outfit with scandalous textures. Everyone else can see it, but you can't.
it creates a layered reality that is something to experience. A friend described to me what is really going on and you'd never know from the room otherwise.
and a majority of venue enjoyers use it. not always for lurid acts, but to show off their vision of their characters for others to see. but not using it, that world is invisible to you.
overlapping realities but being unable to see them is a very sci-fi thing.
John C. Wright has something similar to this in his sci-fi trilogy starting with "The Golden Age". Essentially everyone is always in augmented reality and sees (and hears/feels/etc.) whatever their filter is programmed to show. If you turn off the filters, everything (in reality) is just spam ads.
Overall, I recommend the trilogy - it describes a largely positive way that a post-scarcity society can function without going quite as far as Banks' Culture universe.
https://www.goodreads.com/book/show/207410.The_Golden_Age
Heh, so it turns out that:
-- the gold-plated 747 that the Qatari royal family has Trump so excited about is 13 years old, and has none of the extensive secure-communications and other upgrades that are required for a plane to be used as Air Force One.
-- it would take several years and hundreds of millions of US taxpayer dollars to install those upgrades, a process which amounts to "taking a 747, disassembling it, reassembling it, and then jacking it up to a very high level" according to an aviations-industry expert quoted in the NBC news article I just read.
-- there would be no possibility of that being completed by the end of Trump's current term in office. Whenever he does leave office the plane would be donated to his presidential library (which currently exists only as a placeholder website) after taxpayers had spent many more millions _removing_ the classified-security upgrades.
-- the Air Force is in the midst of a multi-billion dollar contract to upgrade two newer 747s for use as Air Force One/Two (though that work has had delays and those planes may now not be ready before 2029).
"The White House did not respond to NBC's request for comment...."
This is precisely why it's a complete nonstory. Anyone familiar with the mil/demil process and costs knew this couldn't possibly be any sort of bribe, it never even made sense superficially. You can't fly a demil'd plane, it's an inert art installation to be displayed to tourists, like most other demil'd aircraft (some are used for firefighter training or scrapped for parts). To be in service as AF1, it'd be acquired by the wing at JB andrews, militarized, used until Boeing delivers, demilitarized, and towed to the library.
One can't have an informed opinion on the cost of the mil/demil without knowing the cost of the status quo. (747 Mil/Demil Cost + Operating Cost) = (AF1 Operating Cost) is entirely possible.
The diplomatic thing to do when someone gifts you a white elephant, especially when you're doing a "peace in the mideast megatour" is to say "thanks, I'll definitely use this" and then do something else with it.
Mil/demil cost can be whatever they want it to be. The VC-25 is basically an entirely custom plane that happens to look like a 747 (this is why it is so expensive) but you could easily strap a couple of military radios into this one and call it good if Trump really wants to. Demiling basically requires taking the radios out, which is not all that expensive and would leave the plane flyable. There aren't many ex-military modified airliners in commercial service, but that has more to do with the military's tendency to keep stuff in service forever than the inherent problems of making, say, a C-40A ready for airline service. (Also, paperwork, but a low-time 747-8 is a lot more tempting than a 25-year-old 737, even if it's not got a lot of hours by 737 standards.)
You seem to be very certain that the "mil/demil process" will be applied to this airplane. Granted that is the normal procedure when previous Air Force Ones have been retired, but this is a very abnormal administration proposing a very abnormal sourcing of an Air Force One. What is the basis for your apparent certainty that it will be handled in the normal fashion going forward?
If Qatar offers and Trump accepts and swears up and down to every media outlet that he will accept and use the plane for all his official flying for the rest of the term and then keep it for use by his presidential library and that it's an amazing idea for him to accept and it's all very cool, how is it not a bribe?
Even if he can't practically use it in practice during his term, whether he understands that or not (assuming he decides to comply by his security team's advice), wouldn't the proper non-bribe scenario be a very public "Oh you're so generous Qatar, thank you but I couldn't possibly."
No, the standard diplomatic thing would be "thank you, but I can't accept this because as an elected official I am supposed to avoid things that look like bribes." Don't imply that Trump had to accept the bribe to be polite.
Also, while you're right that it's unlikely to be used as Air Force One, I think it's plausible for Trump to say "I can't use it as President? That's fine, it can sit in my presidential library until I leave office and I get it for my personal use."
It would be very unwise to accept it on any terms. I’m sure Trump saw all the gold trim and expressed great pleasure, and seeing as it is a white elephant, as you say, the Qataris were more than happy to offer it to him. It was apparently put on the market in 2020 and nobody bit. Apparently it has only been in the air for 1100 hours so it’s very unused. It’s original cost was 400 million but it seems a regular 747 of its kind would be worth about 50 million in resale. Unclear how much you could get back on the gold plating and trim, etc..
The whole thing is very sordid.
I wish substack had a !remindme equivalent so that in a year if Trump was actually using the plane we could see who was engaging in motivated reasoning and who was being earnest
> You can't fly a demil'd plane, it's an inert art installation to be displayed to tourists
In principle, if we were dealing with a normal administration, I would agree that this is likely a nonstory. But we are not dealing with a normal administration. We are dealing with an admin that has literally run a crypto rug pull, invited a journalist into a signal chat, and turned the white house into a tesla showroom. So unfortunately I don't share your credulity. In the words of a great man, "fool me once, shame on you. Fool me — you can't get fooled again."
I expect that Trump will want to use his pretty new toy basically as soon as he can get his hands on it, past precedent about what you "can" and "cant" do be damned. Who's going to stop him, if he wants it?
(by the way, the goal posts have already massively shifted on this one story. It went from:
- "this is made up" to
- "this is because boeing is behind and the plane isn't coming from the qataris its from a new defense contractor" to
- "its not a gift its just refurbishing a plane that used to belong to the qataris" to
- "its a gift from the qataris but he just wont use it")
This is pretty much what I was thinking as well. (On DSL, I mostly stressed the security risk.) Including the ideal thing to happen: Trump accepts the gift, security quietly wheels it into a hangar, and Trump never gets within a mile of it until maybe after his term is up. Quiet mothballing avoids a small army of snoops getting a look at the USG's mil process, and a small army of actual baddies looking over their shoulders.
Alternately, Trump gets tired of waiting for the official new Air Force One, proclaims he's going to save the taxpayers a yuge amount of money, and orders the United States Air Force to fly him around on the Qatari jet as is - whatever level of communications, security, etc, was enough for a rich Arab prince is good enough for him. As Commander-in-Chief of the US Armed Forces, I'm pretty sure he can do that if he really wants.
Wouldn’t he be abrogating his responsibilities as CoC of our great nation? The nuclear codes and all…
As Paul Botts notes, it is entirely up to POTUS how those responsibilities will be exercised. "I can totally give nuclear launch orders by calling up missile wing commanders via Starlink, reading out the codes on an open channel", would be a mind-numbingly stupid thing for anyone to do, but there's no law saying the President has to make smart decisions and this one is his decision.
I made the point elsewhere that there's also nothing stopping one of his advisors from making the case to him how stupid this would be. I could probably make it, if someone tapped me one morning and said I'd been field-promoted to delivering the PDB.
It's not even like it has to be adversarial. I'm convinced Trump would readily believe someone might try to pull a fast on him in order to get the edge on dealmaking. Heck, I could even make the case in a way that makes him think he thought of it (or at least, make him think I'm assuming he's thought of it).
Somewhere deep inside, don’t you have the feeling that this whole thing is in bad taste? It kind of makes Trump look like repo man.
Yes, that seems to be the case.
Yes it would. And? Who's going to stop him?
The SCOTUS awarded him unconditional blanket immunity for actions that are or could be viewed as part of his "core constitutional responsibilities". Ordering the Air Force to fly him around certainly fits that criteria.
And then he personally sells it for $50 million or more and pockets the money without ever having to pay any of the maintenance or upkeep? That doesn’t seem like a good idea to me.
Q: What do you call minifigures that occur only once, i.e. in a single Lego set?
A: Hapax Lego men.
Take this kudos and also this shower of rotten vegetables for the pun.
The Financial Times published on May 11th an article entitled "Three things we learned about Sam Altman by scoping his kitchen". Altman was a guest for the segment "Lunch with the FT". Analyzing the video made for that segment, Bryce Elder, the author of the article, pounces on the opportunity to satirize the problems that stand out to him and quips that they might be an sign of troubles to come for OpenAI. Looks like a puff piece, feel good article. I was thoroughly amused.
I have some commentary about it in a post I made on LessWrong. If it's OK, I'll publish the link to my LessWrong post. If it's not OK to direct to my post, them I'm happy to mention my thoughts here.
I feel like any journalist who wants to critique someone else's kitchen choices needs to post a full inventory of their own kitchen to let journalists pick over.
This article makes Altman seem a lot more relatable. Picking the wrong olive oil? Yep, that's something I struggle with too, I feel like I never know what oil I should be using. Buying expensive kitchen gadgets that turn out to be disappointing? Guilty.
(Also FT, $2000 is not an expensive coffee machine, sheesh. If I were a billionaire I'd be rocking a La Marzocco or something, not a freaking Breville. Of course I'd also have a full-time barista...)
True, it does make him more relatable - an instance of the Pratfall effect: making a small, harmless mistake can increase the esteem we have for someone, particularly if that person is already perceived as competent or above average.
I'd venture to say that something similar might have happened to Bill Clinton after his affair with Monica Lewinsky became publlic, at least in the eyes of some people.
Oh my, that might be an error, if the article wants to go after Altman (or if anyone reading it wants to go after Altman). Kitchens and UK politicians have not been a happy pairing, just ask Ed "Two Kitchens" Milliband 😁
https://www.bbc.com/news/uk-politics-48268779
https://www.theguardian.com/politics/2015/mar/13/ed-miliband-two-kitchens-use-smaller-one
Now I must try and read that article to find out if Sam only has the one kitchen, or does he only use the smaller of his two/more kitchens? How many ovens? Does he need a cookbook to make toast?
I also remember Elizabeth Warren doing a video in her kitchen where she wanted to share with us all her drinking her OWN beer in her OWN kitchen with her OWN dog and her OWN husband (as distinct from anyone else's husband, but if you're a politician it probably is a good move to emphasise this, given the amount of affairs and sex scandals that hang around them).
https://www.theatlantic.com/entertainment/archive/2019/01/why-elizabeth-warrens-beer-moment-fell-flat/579544/
EDIT: Ah, that article was well-worth it. Sorry Sam, the FT is not impressed by your culinary prowess, maybe you should have consulted one of your tame AIs for advice beforehand!
Oddly enough, due to me watching Youtube cooking channels, I *had* heard about Graza olive oil and the Sizzle/Drizzle/Frizzle variations. And I'm not even a multi-millionaire!
I don't know enough about coffee machines to have an opinion on his model (the fanciest I get is a Bosch Tassimo https://www.bosch-home.ie/en/mkt-product/coffee-machines/tassimo-hot-drinks-machines/tassimo-suny/TAS3102GB). But again, seems the FT isn't impressed.
They sneer at his knife (either really expensive or really cheap, and they intimate it's probably bad quality priced highly, because Altman doesn't know any better except to go by price) and have a *second* go at the bottles of olive oil.
"Maybe it’s useful to know that Altman uses a knife that’s showy but incohesive and wrong for the job; he wastes huge amounts of money on olive oil that he uses recklessly; and he has an automated coffee machine that claims to save labour while doing the exact opposite because it can’t be trusted. His kitchen is a catalogue of inefficiency, incomprehension, and waste. If that’s any indication of how he runs the company, insolvency cannot be considered too unrealistic a threat."
Honestly, Altman would have come out much less scathed had he just gone "Oh, I don't cook, that's what the staff is for" like a proper billionaire and potential emperor of humanity when doing a tour of his kitchen while the cook and scullery maid did the chopping, slicing, frying, etc. before plating up the "pasta-infused garlic" (wait, shouldn't that be the other way round, or are they jeering at the amount of garlic he had on the chopping board?)
The article about Altman definitely packs a sting, that's for sure.
It's not often I get to look down my nose at an oligarch steering the fate of humanity onto the rocks of AI, but the olive oil allowed me to do so! 😀
"We couldn’t find an exact match online. Maybe it’s a one-off piece by an artisan steel forger who shuns tradition. Another possibility is that it’s a Chinese mass-produced blade that’s sold under countless names, usually in sets, often in a fancy presentation box or with fake Damascus patterns etched on the side. There are Sino-Niho-Germanic Frankenknives all over Amazon and AliExpress that look a lot like Altman’s."
I now feel a lot better about my distinctly unhip, purchased piecemeal over the years, some of 'em came with the knife block, kitchen knives! Though I am considering finally breaking down and buying a good knife, but there's a lot of different recommendations out there (usually "pay €€€€€€ for this fancy Japanese knife you will be too intimidated to use at all"). But now I know which to avoid - any knives that look like Sam's knife!
Gentlemen! If you ever somehow find yourselves in the position where you have in some unimaginable fashion totally by accident murdered your wife and buried her body beneath the stairs in your family home, I'm here to tell you that six years later at the murder trial explaining how you put flowers in, and expressing a wish to that it had been roses but you couldn't get them, the grave where you are secretly disposing of the corpse does *not*, in fact, make it all better.
https://www.rte.ie/news/2025/0513/1512580-richard-satchwell-trial/
https://www.irishexaminer.com/news/courtandcrime/arid-41631499.html
"His wife was “a beautiful woman,” he said. But her face was distorted with anger as she lay dead in their home. He wanted to lay roses with his wife’s remains but could not find any so put tulips with her instead, he said."
I'm going to go out on a limb here and say I think he's lying his backside off and that if you accidentally, in self-defence, not at all premeditatedly, killed your spouse you would try calling an ambulance and/or the police instead of going for a stroll around town then back home to dig that grave, then for six years maintain your story that your wife had left you and simply walked out one day and you hoped she would come back. But that's just me, maybe this sounds reasonable to other people.
Scott's back and forth with Curtis Yarvin and the responses to it have made me wonder: What exactly is the existential crisis that many high-profile Trump supporters (especially SV types who all turned from sorta blue to very, very pro Trump almost over night) seem to think would have certainly occurred had Harris won in 2024? I'm leaving aside already die-hard MAGA voters, because it's at least clear to me why Trump losing is an existential crisis to them. I'm talking about people who would have likely been fine had either candidate won, who suddenly became convinced Harris winning would be the end. Not trying to start a political flame war, I'm actually wondering. Many allude to that being some kind of point of no return, existential crisis for the country because *gestures broadly at wokeness*, and I gotta say, I just don't see it to the point that it makes me think I'm missing something significant.
Scott has mentioned this, but woke stuff seemed to be in decline before this Trump term. Joe Biden won 2020 as a straight old white man, and Harris didn't really seem to be a paragon of wokeness herself, probably because the campaign team rightly realized it was going to cost them votes and the best thing to do was moderate. Now, I realize that says nothing of the actual policies her administration would end up implementing, but I haven't seen anyone who thought it would be a crisis moment actually state what policies they were expecting and dreading. From the way I saw some commenters talking about the hypothetical Harris term we would have gotten if we hadn't gotten Trump II, you would think she was going to start rounding up the unwoke and sending them to camps, which (to me) is a completely ridiculous thing to believe.
If I had to guess, her term would probably be much like Biden's. Pretty uneventful, some attempts at progressive "woke" policies that would get shot down by the courts, and some mild lip service for woke causes in a speech every so often. I cannot see the hypothetical Harris term cracking down on speech any more than the median president, implementing martial law, or becoming a dictator. These are all things that, if they happened, would warrant the reaction the people I'm referring to have, but I don't see reason to believe they would.
Personally I'm a centrist that was on the fence before and then ended up voting for trump. What got to me was immigration. From what I see, the left seems dead set on getting as many illegal aliens into our country as possible, and doing everything in their power to keep them here, abusing our welfare system.
I've listened to Musk speak on his idea of the existential crisis of the election, and the idea is something like this. Democrats want to import a ton of illegal aliens into swing states, and push for laws that allow noncitizens to vote (no voter ID). The combination of these two pushes will leave swing states voting democrat forever. Illegal immigrants need the welfare state to live, the welfare state is primarily supported by the democrats, and so these huge numbers of illegals will vote for democrats, locking in swing states. Democrats then get to win every election into perpetuity, which brings with it all the woke stuff, which is catastrophic for the wellbeing of our nation.
Take it or leave it. I find this convincing enough.
> Illegal immigrants need the welfare state to live.
Is that a fact?
If you don't mind me asking, what made this argument click for you in 2024?
I ask because this is not a new argument. This first thing I thought of was Anton's "Flight 93" (1) essay from the first Trump election, where the fundamentals of this argument are already presented:
"Third and most important, the ceaseless importation of Third World foreigners with no tradition of, taste for, or experience in liberty means that the electorate grows more left, more Democratic, less Republican, less republican, and less traditionally American with every cycle. As does, of course, the U.S. population, which only serves to reinforce the two other causes outlined above. This is the core reason why the Left, the Democrats, and the bipartisan junta (categories distinct but very much overlapping) think they are on the cusp of a permanent victory that will forever obviate the need to pretend to respect democratic and constitutional niceties. Because they are."
Were you unaware of this line of argumentation before? Did something happen between 2015 and 2024 to change your mind on this argument? Did Musk give it additional weight?
(1) https://claremontreviewofbooks.com/digital/the-flight-93-election/
Before 2016, there wasn't actually that much difference between Republicans and Democrats on immigration law. Rs might have talked more about it, but there was a general bipartisan understanding that enforcing immigration laws was a basic part of the function of government, and largely acted accordingly. After Trump got elected the first time, liberals negatively polarized themselves into taking "no human is illegal" etc. seriously; under Biden, prosecutions for immigration-related offences cratered to the lowest levels seen in decades.
Personally, illegal immigration is something I don't care much about one way or the other, so this doesn't really matter to me. But if someone considers it to be an issue of supreme importance, it makes sense that they'd take an "anything but the Democrats" stance in 2024, even if they didn't in 2016 or 2020.
I think this is a great example of what I mean when I talk about information diet and epistemology here (https://www.astralcodexten.com/p/open-thread-381/comment/117100524). Elon has fried his brain with ketamine, he may not be the best person to listen to about conspiratorial stuff like this.
But just to respond on the merits:
> Democrats want to import a ton of illegal aliens
You've already walked back on framing this as if this is some intentional scheme on the part of mustaciod democrat villains, which I appreciate. But you go on to say that this is an unintended consequence of Dem policy. Except...even that isn't true. The number of illegal immigrants in the country has remained roughly static at around 10M since 2003. In that time, there have been 7 different presidencies with 4 different presidents. At the end of Biden's term, with all the caterwauling and screaming about immigration we had...fewer illegal immigrants in the country than we had in 2005. So, just empirically, this untethered from reality.
> into swing states
From the data we have, most illegal immigrants live in extremely polarized states in deep blue cities! The states with the largest illegal immigrant populations are California, Texas, Florida, New York, NJ, and Illinois. NONE of those are swing states. They don't even live in swing counties.
> push for laws that allow noncitizens to vote (no voter ID)
As mentioned below, worth mentioning again: there is basically 0 fraudulent voting happening. Trump filed something like 60 cases to try and overturn the election in 2020 after he lost, and they didn't find anything. And now Trump is going after his own appointees because they told the truth about how there is no election fraud: https://www.nbcnews.com/politics/national-security/former-aide-refuted-trumps-false-2020-election-claims-federal-investig-rcna204394
And, again, even if there *was* some small amount of fraud by illegal immigrants, they don't live in swing states or places where those votes matter at all.
> Illegal immigrants need the welfare state to live
Approximately 40% of illegal immigration is visa overstays. Approximately 33% of illegal immigrants are from relatively well off countries. It's true that something like 58% of illegal immigrant households are on some kind of welfare...but 50% of naturalized immigrant households are on some kind of welfare. That delta is only 8%. Which makes this next part...
> and so these huge numbers of illegals will vote for democrats
hilariously, hilariously, wrong.
People on welfare vote for republicans *all the time*. The states that are most dependent on federal aid are all red. I can't find more recent polling data, but in 2012/2013 57% of conservatives were on some kind of welfare. And from 2020-2023, the entire country got some kind of welfare through COVID subsidies. There are tons of people who are on ACA who will happily vote against "Obamacare".
So just to keep a tally here. Your primary concern is voter fraud from illegal immigrants, but:
- most illegal immigrants don't live in a place where their votes would matter
- even if they did, a sizeable % of illegal immigrants don't use welfare
- even if they did, welfare recipients are never single issue voters about welfare
- and, even if they were, there is no evidence that they commit voter fraud
> I find this convincing enough.
If you're serious about this take -- as in, you really are a centrist and this isn't just motivated reasoning -- you should not be trying to approach this from a 'first principles' perspective. Motivations for immigration and, like, people in general are extremely complicated. First principles thinking is useful in the absence of data, but there are countless sources of hard data here that are easily accessible.
If this does not change your mind, what would?
“Illegal immigration was no worse during Bidens term than the past 20 years” is a really wild take
Sorry, do you have anything useful to add? You're welcome to provide "alternate facts" if you want.
Here's pew, showing that the peak of the illegal immigrant population in the US was 2005. https://www.pewresearch.org/short-reads/2024/07/22/what-we-know-about-unauthorized-immigrants-living-in-the-us/
OK, I have graphs too
https://www.pewresearch.org/short-reads/2024/10/01/migrant-encounters-at-u-s-mexico-border-have-fallen-sharply-in-2024/
The level of illegal crossings, you’ll notice, was fairly low when Obama was in office. He deported 3 million illegal immigrants and was known in Mexico as the “deporter in chief.” The left being against deporting illegal immigrants is a fairly new thing.
Sure.
In fact I'll go a step further and say that the left has never been, and even now is not, against deporting illegal immigrants. Given that the context of conversation was explaining that "conspiracy theories are not a good reason to vote for Trump", it seems like we're in agreement. Thanks for helping to make my point!
I'll leave it, this is a conspiracy theory on par with George Soros paying protestors.
George Soros (or anyone) paying protestors shouldn't sound like a conspiracy theory.
Heck, I've been a paid protestor myself, in my student days. Strictly speaking I wasn't paid so much as given a free hotel room and dinner along with a bunch of other folks from the same political group, but when you're a student a night in a fancy hotel sounds just as good as actual pay.
I'm sure this kind of thing happens all the time.
I think the conspiracy theory is that every time you turn on TV and see a large protest, the majority of those people are actors, essentially extras bussed in like a Hollywood set, with no affiliation to the cause and no actual concern for the issue, just mercenaries showing up for a (comically large, if you ask these people to guess) check. You might say, "nobody actually things *that*" but there are people out there who think the parents who go on tv crying after their kids get shot in school shootings are just paid actors, so that's where the bar is.
I incorrectly implied that this is a conscious scheme by the democrats. I think a set of actions taken for different reasons may have unintended outcomes that may be positive for the actor. You are correct, the idea that a cabal of democrats in charge have this in mind is a conspiracy theory. While it is bad Bayesian analysis to give any thought that smells like conspiracy theory a zero percent chance of being correct, I do think it is more likely that this outcome is not intentional.
Okay, well thank you for clarifying.
For your clarified position, I've never been sure how to treat the certainty with which conservatives claim that illegal immigrants use more of the welfare state than they contribute. Its not that I'm certain that they don't, it just that conservatives speak so certainly about this when AFAIK its at least an open question about the long-term costs vs benefits.
The idea that illegal immigrants are voting in elections is obviously unsupported by any data, Republicans have mounted what amounts to a pogrom against supposed illegal voters and have only ever found a handful of people. I guess your fear is that somehow Democrats would reverse this trend in the future?
But even if both those things were true, the idea that illegal immigrants would be out stumping for Democrats just because its in their rational self-interest is the biggest unsupported supposition of all. We know this because the states that benefit most from government welfare the most are some of the reddest states in the country. Rural voters use government aid vastly disproportionately to urban voters and go for Republicans in far, far greater numbers. And even among first generation legal immigrants, we see that the trend towards voting Democrat is dubious at best, because immigrants have a strong cultural bias against a lot of democratic social policies. If anything, if Republicans wanted to sow up every election in the next fifty years, they would be the ones pushing to have a huge influx of uneducated, low-skilled, culturally conservative people voting in every election, because by every metric that is the people most likely to vote Republican. They reason they aren't pushing for that is the same reason those people are voting Democrat, nobody actually cares that much about their self-interest compared to their cultural identity, and Republicans are still pretty hung-up on being the party that opposes an influx of Brown people.
Which is all just to say, if you were being honest and that was really your reason for voting for Trump, it was an extremely poorly reasoned one and you should re-evaluate all of your priors before voting again.
Why not more *legal* immigrants? I mean, if they want more immigrants anyway, why filter for the ones who break the law? Why create a large group of people that can be easily blackmailed by criminal groups? If people already pay money to get illegally into USA, why not let them pay the same money legally?
Policies that make sense, although not each of them for everyone:
* closed borders
* open borders
* select for the smartest / richest / otherwise interesting people
Policy that doesn't make sense:
* select for the people willing to break the law
I absolutely agree with you. It just seems to me that right now the only side interested in some sort of sensible immigration reform is the right. So we end up with closed borders because thats the only good option on offer.
The US still has plenty of legal immigration programs, including ones that do select for smart/interesting people (e.g. O-1) along with people who are just kinda useful (H1-B) along with random people who are just lucky (Green Card Lottery). It's not a perfect or even a great system but it's certainly not "closed borders".
> the only side interested in some sort of sensible immigration reform is the right
Trump killed the bipartisan immigration reform deal.
Well, it wasn’t really bipartisan then was it? Democrats wanted it, Trump and Republicans didn’t, on the grounds that it was less a border security bill and more an amnesty for illegal migrants who had already crossed the border. Anyway, as Trump proved shortly after taking office, America didn’t need a bill to resolve the border issue, it needed a president to enforce the existing rule of law.
I have many friends who were in this position -- they were blue, and turned strongly anti-Kamala. These are smart people who are experts in their fields and who I have a lot of respect for.
A few things have come up repeatedly:
- a belief that the reactions to COVID were unjustified overreaches of government control (mask mandates, lockdowns, etc.)
- a belief that a continuation of Biden era economic and tech policies would be terrible (coupled with the belief that Trump would not implement tariffs). In particular, there is a lot of handwringing about Biden admin attempts to regulate crypto out of existence + their stated intention to regulate AI heavily + they all seem to think that Kamala was going to implement price controls (lol)
- strong disagreement with the incentive structure of wokeness, along with a specific strong aversion to one tenet of wokeness (e.g. one friend is strongly anti-trans, another is strongly anti-affirmative-action, etc.)
I think you are right to be skeptical that together these are not really meaningful signs of the apocalypse. When pushed, my friends will mitigate Trump's actions and play up the hypothetical downsides of a Kamala government; that said, many of those people have quieted significantly after 'liberation day' and find the ICE enforcement pretty bad.
I think you're poking at the wrong thing. You need to go one level up: it's less about their truly held beliefs and more about their information diet and their epistemology. I've written about this before, here: https://theahura.substack.com/p/right-wing-epistemology-and-the-problem
Many of the people in this crew are downstream of independent news sources. They style themselves as independent thinkers. Many are in the startup world, and love people like Marc Andreesen or Tyler Cowen. They are trained to be contrarian. But as a result their epistemic hygiene is...lacking. They are bombarded with information about woke malfeasance (Chinese Robber style) and end up believing that it is a much bigger problem than it really is. LibsOfTikTok and its ilk (which are descended from TumblrInAction and earlier kiwifarms) are extremely damaging to our nation's information landscape. And confirmation bias keeps them in the hole.
Many of these people know nothing about the eastman memos because it never showed up in their feeds. But they know all about Hunter Biden, because that showed up in their feeds constantly. If I had their information diet, I too would think the world would end if Kamala was elected. Unfortunately it is very difficult to spot a problem in your epistemology -- a bit like finding a bug in your bug report form.
As an outside observer, I think it was the covering up around Biden's mental state. Either Harris knew and was complicit in "it's all fine, he's fine, sharp as a tack" until they had to shiv him in the back and suddenly she was the bestest ever possible candidate, or she knew nothing and was pushed aside and had no input at all into the four years of the administration.
Neither possibility sounds good for someone looking to be voted into a globally influential office during a time of tension with Russia over Ukraine and what is going in Gaza and potential trade wars with China/disputes over Taiwan, etc. A nonentity who could have been replaced by a rubber duck during her time in office, or someone willing to go along to get along in order not to rock the boat purely for the sake of personal ambition? Which do you prefer?
That is setting aside any consideration of Trump, qua Trump, if you're looking at the Democratic candidate and being told "you can have any choice that you like - so long as it's Harris".
>From the way I saw some commenters talking about the hypothetical Harris term we would have gotten if we hadn't gotten Trump II, you would think she was going to start rounding up the unwoke and sending them to camps, which (to me) is a completely ridiculous thing to believe.
Is this any different to any other candidate in any other cycle? I remember when Mitt Romney was going to reinstate slavery.
Well, here's what Bill Ackman wrote in October 2024:
https://x.com/BillAckman/status/1844802469680873747?lang=en
Personally, I couldn't get past point 2 which was laughably the wrong way around, based on party platforms and reasonable economic forecasting at the time. I don't know if Ackman believed what he wrote or just thought it sounded good and covered his real reasons.
As to what those real reasons could have been: based on the US economy going into the election, generic Republican had a strong advantage over generic Democrat. Add in the disarray on the Democratic side and there was a very good chance Trump would win again. Further factor in the possibility that he would reward loyalty/punish disloyalty to a greater degree than Harris and it is obvious that endorsing Trump was the better trade.
Yes, that list really reads as largely rationalization. No 3 is particularly laughable, given that the Afghanistan withdrawal was negotiated by Trump (and by the rather predictable Trump cancellation of resettlement of Afghans who aided US forces, and the even more predictable suspension of all refugee resettlement. And #14 looks even worse in retrospect.
Are we worried about the hypothetical 5 Trillion tax cut in the US House bill? I feel like this is a big deal.
Not to put too fine a point on it: Nobody actually cares about balancing the budget. Literally not a single person gives a shit. People that say they do are pretending.
Democrats don't give a shit but occasionally balance the budget anyways as part of the neoliberal project almost by accident; democratic voters don't care or notice.
Republicans pretend to give a shit, then slash income and massively expand expenditures while failing to cut entitlements every time. People vote for republicans to "balance the budget", then theatrically shake their fists at the sky when it doesn't happen and vote republican again next election, because they don't actually care.
What about libertarians, you might ask? What about them. We as a society can do nothing but stand in silent pity and imagine that Sisyphus is happy.
I actually do care about balancing the budget, at least to the extent of keeping the debt-to-GDP ratio from growing any further. I am not pretending. If that was just hyperbole on your point, that's not helpful to the sort of debate we usually try to have here. If you actually believed it, you were wrong and I think foolishly wrong.
You are a kind of probabilistic cloud of possible identities and beliefs to me, as a random name on a discussion thread of a blog. I can't know anything about you but population data and what you say, which doesn't have to be true.
From these datum, I can only draw tentative conclusions until you reveal information about yourself.
For example, you say that you care about balancing the budget. I believe that this might be true, but can't draw a conclusion about it until you narrow down your identity on the graph.
For example: did you vote Trump once? Twice? Who did you pick down ballot?
If you voted for Trump/Trump, I can state you don't care about balancing the budget based on your actions. If you voted Trump/Biden, I can say that you might care. If you voted Clinton/Trump, I can say that you definitively want the budget to be even less balanced.
In any case: If you truly care about balancing the budget, you represent a segment of the population that has no influence on politics except as a marketing tool for the political tendency that wants to make the budget as unbalanced as possible, ie, the non-neocon conservatives who are currently ascendant in the ranks of the party and the state, and who are so committed to unbalanced budgets they got the STATE WITH THE WORLD RESERVE CURRENCY downgraded by Moody's without a government shutdown immanent.
Thus, the "Nobody" in "Nobody cares" can be more accurately stated as: "Nobody who matters has a balanced budget as their first priority vs. IE ending DEI or wokeness or reducing taxes on top earners or some such".
We'll see if that changed, but I think most R's in government who are against the current bill actually want MORE tax cuts and pork, not less.
"From these datum, I can only draw tentative conclusions until you reveal information about yourself."
Says the guy posting anonymously, to the man who has been posting under his own name for thirty years including more than a decade here and on SSC. If you want to know whether e.g. I am a Trump voter, it's not hard to figure out. If you want to posture about how people on the internet are just fuzzy clouds of probability (about whom you are nonetheless willing to make absolute and unconditional pronouncements), go look in a mirror.
I don't actually look at the names of people I reply to on the internet, and if I did I wouldn't research them because this is all shouting into the void in any case.
I reply to these posts for my own entertainment/to refine my own ideas under pressure/to practice my short form writing now that I no longer have to do a quick 200 word explanation of the effects of "A Silent Spring" on the ecological movement, etc etc.
I still haven't read your handle, and I still won't click on it. If you would like to collapse your wave function with some observations, I will go off of those; otherwise I won't.
If you want to address the meat of the comment, IE that there is no political tendency in the US that will sacrifice their Tendies for a balanced budget unless you are willing to mount a doomed struggle ala the Rands, say that and I will respond to that.
Or don't, and I will treat it as a creative writing challenge to whip out a couple lines without going back and editing it, as I find that muy ability to write things correctly on the first go is getting degraded by dictating most of my short messages to the robot that lives in my phone on the go instead of sitting down and hammering them out caveman style.
If you want a nice book about math for kids about 10 years old,
https://www.lesswrong.com/posts/CbxP8x6pk7vTMNEgW/book-translation-three-days-in-dwarfland
FWIW, I contacted a group that publish math books for kids that I thought might be interested. It turns out that they already investigated translating and publishing Levshin's books, but the hurdles to officially publish were insurmountable.
In case you want other books, there are now a lot of great ones for kids, including some that have something of the Dwarfland flavor. I am trying to revive interest in the CSMP storybooks (https://stern.buffalostate.edu/csmpprogram/Storybooks/bygrade.html) which are generally tuned for younger kids, but are also funny and emotionally touching. Natural Math (https://naturalmath.com/goods/) has many good choices. Finally, for now, anything by Raymond Smullyan is bound to be excellent, but probably also quite challenging.
> the hurdles to officially publish were insurmountable.
Please tell more.
Thank you for the links, they seem interesting. I wish we had funny books for kids on every topic, and a library full of such books at every school. (Even better, a torrent of such books, in epub format, translated to every language.) It is such a waste that we don't have that already. Kids are curious about things, they just need it to be accessible for their age.
Also, movies: https://en.wikipedia.org/wiki/Once_Upon_a_Time..._Life
I didn't really get a lot more detail about the obstacles, but it was bureaucracy in Russia with a lot of steps to push against people who don't want to help and couldn't care less about the project ending successfully.
Thank you!
New post up in which I argue that the left and the right are not all hypocrites about the rule of law, even though they frequently seem to display behavior that suggests they are. What we are is unsystematic about our reasons thinking discretionary violations of the law is good. https://hiphination.substack.com/p/are-we-all-hypocrites-about-the-rule
Honestly, your post doesn't seem to be about the rule of law at all. It is about the exercise of prosecutorial discretion, and, yes, people disagree about how that discretion should be exercised.
And when you say, "What we are is unsystematic about our reasons thinking discretionary violations of the law is good," nothing in your post relates to that, because the exercise of prosecutorial discretion is not illegal unless done for an improper purpose. "The rigors of the penal system are also mitigated by the responsible exercise of prosecutorial discretion." Cheney v. United States Dist. Court for DC, 542 US 367 (2004).
I was thinking about how AI "thinks", and it is noticeably different from the way humans think. For example, I have a plant identification application on my phone, which I use by taking pictures of a plant, identifying the part (bark, leaves, fruit, etc.) and it provides a guess as to what plant it is.
I imagined someone showing another person various plants, and identifying them to the other person, especially to a NON-plant person. Something like this:
Teacher: See this leaf shape? You can tell it's a maple tree from the distinctive shape of the leaf.
Student: OK, yes, it's all pointy.
Teacher: Now this leaf is also from a maple tree, but a different type. See how the leaf is broader than the first one?
Student: I guess it's a bit thicker.
Teacher: This is bark from the first tree, and this is bark from the second tree.
Student: The bark looks the same. Isn't bark just bark?
Teacher: No, this first one is much more craggy. This second one has straighter lines to it.
Student: If you say so. They both look like bark to me.
It seems to me that AI is BETTER at finding differences and connections than the average person, for a given individual is unlikely to be specialized enough to care about the differences in most things. Yet something is still missing, so that any given person is probably better than AI at finding differences in SOMETHING, but not everything.
I'm reminded of looking at the AI generated art. I'm no art expert, so it was no surprise I didn't do that well in differentiating between AI art and human art. But experts WERE significantly better at it.
It may be that AI is going to be generally better than humans, but not better than top humans. At least, not until we have another AI breakthrough.
Hey all -- Andrey Fradkin and I are launching a podcast on AI and economics. We're 15 episodes in, I think y'all would like it.
We're launching today, and just had an endorsement from Tyler Cowen:
https://marginalrevolution.com/marginalrevolution/2025/05/new-ai-and-economics-podcast.html#comment-160913290
Check it out!
You are playing a cooperative variant of Rock-Paper-Scissors where the objective is to tie rather than win. You and your partner don't know one another and don't have an opportunity to coordinate in advance nor by side channels during the game.
Consider the following variants:
1. Only one round is played.
2. The game is iterated across, say, nine rounds. You play several rounds with the same partner with the objective of maximizing the number of rounds that end in ties.
3. Also iterated across nine rounds, but your objective is for you and your partner to have won equal numbers of rounds over the course of the series. Does it make a difference if fulfilling the objective is an all-or-nothing thing or if you get partial credit for, say, a 5-4 final score?
4. Instead of playing Rock-Paper-Scissors, you are playing Rock-Paper-Scissors-Lizard-Spock. The rules are:
- Rock smashes Lizard
- Lizard eats Paper
- Scissors decapitate Lizard
- Spock vaporizes Rock
- Paper disproves Spock
- Spock disassembles Scissors
- Lizard poisons Spock
What is your strategy in each case?
1. Rock
2. First round Rock. Remaining rounds, if you tied last round repeat. If you didnt tie last round, theres 2 options: a) pick the one neither of you picked b) 50% own last play 50% partners last play. a) is theoretically the best joint strategy conditional on not agreeing on the schelling point, but never adjusts to constant strategies. b) does adjust, and 50% is the number that most cooperates with itself. Which one I pick depends on how smart I think the other guy is and thinks I am.
3. If youre currently balanced, same as 2. If youre not, repeat with 50% and switch to what would decrease the count if the other repeats with 50%. If you decreased the count last round, you both repeat. Partial credit doesnt matter.
4.1 same as 1.
4.2 The extension of 2a) is determined by putting the options into a pentagon. Then the three unpicked options are either adjacent, in which case you pick the middle one, or two are adjacent and one is separate, in which case you pick the separate one. (Or, always the apex of the isosceles triangle) 2b) is unchanged. Intelligence requirement for a) is slightly higher. I wonder how much difference it makes if youve played 2 before.
4.3 If youre currently balanced, same as 4.2. If youre not, then a) again using the pentagon, to each adjacent pair corresponds exactly one non-adjacent pair that is parallel, and vice versa. You both play the element of the pair corresponding to the last round that makes you win/lose as you should (following this strategy, its impossible to be unbalanced and tied last round, or to be unbalanced by more than one round). 4.3b) same as 3, except because you have two picks you can switch to, you pick the one that would be consistent with 4.3.a) if the last round was not tied, and the one thats not adjacent to the last round if you tied. You use either the a) or b) options consistently.
Question 2 is called Rendezvous Problem or also Mozart Cafe Problem in mathematics, introduced by Alpern in 1976. A common formulation is:
You and a friend have agreed to meet in the Mozart Cafe at noon in Vienna. But arriving there, you find out that there is not one, but n cafes called Mozart Cafe. You and your friend know each other well enough that both of you will go to some Mozart Cafe every day at noon and hope that you both choose the same one.
For n=3 this is the same as your question 2, because after the first tie, both players can continue playing this symbol again. So the question is just how many rounds you need in expectation to reach the first tie, and which strategy achieves this optimum. (Except that you truncate after 9 rounds, which will alter the answer very slightly.)
The question is surprisingly difficult. For n=3 it was only solved in 2012. The answer is 8/3 rounds in expectation.
https://www.cambridge.org/core/journals/journal-of-applied-probability/article/abs/rendezvous-problem-on-discrete-locations/A8E442834F0F4E2F9B0F3523C273C799
The problem is also surprisingly nuanced. For example, if both players can prepare for the situation, without knowing anything about the cafes (in a fictive world with completely symmetrical situation between the cafes), then they can do better by deciding that one player stays put (always goes to the same cafe), while the other tries them out one after the other. But this only works if they can break symmetry between the players. Likewise, if there are other symmetry-breakers (e.g., a river flows through the city and divides the cafes into two subsets) then one can improve the 8/3.
ADDED: wikipedia article at https://en.wikipedia.org/wiki/Rendezvous_problem
Your partner’s choice being public information makes it easier. As pointed out by Schneeaffe, picking the option neither of you chose in round 1 always results in a tie in round 2. (And of course the case where you both chose the same option in round 1 is trivial.)
Oh, I was stupid, you are right.
Variant 1, 2, 4: Play rock every round, let my opponent figure it out.
3 might require thinking, but I might still default to the same strategy.
I would go for rock as well. I think people are expecting their opponent to go on the attack (use scissors) first.
Variant 2: Your opponent threw scissors in the first three rounds while you threw rocks. Still throwing rock in round 4?
Yeah, I don't see a better strategy. Rock is the Schelling point. Variant 3 would involve both people randomly playing scissors until one person loses, and then you both play rock until the final round where you do the opposite match-up to the round that had the win/loss.
Why is rock the Schelling point instead of paper or scissors?
Good ol' Rock.
Nothing beats Rock.
https://www.youtube.com/watch?v=b0SoKWLkmLU&t=8s
I think the cultural prominence of this meme (plus the fact that rock comes first in the name) is enough to tip the choice in its favor.
The latter for me as well. Rock is first in the name of the game itself.
Also rock is the lazy man's choice. Don't need to move my fingers at all.
That's actually the first thing I was wondering when I came up with this. Rock was also my instinct for Schelling point and I wanted to know how universal that was.
I don't know, but it was also the one I thought of. Maybe because it's the first one in the list? Or the 'simplest' shape?
No need for random scissors; if you both win zero rounds, conditions are fulfilled.
If your opponent is also throwing rock every time, then yes. But if your opponent throws scissors or paper in round 1, then that would require the opposite result in another round to balance it out.
Oh right. Yup, rock every time it is.
3. I do not believe he is a "libtard and a coward". If nothing else, he's got the US conservative mindset of what Hanania described as "overwhelmingly motivated by cruelty and hate" when it comes to immigration and immigrants, as part of the broader "suspicion of outsiders" that is the most common attribute of conservatives. It seethes through every point of his reply chain in the click-through link.
In any case, I agree with you on Trump - mostly. Pretty dumb, but I wouldn't underrate just sheer corruption in what the Trump folks do as well (some of them are very likely trading on the impact of his pronouncements and reap rewards every time he walks a tariff back or such).
I'm laughing at this news story from my country today:
https://www.rte.ie/news/ireland/2025/0512/1512488-ireland-us-education/
https://www.irishtimes.com/politics/2025/05/13/ireland-seeks-to-poach-us-academics-with-new-funding-plan-and-talent-scouts/
First, are there top Top Men fleeing from US universities? Second, if I'm a Top Person fleeing the US, I think I'm more likely to head to the UK or to continental Europe than Ireland (unless I'm linking up with an American multi-national already established here).
Third, if the US academics are going to be paid higher salaries, this will kick off unrest and protests from Irish academics about two-tier systems and unequal wages and court cases about discrimination.
Fourth, what is viewed as a 'high salary' by Irish terms may (or may not) be buttons by US standards. If they are really looking for Big Names, those people with established reputations and careers are going to have expectations around pay and conditions. How long will the Exchequer fund the higher salaries? Are we really going to poach Professor Bigname from Harvard?
Now, to be fair, I have no idea what the average salary of a Harvard professor is, or the average salary of an Irish academic, and how one stacks up against the other. Unreliable online searching tells me Harvard ranges from $165-282k which is around €148-254k, and average Irish academic salary is €153k. So depending on how Big a Bigname they are going after, the pay could indeed be competitive. But I think that if you really want to advance your academic career, you're going to Europe or back to the US, not staying in Ireland.
I think there might be some chance of coaxing bright young academics who are research students or junior faculty from non-EU countries, but I do wonder how many Harvard exiles are going to want to work at Munster Technological University (formerly Institutes of Technology, formerly Regional Technical Colleges) for one?
ireland might be popular because irish-americans can fetishize their roots some.
I get what you’re saying here, but Ireland might make some sense if you’re planning on some major industry collaboration with some company that is there for *cough* tax reasons.
I think that is what our government is hoping for (Professor Bigname has a hot new idea, has lined up funding from American Megacorp, and all that is needed is to roll it out in Ireland where the new factory will be located), but how that would shake out I don't know.
Like in the Computer Security business, Michael Purser was at a trinity College Dublin, etc.
Also, if you’re a pessimist about current political situation, you might guess that the Irish government is less likely to put you a death camp, and this overrides the salary concern.
I know Ireland probably seems boring to you, but when I was living in Michigan I spent a week in Ireland and it was f***ing amazing. It's so mind-bogglingly gorgeous and everyone has cool accents and everywhere I went the food was delicious.
If one day the President cracks down on US-based software engineering and I have to flee to Dublin, I will be pretty chill about it.
Imagine an American university president in 1930 saying “who would want to flee the University of Vienna to come to UCLA or university of Minnesota?” Turns out, a lot of people did, and it put these places on the map.
It would make a great remake of the Quiet Man, though!
Damn it, now I want another one of those terrible Paddywhackery movies about an American moves to Ireland, only set in Dublin where our Ivy League Professor of Big Numbers moves to Trinity and it's full of spud-in-the-mouth Paddies and horse Protestants.
The cognitive dissonance between "this is not the hell what Dublin in 2025 is like" and the fakery of "Sure aren't the Cliffs of Moher and Skelling Michael just on the doorstep of dear old dirty Dublin?" would be both enraging and highly entertaining.
They would have a lot to live (or down) to in the tradition of "what the hell is wrong with the Yanks?" movie-making about Ireland:
https://www.vulture.com/2020/12/why-the-irish-love-making-fun-of-wild-mountain-thyme.html
My personal favourite in that genre is "Blown Away" where I saw it in the local cinema and at the flat pint scene every single person, including myself, groaned out loud. It's a fleeting scene where Jeff Bridges and his father, Lloyd Bridges, are playing an Irish-American father and son (edit: sorry, having read the synopsis, they are both Irish-American cops) having a drink together and they both light up with anticipation about the lovely pints they are about to consume. Except the pints are flat, which would not cause a genuine Irish pint drinker to be delighted if those were put in front of them to drink.
https://en.wikipedia.org/wiki/Blown_Away_(1994_film)
My Irish wife always complains about the geography of Ireland being ignored in movies, but that is true of everywhere. People stroll impossible distances in New York, leave London at noon to be in Scotland by the afternoon, go east to go west, and take a one way the wrong direction. This happens all the time and only locals know it v
Yeah - here in the UK perhaps the most famous one is Robin Hood: Prince of Thieves showing our heroes traveling from Dover to Nottingham via Hadrian's Wall. And the (much, much better) American Werewolf in London appears to show a man waking up in a London hospital after being attacked on the Yorkshire Moors (which you could justify by saying they took him down there to get access to specialist equipment, but the movie doesn't).
It didn’t seem like it took Bond very long to dive M to Scotland with his Aston Martin DB5.
As someone who moved from a US University to an Irish one, the pay is competitive - most places in the US are not paying as much as Harvard. On the other hand, there is a major funding squeeze on Irish Universities. The 2009 austerity measures seem still to be in place - permanent staff do well, but it's difficult to get new positions approved (I'm currently in a position which will end in the Summer, mostly due to inertia and bureaucracy). It's infuriating that permanent posts for people on renewing one year contracts will likely be displaced by this initiative, which probably isn't going to attract top talent. The talk of sending out talent scouts to the US strongly suggests a closed nomination process where the universities themselves will have limited say in who lands where - it seems unlikely to me to work well.
Yeah that's the sticking point I see here - where are all the hundreds of millions to be poured into research going to come from? I can see from our government's point of view that they're hoping "smart/big name academics come here, already have idea that can be spun off into profitable commercial product, we fund start up, profit!" but they are not really contemplating "smart/big name academics come here, we have to match US standards of pay and conditions, plus match US levels of funding for 10+ years as they explore 'maybe this will result in commercial product'".
The end goal will be an increase in successful funding applications to European programmes. Ireland doesn't perform well on this front - this has been directly linked to longterm underinvestment in research here. As a personal anecdote, I finished my PhD in 2011, and applied for funding in 2012; the application was returned unassisted, because they weren't funding my area that year. Things have improved, but only slightly.
Whether it will work remains to be seen. It will certainly be the case that the majority of those employed will bring in no external funding at all. Success in these schemes typically runs at around 10%, with a strong bias toward funding applicants with a history of success already. I imagine the most likely outcome will be a small number of large grants being awarded to researchers in this scheme, not nearly enough to cover the costs of the scheme.
" I imagine the most likely outcome will be a small number of large grants being awarded to researchers in this scheme, not nearly enough to cover the costs of the scheme."
I think that is the most likely outcome. I imagine the government expects "big name academics and/or whizz kid junior academics with connections who can draw funding from American multinationals with them to spin off commercial applications of their research" so that the scheme will pay for itself/be a job creation scheme, but how that translates into reality is a different matter.
There's already far far more good researchers than research jobs on both sides of the Atlantic. Maybe not "great" researchers, but great researchers these days are characterised more by their ability to write grant applications, not to actually do work.
If you want to recruit top talent from US universities then you not only need to provide them US-level salaries but also US-level funding for their research. Nobody on that side of the Atlantic is going to be willing to do that. All these numbers are routinely two or three or four times higher in the US than in Europe.
Well, you see, the US has decided to cut their funding to the bone. So really, Europe doesn't need to do anything. The US will do all the self-sabotage Europe needs to benefit.
Are academic salaries decreasing threefold or fourfold? Genuine question.
And if they really are three or four times higher than in Europe, then it's not actually clear if it's higher expected to value to stay in the US, get nothing for four years and then return to normal, or to get European normal per year during that time. And that's assuming moving back to your old job is trivial.
Fourfold differences are that significant.
They decrease to zero if you get fired because the university no longer has the money to employ you. And most academics getting this kind of salary are the ones that actually run the labs, and if there are no grants available they will not hire the postdocs and PhDs that do the research (and are generally far worse paid).
>First, are there top Top Men fleeing from US universities?
Other Euro countries seem to believe so, too:
https://www.dw.com/en/dear-us-researchers-welcome-to-germany/a-72058292
> There are no reliable figures yet on US researchers interested in moving to Europe. But 75 percent of US researchers who took part in a Nature survey are thinking about leaving the US. Of the 1,600 participants, more than 1,200 scientists are considering moving to Canada or Europe. The trend was particularly pronounced among researchers who are still at the beginning of their careers. Of the 690 postgraduate researchers who responded, 548 were considering leaving; 255 of 340 PhD students said the same.
> Leading German research institutions are receiving more applications from the US, including some top researchers they would love to attract.
> In early February 2025, the president of the Max Planck Society, Patrick Cramer, reported that applications from the US had at least doubled, and in some cases, even tripled.
PhD students and postgraduate researchers are always considering moving, because permanent jobs are very hard to get even with a stellar record and no geographical constraints. Those with geographical constraints are almost guaranteed to not find a job and have to leave the field.
I imagine that if you're a top tier US academic, you'll be looking at the likes of top tier European universities, not Ireland. This ranking site has Trinity at no. 26 which is a very respectable showing, but the other Irish universities are way down the rankings:
https://www.topuniversities.com/europe-university-rankings?sort_by=rank&order_by=asc
That may well be the case, but I was answering your first question - "Top Men" of Science indeed seem to be leaving the US at a greatly increased rate, looking for greener pastures. Also, money (as in US vs. non-US) is important, yes, but academic freedom, let alone personal freedom (as in "I don't want to be deported to a Venezuelan mega prison without recourse") should not be discounted out of hand as a potential contributing factor.
Suppose a time traveler from the year 1900 materialized in the modern-day US. I would expect him to be arrested in fairly short order for innumerable reasons, but what would happen to him after that? He has no documentation and, as far as the government is concerned, doesn't exist. Does he get deported? To where?
> him to be arrested in fairly short order for innumerable reasons
So innumerable you couldn’t mention one. Although zero is a number I suppose.
Correct! As the reasons are innumerable, it is quite impossible for me to name one. You see, "one," or "1," is what mathematicians call a "number."
Ok, Imagine a current living person shows up somewhere, with no documentation. Looks and sounds American.
What would happen to him?
How different this is from most Homeless people?
Most homeless people I've encountered are smelly, dressed for sleeping outside in clothes that look like they haven't changed, and appear mentally ill or suffering some sort of drug addiction. My central example of a time traveler from 1900 would not have these properties.
Interestingly, I think what we care about here is really "appears homeless upon a two-second inspection", as opposed to "actually homeless". Someone who appears to have showered recently, is wearing clean clothes, lucid, and not casting about frantically for a bottle or needle, is not going to strike me as homeless unless it comes up in conversation, and if it does, they're probably going to strike me as "temporarily embarrassed" and might indeed be treated similarly to an actual time traveler from 1900.
I don't think anyone is thinking about deporting them though, documented or not.
I see a whimsical Jimmy Stewart movie in this. But we somehow have to bring him from. . . the past.
Are there any cool diseases we could extract from him? Like small pox?
Because of the "to where" question, I think the answer has to be that no, he doesn't get deported. If the Trump admin makes an agreement with a third party country to just take anyone not in the US legally, then that would become a possibility. It seems that is not currently in place and has not been previously. Third party countries have very little reason to want to take on random people being expelled from another country.
If he makes a nuisance of himself, there is a good chance he spends time in jail or similar, but generally there wouldn't be grounds to hold him long and a bare claim of US birth might be enough to preclude attempts at sending him out of the country.
Even assuming they don't accept his claim of time traveling, what do you do with a guy who is clearly spouting nonsense but seems otherwise mentally whole and responsible? You arrest him if/when he commits crimes, but you otherwise let him go free because it's too much of a hassle to deal with him.
If he makes enough noise about this and gets broader attention, there must ways to ascertain that he is indeed a time traveler. Such as measure the concentration of microplastics in his blood. Something. There must be some biomarkers that would unequivocally support or disprove time traveling theory
Lack of PFAS in his blood (if he hurries).
Bone assay to show he lived his life before atmospheric nuclear testing?
Maybe, but he's hampered for a lot of reasons by coming from the past instead of the future. Anything that modern man could verify historically is something he also could have looked up. If modern man cannot verify the information, then they also can't tell the guy didn't just make it up. Someone from the future may be able to predict something to verify his knowledge. There's also the question of how time travel was invented in the past but we never heard about it and no one reinvented it since, even with a supposedly better understanding of time since then.
I don't know the science behind it, so maybe there are means to prove certain age bands. Another thread someone mentioned being able to prove post-1955 by radiation in teeth. No idea if that's possible. Obviously that wouldn't help him very much if he was coming from 1960 instead of 1900, but perhaps there are ages that it would be clearer. I fear that many/most forensic approaches would be hard to falsify. Maybe he has no microplastics because he's from the past, or maybe there are dozens of other explanations. I strongly suspect that we would be very much more likely to come up with such an alternate explanation than to believe in time travel.
>There's also the question of how time travel was invented in the past but we never heard about it and no one reinvented it since, even with a supposedly better understanding of time since then.
They tested it on this guy, and he exploded in a ball of fire. They concluded it was a flawed design, and also were arrested for blowing the guy up.
This is totally tangential to your question, but the scenario reminds me of the Norwegian series "Beforeigners", in which some sort of time portal has dumped crowds of Stone Age, Viking, and Victorian time travellers in Oslo. The somewhat clever twist to the setup is that 'beforeigners' get the same treatment as illegal immigrants, with the non-Victorians mostly living in ghettos where their 'beforeign' ways make them subject to xenophobia - despite them not actually being from somewhere else, just 'somewhen' else. The Victorians integrate a bit better, given their similar culture.
It's Nordic, so the main story involves a pair of cops (one modern, one a Viking) trying to solve a series of murders.
And I'm reminded of a satirical novel, _Look Who's Back_. Hitler appears in modern Germany published in 2012. Hitler appears in 2011. He's in a uniform, but it happens to be one without swastikas.
This is not a science fiction novel-- there's no interest in how it happens. Hitler struggles to understand enough of the world he's in, and is picked up by the media as a brilliant Hitler impersonator.
When I read it about 10 years ago, I thought it was a lot of fun. Maybe not so much now.
It was cool that Hitler would say "Jews are not a laughing matter". He meant it one way and his audience heard it differently.
There's a well-made German-language film adaptation of Look Who's Back which came out in 2015. I watched it a few years ago when it was on Netflix with subtitles. I didn't know there was a book; I'll have to check that out.
I remember the "Jews are not a laughing matter" line being in the movie, too. IIRC, it happened when he was prepping for the first episode of his TV show, and a network executive (or maybe the show's producer?) pulls him aside and asks him (awkwardly and somewhat indirectly) not to make jokes about Jews, which Hitler solemnly agrees to.
There are several scenes in the movie where Hitler is traveling around Germany in uniform and striking up conversations with random people he meets who think he's an impersonator. These scenes hit very differently in hindsight when I found out later that they were real Borat-style unscripted street interviews.
The best part of the movie (scripted, however) is when he meets the leadership of the modern heirs to the Nazi Party. They're clearly very uncomfortable with his presence and he rips them a new one for their worthless, imbecilic "management of the National cause".
If he's a smart time traveler, he'd show up with some kind of proof to show that he's actually a 19th century person and not just a crazy guy pretending to be one. Drop a time-capsule somewhere or do the Marty McFly thing on a sealed letter from Back to the Future II.
I don't think he gets deported unless he claims to be from a particular country that still exists in some fashion, but he gets treated like a crazy guy unless he does the above.
The criminal justice system processes plenty of people who won't give their legal names/records. For a minor offense they'll just take down whatever information is given, process him and then release him. This is the procedure for 'nuisance' crimes like trying to sleep on park bench.
For a more serious offense, he'll stay in jail and they'll spend more time trying to establish his legal identity and then shrug and give up when they can't figure it out. His lack of legal identity will make getting bail impossible but not otherwise affect proceedings much.
Legally speaking they can't deport him without figuring out his native country (and getting the country to accept). However ICE is often cavalier about the legal standards they're supposed to meet. Since he will not be able to provide proof of citizenship while he shouldn't be deported, if he is deported he's going to be pretty out of luck trying to get back in.
https://en.m.wikipedia.org/wiki/Zadvydas_v._Davis
https://en.wikipedia.org/wiki/Lyttle_v._United_States
In the long term the problems he'll face will be pretty much identical to most illegal immigrants, with the same solutions (working in cash or getting a false identity).
Now I'm imagining them deporting him back to the 1900's, and then a judge ordering that they've got to try to bring him back.
But he is back! Not the government's fault if he happens to be six feet under now...
That would be a good premise.
There's an Alfred Bester story about people in the future being deported to the past. A lot of homeless people are actually forced time travelers who can't manage.
In the 2013 movie Only Lovers Left Alive, some several hundred year old vampires travel between Tangiers and Detroit by taking nighttime flights under false identification. It struck me that by 2013, the idea that you could take a commercial flight with false identification was already pretty totally dead, but it made me realize that even just a few years before then, it might have still been alive.
Lots of 20th century media involves the idea of running away and starting a new life somewhere else under a false identity. How much is that still possible if you don’t try to do anything like fly or vote?
I'm spacing on the name, but someone in trouble for crimes against humanity(?) went on the run and hid for years (decades?). He ran some sort of new age-related business when he was hiding.
If you don't try to do something that requires someone to ID you - vote, get a DL, buy beer - then how often do people really get asked for ID? At that point the only required uses I can see would be if you get stopped by the police, which is not so infrequent as to ignore, but I haven't been asked for my ID by the police for ~17 years?
That said, getting a job would perhaps be required for many people and is supposed to require some pretty clear ID. That should be very hard to fake and would end the experiment right away if all employers were really asking like they should be. I don't have much experience with it, but it seems obvious that a number of employers aren't even trying to get ID from employees. Less so if you're doing odd jobs or otherwise self employed and get paid in cash. There's plenty of businesses that prefer to get paid in cash for labor so they can decide how much their taxable income is.
Buying or legally renting housing should also require ID, but there's lots of ways around that as well.
The ID itself need not be false, but the documentation supporting it (birth certificates, etc) would be much easier to fake.
1900 isn't so far back as to be beyond historical recovery, he might be able to prove his identity through census records and town archives and such.
He can prove that the person he claims to be existed and that he knows certain biographical details, but proving that he is that same person is a much taller order.
Fingerprints would probably be the best evidence if you could compare a set of pre 1900 records to current prints. I don't know when fingerprints started getting taken and kept for official records, but my guess is 1900 is far too early.
"I would expect him to be arrested in fairly short order for innumerable reasons"
This doesn't seem obvious at all. I'm pretty pessimistic about the authoritarian nature of the U.S. these days, but unless it's more quickly and more thoroughly than I've heard, there still aren't watchful cops lurking behind every bush, looking for people to snatch up. A particularly oblivious, foolish or belligerent time traveler might quickly attract the sort of attention that would lead people calling the cops. But merely being dazed and bewildered isn't likely to get you there. Having no ID or job history and thus no prospects for legitimate work is definitely a potential problem, but that's a medium-term problem not a short-term problem, and one that plenty of real-life people successfully navigate without running afoul of the authorities.
To address the actual question: now that you bring it up, it seems like there has to be a procedure for this, doesn't there? Not the time traveler part, mind you, but the "personal with no official, legal existence" part. In a population of 300 million people, it has to happen from time to time that the police arrest somebody whose records simply don't prove findable. But I have no idea what that procedure is.
Some more fictional evidence for a hypothetic situation: In the 2001 movie Kate & Leopold, Leopold (Hugh Jackman) gets in conflict with the law by not picking up his dog's droppings. Luckily, he is a rather modern-thinking gentleman for his era, so he doesn't disrespect the policewoman (Viola Davis) or call her a racial slur. This scene could have ended very differently.
The case where there's someone with no official legal existence is probably a lot less common than the case where "this guy might have legal existence somewhere but he's not carrying any ID and he's not usefully answering questions that would allow us to prove his identity".
In which case the procedure, if they haven't committed any serious crime, is to deem them too much trouble and let them go.
If he can persuade a responsible official to take his claim seriously as a real possibility, there are well-established ways of testing his claim. If he will consent to having a tooth extracted, then radioisotope testing of the enamel can establish with a very high level of confidence that that tooth grew before c. 1955 when large-scale above ground nuclear testing started making a huge spike in background radiation in the atmosphere.
If he's newly arrived, you can probably also do the tests with a lock of hair or a bone biopsy, but teeth are the most precise and convincing because of established protocols and because tooth enamel grows at a very specific time in childhood and isn't remodeled like bone or continuously growing out like hair.
If the person persistently maintains they were born in the 19th century and seems confused by modern objects, they might be temporarily held for psychiatric evaluation. But supposing the outcome of that evaluation is that the person poses no danger to themselves or anyone else, and the initial cause for their arrest is sufficiently minor to pose no interest to the DA, they'll have to be released.
If there is evidence of foreign origin, such as limited English-language skills, and if the police department is interested in referring the person to ICE custody, deportation might happen, but more likely the person is simply put back onto the street, or possibly connected with advocacy resources who could assist in re-establishing a contemporary identity.
Assume they are native to the US of 1900.
A few other folks have hinted at this, but I want to be direct about it: @Scott there are so many people in the world who have legitimately interesting things to say. And you only have one second per second. As they say, don't feed the trolls -- why do you spend any time at all on Yarvin, who seems to basically be an edgy teen in arrested development?
There are obvious consequentialist impacts of giving airtime and legitimacy to arguments that are both heinous and stupid -- cf your own writing on Crazy Like Us. If you read the comments under his thread, there are countless people who are discovering Yarvin for the first time and are thoroughly enthralled by his bullshit philosophical justifications to destroy society. And it seems equally obvious that Yarvin is either arguing in bad faith or is so deep in an epistemic pit that your response cannot possibly push back on all of the crap at once. Like, he straight up says "[liberals] invented a pandemic and killed 20 million people" and it just slides by, you do not make the time in your response to call that out as obviously false. The result is that all of those people who are reading Yarvin for the first time will see that "fact", see that you do not respond to it, and go "O, I guess COVID was a farce. Yea, fuck democracy."
This is really bad.
I'm purposely leaning heavily on consequentialist framing here because you're an EA. Do you think that the upside of responding to / arguing with Yarvin is worth it?
EDIT: adding a comment I made from deeper in the thread.
As of this edit, Scott's response to Yarvin got just under 215k views. 1k likes, 65 retweets. Yarvin, by contrast, got 380k views, 2.6k likes, 175 retweets. There are 165000 people who only saw Yarvin's tweet thread, and who never even saw Scott's response and maybe never saw the original piece that kicked off the whole thing *since Yarvin doesn't link to it*. Just to put some numbers on this.
You've made so many long comments in this subthread saying almost exactly the same things in slightly different ways, so forgive me if I missed it, but multiple people have pointed out that mockery and oxygen-deprivation are symmetric weapons that can be equally well used against good ideas and bad ones, and I don't think you've addressed this at all. I agree with you that Moldbug/Yarvin/whatever-his-name-is-now is insane. Probably almost everyone in this thread agrees that he's insane. But your attitude is dangerous and terrifying, because your calls to mock or ignore him are *not based on an objective standard proving that he's insane*. You have not said Scott should mock or ignore all people who 80% of the population agrees are insane, or that Scott should mock or ignore people who have spent time in a psychiatric facility. No, you instead say Scott should mock or ignore Yarvin because clearly he's insane. Which is just another way of saying because *you* think he's insane.
Can you honestly not see how symmetric that is? How it is *just as easy* for someone on Yarvin's blog to say "don't engage with Scott or acknowledge his arguments, he's clearly insane"? How it's just as easy for me to say you're insane so everyone should mock or ignore you?
People here aren't necessarily averse to rejecting or deplatforming some arguments. They *are* for the most part absolutely averse to doing so *solely on the basis that the person calling for said rejection/deplatforming is really REALLY sure they're bad arguments*. Because, you know, they've seen this done over and over to arguments that turned out to be true! Ranging from "your rape accusation statistics have not the slightest connection with reality" to "there's some evidence covid may have been a lab leak".
Making a comment saying that Yarvin is clearly "heinous and stupid" (and so should not be engaged with) is utterly unpersuasive as an argument, because you could say that about anyone. It's based entirely on your own claimed authority to make this pronouncement. But there are two ways that you *could* give people strong, asymmetric-with respect-to-truth-and-falsehood reasons to agree that Yarvin is heinous and stupid. The first is to use some totally or near-totally objective standard to demonstrate this, that doesn't at all depend on your or any other person's personal judgement. Examples might be "Yarvin has been successfully sued for libel five times" or "Yarvin has been clinically diagnosed with paranoid schizophrenia". You didn't, as far as I can see, provide anything like such objective evidence in anything you've written for your claims about what he "clearly" is. The second is to actually engage with his arguments and rationally demonstrate that they don't hold logically and/or that they're dishonestly or maliciously motivated.
Which is...exactly what Scott is doing!
> You've made so many long comments in this subthread saying almost exactly the same things in slightly different ways
apparently i haven't been clear enough 😂
> Making a comment saying that Yarvin is clearly "heinous and stupid" (and so should not be engaged with) is utterly unpersuasive as an argument
> I agree with you that Moldbug/Yarvin/whatever-his-name-is-now is insane. Probably almost everyone in this thread agrees that he's insane.
I don't understand the confusion! It really seems exactly right that everyone in this thread agrees he is insane! So what is the point of me spending a bunch of time explaining what everyone in the thread seems to already know? Like, not a single person in this thread has said "actually I think Yarvin has some good ideas". I have explicitly asked people to make this argument and no one has.
So I keep coming back to Alex Jones.
Scott doesn't write about Alex Jones -- is this 'depriving of oxygen'? Scott doesn't spend time refuting whether or not the Sandy Hook Shooting happened. Is this "deplatforming"? No! There is no platform! This is a blog. It's not like, facebook, or substack, or twitter. It is absurd to make the case that everyone who Scott doesn't write about is somehow being deplatformed! Like, what?
And even if this was somehow the case, I don't think I need to spend time explaining on the merits why Alex Jones is insane. I can just, like, point to his body of work and say "It is self evident" and everyone here would hopefully understand that, and if someone did not understand that they are welcome to say "I actually think Alex Jones has some good ideas".
If everyone agrees that Yarvin has lost his mind, what are we even talking about? And yes, I do think he is heinous -- the dude is explicitly, openly racist (https://x.com/curtis_yarvin/status/1922234238444711949) in addition to everything else. Do I have to sit down and explain, in detail, why this tweet is racist? Or does it stand on its own as something that is obviously horrible and *the person who sincerely believes this should not be treated seriously*? In my opinion, no, it is not worth my time explaining why this tweet is racist -- the body of work makes this self evident.
> Can you honestly not see how symmetric that is? How it is *just as easy* for someone on Yarvin's blog to say "don't engage with Scott or acknowledge his arguments, he's clearly insane"? How it's just as easy for me to say you're insane so everyone should mock or ignore you?
Yes, I see the symmetry. I do not think it matters. I am not making a platonic argument about good discussion norms, even though several people have tried to put me in that box. I am asking a much more mundane question: "Hey, Scott, are you sure this is worth your time?" If someone, somewhere else in the world, was like 'dont engage with Scott', power to them. That is their right. If someone, somewhere in the world, said 'dont engage with theahura', that is *also* their right. There are billions of people in the world, you cant engage with everyone.
But as I said elsewhere in thread, if a *really famous person* told their audience 'dont engage with theahura', my standing would immediately increase overnight. If Trump said, on air, 'theahura is the worst, no one should read him', I would get hundreds if not thousands of subscribers immediately. So if you think my ideas are bad, and you have more reach/standing than me, it would be better to just not engage at all, at least from a consequentialist framing. This isn't some enlightenment salon where only 10 people are around, after all. This is the public web.
> The second is to actually engage with his arguments and rationally demonstrate that they don't hold logically and/or that they're dishonestly or maliciously motivated.
Rationally engaging with someone who a) is not acting in good faith and b) thrives off bad publicity is a great way to make Yarvin more well known. I think this is bad. You could rationally engage with a 5yo throwing a tantrum, but it won't get you anywhere, and instead will give the 5yo more attention which is exactly what they want. More generally, you are making a category error -- you admit that Yarvin is insane but think that we should still engage with his arguments. Why? How do you rebut an insane person?
I go back to the numbers I posted in OP. As of right now, Yarvin's thread -- full of blatant misinformation like "liberals created COVID to kill 20 million people" -- is sitting at nearly 400k views. Scott's response is sitting at 240k views. 160000 people have *only* seen Yarvin's show boating, irony filled, steaming pile of bullshit. And unfortunately, Yarvin is a showman. So he knows how to write things in a particular kind of tumblresque way that seems convincing if you don't stop to think about it for 5 seconds, which most people won't. If you think he is insane, you have to deal with the fact that Scott treating him seriously has led to more eyeballs on the work of a madman. This isn't a hypothetical, we can see this objectively. So if you are a consequentialist, I ask a simple question: is this good?
Scott may have other reasons for responding to Yarvin in a serious way, but that is why I framed my question the way I did -- as earnest inquiry. Maybe he thinks Yarvin *does* make some good points, or maybe he thinks that it *is* worth responding to Yarvin despite the negative effects of spreading Yarvin's shit around. Maybe he thinks Yarvin can be reformed. Idk. He's been silent on the matter and I suspect we aren't going to get a response, which is also fine.
A little beef between rappers helps them both sell albums, as long as nobody gets shot.
You're right, I think... or if you're wrong then I'm also wrong, but... and hear me out on this, it's bad that you're right.
Why? Because basically the upshot of all this is that there are some topics that Scott, or anyone with a substantial audience really, is forbidden from discussing in good faith. Instead these topics must, at least in public, be approached with an attitude of ridicule from the get go. Any failure to submit a sufficiently scornful general disclaimer imposes an extra burden of explicitly calling out every individual point of difference and endlessly caveating every point of assent to forestall any impression of general agreement that may be fermenting in the mind of our impressionable reader.
> it's bad that you're right
I agree with this.
I'm reminded of Scott's writing about the culture war thread (https://slatestarcodex.com/2019/02/22/rip-culture-war-thread/)
> The thing about an online comment section is that the guy who really likes pedophilia is going to start posting on every thread about sexual minorities “I’m glad those sexual minorities have their rights! Now it’s time to start arguing for pedophile rights!” followed by a ten thousand word manifesto. This person won’t use any racial slurs, won’t be a bot, and can probably reach the same standards of politeness and reasonable-soundingness as anyone else. Any fair moderation policy won’t provide the moderator with any excuse to delete him.
Scott doesn't quite go this far in his post, but it's reasonable to ask: what *should* be done about this kind of person? You can't, like, do a beat by beat takedown of this manifesto every time it gets posted. Even if you could, you basically immediately elevate this person's standing, while possibly tarring your own.
It would be great if we lived in a reality where the public sphere became a place of enlightenment and earnest clash of ideas. In practice, only the worst ideas get spread, everything is taken out of context, the confirmation bias + Chinese Robber Fallacy dooms us, and our own engagement algorithms amplify our worst traits such that more reasonable takes never even see the light of day. I'm actively trying to think about how to solve for this, but right now no one has a solution
This is only a problem on the internet, and can be added to the long list of downsides that, in my opinion, tip the scales to the side of being a net negative. For all the good on here, I'm becoming increasingly convinced that humanity would be better off if the internet ceased to exist.
Eliminate all of the polarizing pseudo-debates in which both sides make claims and their respective supporters fawn over them and deride the other side in the comments. Create a real physical forum for meaningful political debate between public figures. The current state of the information world is just awful.
Social media, not internet, right?
Regular ole websites, file transfers, streaming, dont have the same set of issues. Its just social media.
> I'm becoming increasingly convinced that humanity would be better off if the internet ceased to exist.
Increasingly of the same opinion. I think there is regulation that could make this better -- a very low hanging fruit example is "Make all algorithmic personalization opt-in instead of opt-out". A less low hanging but still valuable approach is something like requiring all public social media above a certain size to have community notes, or to require that they all show some measure of life experience / expertise.
In general, government exists to try to internalize these externalities. I think there is a way forward, but not if everyone continues blindly following their incentives. This is prime moloch territory.
Another solve is limit the number of posts you can see or make each day. Select for quality not quantity.
o, also a good one! though i think that may have 1A problems if it was federal regulation
Funny that you bring up Moloch here because the only solution to Moloch is monarchy, and isn't the reason that you brought this up in the first place because you're scared of the monarchist?
I'm sorry, this is an insane gigabrained take. I know Scott said this in his original post, and I thought he was wrong when he wrote it.
How do you think the EPA was created? Or, like, cigarette legislation? Or insider trading laws? Or GDPR? Or any of the other hundreds of coordination problems that have already been solved by democratic governments? Companies often request additional regulation in democracies because they do not want to be caught in competition traps! Saying "well the only solution is monarchy" is exactly the kind of edgy bullshit that Yarvin thinks he's so smart for saying, without actually doing the hard work of cracking open a history book or even the daily news for 30 seconds. He created an extremely shallow and incorrect model of 'human behavior', then reached a nonsense conclusion, and said "welp I'm done, I've solved political philosophy". Starting from the premise that absolute monarchs don't have to deal with incentives is so far off it's not-even-wrong. "Democracies have all these real world practical problems, but if you pretend like absolute monarchs are all idealized and don't have real world practical problems then absolute monarchy is better, and also cows are spheres". If you believe this, congrats, you have gigabrained yourself.
Also, like, for people who supposedly care a lot about science, this is just such a profoundly bullshit take just on the empirical level. There are something like 150 democracies in the world, representing billions of people. Assume theyve been alive for 50 years on average, thats 7500 democracy-years. You think all of that is just decline? What? There are absolute monarchies in the real world that you can compare against and they all suck! Like *clearly* Oman is worse than the Netherlands.
Government *is* a solution to moloch. All governments exist to solve coordination problems, and also have to deal with the incentives of the people that make up that government. Concluding that therefore you can only have monarchy is maybe an interesting take for an 8th grade history essay, a good learning opportunity. But Yarvin -- and you -- should know better, no matter how many layers of clever weaponized irony he tries to hide behind.
It's definitely Moloch territory. Destructive competition and bad incentives become more relevant every day and I can't wait for the pendulum to swing back in the other direction.
There's one regulation that I'd most like to see enacted and I think most people (of voting age) could agree on, and that's a total ban on algorithms for minors. It was bad enough when we were just publicly ranking our friends and comparing ourselves to others. Giving children their very own Skinner box that's available 24/7 is one of the more deranged things widespread in society today.
> total ban on algorithms for minors
extremely strong +1
Well, the administration seems to have a new idea on how to solve this issue. With force. It seems you and they are in agreement that bad ideas need to be silenced. You're only in disagreement about what those ideas are.
*taps the sign* Ignoring and silencing are not the same thing
When your concern is "social contagion", yes, they are the same thing. When your goal is to prevent a person from having any avenue from expressing their opinions by pressuring everyone into deplatforming them, there is functionally no difference. Of course, liberals did half-ass this effort. Perhaps they should've pulled the trigger first... but it's too late for that, isn't it?
Please do not strawman.
I said nothing about preventing a person from having an avenue from expressing their opinions. I also did not say anything about deplatforming. At no point have I said that Substack or X should remove Yarvin. I'm not really sure what your political persuasion is, but you seem to be bringing in past scar tissue that is not relevant here.
What I *did* do is ask Scott a question: "why do you spend any time at all on Yarvin, who seems to basically be an edgy teen in arrested development?" And then I gave my reasons for why I spend no time on Yarvin, and why it may be harmful generally for Scott to do so.
If you disagree that Scott's actions are harmful on the merits you are welcome to make that argument. For example, you could take any of the following positions:
- Yarvin actually has ideas worth debating
- We should actually debate everyone regardless of how bad faith they are
- Consequentialist framings are incorrect in this setting, we should instead be looking at this through <some other lens>
I am open to hearing any of these arguments
Maybe Scott does it because he enjoys it, did you think of that? Scott is not a political operative or a lobbyist, so why are you demanding he act like one?
Why are you asking people who have not made any demands why they're demanding things?
Power to him. I'm not demanding anything.
I'll rephrase. Why are you passive aggressively demanding he act like one?
I'm not passive aggressively demanding anything either. I asked a question: "why do you spend any time at all on Yarvin, who seems to basically be an edgy teen in arrested development?"
I assume there is something that I am missing.
To me, Scott is arguing with the intellectual equivalent of Alex Jones. I legitimately do not understand why he would do this. I provided my reasoning for why *I* would not engage with Yarvin, which is that engaging with him seriously has negative consequentialist effects.
Like it or not, Yarvin has influence and is an important person. Rising up to the challenge is a far greater use of time rather than shirking back and hoping his ideas die on their own.
This constant idea you generally see from the left of censoring/ostracizing bad ideas has clearly *not worked*. You need to be able to give people tools on how to counter them.
It's amazing that the exact opposite of what you said is true. Engaging too much with bad ideas is what has given them fuel in the modern information environment. There is no free market of ideas.
You can’t kill an idea man.
Then why engage with it? There’s just no evidence that debating crazy ideas kills them. The hope that a neutral party looks at two sides of a debate and thinks “I’m going to believe the correct side” is not reliable. When you engage with crazy it has the effect of normalising crazy. The best thing to do is suppress bad ideas into very fringe corners of society. We don’t want the world to be run by people who think measles is curable with chicken soup and we have no time to debate them, we need to use every tool available to determine and shun ideas which are dangerous and wrong (to our best imperfect abilities). It worked extraordinarily well up until unregulated media took hold.
Heliocentrism. Four humors. Phlogiston theory. The rain follows the plough. Divine right of kings. We may have the corpses to admire and study, but many bad ideas are very much dead.
Some bad ideas, though, are /deliberately kept alive/. They keep coming back relentlessly, like superbugs in a dirty hospital ward after each round of antibiotics.
The best takeaways from the rationalist movement, the most useful parts of the sequences are the explorations of the edges of human reasoning - the explorations of when, how and why we fail to arrive at truth, and strategies for reasoning through problems in ways less likely to fail than just trusting our instincts. Universities generally attempt to teach similar things, though success varies.
The crowd here is more likely to have either or both sets of tools available than the average internet denizen; and yet every open thread is filled with discussions of bad ideas that just won't die.
We live in a world of gish gallops and similar firehose-of-concepts strategies, where a few ideas that are bad yet particularly attractive to some are repeated faster than each instance can be dismissed. Ostracising may not work, but neither does giving people tools to reason things out for themselves.
While they remain abstract internet arguments, walking away from the darker internet corners remains a reasonable response. But when there is a direct pipeline from those corners to daily life via the highest levels of government, the situation becomes untenable.
In a world that is tired of experts, how do /you/ suggest we deal with bad ideas?
Alex Jones has influence and is an important person, and arguing with him is a total waste of time. The only thing that arguing with Jones does is give him legitimacy. The same is true here.
More generally, if we believe that social contagion is at all real, then the worst thing that we could do with these ideas is *spread them*.
I am *not* saying "censor" nor am I saying "ostracize", because both of those actions take the underlying argument as something that is serious and necessary to respond to. I am saying "do not treat him seriously". He is a clown. A joke. Not even worthy of a footnote. If other people want to talk about him, sure, whatever. People talk about all sorts of things, like celebrity gossip or which superheroes could beat up other superheroes.
But Scott doesn't go and write 5000 word essays explaining why Iron Man actually could take Batman in a fight. He also doesn't go and write 5000 words explaining that, actually, the Sandy Hook shooting really did happen, or that the water does not actually turn the frogs gay -- both things that Alex Jones has continued to push.
Scott owes no one an explanation for how he spends his time, except for his wife and kids. Certainly not to me. But looking at this from the outside, I'm sad that someone of his talent is using his time to do...this.
What do you want him to do, kill him? Obviously debating him won't amount to anything, but surely you see value in studying this new ideology that has seized control of this country. It's also a little bit late to be worrying about "social contagion". The damage has already been done.
I feel like I've been very clear about what I want him to do, which is nothing
> Obviously debating him won't amount to anything
Unfortunately this is part of the problem. In my OP I say "And it seems equally obvious that Yarvin is either arguing in bad faith or is so deep in an epistemic pit that your response cannot possibly push back on all of the crap at once. Like, he straight up says "[liberals] invented a pandemic and killed 20 million people" and it just slides by". Debating him is actively harmful.
> surely you see value in studying this new ideology that has seized control of this country
I do not, because I do not think there is anything here to 'study' -- there is no intellectualism here, only intellectual people losing their dignity. I prefer Hanania's approach on this, which is to mock it relentlessly.
Is your concern that by engaging with Yarvin, Scott is legitimising him/his arguments? The old "oxygen of publicity" argument?
https://waccglobal.org/media-platforming-and-the-oxygen-of-publicity/
That to me seems to be what you are arguing about, because otherwise 'this is Scott's blog, he can do what he dang well likes and if he wants to write 5,000 words about gay frogs well that's Hidden Thread 999'.
In the OP I ask Scott a question -- "why do you spend any time at all on Yarvin?" This is legitimate curiosity. Scott is a brilliant person, and I have a lot of cognitive dissonance from him spending time taking a troll seriously.
I back that question up with what my motivations are for NOT spending any time at all on Yarvin, which includes 'oxygen of publicity'. Scott hasn't responded, so until he does I'm more or less left to explain the oxygen of publicity position.
(There are other concerns I have too, like "Yarvin is a bad faith actor who will purposely twist your words to score internet points and cannot be convinced or won over" or "I would like to never have to think of Yarvin again and I wish it did not appear on my feed through ACX", but the former is likely less compelling to Scott and the latter is just me being selfish)
Did anybody, anywhere in this thread, at any time, remotely imply that this isn't Scott's blog and that he can't do what he dang well likes?
Seems to me that there's a middle ground between "earnestly engage with Yarvin as though he didn't manifestly have end-stage brain worms" and "pay him no mind whatsoever even though he still has influence over people whose hands are on the levers of power."
I mean, people study Qanon!
Yea, I think that's what Hanania does -- a good middle ground between mockery and actually pushing back on bad argumentation. Scott is unlikely to be as vicious as Hanania is, though.
I saw a Note the other day where somebody was bragging about his Substack having reached #4 in the History category, and right below him in the rankings, to my surprise, was Yarvin's Grey Mirror. So he has a decent sized readership, it would seem, and those readers (this being Substack) are probably a good deal smarter and more educated than Alex Jones', and thus might be more persuadable, so I think I'd disagree with the idea that engaging with these ideas is a total waste of time.
Copying from a few of my other responses:
> And it seems equally obvious that Yarvin is either arguing in bad faith or is so deep in an epistemic pit that your response cannot possibly push back on all of the crap at once. Like, he straight up says "[liberals] invented a pandemic and killed 20 million people" and it just slides by
> To be clear, *my* position is "If you must talk about such things, do so with the disdain and mockery that such positions deserve -- both because that is the correct response and also because it will help dissuade people from taking such positions seriously."
> I prefer Hanania's approach on this, which is to mock [such things] relentlessly.
> But Scott is kinder than I am, and I know that he will not do this. Given that, the next best thing is to not engage at all.
Mockery isn't as persuasive as rational argument.
Its nice to pretend that these debates are all happening in an Enlightenment Salon where only those committed to truthseeking as a project will hear the arguments in the proper context.
Unfortunately, these debates are happening on Twitter.
Scott's response to Yarvin got just under 215k views. 1k likes, 65 retweets. Yarvin, by contrast, got 380k views, 2.6k likes, 175 retweets.
There are 165000 people who only saw Yarvin's tweet thread, and who never even saw Scott's response and maybe never saw the original piece that kicked off the whole thing *since Yarvin doesn't link to it*
Mockery is one of Yarvin's favorite tools.
But if you treat Yarvin’s sketchy premises as axiomatic you have to admit the whole thing more or less hangs together. Kinda.
He’s rockin’ the leather jacket and unkempt hair, no? The boho intellectual look.
And his command of *italicized irony* is second to none.
But unironically, I think he is a just a very clever, well read conman.
I don't think it does. He's much, much better as a critic of liberal "democracy" than he is at outlining a superior alternative.
I think his premise is solid but his conclusions don't follow. Still, I view him as one of the leading intellectuals on the right, so he's worth critiquing for that reason alone.
It's actually pretty common for intellectuals to have a very coherent and sensible critique of the current system and not a very well-thought-out proposal for a better system.
The list of things that "more or less hangs together, kinda", includes the pre-implementation vision of communism. And many others of the same ilk. Do we want a policy of never speaking against such things until *after* they have been implemented and then fallen apart?
Yes, talking about them acknowledges that they are worth talking about. The bit where other people are already talking about them, already established that. They *will* be talked about, and if they're at all clever they'll couple all that talking to the real problems of the current system, and if those problems are substantial, who's going to win? The people who are being talked about because of their bold, decisive plan to fix the problems, or the people who are, what, talking about nothing at all? Talking about boring incremental technocratic measures to mitigate the problems?
Someone needs to be talking about all the horribly bad ideas that "more or less hold together, kinda, until they're actually implemented". This needs to be talked about anywhere people ideologically adjacent to the proposed whatever-ism gather, and for Yarvin's particular whatever-ism, that includes here.
To be clear, *my* position is "If you must talk about such things, do so with the disdain and mockery that such positions deserve -- both because that is the correct response and also because it will help dissuade people from taking such positions seriously." **
But Scott is kinder than I am, and I know that he will not do this. Given that, the next best thing is to not engage at all.
**Caveat that there are of course times and places to respond gently as a teaching moment -- for example, when talking to a young student, or someone who we have good reason to believe is making an earnest mistake. Yarvin is neither.
You want Scott to sneer like John Oliver. That isn't his style and his rejection of that crap is why Scott got popular in the first place. Would you have him delete the Anti-Reactionary FAQ from the archives?
No, I don't want Scott to sneer like John Oliver. *I* would probably sneer like John Oliver if I was going to write about Yarvin, but I doubt I will ever do so. Scott is not me. I literally wrote:
> "But Scott is kinder than I am, and I know that he will not do this. Given that, the next best thing is to not engage at all."
I...pretty clearly think it would be better for Scott to not engage at all. I don't know how you got "You want Scott to sneer like John Oliver" when I straight up said "the next best thing is to not engage at all".
As I just responded to your other comment, maybe there is something that I am missing about why Yarvin is worth a passing thought.
> Would you have him delete the Anti-Reactionary FAQ from the archives
No.
You wrote that Scott should either sneer or shut up. I'm not sure how you found this blog but I don't think it is for you.
The problem here is that disdain and mockery work about as well for dismissing smart or good ideas as they do for dismissing dumb or evil ideas. If Moldbug has some ideas worth discussing, then we should actuall discuss those ideas, honestly engage with them, and maybe then we decide they're crap and don't support them. Of course, Moldbug's literary style of smoke screens and squid ink makes it hard to actually engage with his ideas (that's one reason I'm not overly fond of his work), but the point is to get smarter, not to consign the stinky outgroup members to the nerds' table.
I don't think Yarvin has any ideas worth discussing, in the same way that I do not think Alex Jones has any ideas worth discussing
Suppose Alex Jones suddenly appeared and asserted that he did not think you had any ideas worth discussing. Suppose further that he happened to do it first. Should we just as righteously block you from further engagement?
I've noticed you're employing this tactic that ought to be in a 101-level course on What Not to Do in a Discussion, where you assume you possess the authority on what ideas are worth considering and follow through to which sources are worth considering and proceeding next to a total shutdown, without apparently realizing that you don't naturally possess that authority, and that it could therefore be used against you. And then you compound the problem by abusing everyone else's sense of not employing this tactic, by forcing them to spend their time arguing with you about whether this is a good tactic. At that point, one of the few natural heuristics left to people is to consider which person appears to have rhetorically run amok, and is therefore a threat to the consideration process itself, and shut *that* person down.
I didn’t say he shouldn’t be talked about. I said I think he is a yutz. I was using the same sort of ironic mockery that he employs so well.
This is the way
Hi guys,
I thought I'd give a shout out to my son's school, Achievement Unlocked in Brooklyn, NY, which is currently recruiting for the next school year. It is a twice-exceptional school for gifted neurodivergent children and it is run by a rationalist. It has comments on the wall like, "The map is not the territory,' and "Never ascribe to malice that which can be explained by ignorance." I figure there may be others on this list seeking something like this. I'm attaching a flyer for their upcoming openhouse, and of course if you have more specific questions about my experience you can contact me directly.
This is what another parent wrote about it:
"
Achievement Unlocked (AU) is a small private school in Brooklyn that serves students in 1st-8th grade that have struggled in other schools. AU has a neurodiversity positive environment, reflecting that many but not all students are neurodiverse. AU is not a school that turns kids away who have had behavior issues nor is there any specific diagnosis that they are looking for. They really seek to serve students for whom other schools, both public and private and often both, have not worked. It is so worth scheduling a conversation with them if you are interested but unsure your child would fit there.
There are a max of 25 students. Classes are currently 5 students or less with a max of 6 students at any time. It seems to me there are 2 to 3 teachers/specialists present in every class/subject. My son has differentiated groups - he studies with the youngest kids in ELA and with older kids in math and even within these groups the work is individualized. We get daily report cards showing how he did with academics and behavior in each subject. Also, in addition to their regular academic classes every student takes Independent Study and gets to work on projects of interest that help them stretch their skills and experience and has a choice of many enrichment classes.
Social skills and support is scaffolded through the whole experience - for example, each student works on a self-skill plan in which they pick a skill to develop and negotiate with teachers and staff how to have that skill supported. This can be something they want to develop or something they want to be able to do. The school uses the Collaborative Problem Solving model proactively to do this. They help students proactively figure out how to make something happen and solve problems. Another example is they also have a system in which students are always earning chances to make "reasonable," (easy to earn), "less reasonable" (harder to earn) and "unreasonable" (very hard to earn) requests. A reasonable request might be staying indoors for recess or spending 5 minutes in the quiet room. An unreasonable request might be "the whole school doesn't get homework for a day." Students earn these by doing well, helping others, and accomplishing tasks. To earn an "unreasonable request." students have to do something quite challenging like memorizing the periodic table or beating the school director at chess. This way students are constantly learning about advocating for themselves and what is reasonable to request. My son LOVES this system. Homeroom teachers also regularly help support and prime students for social interactions and choices they might make throughout the day. For example, my son's homeroom teacher always thinks with them about what they want to do at recess and how they can accomplish it, which has really helped him.
Finally, the staff is tremendous! The school founder is an amazing role model, an active participant in developing each child's plan and experience, and the leader for developing the systems and atmosphere of the school. He ensures the school uses research-based approaches and collects a lot of data about how the kids are doing, like how many times they ask for a break and their daily behavior and academic scores. The classroom teachers, related service providers, and support staff and the admin team are also dynamic, talented, and caring.
In parent meetings recently, it struck me that as parents talk about the school, the most common phrase I hear is "no matter what" - for example, "AU will support your child no matter what," "AU staff will welcome your child with open arms, no matter what happened the day before," and "AU staff will never be disrespectful or mean to [parents], no matter what." I have never before encountered a school before that actualized the principle of unconditional positive regard as a whole organization - but AU definitely does."
I had to look up what "twice-exceptional" was because it's a term I've never heard of before. Another example of "all American geese are swans":
"A twice-exceptional school is an educational setting that caters to gifted and twice-exceptional students."
Okay, so "What does it mean for a student to be twice-exceptional?"
"The term “twice-exceptional,” also referred to as “2E,” is used to describe gifted children who have the characteristics of gifted students with the potential for high achievement and give evidence of one or more disabilities as defined by federal or state eligibility criteria."
So kids with special/additional needs who do not have an intellectual disability? Good luck to the school and the kids, but we are getting on the euphemism treadmill here!
To be clear, I'm not criticising you, Laura, but there seems to be this push for "you don't have a disability, you have a superpower! you're not just exceptional, you're twice-exceptional!" around disability (both physical and mental) that is well-meant as encouragement particularly for children so as not to consign them to the basket of "disabled and incapable", but which hampers the people suffering these problems. No, my autistic tendencies are not a superpower, they don't make me super-duper-extra-mega-with a cherry on top exceptional, they've held me back and are something I struggle with daily, and if I had had to handle the rah-rahing about "twice as great because of your problems" on top of that, I'd have melted down a hell of a lot more than I already did and do.
> So kids with special/additional needs who do not have an intellectual disability?
No, not that at all. It's kids who have special/additional needs in addition to being exceptionally bright and not merely "not intellectually disabled". The kids need to meet the gifted cut-off in addition to having additional needs.
Referring to that as "twice exceptional" is one of the most obnoxious things I have ever heard
I'm sorry you have such a big chip on your shoulder. Twice exceptional is an important term used to differentiate schools catering to children with disability and normal or high IQ from those catering to children with disability and low IQ. This is an important distinction when you are visiting schools. What is the curriculum? What is expected of these students? The public schools stick all kids with disability in the same classroom - so you can have a child with mental retardation in the same room with a kid who has some behavioral difficulties but can do calculus, as well as a normally developing child who has a phyiscal impairment. This is not an acceptable situation. The school doesn't blow smoke up the kids asses about them being 'exceptional,' though I know there are parents who cling to this - but they would anyway.
Twice-exceptional just sounds like such a round-the-houses way to try and avoid at all costs acknowledging "the kids have problems that interfere with their normal development".
Maybe I have a chip, but this kind of "no no there is no *problem* there is only *opportunity*" approach can leave people twisting in the wind - if there is not a problem only a second heaping helping of exceptionality, then there is no need to do anything to address or support the problem which does not exist, now is there?
I'm glad the school is very supportive and seems to be doing the proper job, but the euphemism treadmill seems to be in full swing (you're not retarded, you're special - oh darn people are using 'special' to mean 'retarded' - well now you're not special, you're exceptional - wait, what? okay - you're *twice* as exceptional as the ordinary exceptional kids, by which we mean the ones who are exceptional and not, you know, 'exceptional')
Yeah, I'm with Laura on this. I think maybe it's just that Deiseach isn't as immersed in the context. In North American educational circles "exceptional" is not necessarily a compliment or a euphemism. It's a neutral term for kids who do not fit the mold for whatever reason, be it positive or negative. This is the most literal meaning of the word "exception" after all.
Also, as a neurodivergent myself, I recognize that my autism and ADHD ARE both a disability and a superpower. I'm not trying to sugarcoat reality, this is truly the most accurate assessment I can come up with. It's not all good or all bad but it's different, i.e., exceptional. Basically, almost everyone has a car and I have a motorcycle. Is it an advantage of a disadvantage to have a motorcycle rather than a car? There isn't one answer to this, it depends on the situation and your criteria.
I (UKl also took "twice exceptional" to be intended as complimentary. Perhaps it's a US vs UK/EI thing - if I referred to a kid who was struggling or acting up as "exceptional" I would sound like I was being sarcastic ("wow, that kid sure is something"). Thanks for the clarification!
It does seem to be a US English versus British English thing. Maybe they are so used to "This term that seems complimentary is, in fact, whitewash on a dungheap because we all know it's code for 'got problems' but you can't say that out loud" that they recognise "okay, twice-exceptional means twice the problems".
Trust me, no one is under the illusion that these kids don't have problems.
In his thread, Yarvin suggests that liberal elites are worse than Facebook moms at medical policy because they caused COVID, and also says they turned America's cities into desolate and dangerous places, and in your response you say "I agree there are many bad things about the liberal American regime."
Do you actually, materially agree with Yarvin? Or do you just agree with him in the sense that I might agree with somebody who thinks Joe Biden is a lizard-person (we both would have preferred if he hadn't run for re-election)?
He's either a grandstanding bullshitter or he's delusional. Cities are safer now than they've been since the early sixties, and theyre much more livable -- better urban design and less pollution. And how pray tell did the liberal elites cause COVID? Honestly, his untruths are mendacious, and he's practicing the Paranoid political playbook as described by Richard Hofstadter back in the 1960s. Yet somehow people seem to take him seriously around here.
Even acknowledging that cities are much better than they were in the past (especially the fairly recent past of the 1990s), it's still worthwhile to note negative trends as they happen.
I don't read Yarvin's stuff, but I presume he means either the gain of function research or things like the covid lockdowns that caused significantly damage to people and the economy.
Hey, don't knock the lockdowns. NPIs (non-pharmaceutical interventions) made a significant statistical difference in the mortality rates between states that used them and states that eschewed them.
Also, I got a big chuckle out of MTG who claimed the Dems were giving the Republican caucus COVID. The Dems mostly masked, while the Republicans refused to mask during the B.1, Delta, and Omicron waves. The fact that they kept catching COVID was sinister plot by the Dems according to MTG.
Do you have a source for that? Because according to the CDC Florida had a lower mortality rate than California https://www.cdc.gov/nchs/pressroom/sosmap/covid19_mortality_final/COVID19.htm
Furthermore, you would have to have a pretty massive reduction for me to be willing to conclude that a multi year lockdown was a good trade off.
Being a data-driven person, at the outset of the pandemic, I started charting cases and deaths in nine counties of the SF Bay Area (where I live), the Seattle area, and NY City. Dr. Sara Cody of Santa Clara County persuaded the health officers of the nine counties in the SF Bay Area to push for work-from-home policies, restrictions on public venues, masking, and social distancing. Likewise, the public health officers in the Seattle area urged the same NPI programs. These were recommended the last week of February 2020. My employer and most of the other tech companies in the Bay Area had workers work from home. Mayor London Breed of San Francisco announced a shelter-in-place policy on March 16th, but the freeways and cities of the Bay Area were already ghost towns by the first week of March.
OTOH, Mayor De Blasio (sp?) of NY City followed the initial CDC and WHO guidance (wash your hands frequently and go about your business normally). From March through April, NYC's cases and deaths climbed quickly in NYC. Hospitals were overwhelmed. Refrigerator morgues were trucked in, and the dead were carted off to mass burials. While in SF Bay Area and Seattle, cases and deaths climbed at a much slower rate. Emergency rooms and Critical Care facilities were stressed, but we didn't require refrigerator morgues and mass burials. Gavin Newsom was persuaded a couple of weeks into March to apply Sara Cody's NPI recommendations to all of California. By that time cases and deaths in LA were starting to, and their healthcare system was overwhelmed. And a few days later, the Governor of NY (whose name escapes me at the moment) ordered lockdowns across NY State (over De Blasio's objections). Too bad I can't post graph I created in substack comments, because it's pretty stark.
However, serious researchers also noticed a difference between counties that used NPIs and those that did not...
IS THE CURE WORSE THAN THE DISEASE? COUNTY-LEVEL EVIDENCE FROM THE COVID-19 PANDEMIC IN THE UNITED STATES, by Catalina Amuedo-Dorantes et al
https://www.nber.org/system/files/working_papers/w27759/w27759.pdf
>Our estimates suggest that advancing the date of NPI adoption by one day lowers the COVID-19 death rate by 2.4 percent. This finding proves robust to alternative measures of NPI adoption speed, model specifications that control for testing and mobility, and across various samples: national, restricted to the Northeast region, excluding New York, and excluding the Northeast region. We also find that the adoption speed of NPIs is associated with lower infections, as well as lower non-COVID mortality, suggesting that these measures slowed contagion and the pace at which the healthcare system might have been overburdened by the pandemic. Finally, NPI adoption speed appears to have been less relevant in Republican counties, suggesting that political ideology might have compromised their efficiency.
And red states that valued personal freedom over public health fared worse that blue states...
https://jamanetwork.com/journals/jamainternalmedicine/fullarticle/2807617
Of course, then Omicron came along in December of 2021. Omi was more infectious than any of the previous strains. NPIs were being loosened and people were being vaccinated. After Omicron the death toll became more evenly distributed per capita across the country.
>Cities are safer now than they've been since the early sixties,
I wouldn't be so quick to gloss over the very noticeable increases in street crime and general public incivility since 2020 nor the even more noticeable increase in the number of visibly homeless and unwell people since 2010 or so.
But even so, it's deeply unhinged to jump from there to "American cities are desolate and dangerous places." There was a huge amount of room between the state of the American city in 2009 and "desolate and dangerous," and even though the movement since then has been in the wrong direction, there's...still a lot of room between 2025 cities and the hellscapes that people like Yarvin imagine our cities to be.
How do you measure public incivility? But it doesn't look like cities are the hellholes that Grima Moldbug claims (see below). Of course, if you consider homelessness to be a crime, then indeed the problem has gotten much worse over the past 25 years. And it doesn't seem to be localized to US urban centers. It's becoming a big problem in Canada and the UK. Canadians blame substance abuse and psychological problems is the cause. UK claims the cost of living drives people to homelessness. I think these explanations are incomplete.
</ChatGPT warning>
Over the past 25 years (roughly 2000–2025), non-violent crime in U.S. urban areas has generally declined, though with some fluctuations depending on the type of crime and time period.
1. Long-Term Trends (2000s–2010s)
According to FBI Uniform Crime Reporting (UCR) and Bureau of Justice Statistics (BJS) data:
Property crimes (which are non-violent and include burglary, larceny-theft, and motor vehicle theft) declined steadily in most urban areas from the early 2000s to around 2014–2015.
Examples:
Burglary and larceny saw consistent drops.
Motor vehicle theft dropped significantly through the 2000s but started rising again in the late 2010s and early 2020s.
2. Recent Trends (2020–2024)
There were increases in some types of non-violent crime after the onset of the COVID-19 pandemic.
Motor vehicle theft rose sharply, especially in cities like Chicago, New York, and Washington, D.C.
Some cities saw increases in retail theft/shoplifting, leading to headlines about “crime waves,” though these increases were often localized or exaggerated.
Despite these spikes, overall non-violent crime levels remain significantly lower than in the 1990s or early 2000s.
3. Caveats
Urban vs. national data: Most national statistics are aggregated and may mask local variations.
Data quality: After the FBI switched to the National Incident-Based Reporting System (NIBRS) in 2021, fewer police departments reported data, making recent year-to-year comparisons more difficult.
Bottom Line:
No, non-violent crime in U.S. urban areas has not increased over the past 25 years overall—it has decreased significantly since the late 1990s and early 2000s, though there have been recent upticks in specific crimes like auto theft.
Let me know if you want a chart or links to specific crime databases or reports.
</end>
What you are not taking into account is that the violent crime rate for 90% of the population of a major city has stayed the same since the 90s at basically 0. The violent crime is concentrated in very small parts of the city, so unless you live or go there often (which there isn't any reason to if you don't live there) this doesn't really play into your perception of safety.
Most city residents experience what I did living in SF. Violent crime was never a threat to me and still isn't, but I notice a massive increase in homelessness and public drug use. I go out to the suburbs to go shopping because the CVS down the street locks up half the items and makes shopping a pain in the ass. This didn't happen 20 years ago. And what makes it 10x worse is that there is a degenerate city government that refuses to enforce the law.
If your degenerate city government enforces the law, where will your homeless go? They'll be pushed out into the suburbs where I live. But I live in one of those suburbs you're talking about, and my CVS has locked everything up. We've got a serious homeless problem in our town, and our city council just made sleeping on the street a crime. I still see tents along the sidewalks, though. I wonder if the homeless numbers per non-homeless aren't the same as in SF (I don't know). Anyhow, unless we're willing to fund massive internment camps on the scale of Manzanar, the homeless problem is not going to go away whether our government is run by degenerates or by generates.
Please note: I'm a degenerate libtard and I don't advocate the internment camp solution for homelessness. I'd advocate a massive increase in social services. I'd swamp homeless encampments with case workers to cajole or coerce the homeless into housing that we'd either build or confiscate (something like 1/10th of San Francisco housing is empty now). And I'd have case workers on their asses to make them take their meds (if required) and keep them off of drugs and alcohol. Of course, I'd tax the wealthiest 5% at higher rates to subsidize this.
If you're a genarate repturd, you'd probably opt for internment camps. And given how much it costs to incarcerate prisoners, I'd suspect they'd be just as costly or more so than the libtard solution.
But otherwise, we'd just be playing whack-a-mole with the homeless problem.
>How do you measure public incivility?
No idea, and not much interest. What I said was not that there exist pieces of paper in some official repository, extensively documenting and quantifying the change in rate of public incivility. I said that it was *noticeable*, which is to say, I've noticed it, and quite literally everybody I've talked to about the subject has also noticed it.
></ChatGPT warning>
Genuinely, thank you. I stopped reading at that point. (Just not doing it at all would be better, but warnings are better than nothing!)
Mendacious Moldbug
The Grima Wormtongue of Substack.
Liberal elites created a globalized economy with accessible air travel. Good luck getting a global pandemic in the 13th century!
Will 14th century work? That’s when Black Death happened. Or did Bill Gates caused that plandemic too?
In this context, "plandemic" is a happy accident 😁
Yes, it came out of Central Asia, ripped through Europe (well, at Medieval travel speeds), killing between 30-50% of the population (estimates range between 25 and 50 million people died). Northwest India got hit bad, and.I suspect the rest of India did, too. It spread down into Africa (at least in the places we have records — Ethiopia and Muslim West Africa). And it moved along the trade routes to China, killing at least 13 million people in the first wave. Then it sloshed around Eurasia for a while, resurfacing in the following decades until it burned itself out.
So, the semi-globalized pre-neoliberal economy of the 14th Century was connected enough for the Bubonic Plague to spread almost everywhere along its trade routes. The Americas were spared because no Europeans visited in the right time frame to light the fire there. The Vikings predated the plague, and Columbus was a 150 years later. But then the Europeans brought Measles to the Indigenous people of the Americas, which was as deadly or more deadly to their unexposed populations than the Bubonic Plague was to the Europeans.
As an aside, the current Amazonian rainforests may be a result of the 16th-century neoliberal expansion of Europe. Unexpectedly, archaeologists have discovered vast networks of roads and cities beneath the jungle floor, which were intensely farmed. The density of the communities suggests that the Amazon basin supported 10s of millions of people. The Amazonian civilization seems to have collapsed after the Europeans arrived in the Americas. </ChatGPT alert> Gaspar de Carvajal, who sailed up the Amazon (1541-42) reported seeing large, well-organized settlements along the river, with extensive agriculture, including fields of manioc. Carvajal described a nearly continuous series of villages and towns along hundreds of miles of river, with signs of advanced societies, including fortifications and coordinated resistance to the expedition. </end>
Modern historians had discounted or ignored these reports until the archaeological finds of the past decade. But just as the jungles of Guatemala and Yucatan grew up after the collapse of intense agricultural exploitation by the Maya, the jungles of Amazonia evolved after the collapse of Amazonian agricultural civilizations. I don't know what kind of ecology existed before humans started cultivating the Amazon. If anyone has any links to palynological evidence, I'd be interested in seeing them. But the Anthropocene started a lot earlier than most people realize.
> it came out of Central Asia, ripped through Europe (well, at Medieval travel speeds)
Someone will blame the immigrants, and the 14th century liberals who invited them...
During the actual pandemic of the 21st C we all blamed the immigrants, and closed the borders. If anybody wanted open borders then, they were the fascists.
A lot of things were topsy turvy back then.
It was that goddamn silk road! If the decadent European elites hadn't indulged in that fancy foreign fabric, it would have never happened!
Purely for clarification, he's saying it's a lab leak and therefore caused by the scientific establishment (who he obviously does not credit for anything else or they'd still have a great track record, medically speaking).
Does anyone have a recommendation for an immigration consultant or lawyer? I want to go to a knowledgeable person and ask: “here are my skills and ethnic background, and here are my husband’s skills and ethnic background. Who will have us, and at what price?” And I want to hear good advice back. Thanks!
I feel like there *must* be a Subreddit all about this where they can point you in the right direction. Most immigration lawyers are focused on immigration TO the country where they practice.
Are you asking which of the many countries in the world are likely to give you and your husband a work visa? Or are you asking about your options for a particular country?
Nope, looking for a recommendation for someone who could advise me.
About what, specifically? Where are you now, and where do you want to go? it is impossible to recommend an attorney unless we know what your goal is.
Advise about what? Immigration or emigration? Those are different things.
I wrote my first Substack post today - https://worriedbutch665291.substack.com/p/read-these-books-if-you-want-to-understand
It's a book-review-ish piece about 4 mental illness memoirs that combined give a good view of a variety of moderately-debilitating mental health issues. I've read these over the years and quite like them, and I think they will all help people understand mental illnesses that fall between depression/anxiety and crazy homeless person on your bus level of severity/prevalence.
It was an interesting experience to write it because I haven't written anything that long since graduating college and everything I've written before over 750 words has been term paper with MLA format citations. I've written plenty of 250-500 word things on reddit and substack, but no real blog posts. I think I fell into a lot of term paper ish traps and definitely don't have my writing voice figured out yet, but it was nice to try my hand at it. It definitely helped me get thoughts together in a way I haven't in a while.
I think I'll do a couple more 1k-3k word book reviews just to see what the process is like. I'm curious if any of y'all have gone through the process of starting to write longform pieces and whether you have any suggestions on the process.
This is going to be foggy, but has AI produced any memorable art? Things that aren't just kind of okay, but that people particularly like?
I wasn't just looking for AI art that people like individually, but for art that is at least reasonably popular on a large scale.
I really like the SF apocalypse art created by the Bluesky user "RevHowardArson" (formerly Theophite on Twitter), which has a distinctive but colorfully weird aesthetic. It's a pity he's got a self-hating complex about AI art because he moves in liberal circles where you have endless whining capital-A Artists who dogpile anyone who promotes AI art on social media (although they're thankfully finally leaving Ethan Mollick alone over there, mostly).
There's two pieces of AI art that have stuck with me as things I appreciate having seen.
The spiral piece that Ming posted is AFAIK the first AI artwork that seems to be original in a meaningful way.
I'm still a big fan of the Pope in the puffer jacket.
I thought of one from a few years back. It was labeled as a building in Budapest, but no one could locate it. Art nouveau with big windows reflecting a forest. Eventually word got out about details not making sense, but there were a lot of people who wished it was real.
There was a lot of prompting by a human involved, and when I saw another picture by him, I recognized the style.
I'm a big fan of Bad Apple Played On AI Paintings as an example of something cool that could *only* be done with AI style transfer: https://www.youtube.com/watch?v=E58aMjthQCM
I like how if you pause it on any frame it just looks like a normal painting, you only see Bad Apple when it's in motion.
That's clever, but the very fast shifting made me feel ill. It might well have a fandom of people who are neurologically different from me.
Refik Anadol's Machine Hallucinations! https://youtu.be/5Y384U-bOJo?feature=shared
Maybe not memorable from an aesthetic perspective, but I think if you had to pick an AI image that captured the "vibe" of the early 2020s AI acceleration, it'd be the DALL-E 2 "astronaut riding a horse" image -- not particularly good, not emotionally moving, but historically important in the sense of "AI can compose something novel and unlikely to have ever been seen before." So, I guess you could call it memorable in the same sense that Muybridge's galloping horse images were memorable.
https://images.ctfassets.net/kftzwdyauwt9/5GOIjwbUjLZHoGhX6q5oQg/d2984681d2a9466b71b7ca7632a8481c/Anastronautridingahorseinaphotorealisticstyle0.jpg?w=1080&q=90&fm=webp
I loved John Walter's fractal lizards series. Some of them show up if you do a google image search of "John Walter fractal lizard." The rest are probably not hard to find online.
I have a series of images Dalle-2 made in response to a prompt that was deliberately confusing. I like them, and Beowulf did. I'm not sure if any quite make it to the level of memorable art, but I'd say they are very enjoyably striking and novel images. There's a batch here on Imgur: https://imgur.com/a/6vaTrHZ
Would be interested to hear how you react, even if you don't like them. If you find them amusing or appealing, I had a couple projects where I actually made things using images of this kind, but I'd feel sort of silly pulling them out if you're not a responder to this sort of thing so try the litmus test first.
Passable surrealist art, nothing I want to see again or more of. They're more disquieting than I like.
It's funny because I do like a lot of Cyriak's work.
This spiral comes to mind: https://x.com/ptrschmdtnlsn/status/1919434568568123534
Also the person who used AI to complete Keith Haring's intentionally unfinished painting, as ragebait.
At least passable, and they're right that having this somewhat pleasant stuff in quantity does make it less individually impressive. I'm reminded of surrealist images made by combining jigsaw puzzles.
yes: https://cdn.mos.cms.futurecdn.net/z68jgeV7JFKXGiu8B5jAuZ-970-80.jpg
This one won a prize at the Colorado State Fair!
https://www.nytimes.com/2022/09/02/technology/ai-artificial-intelligence-artists.html
I like that one.
It’s stunning, I kept staring at it when I saw it the first time, something Iain Banksian to the whole scene was just mesmerizing.
Are you supposed to have gotten an email notification after submitting to the contest?
Nope. I didn't get one, either.
If you're concerned, you can always email Scott directly - scott@slatestarcodex.com - to confirm it made it.
Good luck!!!! I am so excited to read all the silly meta topics this year!
Thanks! And likewise :-)
Has anyone tried to model bipolar disorder as a consequence of dopamine receptor upregulation? If depressive states lead to decreased dopamine production then I could imagine homeostatic factors causing the receptors to upregulate. Then if the depression lifts slightly and dopamine levels return to normal then you get this crazy surge of activity, i.e. mania. Is this plausible?
Doesn't sound plausible to me because I think the feedback loops involved in brain cells influencing each other are much more complex than that. Seems like your model just involves dopamine sensors and dopamine producers, with the pair of types working ok to upregulate when there's depression, but doing a bad job of sensitive adjustment when the depression starts to lift. The real loops involve neural nets and many nested feedback loops.
Oh certainly my model is simplistic but don't most models of mood disorders focus on neurochemical dynamics as opposed to the connectome? Isn't that why mood stabilizers are much more effective at treating bipolar than CBT?
Lithium et al are definitely more effective than CBT, and anyone who has dealt with a manic person. comes away with the feeling that what is going on is very powerfully driven by some biochemical process, and is impervious to even quite clever attempts to access and tweak the person’s state of mind, cognitive biases etc. I don’t know a thing about how lithium works, and maybe it does have a whole different mode of action from what i’m picturing. I don’t see how, though. Those neuro chemicals are the stuff the cells in the neural net squirt out and respond to. Also, last I heard nobody knows how antidepressants work. By the time the person begins to feel better, they are no longer doing the thing people thought would have an antidepressant effect (for SSRI’s that would be slowing down the process by which the cells slurp the serotonin back up). MAOI’s, which are actually the most powerful antidepressants, came to be used for depression not because somebody cooked them up with the idea that tweak brain chemicals in an anti-depression direction but because people who gave the stuff for tuberculosis. which it combats pretty effectively, noticed that many people taking it had big mood lifts.
I don't know that that would be a helpful model. Depressed then manic isn't a particularly common pattern of symptoms - the classic pattern is manic then depressed.
There's also the usual wastebasket diagnosis phenomenon where a bunch of different illnesses are being lumped together, particularly in community mental health settings. A shocking percentage of bipolar diagnoses in datasets are labels for "was a teenager" or "has borderline personality disorder" or "saw a psychiatrist who read too much Ghaemi" or "was on meth." And even once you filter it down to people who clearly have the disorder you run into the division between the classic Cade's Disease lithium responsive cohort and the various other phenomena.
My preferred model for bipolar disorder is clock genes and circadian rhythm disregulation. In my family with classic bipolar 1 (we've got 4 of use who score a perfect 100 on https://www.ohsu.edu/sites/default/files/2021-10/Bipolarity%20Index.pdf), the people with bipolar disorder have notably deranged circadian rhythms and sleep cycles, even before phones and late night blue light. Sleeping 16 hours when depressed, sleeping 0-4 when manic, non-24 sleep cycles when unmedicated or in mixed episodes.
>the classic pattern is manic then depressed.
Really? I thought the typical pattern, at least for bipolar 2, was long periods of depression punctuated by shorter hypo-manic episodes. Does bipolar 1 typically start with mania?
Yeah I'm sure it's more complicated than what I said. It was just a brainstorm that seemed like it generated the swings from simple homeostatic considerations.
There's a whole mess as to what bipolar 2 is - it's been used as a wastebasket diagnosis for so long that there's some amount of debate in the field as to whether it's a real phenomenon and what that real phenomenon even is.
Bipolar 1 has a specific pattern for most patients, and our manic episodes are extremely easy to distinguish from any number of other psychiatric phenomena, and the longitudinal course is very different from the psychotic disorders mania can be confused with. In the 80s, researchers noticed that there were people with recurrent depression and subthreshold versions of manic symptoms whose depression responded well to medication commonly used for bipolar 1 and did not respond well to antidepressants, and thus the bipolar 2 label was born. https://www.ohsu.edu/sites/default/files/2021-10/Bipolarity%20Index.pdf gives an idea of what the classic phenotypes of the bipolar disorders look like.
However, it turns out that "checks all the boxes for a history of 1+ depressive episodes and 1+ 4 day period of mild manic symptoms" is not particularly specific for being a responder to lithium/lamotrigine and not SSRIs or therapy. Bipolar 2 as the label is actually used includes anything from "was a teenager" to "has borderline personality disorder" to "self reports mood swings" to "didn't tell my therapist about my stimulant habit" to "depressed person who remembers being happy once," and relatively few people with bipolar 2 in their chart cleanly fit the diagnosis.
Scott's tweet says that America today is a weird place for a revolution, unlike historical revolutions. On the contrary, it looks to me like France 1789.
Not even close. France in 1789 was literally two days from royal bankruptcy at one point, and had been through one of the worst years in centuries for the harvest in 1788. People were hungry, and it had become clear even to the ultra-royalist reactionaries that something had to be done.
US in 2024 had a 3% unemployment rate, a booming economy (so strong that it absorbed a couple million extra immigrants quite effectively), rapidly advancing technology, declining crime rates, and even heavily declining inflation.
We aren't two days from bankruptcy, but we aren't two decades from it either. Our debt levels are higher than during WWII.
Much lower than Japan though, which has had higher debt loads than ours for more than two decades…
I've been thinking about US debt lately. It occurs to me that Trump had a very specific and successful strategy as a businessman -- he would talk a great game (endless evidence he is good at convincing people, ie he has a powerful reality distortion field). He would use his talents of persuasion to leverage so heavily, with such unrealistic promises and projections, that his deals all eventually went bust. Now his business partners start thinking rationally, and rather than sue him to oblivion (he's so leveraged they wouldn't get much) - they basically have no choice but to take cents on the dollar. (ref Getty, "If you owe the bank...)
Isn't he basically doing the same thing on a worldwide scale? I suspect the budget will get passed with the tax breaks and all the candy he wants; Social security and Medicaid won't get more than a pro-forma "waste fraud & abuse" reorg, and the difference will be massive deficit spending as far as the eye can see.
At which point, he can run his game: basically take the country into bankruptcy, through a combination of inflation and possibly direct negotiation with our biggest creditors (can you say China?)
The tariffs start to make some sense: they are a stick he can use to beat countries into submission, and pit them against each other to the point where they are begging him to give them a few more cents on dollars owed than the next guy.
Frankly, I think the guy is a sociopathic, evil genius. His evil plans just might work.
In what way is the contemporary United States like France in 1789? I am honestly not at the moment seeing even a single similarity between the two, at least in the ways that are relevant to revolution. (I’m ignoring the trivial things like “both of these are a place at a time” and “both of these had a strong majority of the population speaking an indo-European language”).
I don't agree with the analogy, but for fun, here are some similarities.
- People are complaining about the cost of gas and groceries. It's not at the level of near-starvation which led to rapid increases in the price of bread in 1789, but one could argue that people's awareness of the increases and ability to compare their lot internationally would lead to the average person being unhappy with their lot.
- There's an excessively rich and out of touch political elite in the US, arguably built on an out-of-date and unresponsive political system. Could one imagine Melania saying 'let them eat cake'? Or Donald surrounded by courtiers in Versailles ignoring a developing political crisis?
- It's been too long since I read 'The age of Revolution' by Hobsbawm, but I remember him discussing rapid technological change and the economic imbalances brought on by the Industrial Revolution and the move to urban living. I'll wave my arms here and say one could draw analogies with globalisation, the digital economy, and AI leading to a disenfranchisement of workers and concentration of wealth.
People have complained and there have been rich out of touch elites since roughly 1890. One could even imagine whoever's leading the country saying or doing dumb things that the poorer people would find outrageous. So this by itself doesn't convince me there's anything special about 2025.
You really don't think there's any difference in the public perception of the Trumps in 2025 and, say, the Obamas in 2015?
I think there's copious difference between perceptions of the Trumps and the Obamas from the perspective of individuals. I also think that if one could compute the integral of such perceptions over the entire population, there will be amounts of "they're one of us" and "they're out of touch elites" for both, and they'll be close enough in magnitude that anything one wants to say about one, ought to apply to the other.
I checked - Trump has 44% approval, in May of his second term, Obama had 47% approval. So I guess your prediction on US perspective of the two presidents is right. International coverage isn't comparable.
Are you seriously comparing current USA with France in 1789?
How so? No one starves to death here. Do you think the peasants in 1789 would've revolted if they'd had iPhones and YouTube?
We're not gonna revolt, we're gonna collapse. America is in the second half of Atlas Shrugged, not the French Enlightenment.
Is there a real-world model that you're envisioning for this "Atlas Shrugged" type collapse? Which country has experience it in the past?
All of Europe, 1900-present?
I'm trying to figure out how you could describe a time period that includes two world wars as an Atlas Shrugged-esque slow decline. (Or a continuous movement in *any* direction, given that both wars caused dramatic economic and political shifts.)
Like, if your definition of "collapse" covers "winning the war against the Nazis," then I think you need to explain your definition a bit more.
Look, if you want to define "Atlas Shrugged Style Slow Collapse" sufficiently narrowly then nothing will ever be sufficiently slow or sufficiently Atlas Shrugged-ish.
Europe's decline from unquestioned masters of the globe to an economically stagnant but moderately pleasant corner of Eurasia that's good for holidays did go in fits and starts, with World War 2 being one of those fits, a dumb socialists-vs-socialists war that managed to destroy a huge slice of what was still good on the continent.
Look, it's a better analogy than "present day USA is like France in 1789" anyway.
If *I* want to define it? *You* are the person who made the claim, you're the one who needs to define your terms. What do *you* think is the defining feature of an "Atlas shrugged style slow collapse," as opposed to other forms of collapse, or other forms of decline that are not collapse?
>Look, it's a better analogy than "present day USA is like France in 1789" anyway.
That is the lowest possible bar you could set for yourself.
I mean, they just got eclipsed by much greater powers in the US, Russia, and now China and probably India. There was no realistic way around that unless they actually went full "United State of Europe" with a pan-European state and common identity with US-level economic integration, and they couldn't bring themselves to do that.
Rome?
The western Roman Empire collapsed because it was an over-centralized military dictatorship that couldn't handle what were actually not THAT large of threats because the western Empire was constantly being undermined by civil wars and a weak center in the Imperial Court.
And because of bad luck. If one or two things had gone slightly their way with the Vandals, for example, we might be talking about the "Crisis of the 5th Century" instead of the fall of the Western Roman Empire.
There were civil wars going all the way back to the Republic, that was just how politics operated back then. The reason they became so much more destabilizing in the 3rd century is because of economic strife. This is evidenced by the heavy debasement of the currency prior to the 3rd century crisis. There's a fairly extensive literature on this.
Look, anything as complicated as the decline of the Roman Empire is going to be multi-factorial, but I think the minimal explanatory model is one of disordered price signals. Those have both economic and social dimensions, as moral norms and social mores function as informal price signals for the allocation of social and political capital. When coin content no longer reflects intrinsic value, or when patronage replaces merit as the currency of favor, the system’s feedback loops break down. From an abstract perspective society is just an information processing machine that coheres and directs itself via mostly-decentralized signaling. Even centralized institutions are only as good as the information they have access to. When those signals become systematically detached from the underlying realities then the system will inevitably collapse. In my view this is what Nietzsche was talking about when he complained about slave morality being responsible for the decline of Rome, but of course he thought about it in moral rather than economic terms. (Although I think that's a superficial distinction as I believe that morality is more deeply tied to economics than most people appreciate.)
Interesting; I'm not arguing but I'd be curious to hear why you think economic disorder is the minimally explanatory model.
To me, each of: the successive Antonine and Cyprian plagues; the rise of the Sassanids; and the increased pressure from Germanic tribes seem like more natural candidates for the minimally explanatory factor.
As another point, it feels to me like the crisis starts to end before the economic situation is stabilized: Aurelian is the clear turning point, and but don't Diocletian's price controls indicate that the economy was still fried a quarter century later?
As you say, it's multi factorial and obviously economic collapse is part of it, but it's easier for me to see those other factors as more important, and limiting the ability to restore economic stability because of the breakdown of trade, etc.
You think Rome collapsed because people stopped being self-interested enough and turned into unproductive socialists...?
Not so much that, but a dictatorship took over can made it too hard for people to effectively pursue cooperative competition to mutually pursue their interests.
The dictatorship happened because a sufficient percentage of high status people didn't want to do useful work.
Not socialism specifically but cultural and moral decay combined with disastrous economic policy. So directionally, yes.
What if neoliberal capitalism is the disastrous economic policy that will destroy the west.
Is this really much more specific than saying "in the sense that they were both bad"?
It wasn't the starving peasants who revolted in 1789.
Certainly the Flour War 15 years earlier was relevant. Certainly Scott is right to say that bread lines were relevant in 1991. But food wasn't definitive in those places, either. Soviets were rich enough to drink themselves to death, with life expectancy peaking in 1960 and declining until Gorbachev restricted alcohol in 1985. Wealth leads to revolution.
Urbanization leads to revolution; the number of successful peasant revolts can be counted on one hand. Once you have cities, material deprivation becomes VERY important.
No, the expectation of gain combined with enough poverty to tolerate the high variances of war leads to revolt. Only people who have nothing to lose are willing to risk death for a political cause.
In France the third estate paid > 90% of the taxes. In the US, the top 10% of earners pay 70% of all Federal Taxes. Who's gonna revolt? The rich are the ones being exploited but the rich have too much to lose. They'll just leave, much like in Atlas Shrugged.
This is a weird comment. You mention Ayn Rand, but doesn't she say this is all wrong? In Atlas Shrugged it wasn't the rich who left, but the productive. The whole point is that they aren't the same. Taxes to the central state weren't a problem in either 1789 or today. The problem was that the state allowed parasites. 1789 wasn't a revolt against the king, but against the nobility. The money that the state steers to parasites was and is much bigger than the money that the state collects. Prevention of production is much bigger than that.
Elon Musk doesn't pay taxes. He didn't leave California because of the taxes. He left because it wouldn't let him work. His complaint about the federal government is that it stops him from being productive, not that it exploits him.
I don't think Musk would be a hero in Atlas. At best he'd be a complicated ambiguity - a cynical con man who exploits the diseconomic cracks in the system. I don't think Musk is productive. I think he represents a shallow kind of performative capitalism of the kind that Rand would've hated. He built his empire on the back of government subsidy, a cult of personality, and meme stock investment fads. Tesla is/was probably the most overvalued company in history, the archetype of an economic Potemkin village. And actually Trump is exactly the same. He isn't a businessperson in any meaningful sense: he inherited a billion dollars and used it to purchase a real-life Monopoly set, then got famous for being an archetype of the thing that he was only performing. Trump and Musk aren't successful capitalists any more than Elizabeth Warren, Kamala Harris, or Barak Obama are oppressed minorities. Both parties now worship flimsy cutouts that symbolize the values they claim to care about without thinking for a second about what those values really mean.
As for your analysis of the French Revolution I just disagree. Extractive taxation and parasitism aren't really separable in that context - one leads to the other. And American parallels to Atlas are tilted slightly: I think wealth and productivity are still reasonably correlated here, the problem is that the political class has become completely decoupled from notions of fundamental value. And yeah you have to squint a little to put all three things in the same box but what do you want, it's just a metaphor. The thread that ties them together is a delusional elite that worships irrational diseconomic values not understanding that they're slowly poisoning the garden that feeds them.
They pay only about 50% of all federal taxes. Include all taxes, state and local, and it's much less progressive. As to how it feels, consider declining marginal utility. The people I know in that bracket seem materially happy.
Yes, that's why they won't revolt.
LOL to "only".
The top 1% types that I personally know are uniformly resentful of the US tax system and many are planning on moving abroad in the near future. Not that the European tax structure is less progressive but they spend the revenue more intelligently. Here it's all transfer payments to ineffective entitlement programs. There's higher social cohesion over there and therefore less social fraud. At least you get something for your tax dollar there.
The US is perilously close to having a voting majority that pays zero income tax. That's a disaster, particularly in the context of a culture that feels entitled to treat the successful as piggy banks that can be used to underwrite the poor life choices of the hoi polloi. Once that balance really tips over the wealthy will flee like rats from a sinking ship. Given our rising debt levels and changing demographics I think that's a realistic scenario sometime over the next 20 years.
I've learned somewhat not to use my imagining of the future, short or long term, as feed for confirming my takes. As yogi may have said, "predictions are hard, especially about the future".
I remember some previous Open Threads from early in Russia's Ukraine debacle, where some doubted that drones would be able to blunt the massive deployments of Russian armor. Does anyone still doubt that drones have permanently changed the nature of warfare?
1. First off, an interesting thread compiled from the commentaries of Russian bloggers on how the insect screech of drones has become the defining sound of the Russia-Ukraine war. Drone noise has become a form of psychological warfare.
https://t.co/uFgw74atan
2. Ukraine is laying down a 15-20 km “kill zone” of autonomous drones along their defensive lines to deny Russian forces the ability to move undetected across the front. And it's trying to extend it to 40km. And it seems to be working. I'm having trouble getting Threadreaderapp to unroll this thread — apologies to those who are allergic to X.
https://x.com/DVKirichenko/status/1915439099588329809
3. Because of drones, tanks have been relegated to the role of mobile artillery.
https://cepa.org/article/the-era-of-the-cautious-tank/
And this made me snicker: "Drawing lessons from the Ukraine conflict, Driscoll emphasized the need for tanks to adapt to the increasing threat posed by inexpensive drones. This strategic pivot underscores the Army's commitment to *preserving its armored capabilities* while addressing emerging battlefield challenges" [emphasis mine]. US strategists still don't seem to have internalized the idea that tanks may be going the way of the Bronze Age chariot. But Driscoll still has confidence in his chariots.
https://armyrecognition.com/news/army-news/2025/exclusive-us-army-reevaluates-tank-warfare-strategy-amid-drone-threats
4. Ukrainian sea drones have driven the Russian Navy off the Black Sea back into their ports. And now sea drones have used AIM-9 Sidewinder missiles to shoot down two Russian SU-30 fighters.
https://cepa.org/article/ukraines-marauding-sea-drones-bewilder-russia/
https://www.twz.com/news-features/two-russian-su-30-flankers-downed-by-aim-9s-fired-from-drone-boats-ukrainian-intel-boss
I would be wary of claims/dreams of “Wunderwaffen”. Russia is advancing every day:
https://x.com/Kalibrated_Maps
Counter example 1) Iran twice has launched multi-hundred drone assaults against Israel to no avail. 2) the recent India - Pakistan unpleasantness also saw multiple drone swarm attacks. It’s less clear what happened there because of fog of war, but it seems to me that on both may 8 and 9 Pakistan launched multi hundred drone assaults where none of the drones got through. Most of the Indian drones launched at Pakistan also appear to have been neutralized, although not all.
Maybe drone warfare is mostly just a nuisance against a competent air defense? (Corollary, Russia does not have a competent air defense).
These are different things. Drones used for some kind of stand-off strike with no ground component, drones used to facilitate a ground assault, and drones used to deny an enemy's ground assault.
The first two have been pretty underwhelming so far, the third has been transformative.
Why the asymmetry between (1,2) and 3? Is it because advancing forces in a ground assault don’t have embedded air defenses against drones? Why can’t they? Are we over indexing on data from a single conflict against a single somewhat incompetent adversary? Like, maybe the Russians problem is just that they are unable to do combined arms warfare, where ‘combined arms’ now includes drone defense.
Even in the Iran case, you cannot really be sure what happened. Of course Israel denies any serious losses, whether they happened or not (I mean, they do it all the time, admitting only what is impossible to deny), and the US, which is otherwise happy to publish satellite images etc. on Russian or Iranian losses, remains silent when it is about Israel.
Don't get me wrong, this is all fine and understandable, but it also means that public information is limited. Sarcastosaurus, for example, concluded that the Iranian attack was successful in the sense that it caused enough damage to satisfy the Iranian public, while being moderate enough to avoid unwanted escalation of the conflict.
link?
IIRC these were the ones (but it was a year ago).
https://xxtomcooperxx.substack.com/p/israeli-iranian-soap-opera
https://xxtomcooperxx.substack.com/p/on-terrorists-and-fascists
I cannot support it with a link anymore, but I vaguely remember that the US alone spent like USD 1M in this particular operation.
PS very interesting blog.
I agree.
Can't say it is not biased, but it is not extreme.
thanks
I was thinking about this. It seems to me that the fact that drone swarms are what is being tried here is already a big part of the change. This is very different from Obama era drones that worked much like a conventional aircraft bomber. Even if it’s mostly not getting through, it’s very different in kind, and it’s cheaper.
The Obama era drone use was against opponents with literally zero air defense. WW1 era aircraft would have worked as well.
Israel has used MBTs to great effect in urban combat in Gaza. By all accounts the tanks were decisive in many battles.
Active Defence systems like Trophy have made a huge difference with regards to AT weapons like the Kornet.
Admittedly, Hamas' drone technology is probably not the best, but if you want to discuss the future of MBTs in the battlefield, you should discuss the way Israel uses them, not the way Russia does.
In early 2000, the IDF cut its Armoured Corps in half. Following the current war, it has admitted it was a mistake and is backtracking and expanding the tank brigades considerably.
Drones are absolutely changing battlefields. Including against said Russian armour and not just that.
"'No Russian Tank Would Survive': German Leopard 2A4 Withstands 10 FPV Drone Strikes in Ukraine"
https://daxe.substack.com/p/no-russian-tank-would-survive-german
"'FPV Is Close.'"
https://daxe.substack.com/p/fpv-is-close
"Factory-to-Frontline Pipeline"
https://www.warquants.com/p/factory-to-frontline-pipeline
"The Bears Robot Soldiers – Russian Drone Development"
https://modernwarfare.substack.com/p/the-bears-robot-soldiers-russian
"An Armed Russian Drone Shakes Latvia’s Defense System"
ttps://fpribalticinitiative.substack.com/p/an-armed-russian-drone-shakes-latvias
That said, drones are not the only component of current and future battlefields.
"Last-Mile Targeting in Ukraine’s Drone War: AI, Edge Computing, and the Limits of Autonomy"
https://xxtomcooperxx.substack.com/p/last-mile-targeting-in-ukraines-drone
"Drones Are Winning on YouTube, but…"
https://blackcloudsix.substack.com/p/drones-are-winning-on-youtube-but
"Droning On About Drones"
https://crackingdefence.substack.com/p/droning-on-about-drones
I think the general points in this article are still relevant, despite it being a few years old: https://www.navalgazing.net/The-Drone-Revolution
(Although note the post-script where he acknowledges drones having more of an impact in infantry operations.)
A persistent factor in the Russia-Ukraine conflict is that Russia's military is surprisingly bad at stuff. They've lost a ton of soldiers and been forced to come up with new (and therefore inexperienced) ones, and they keep relying on meat-grinder assaults to make progress despite heavy losses. This is bad if you want to avoid getting blown up by drones, but also bad if you want to avoid getting blown up by artillery or airplanes.
I can imagine a future totalitarian state policed by drones.
I expect tanks to morph into self-driving/FPV combo land drones.
Keith Laumer had a lot to say about the concept!
Good pull.
Today's main battle tanks are (to a significant degree) meant to fight other MBTs. This dictates their size. If you want to prevail in a direct fire pissing contest, you generally do not want to send twice as many tanks with half the armor and gun penetration.
I do not see a lot to be gained by eliminating the crew. In fact, MBTs have a crew of four or so, including a loader who is presumably employed for mechanical rather than cognitive tasks. (That being said, many MBT designs are from the 80s, when "can we build a machine to replace the loader and then replace the crew cabin with a somewhat smaller housing for a computer" was SciFi.) Sure, eliminating the crew will allow you to reduce the armored interior volume, which is expensive, but you will still need to store the ammo, the loading mechanism, the computer, the fuel tank and the engine somewhere, so there are no huge gains to be had, I guess.
If autonomous tracked vehicles gain dominance, I do not think they will be an evolution of traditional tanks. I am not sure they generally have much of an edge over flying drones, though. Sure, they can operate indefinitely instead of an hour (or however long the batteries will hold) and can pack much more firepower at cost parity. But they also have much lower mobility. The problem of flying a thousand drones through the forest seems much more solvable than the problem of driving a thousand drones through rough terrain.
I kind of envisioned a crew-less tank, still a huge beast but with the ammo and robotics taking the crew's space. What is gained is not having to worry about personnel losses.
With modern Western tanks, you already don't have to worry about personnel losses. You can if you want to, but if you're fighting an actual war there are so many bigger things to worry about that this would be silly. There are weapons that can *disable* an M1A2 or a Leopard 2A6 or a Merkava, but weapons that can destroy one outright are cumbersome enough that almost nobody ever bothers, but settles for the mission-kill. And the crew almost always walks away from that.
Russian tanks (and most of Ukraine's tanks are based on old Russian designs), are an entirely different matter. But the *reason* Russian tanks blow up forcefully enough to loft their turrets two hundred feet skyward any time the armor is penetrated, is because Russian designers a generation back were over-enamored of automation, and the automated gun-loading systems were rather less robust than a guy who knows what he's doing and would really rather not die today.
Interesting, I admit my understanding of tanks’ vulnerabilities is largely based on watching too many videos of Russian tanks blowing up in Ukraine. But still wouldn’t it be nice not to have a human crew at all?
In the sense of "wouldn't it be nice to not have soldiers at all, just turn wars entirely over to the robots", perhaps. But there are problems with that approach.
If you are going to have soldiers anywhere in a war zone, the inside of a modern western main battle tank is one of the safest places for them to be. Possibly even safer than a command bunker fifty miles behind the lines, because that bunker is going to be a much higher priority target and not that much harder to find and hit.
If British tanks didn't have human crew then who would use the tea kettle?
The real interesting part is that, if anything, I see drones getting more dominant in the near future. Some of the most popular current anti-drone techniques are jamming the signal from the operator, or jamming the GPS connection, but even with current "dumb" AI tech that can run on a smartphone, it seems trivial to operate based on pre-specified instructions and pre-loaded satellite images ("fly to this area and if you find a tank, blow it up").
Some of the *other* popular anti-drone techniques involve automated kinetic weapons (simplistically and sometimes literally "guns") that shoot drones out of the sky before they can come close enough to do any real damage. Drones are slow, fragile, easy targets, rather like skeet. These systems are heavy enough that you'll really want a large vehicle to carry them, and the potential for collateral damage is enough that you'll want your people to be under armor when that happens.
So, the future of tanks is, drones keep exploding every time they come within fifty meters of one. The future of everything that *isn't* a tank, is that it explodes as soon as a drone operator realizes "to hell with that tank, I can actually *kill* this SOB". A bit of an oversimplification, but I think closer to reality than your version.
I assume they'll just have an anti-drone* vehicle as part of the tank's screen forces, which would cover both the tank itself and any APCs moving along with it.
* Or just generally a vehicle to provide active defense against threats. I think tanks are going to go the way of battleships and lose some armor for greater speed and range this century.
An anti-drone system that can cover an entire mechanized company's front, is going to need to have a much longer effective range than an anti-drone system that just defends a single tank. And the latter can probably do double duty as the tank's defense against anti-tank missiles. Meanwhile, the dedicated anti-drone vehicle will be vulnerable to everything that *isn't* a drone, and it will be constrained in its deployment by the fact that it needs to maintain constant line of sight to all threat axes against all of the vehicles in its charge.
If at all possible, and I'm fairly certain it's possible, a short-ranged anti-drone-and-missile system on every tank (and maybe IFV) is I think preferred.
Tanks have already been written off in 2022 and 2023 when Javelin & Co got all the praise. It was too early to write off tanks then, and it's still too early today. The combination of mobility, firepower, and protection is a potent one and will always have a place on the battlefield; tracked and armored fighting vehicles will be replaced if and when some new technology makes them obsolete, but not before. Right now drones have the upper hand, but that's just the nature of warfare - it's an incentive to invest more into anti-drone capabilities (new tech & tactics), and that in turn is an incentive to invest in anti-anti-drone capabilities, and so on.
How could tanks defend themselves against drones? Heavier armor reduces the range and/or speed of tanks, but current APVs can carry 5 Kg of explosive. According to ChatGPT (AI warning!), 5 Kg of C4 can puncture armor that's 10 cm (4 inches) thick. The thickest armor an M1 Abrams is 700 mm thick. Likewise, reactive armor could resist a 5 Kg C4 explosion, but it can only resist it once.
Drones are low-cost compared to tanks, and the enemy can send in dozens to swarm a tank to overwhelm the tank's defenses. Videos show Abrams tanks being taken out of action by Russian drone attacks in Ukraine, which demonstrates the effectiveness of this tactic. Ukraine has developed an interesting net-ejector system that tangles drones as they approach, but, at least right now, that can only be used once. It wouldn't work against a swarm of drones.
That's why you shouldn't operate a tank unsupported, as both Russia and Ukraine are prone to doing due to their Soviet upbringing and somewhat limited capabilities. If you expect the enemy to use drones or drone swarms and your tank can't handle that on its own, you give your tanks the support to deal with that threat, such as an ECM suite, airburst anti-air [1] for hardkills, or perhaps a more futuristic, energy-based defense system like lasers, should that prove practical on a battlefield.
[1] https://en.wikipedia.org/wiki/AHEAD_ammunition
Seems like something like a sensor-driven shotgun could work. Kind of like the Navy's Sea Whiz anti-missile defense but lower caliber.
How can tanks defend against anti tank missiles?
How can a man defend against bullets?
Just because something can be destroyed does not mean it's not useful.
Out of curiosity, how do tanks defend against Javelin missiles?
For Javelins specifically, it mostly comes down to A: use artillery to suppress the people who would be launching Javelins against the tank and B: if that doesn't work, hope that your active protection system (https://en.wikipedia.org/wiki/Active_protection_system) can shoot it down before it hits, and C: if you don't have one of those (they're fairly new) hope you are fast enough to hit the button that fires a bunch of smoke grenades to create an insta-smokescreen that will block the missile's guidance system.
Against pretty much every other sort of antitank missile, there's also really, really tough armor. It is possible to make the *front* of a tank nigh-impenetrable to all but the heaviest antitank weapons, and mitigate the after-armor effects of those to the point where the crew probably still walks away. Javelin specifically can fly a top-attack trajectory, and it's not practical to armor the entire top surface of a tank to that level.
Thanks for the excellent answer!
Option A seems like it would be hard for a missile that's man-portable, can be vehicle mounted, and has a range of several kilometers. What use is the tank if you're already destroying threats before they get within a few kilometers of it? Would they be able to reliably spot the Javelin missile launchers and pick them off at that distance? Are the missiles just rare because they're expensive or heavy?
Reactive/explosive armor, lasers, jamming, chaff/smoke, combined arms doctrine that doesn't leave tanks by themselves.
Mostly the US just designs them to be very sturdy with redundant armor and sealed compartments that can explode outward instead of inward.
Javelins are a bit different because they are top attack, but they can be disrupted and reactive armor can be mounted on the top surfaces.
It seems the Javelins have two warheads in them, the first of which disables reactive armor.
Apparently smokescreens aren't that effective either because you have to both detect the missile early enough, and move the tank before it hits.
For reasons I don't fully understand, current APS systems can't target the Javelins' top-down attack.
From what I'm reading, the main defense against Javelins is just to not let the missile crew acquire a lock to begin with.
I am not sure that it is inevitable that the anti-drone tech will catch up.
Consider the cavalry charge. For millennia, it was a staple of land war tactics. Then one day, the machine gun was invented, and cavalry seemed obsolete. Some people might have claimed that it would be inevitable that cavalry would catch up -- armoring horses and riders or something -- but that never happened. It turned out that the armor required to withstand machine gun fire is not feasible for a horse, but it is in fact feasible for a motorized vehicle.
(Naturally, history is likewise full of just-so stories about weapon systems which were declared obsolete by someone only to be proven terribly wrong, e.g. "we can win wars through air power alone".)
My guess is that the main advantages of tanks are that they are mobile, impervious to small arms and a very cost-effective way to deliver firepower (compared to missiles).
If the best way to defeat a tank is another tank, then tanks will stay dominant. However, if you can defeat tanks with other weapon systems which cost a fraction of what a MBT costs, then the MBT will eventually die out.
A MBT is, from my understanding, ruinously expensive. The enemy has every incentive to outperform you in gun and armor performance, so you want to go big or go home. (Also, personnel is expensive, and generally will not survive the defeat of their vehicle.)
I will grant you that in principle, defending a tank against drones and missiles is easier than defending it against pure kinetic penetrators (fired from other tanks). Perhaps drone defenses will catch up and tank vs tank will remain a thing. Or perhaps the economics favor overwhelming the drone defenses with thousands of cheap expendable autonomous drones. Or perhaps someone will build a rocket-based kinetic penetrator which can be safely fired out of sight of the tank using real time satellite images.
There is a saying that armies prepare for the last war they were in. Thankfully, for the last 80 years, we have not seen open conflict between superpowers. We certainly did not have any large naval or armored vehicle battles using cutting edge tech. While this is nice for the humans involved, it also means that to a large degree, we rely on extrapolations and models. Both the military-industrial complex and branches of the military tend not to be perfectly aligned with the defense interests of their host nation in peace time. For example, if the future of air superiority belonged to unmanned drones (which can react faster than a human, can handle much higher accelerations and whose brains are cheaply replaced, at the cost of less versatility), I would not expect the US air force to realize that fact in the absence of peer conflicts.
Quite frankly, if we had stopped doing great power wars in 1870, I would half expect today's militaries still to drill cavalry charges.
I recently saw a Ukraine war video of a stealthy drone following a tank bank to its hiding place, then taking out the tank, the truck next to it and a lot of trained tankers and motor pool guys.
If that becomes more common, tanks will be a lot less useful.
It's better to think in roles and capabilities rather than counters. Economics are important, yes, but the first job of a military is to win a fight. To be cost-efficient is second to that. There is a job on the battlefield that is best fulfilled by tanks and that cannot be fulfilled by drones, and as long as that job is militarily relevant, the military will evolve to enable the tank to do that job, or find a replacement that can - if you simply stop using tanks, the job remains unfinished and there is a hole in your capabilities.
Also, drones are a general threat to everything on the battlefield and beyond, so the imperative to defend against them goes well beyond just to protect tanks. There already exist possible solutions to drones and even drone swarms (e.g. air-burst munitions), plus whatever people come up with when facing an existential threat, and it's only a matter of time until they are deployed at a scale that allows tanks and other vulnerable assets to do their job again. Maybe not every country will be able to afford these solutions at the necessary scale, and yes, those countries might be as screwed as if they showed up with chariots. But everyone who can afford to can and will invest in anti-drone capabilities, and the cycle of offense and defense will continue.
As far as I understood, WW2 was mainly won by outproducing Germany. Cost effectiveness is still a thing.
Of course, if your military can afford to outspend the enemy in 100:1 ratio (like Izrael can outspend Hamas), you don't have to worry about it that much.
Thanks! I hadn't thought of airburst munitions. But once they're expended, the tank would be a sitting duck. Swarm it with cheap sacrificial drones until it expends its airburst munitions, and then drop 5 Kg of C4 on the turret and see what happens.
Maybe tanks would be flanked by anti-drone vehicles loaded with sensors and anti-drone weapons, kind of like how carriers are flanked by a battle group.
Or the tank could travel with its own swarm of anti-drone drones that suicide detonate into an oncoming attacker.
But that begs the question: Why not just have swarms of cheaper suicide drones take out enemy positions? I'm still unclear about the unique function of tanks in 21st-century tactical engagements. My understanding is that tanks were used to overwhelm and/or outflank established enemy emplacements, and to make a path through the enemy that mobile infantry could follow. Seems like drones could perform the same function for less money. But maybe I'm missing something obvious.
https://www.tumblr.com/andrewducker/783048372986707968?source=share
For much of the Middle Ages, the Catholic Church wasn't hunting down magic users. The official view was that magic didn't work, and anyone who claimed it did was a fraud.
As might be expected with any high-minded, abstract, and contrary to popular opinion command, compliance wasn't happening.
A lot of magical practice was something congregations insisted on priests walking a host around fields to improve harvests, and compliance with the hierarchy just wasn't worth it.
The later, witch-hunting period was real and horrifying, but it didn't happen in the middle ages. The Catholic Church isn't just one thing, and I wouldn't even say its behavior is especially predictable.
From what I remember of my history classes, witch-hunts were more popular in Protestant areas or areas with mixed religions than in Catholic areas. (James VI & I famously wrote a manual on hunting witches, for example - I don't know of any equally famous Catholic who was as obsessed.) But they did happen everywhere.
I don't think that any witch-trial has been as well-studied as Salem, but I wouldn't be surprised if many of the European ones had similar features: that the cases were brought by members of the community rather than by the religious or secular authorities, that they came during or after times of intense hardship and distrust of neighbours, and that they were (within limits) tolerated by the authorities as means by which the communities blew of steam by scapegoating outsiders. I think you could bring René Girard into the discussion at this point...
"A lot of magical practice was something congregations insisted on priests walking a host around fields to improve harvests, and compliance with the hierarchy just wasn't worth it."
Oh, yeah. Folk religion shades into and out of magical practices very easily. Plus we are a material religion, so we do believe in the efficacy of things like prayers for rain. Blessing the fields and combining that with a Eucharistic procession wouldn't have pinged anyone's radar as "this is magic not religion" (until the Protestants got all neurotic).
My own pet eccentric theory is that the witch-hunting really kicked off due to Protestant neuroticism; the world was haunted by devils, your only bulwark was faith, and if your faith was insufficient you were vulnerable. All the supports, the pseudo-magical protections, had been stripped away and what was left was the bare word of Scripture (hence why Scripture itself began to be treated like a magical text). At the same time, you were told the Devil prowls about like a roaring lion seeking whom he may devour, and it was possible to make bargains with evil spirits to cause harm. How could you protect yourself, except to root out the malicious evildoers?
People seem to find it easier to believe in demons than gods. I don't know how many modern horror/dark fantasy stories I've read where the conventional belief system of Christianity is for the birds, but you bet the monsters or spirits of the old pagan faiths are real and ready to rip your lungs out.
>People seem to find it easier to believe in demons than gods. I don't know how many modern horror/dark fantasy stories I've read where the conventional belief system of Christianity is for the birds, but you bet the monsters or spirits of the old pagan faiths are real and ready to rip your lungs out.
Reminds me of a review of Buffy the Vampire Slayer I read once which complained that none of the characters seemed religious. After all, they live in a world where there are literal demons from hell which are defeated by Christian religious paraphernalia (crucifixes, holy water, etc.), and not once do they say, "Hey, if Christian objects are so powerful, maybe Christianity's true and we should all start going to church."
If I can believe the article, the Church did believe that walking a host around a field to improve the harvest *was* an attempt at magic and wouldn't work. They thought it was fraudulent.
I thought the Catholic belief in the power of Satan rose before the Reformation, but I could be wrong about that.
"If I can believe the article, the Church did believe that walking a host around a field to improve the harvest *was* an attempt at magic and wouldn't work. They thought it was fraudulent."
Not so much that (because processions and 'beatring the bounds' and the likes were embedded into parish life) but as mentioned in the linked Tumblr post:
"In some cases he also willingly shared a bit of the communion wine to the local cunning woman (who was also probably the midwife), since that was the best thing to mix into ointments and potions for everyday ailments. But only a little, of course, as it was precious stuff."
Absolutely you do *not* do that if it's consecrated wine, as that is the Blood of Christ and there are *very* strict rubrics about how to dispose of any leftover wine (the priest is supposed to drink it) and what to do if you wash out wine stains from altar cloths. Giving any dilution of the consecrated wine *is* magic and will get you into big trouble. If you're giving out the *unconsecrated* wine, that probably will get you into trouble as well, but not as much. I think maybe what they are talkigna bout here is not so much "a bit of the communion wine" as "water used to rinse out the chalice" and that too is a grey area.
For instance, when rinsing out communion vessels the water is supposed to go directly into the ground, not into the normal drains/sewerage (see "sacrarium"):
https://en.wikipedia.org/wiki/Piscina
"The purpose of the piscina or sacrarium is to dispose of water used sacramentally, by returning these particles directly to the earth. For this reason, it is connected by a pipe directly to the ground; otherwise presumably, a basin was used. At times the piscina has been used for disposal of other items, such as old baptismal water, holy oils, and leftover ashes from Ash Wednesday.
In the Catholic Church, pouring the consecrated wine, the Blood of Christ, or the Host down a sacrarium is forbidden. Extremely rarely, the Eucharistic species spoils or becomes contaminated such that it cannot be consumed. The host is then dissolved in water until it disappears, and then the water is poured down into the sacrarium. In accordance with what is laid down by the canons, "one who throws away the consecrated species or takes them away or keeps them for a sacrilegious purpose, incurs a latae sententiae excommunication reserved to the Apostolic See; a cleric, moreover, may be punished by another penalty, not excluding dismissal from the clerical state."
So if in the first instance we mean not "having the full bells and whistles procession of the Host in the monstrance" but rather "giving crumbs or pieces of the Host to be scattered on or buried in the fields" hoo boy yeah, magic, big big deal.
It’s kind of amusing that in contrast to the Chritianity, in Vajrayana Buddhism, crumbs of terms left over after took are to be fed to birds.
This is correct. Until the Renaissance, witch trials were rare and were almost exclusively conducted by civil authorities. In general, the medieval witch trials were otherwise-ordinary murder trial where the killing was alleged to have been done by means of magic.
Even in the Renaissance, it wasn't just Catholics doing witch hunting. Protestants did a lot of witch hunts, too. And at least one Orthodox country (Russia) did thematically-similar stuff in the same time period, but they were hanging sorcerers instead of burning witches and didn't generally consider their sorcerers to be devil-worshippers the way Catholic and Protestants thought of their witches. And in this time period, it was still usually civil authorities doing the burnings: the Dominicans in Catholic countries and various religious groups in Protestant countries did search for witches and do torture to force confessions and compel testimony against other witches, but they almost always turned unrepentant witches over to civil authorities for trial and execution, and a lot of civil authorities did witch hunts on their own.
Popular conceptions about witch hunts, especially in English-speaking countries, seem to be rooted in four sources:
- The Murray Thesis. This is a theory proposed in 1921 by Margaret Murray, an archeologist specializing in ancient Egypt. The theory was that only elites converted to Christianity in Roman times while commoners mainly remained pagan through the Renaissance and well into the Early Modern period and specifically practiced a common Europe-wide pagan tradition centered on a "horned god" which Christians identified with Satan. Under this thesis, the witch hunts were attempts by elites to destroy this pagan tradition. The Murray Thesis got taken seriously by mainstream scholars when it was first proposed, but it has since been thoroughly discredited. In the meantime, though, Gerald Gardner relied heavily on Murrays books when formulated Gardnerian Wicca in the 1950s, which he based on the assumption that modern English folk magic, early 20th century English mysticism, and classical Druidism were all parts of the "horned god" tradition which he aimed to reconstruct and codify. Thus, Wiccans and Wicca-adjacent neopagans (many of whom are terminally online) often take the Murray Thesis as an article of faith. The idea of the "Burning Times" and of witchcraft being "The Old Religion" are references to the Murray Thesis by way of Gardner
- The exploits of Matthew Hopkins, the self-styled Witch-Finder General of England. He was a Puritan con artist who operated during the English Civil War. His technique was to turn up in a village, announce that he'd been appointed by Parliament to root out witches throughout the country, conduct a highly dramatic investigation, and hang whatever "witches" he found that hadn't been killed by his trials-by-ordeal. Along the way, he'd also collect a "tax" to pay for his services. In the lawless chaos of the Civil War, he was able to keep this for several years before dying of natural causes. About half the tropes (at least in Anglo-American culture) of how witch trials were conducted, stuff like "swimming the witch" and searching an accused witch's skin for distinguishing marks, comes mostly from Hopkins. The Witchfinder Sergeant character from "Good Omens" is a pun on Hopkins's title.
- The Black Legend, a body of anti-Spanish and anti-Catholic propaganda originated by Queen Elizabeth I during the 1585-1604 Anglo-Spanish War, and which took on a life of its own afterwards. The Black Legend featured highly graphic exaggerations of the Spanish Inquisition techniques and practices, painting Spain as a totalitarian theocracy. A lot of elaborate torture techniques that have little or no historical basis (the iron maiden, the pear of anguish, etc) come from Black Legend propaganda. The truth of Early Modern Spain and the Inquisition is plenty awful, but the Black Legend exaggerated it enormously. The Inquisition primarily targeted Jews, Muslims, and Christian heretics (especially Protestants), but it also hunted witches and was conducted by the same religious order (the Dominicans) who also did witch hunts elsewhere in Europe in the 17th century, so the Black Legend spilled over into popular conceptions of witch hunts. Most of the other half of the Anglo-American pop culture image of witch trials come from the Black Legend.
- Late 20th century popular edutainment tropes about the history of science in the Renaissance, which tend to heavily emphasize the trial and imprisonment of Galileo for heresy by the Roman Inquisition (also conducted by the Dominicans).
Did witch hunts primarily target women, and if so why?
Primarily but far from exclusively. There are a lot of proposed factors, but the big one I remember is that witchcraft was associated with folk magic. Folk magic isn't the same think as witchcraft (this being one of the major errors of the Murray Thesis), as there was a fundamental distinction between beneficial or neutral magic on one hand and harmful of malicious "black magic" on the other. But even though they're separate, people did make the obvious connection that if it seems someone has been doing black magic in the village, then the local wise woman whom everyone goes to for herbal poultices and protective amulets is going to be near the top of the suspect list. And that type of folk magic that got culturally associated with witchcraft was, in most of the relevant cultures, primarily a thing women did.
This is highly culture dependent, ranging from about 60% of accused witches in our records being men (Russia and Estonia) to around 90% women (Southern England). It looks like the norm in most places we have data from is about 70-80% of accused witches were women. The gender ratio for people executed for witchcraft, compared to that for people accused including those acquitted or otherwise not executed, ranges from about the same to somewhat more likely for women to be executed, in the relatively few areas that we have detailed data for both categories.
Other proposed reasons include: societal antipathy for norm-breaking women, anxiety by male decision makers about female sexuality (a lot of the lore around witches being highly sexualized), and the peculiar position of widows and never-married woman without surviving relatives in the Early Modern social order that made them both suspect and vulnerable.
If you want to read whole (relatively short) book on the subject, one of the texts for my witch hunts class focuses pretty heavily on gender dynamics. It's "Witchcraze: A New History of the European Witch Hunts", by Anne Llewellyn Barstow, published 1994.
The other two texts were:
- "The Witch-Hunt in Early Modern Europe", by Brian Levack, 1987, which focuses more on how the witch hunts were conducted and the legal, religious, and political factors influencing the witch hunters.
- "Witchcraft in Europe, 400-1700, A Documentary History", edited by Alan Kors and Edward Peters, 2001. This is a compilation of selected primary source excerpts from the period in question.
The book we were assigned in college history was _The Devil in the Shape of a Woman_ by Carol F. Karlsen. Amazon reports copyright as 1998, but that must be a later edition - I still have mine, and read it almost a decade earlier. It also focused on the gender dynamics, as well as economic (IIRC, most of the accusations were of roughly middle-class women by poorer women, implying that these were the richest rich they could eat).
A thing I've read (it may be hard to find the source) claimed that county by county research found that witch trials corelated with weak government authority rather than with a region being definitively Catholic or Protestant. Is that looking solid?
Governments actually don't like witch trials. It's too easy for anyone to be accused.
There's definitely a correlation, yes, but my professor's take was that the actual trigger of the witch hunts was large-scale social disruption. Witch hunts tended to occur following or concurrent with wars, plagues, and famines and to be concentrated in the worst-affected areas.
Whether or not governments liked witch trials depended on the government and the trials. Some rulers sincerely believed in the narrative that witches were a vast conspiracy that posed an existential threat to Christian civilization, and these rulers tended to push witch trials regardless of other factors. James VI and I of Scotland and England, for example, was conspicuously pro-witch-trial and one of the first things he did after taking the English throne was to vastly expand England's relatively narrow laws criminalizing certain specific forms of black magic.
Another interesting bit I remember from my Witch Hunt class is that Protestants and Catholics had a notable tendency to burn their own witches. That is, the defendants at witch trials conducted by Catholic authorities were overwhelmingly other Catholics and similarly for witch trials conducted by Protestant authorities. Catholics in Protestant countries and Protestants in Catholic countries did face persecution, but they were persecuted as heretics or recusants, not as witches.
I'd clarify a small point, that when the Dominicans tortured and forced confessions, it wasn't the *Dominicans* who were doing the torturing, oh no sir. Besides the bad look it would be for priests to engage in torture, the church didn't have the local resources to detain or imprison people (outside of church strongholds like Castello Sant'Angelo in Rome). That was the domain of the local, secular ruler, and a strong inquisition was only possible with the collaboration of secular authorities. That's why the Inquisition got so bad in Spain (very enthusiastic financial support and authority from Isabel and Ferdinand) but inquisitions elsewhere in Europe at the same time had no teeth -- the local rulers refused to lend their dungeons and torturers and police forces.
That's how you get Pico della Mirandola living it up it Florence even after his excommunication and a clan of barely-disguised pagan sun worshipers operating with abandon in Rimini for decades.
I'd forgotten that aspect of it, thank you. I remembered the Dominicans turning over unrepentant heretics and witches to be tried and burnt by secular authorities, but I didn't remember that the also delegated the torture.
The other important thing about the Inquisition's use of torture is that the Medieval Inquisition knew the failure modes of torture victims trying to guess their interrogator's passwords instead of telling the truth. They'd developed a big set of rules around the use of torture designed to mitigate this. To the best of my recollection:
- Torture was a last resort. Use persuasion, psychological manipulation, threats, and bribes first.
- Getting permission to torture someone required hearings and a fairly high standard of evidence against them, something like the modern standard of "clear and convincing evidence".
- There's a formal escalation sequence before torturing someone, including "showing them the instruments", where you give them a tour of the torture chamber and formally warn them that they will be tortured if they don't cooperate.
- Only one ride per customer. If a suspect is tortured and you don't get what you want out of them, too bad. You can never torture them again.
- Statements made under torture only count as evidence if the subject repeats them later on, under oath, in a formal court setting. Combined with the "one ride per customer" rule, this allows them a chance to recant what they said under torture without fear of being tortured again.
- Only specific techniques were allowed.
- Never lead the witnesses. Ask them questions in ways that don't give hints at the "correct" answers.
- Mix the real questions with questions that you already know the answers to. Ideally, the torture victims shouldn't know which are the real questions and which the controls.
These worked pretty well (by medieval justice standards) when applied to large-scale heresies like the Cathars, Lollards, and Waldensians. But when a vast underground network of Satan-worshipping practitioners of black magic turned up on the Inquisition's radar in the Early Modern period, they turned their sights on it and found, much to their consternation, that the safeguards were stopping them from getting the evidence they needed to dismantle the network of witches that they "knew" existed. So, given the dire threat the witches posed, they rationalized ways to bend practically all of their rules until enough accused witches started confessing and giving up the names of other witches.
Suspiciously similar, actually, to the Iraq-war anti-terror torture (actual) practices of the CIA. If you go back and read the official documentation when they approved "enhanced interrogation", the fascinating part of it all was that on its face almost everything written in the white paper doesn't seem all that absurd, or at least there was a seeming attempt to set some rules -- at least in the original source white paper I read, the vibe is very much about how these "enhanced" interrogation techniques are really only designed to be done once or twice at most, and there wasn't much point to doing more. Whether you believe this to be honest or not is a separate question, but the psychologists in question at least had no experience with torture to start with. The context for this was actually like, one specific guy that the CIA was convinced wasn't talking (since, some have argued he wasn't even an AQ member, but how much of this is on a legal-proof standard vs common-sense probability is still up for debate), it wasn't even a general policy goal.
What IS clear, is that despite the official wording and justification calling for its use sparingly and in extreme situations only, they went on to waterboard the guy something like 80 times in a single month, and of course to rely more and more and more on black sites, and to have "known" terrorists give up the networks they "knew" existed. One might say, "given the dire threat terrorists posed, the rationalized ways to bend practically all of their rules until enough accused terrorists started confessing." Got 2-3 people who know each other blaming each other, occasionally, under extreme torture? Baby, you got yourself a network going! Add some cash bounties to the mix?
There is a very interesting paper collected by Gwern that compares the use and effectiveness torture by the Spanish Inquisition vs. American intelligence:
"Inquisitors tortured for different reasons, with different goals, based on different assumptions, and in a social, political, and religious setting entirely alien to that of modern interrogators…
The Inquisition put in place a vast bureaucratic apparatus designed to collect and assess information about prohibited practices. It tortured comprehensively, inflicting suffering on large swaths of the population. It tortured systematically, willing to torment all whom it deemed to be withholding evidence, regardless of how severe their heresy was or how significant the evidence was that they were withholding. The Inquisition did not torture because it wanted to fill gaps in its records by tormenting a new witness. On the contrary: it tortured because its records were comprehensive enough to indicate that a witness was withholding evidence.
This torture yielded information that was often reliable and falsifiable: names, locations, events, and practices witnesses provided in the torture chamber matched information provided by those not tortured. But despite the tremendous investment in time, money, and labor that the Inquisition invested in institutionalizing torture, its officials treated the results of interrogations in the torture chamber with skepticism. Tribunals tortured witnesses at the very end of a series of investigations, and they did not rely on the resulting testimony as a primary source of evidence.
This systematic, dispassionate, and meticulous torture stands in stark contrast to the “ticking bomb” philosophy that has motivated US torture policy in the aftermath of 9/11… US interrogators expected to uncover groundbreaking information from detainees: novel, crucial, yet somehow trustworthy. That is an unverifiable standard of intelligence that the Inquisition, despite its vast bureaucratic apparatus and centuries of institutional learning, would not have trusted.
The Inquisition functioned in an extraordinary environment. Its target population was confined within the realms of an authoritarian state in which the Inquisition wielded absolute authority and could draw on near-unlimited resources. The most important of these resources was time… It could afford to spend decades and centuries perfecting its methods and dedicate years to gathering evidence against its prisoners… Should US interrogators aspire to match the confession rate of the Inquisition’s torture campaign, they would have to emulate the Inquisition’s brutal scope and vast resources… one cannot improvise quick, amateurish, and half-hearted torture sessions, motivated by anger and fear, and hope to extract reliable intelligence. Torture that yields reliable intelligence requires a massive social, political, and financial enterprise founded on deep ideological and political commitments."
Ironically, the "ticking time bomb" scenario that is usually given as justification for torture is also the one in which torture is least likely to be useful.
I think I took the Witch Hunts class in 2003 and a class on Renaissance and Reformation Europe from the same professor (Paul Hiltpold at Cal Poly in San Luis Obispo) in 2004. I remember him remarking that the then-public information about enhanced interrogation was following patterns he found very familiar.
That is really interesting. Thank you. You say early modern period. Would that be under what we call the enlightenment for instance? In other words was this a reactionary backlash against the age of reason?
What Erica says in her comment - the heyday of magic and mysticism was with the scholars, not the common folk engaging in superstitions and traditional practices. This was always there, you had them drawing up horoscopes and trying to straddle the line between permissible and impermissible (not for fortune-telling or trying to discover the future) usage of same, as well as sailing damn close to the wind where evoking angels (to help you get rich and get women) gradually over time starting evoking devils/necromancy (to help you get rich and get women).
It really kicked into high gear during the Renaissance, when all the old Classical texts were being rediscovered (including Classical era magic and religious systems which were then reworked into philosophy and mysticism - see the revival of Neoplatonism) and the Early Modern Period - the tale of Doctor Faustus comes to us via Christopher Marlowe in 1594:
https://en.wikipedia.org/wiki/Doctor_Faustus_(play)
That leads us into the reign of James I who had a lively belief in the supernatural and who, in the wake of stormy weather which prevented his new bride from sailing to Scotland and the subsequent Danish witch trials, set up his own tribunal of investigation into the alleged witchcraft plot:
https://en.wikipedia.org/wiki/North_Berwick_witch_trials
He was interested in witchcraft because of his interest in theology, which probably arose from his upbringing by those determined to make him a good, godly, Protestant (Presbyterian) king who would respect the clergy and the limits of the power of the monarchy. So he had it dinned into him about Scripture and the Devil from a very early age.
https://en.wikipedia.org/wiki/Daemonologie
"King James wrote a dissertation titled Daemonologie that was first sold in 1597, several years prior to the first publication of the King James Authorized Version of the Bible. Within three short books James wrote a philosophical dissertation in the form of a Socratic dialogue for the purpose of making arguments and comparisons between magic, sorcery and witchcraft, but wrote also his classification of demons.
In writing the book, King James was heavily influenced by his personal involvement in the North Berwick witch trials from 1590. Following the execution of an alleged sorcerer in the year 1591, the news of the trials was narrated in a news pamphlet titled Newes from Scotland and was included as the final chapter of the text. The book endorses the practice of witch hunting in a Christian society. King James may also have been prompted to write the Daemonologie by an anonymous critical pamphlet John Upaland which circulated in 1597 and alleged the Scottish court was bewitched."
At this point we are in the late 16th century, well into the Renaissance and Early Modern Europe, and far from the mediaeval period. The Enlightenment is peeking over the edge of the horizon, though it is still a long way away. The foundations of what will become the scientific method are being laid, as is the ferment about society and government that will spill over into civil wars at a later period.
"Early Modern" means the whole period after the Late Middle Ages and before the Industrial Revolution. It includes the Renaissance, the Age of Discovery, the Protestant Reformation, the European Wars of Religion, and the Age of Enlightenment. There are different schools of thought on its endpoints, but generally from some time in the 15th century until within a decade or two on either side of 1800. 1453 (Ottoman conquest of Constantinople), 1485 (Battle of Bosworth Field), and 1492 (Columbus's first expedition to the Americas) are popular start points. The American Revolution (1776) the French Revolution (1789), and the Congress of Vienna (1815) are popular end dates.
The witch hunts weren't a reactionary backlash as such. Nearly the opposite in some respects: the ideological underpinning of the witch hunts was downstream of Renaissance-era changes in how Europeans conceptualized how the world works. As Nancy alluded to in her original comment, the medieval Catholic view was that Satan was essentially powerless in the world except as a tempter and deceiver and that society is relatively static and exists as God ordered it. In the Renaissance, this shifted towards attributing to humans more agency in shaping the world, and at the same time Satan came to be seen as a more active and dangerous entity. This is essential to the fundamental idea behind the witch hunts, that witches represented an existential threat to European Christian civilization rather than (as they were thought of in the middle ages) small-scale local problems whose activities at worst endangered their own souls and their neighbors' health and safety.
The other significances of the early modern period are:
1. Massive social disruption. As I said in the other branch of this thread, witch hunts tended to happen in times of suffering, disruptions, and uncertainty. The European Wars of Religion were a really enormous factor here, as there was a period of about a century there were a bunch of overlapping international and civil wars that were massively destructive and wrecked lives, property, and social order in affected areas. More countries than not spent at least a decade (sometimes several decades) being fought over:
a. The 80-years war (1568-1648) in the Low Countries, also involving England, France, and the Habsburg realms. The Anglo-Spanish war (1585-1603) was an offshoot of this.
b. The French Wars of Religion (1562-1598), which also included England, Scotland, Spain, and several Italian states.
c. The Thirty Years War (1618-1648) in Germany, which also included the Habsburgs, France, Sweden, and Denmark. This also had several notable foreshocks of lesser scope and scale.
d. The Wars of Three Kingdoms (1639-1653) in the British Isles, including the English Civil Wars, the Scottish Covenanter War, the Irish Confederation Wars, and the Cromwellian Conquest of Ireland.
There were probably others I'm overlooking.
2. Big increase in state capacity. Secular governments got more powerful and more centralized, allowing large-scale witch hunts to be organized.
Thanks for this; I am digesting….
That was a lot of information you just gave out.
“witch hunts tended to happen in times of suffering, disruptions, and uncertainty.”
Yeah
I thought your response to Yarvin's response was very good.
Like you I was slightly cheered to learn that the Dems have started an Abundance Caucus and that as of last Friday it already had 30+ members. An equally positive sign is that the knee-jerks at places like The Nation absolutely haaaate it.
Meanwhile since Thursday afternoon,
-- another Trump-appointed federal judge ruled that the administration must immediately bring back another specific deportee from the El Salvador dungeon, and this judge added that she will start placing administration officials under oath for updates on their work to comply with her order;
-- Trump when asked live on-air whether he is obliged to defend the Constitution replied "I don't know";
-- the FBI Director denied under oath that the Constitution's guarantee of due process applies to noncitizens despite having Antonin Scalia's ruling to that effect read to him;
-- Trump appointed yet another Fox News personality to a high federal office;
-- Stephen Miller is convening White House meetings to explore a national suspension of habeas corpus;
-- Trump defended accepting a personal gift from the ruling family of Qatar of a half-billion dollar glammed-up 747, which he would graciously lend to the government for use as Air Force One until he leaves office at which time the plane goes with him. [Never mind that the Constitution's Emoluments clause _specifically_ bars the acceptance of gifts from any “King, Prince, or foreign State” without prior Congressional approval.]
I don't know the details of this case, but "due process" means different things in different contexts. Someone in the country illegally may be subject to due process, but that may only mean in practice that the government follow a standard practice in immediately deporting them with no trial or legal counsel. In fact, that happens a lot at the border. Illegal crossings where the people are rounded up and sent back, with no courts involved at all.
Do you have more details on the court case of the guy in El Salvador? My understanding of the SCOTUS ruling in the other case was that a court can't order the Executive to engage in any particular type of foreign affairs, negotiations, etc. At least your summary of the ruling seems to directly contravene that SCOTUS decision.
Banned for 24 hours.
Yes of course "due process" means different things in different contexts, and every federal judge ruling on this topic has noted that. The courtroom arguments about whether due process was honored are always facts-specific.
Your understanding of the various SCOTUS actions on this topic is muddled, which is entirely understandable and I've given up trying to read their tea leaves. The important point is that the SCOTUS has thus far issued only temporary orders related to this topic, allowing full judicial review to play out in the lower courts. That's a totally normal SCOTUS approach given that this isn't part of SCOTUS's constitutionally-specified "original jurisdiction". But the mediots mostly fail to understand any of that and persist in writing stuff up as if the Court has settled a big question when it hasn't at all.
Presumably at some point SCOTUS will have to consider and make a substantive ruling on the AEA and/or what exactly a district court can and cannot order the executive to do. That though is a ways off yet at best.
In the meantime this specific court case is not related to any of that. This case is about the current administration refusing to either honor or try to undo an existing federal court settlement. [Trump wants to be able to just do whatever without bothering about any legal or constitutional considerations, he thinks being POTUS is supposed to be like being a king.] This judge wrote, "this is just a breach of contract case. It is not a habeas case, or a case assessing the propriety of the government’s recent invocation of the Alien Enemies Act (“AEA”)."
https://storage.courtlistener.com/recap/gov.uscourts.mdd.457483/gov.uscourts.mdd.457483.253.0_2.pdf
https://www.politico.com/news/2025/04/24/trump-el-salvador-deportation-venezuelan-minor-judge-00307336
https://www.politico.com/news/2025/05/06/daniel-lozano-camargo-deportation-hearing-00331228
From a first reading of this situation, it does sound similar. A quiet fight between the Executive and Judiciary on exactly what a court can require a president to do. I don't think anyone is quite sure where the limits are. As with the previous case, there's an order to "facilitate" the return of someone the court says was illegally removed. No one seems to be arguing about whether the removal was actually illegal, but taking that as given. A more detailed definition of what that would require could easily run afoul of what SCOTUS did rule, while a less detailed one leaves a lot to the imagination.
Would you agree that a court order for the Executive to enter into negotiations with a foreign government to procure the release of a prisoner would be foreign affairs? Would you agree that it would therefore overstep what SCOTUS was willing to lay out? Is there an alternative order a court could implement that actually leads to the release of Daniel without requiring something along those lines? Maybe someone can be really creative, but I'm not seeing anything obvious and haven't heard any suggestions.
The only material difference I see in this case is that the DOJ tried to go through the process to make the case moot, which the court essentially said would not be sufficient. There's still the open question about what remedy the court can order.
US contract law, like the older English contract law that it's rooted in, is very clear that the party violating a contract must cure that violation; and, that courts are supposed to prioritize very highly the enforcing of that requirement.
And then quoting from this judge's most recent order:
"Of course, [her previous ruling] puts this case squarely into the procedural morass that has been playing out very publicly, across many levels of the federal judiciary, in Abrego Garcia v. Noem, No. 8:25-cv-00951 (D. Md.). As of the date of this order, discovery is underway regarding the government’s efforts to comply with court orders (including from the United States Supreme Court) to “facilitate” Mr. Abrego Garcia’s return to the United States. See id. at ECF 79. This Court is mindful of the Supreme Court’s reminder to afford the “deference owed to the Executive Branch in the conduct of foreign affairs.” [Given that this is a Trump-appointed judge quoting an order issued by Trump-appointed justices, I'm sure that she is genuinely "mindful" of that SCOTUS order.]
This judge then reiterates that " 'Facilitate’ is an active verb. It requires steps to be taken as the Supreme Court has made perfectly clear….Standing by and taking no action is not facilitation. In prior cases involving wrongfully removed individuals, courts have ordered, and the government has taken, affirmative steps toward facilitating return." [citations listed]
Therefore, "this Court will [now] order Defendants to facilitate Cristian’s return to the United States so that he can receive the process he was entitled to under the parties’ binding Settlement Agreement. This Court further orders that facilitating Cristian’s return includes, but is not limited to, Defendants making a good faith request to the government of El Salvador to release Cristian to U.S. custody for transport back to the United States to await the adjudication of his asylum application on the merits by USCIS."
Does that new order represent a court trying to direct foreign policy? I have no idea and neither yet does anyone else, pending the administration appealing this new order up to the SCOTUS.
Yeah, sounds like we're in agreement then. We have no idea if this order, or any other, actually meets the requirement to not interfere with foreign policy.
I suspect, when push comes to shove, that the courts will ultimately be unable to require the return. Perhaps if SCOTUS were 5-4 the other direction they'd find a way to make that order, but even then that's a tricky thing to require. It would mean that the courts have some identified right to overrule the executive in foreign policy, which is extremely ripe for abuse. An easy comparison would be a conservative Texas or Florida judge ordering the Biden administration to enter into negotiations with Mexico or another foreign source of illegal immigrants in order to "facilitate" the deportation of illegal immigrants. I would imagine that would not be palatable to many of the people who want the courts to overrule Trump now.
Sorry, then I misunderstood your point. Don't agree at all with that legal/constitutional analysis.
(Nor your "easy comparison", I mean come on now...if the administration tries to pose that equivalency in SCOTUS oral argument it will be a gift to the other side.)
>Trump defended accepting a personal gift from the ruling family of Qatar of a half-billion dollar glammed-up 747, which he would graciously lend to the government for use as Air Force One until he leaves office at which time the plane goes with him. [Never mind that the Constitution's Emoluments clause specifically bars the acceptance of gifts from any “King, Prince, or foreign State” without prior Congressional approval.]
In other news
>May 1 (Reuters) - A stablecoin launched by Donald Trump's World Liberty Financial crypto venture is being used by an Abu Dhabi investment firm for its $2 billion investment in crypto exchange Binance, one of World Liberty's co-founders said on Thursday.
Thanks for the update, I was going to come in and post roughly the same thing. The only things I would add include: DHS has threatened to arrest democratic house reps for visiting an ICE facility that they have every authority to visit, claiming the house reps "assaulted" someone.
> the FBI Director denied under oath that the Constitution's guarantee of due process applies to noncitizens despite having Antonin Scalia's ruling to that effect read to him;
Does that ruling only apply to non citizens in the US, does being in the US apply special status? Because there’s been a lot of extraordinary renditions prior to this.
And then there’s Maher Arar from 2002. Deported from JFK to Syria, back when Syria and Assad were not on the naughty step.
Yes that SCOTUS ruling in particular, and the Constitution generally, applies to people within the US. (With the caveat that "within the US" is not legally limited to the 50 states...."the US proper plus territories currently controlled by it" would be the fuller answer.)
"I was a libtard and a coward. I still am. I’ve just recovered a bit more"
Up is down, war is peace. You can't seriously engage with postmodernism.
Well, the way Yarvin is expressing himself might be overly poetic and flowery for my tastes, not to mention profane, but it's not that hard for me to mentally translate it into what I would consider plain english, at which point I could engage with it in the same way I would engage with anything else.
Assuming it's Yarvin you're accusing of being postmodernism, I'm not sure.
I, also, can translate the words of Yarvin into something plain and reasonable. But he chose to translate his thoughts into words hateful and profane, so why would I want to do such a thing?
It is he. Being postmodern is not contingent on words being indecipherable.
After seeing Scott recommend the Center for AI Safety free online course on AI Safety, Ethics, and Society, I took the course, and can report that it was great! Good weekly discussions, the textbook has good coverage of the relevant topics, and the project phase at the end was a great opportunity to dive deep into an area of interest. For my project I wrote a post clarifying the conclusion I have come to after reading much of Scott and Zvi's writing and taking the course - namely that we absolutely are not ready for AGI and it's kind of crazy that building it is still the stated goal of many of the top AI companies, and we should thus work to build consensus on (1) that this is stupid and dangerous and (2) what to do about that fact given the systemic forces pushing us forward. See the post here, I welcome any feedback or corrections! https://natezsharpe.substack.com/p/what-if-we-justdidnt-build-agi
It seems one of the most popular arguments for imminent AGI goes "skeptics said AI would never do x and then AI did x! They said it would never do y and y happend a few weeks later! Then they said those things didn't really count! When will you all wake up?????"
I think this is a bad argument. Bear with me.
ALICE: AI may solve equations, but it will never write a story.
BOB: Actually, AI has just written a story. "A man planted a tree. The tree grew. The end."
ALICE: I meant a real story. Like, with description and a multi-part plot.
BOB: Lol, moving the goalposts. But as it happens, AI has just produced such story.
ALICE (*reading it*): That's...just a copy of an existing story, with some paraphrasing and word-swaps. That's not what I meant.
BOB: Ha ha, you just keep shifting your claim. Be precise.
ALICE: AI will never produce a story that will win an award.
(A month later)
BOB: An AI story has won a prestigious award.
ALICE: Only because the judges wanted to "take a stand against ableism and human supremacy". That doesn't count.
BOB: I can't believe it. You'll never ever admit you were wrong.
The point is that Alice does, in fact, have something she means when she says "story", even if she can't describe it in advance. Even if she gives ten conditions, and gets presented with something that satisfies all of them, she can still say "that's not quite right, that's not what I meant" and be telling the truth. "I know it when I see it, and this is not it".
Yes, there is some Carol who really is satisfied by the stories AIs produce, and some David who, even when presented with exactly what he was thinking of, will lie and say that wasn't what he meant to avoid admitting he was wrong. But Alice is not Carol or David. There *is* some story that were an AI to produce it, she would be completely satisfied, and happily admit she was wrong. But that simply hasn't happened yet.
And I think there are a lot of Alices in the AI debate, even though there are a lot of Davids and Carols as well. And so there's a fundamental circularity in the pro-AI argument. The other side say that intelligence is really subtle and not something we understand or can model or describe. The AI side say that if they think that, they need to model and describe exactly what things AI will not be able to do, in advance. Which is just assuming that the nature of these things ("art", "emotion" etc) are things that can be modelled and described. Which is precisely what's at issue.
I think it's a bad argument because it's a subspecies of the old argument "People have been wrong in the past, therefore you're wrong now".
I feel "People have been wrong in the past, therefore you're wrong now" is maybe too strong a statement, but I think there's _something_ to that logic.
In some instances it seems like a legit argument. If tell you "I'll buy a lottery ticket because my horoscope said it'll be my lucky week" I feel it would be fair to respond with "good luck with that, but have you seen how it turned out for everyone else who tried that". I could answer with "what have the failures of past people got to do with me?" but I'd be wrong.
In other cases it's clearly an error. If I told the Wright brothers that "people have tried building flying machines, it never works, give up, man is just not meant to fly" then I'd be wrong.
I haven't yet found the good rule when the argument applies and when it doesn't. At least not in a way I could articulate. I tend to think that "people have been wrong in the past" is evidence but not proof, though it depends on the details ofc.
It's a bad general heuristic, but there are much better heuristics that can be used instead. "Horoscopes are fake" is the better heuristic. "The odds of winning the lottery are incredibly rare" is a good general heuristic.
Betting that nothing ever changes just seems like a worse argument than individualized arguments against the specific claims.
I mean isn't the horoscopes are fake heuristic basically just the "people have been wrong in the past" heuristic?
The two things I know about horoscopes is that a) we don't know of any mechanism for why they would work and b) so far they've never worked (better than luck)
But it seems that (a) is not load bearing, in the world in which horoscopes reliably predict the future we would say "they work, we just don't know why", argument (a) notwithstanding.
And if you'd ask me for a reason for why I don't buy horoscopes I'd answer something like "It's been tried, systematically, and it's never worked. Why would it work now?" which is basically exactly the general heuristic.
Also the general heuristic has the upside that you can use it on things you don't understand. If you offer me a game where I pay $2000 and may win $2million and I observe that none of the first 300 people who try ever seem to win anything, then I won't play your game and I believe that's reasonable. And that works without any knowledge about what your game is.
$2 million divided by $2,000 is 1,000. You would naively expect at least greater than 1,000 people to play the game before reaching a winner. Otherwise the game clearly loses money and in my mind would be even more suspect. I still wouldn't play it, because $2,000 is a reasonably significant amount of money and 1/1,000 odds is bad odds (the lottery would be something like 1/2,000 odds for the same payout).
You don't need to use "people have failed to win before, so it must be rigged" to conclude that this game is a bad idea. You can do the math on that one, as well as the lottery.
In terms of a blank slate review of horoscopes, I would agree with you. With that said, we don't actually have a blank slate. We have a rich and recognized history of scams and shady behavior related to such things. You could actually argue that we do live in the world where your "a" is true about horoscopes - they are reliably correct. Millions of people can read the same horoscopes from the newspaper and conclude that they were correct. They do this by being so vague as to be unfalsifiable. "You're going to meet a stranger" or whatever is so broad as to apply to almost everyone, and actually ends up conveying no information at all. I look at them and conclude that even if they were correct, it would offer no meaningful information or change in my life.
I should've added "(better than luck)" to both conditions. I'm with you that horoscopes are so broad that in some sense they work. But not better than luck. Or, perhaps more precisely, that to the extend horoscopes predict things, these predictions come true about as often as statistics would predict anyway.
And you don't have to go to "people failed it must be rigged", I agree. My point is that even absent all other knowledge it's reasonable not to pay 2k to play this game, because what makes you think you're the special one that'll win it, when the other 300 weren't?
I believe that stance is reasonable without requiring any further assumptions about how the game works, how likely it is this is a scam or anything of the like.
Doing some back of the envelop math mostly projects your own assumptions unto the scene, they might well be accurate but they might also be wrong. If you can find a reasonable way of action without having to make further assumptions, then all the better.
Which is exactly the usefulness of general heuristics. Yes, when you have accurate info then I agree, better to math it out. But for enough scenarios irl it's just not that realistic to gather all the facts, and I'd argue "has been wrong before, often, so is probably wrong now" is quite useful then.
Horoscope.com for today - This isn't a good time to make any big relationship decisions. If you're thinking about starting a relationship, moving in together, or getting married, you may feel a bit troubled by your decision.
Even outside of the "If" in the second sentence, that's insanely vague and of obvious applicability in just about any relationship where people are considering a major life step. And even then it's "may feel" a bit troubled. It's plausibly true for everyone, and therefore worthless.
It's much simpler, actually, it's a bad argument because it fundamentally misrepresents not only what actual "skeptics" (as in, people interested in the technology, not some rando human exceptionalists on the internet) say, but [literally everything we actually care about].
The key trick is "be precise". The hypesters insist on easily demarcated goals so that they can either a) brute-force something that meets them out of the AI via infinite monkey theorem b) teach it to the test. Then they just won't accept "uh, this obviously does not generalize" for a rejoinder, and at this point I'm inclined to believe many genuinely do not understand why it matters.
Agreed. Like the r's in strawberry thing - of course no one is going to be impressed that hundreds or thousands of man-hours and millions of dollars will allow some of the most powerful companies in the world to get a computer to spit out the number "3" when asked that specific question. Maybe along the way they also figure out how to get it to count letters in other words. That's far from impressive to anyone. It was embarrassing that the systems would consistently get it wrong, but not praiseworthy to get it right.
I'm much more interested in generalized ability and reliability. A system that counts basic information wrong is not generalized or reliable, even if the specific examples can be fixed.
Does generalized ability have to mean ability at every task? Why can't computers, like humans, have their own peculiar strengths are weaknesses? Maybe AIs will be able to run Fortune 500 companies and come up with Nobel Prize winning theories of physics without being able to count letters in words. It's just as unfair to call a machine unintelligent because it can't count the number of r's in "strawberry", as it is to call a human unintelligent because he can't multiply 100-digit numbers in 2 milliseconds.
The question is whether the AI can in fact "run" a Fortune 500 company, or whatever other task is set before it. There's an amazing level of detail required to work in such an important position. The impression that people get from a system that cannot do a simple task is that there are one or more significant flaws in that system. Maybe the system would do well running the company most of the time, and then once a year might do something really bad like sell off all the stock for a completely unknown reason. That would be really bad, bad enough to never use the system even if it was 100% perfect the rest of the year.
If we tried to get an LLM to run a company right now, it would be terrible. It would fail in very obvious ways, but worse, also fail in many non-obvious ways. We don't actually know what it would take for current LLMs or a new approach to AI to be able to run such a system, so it's difficult to know how close or far we are away from being able to do it. Small details like miscounting letters (a task many humans do easily) is potentially an important clue. Even knowing that the reason LLMs struggle with this has to do with tokenization (instead of stupidity) doesn't help much, because it's impossible to determine how many things the LLM does might also be related to tokenization. Brute forcing a system to know how many letters are in a word doesn't give us any assurance that the actual generalized problem is or can be fixed.
Yes, by definition. AIs that do a specific thing well, including much better than humans, are just, well AI, G not included. They can be and are immensely useful, but singularity they won't bring.
Humans' intelligence is general not because we can multiply in 2 miliseconds, but because we can (teach ourselves to) do it at all.
(Though I have to say, contra the other commenter, the reasoning models' ability to count letters, from what I've seen, does generalize to performing more complex tasks on character strings described in natural language. I'm not sure how impressed I should be with that, since I'm pretty sure they've been specifically trained to perform that particular task in response to generative models being a running joke in this area, but it does look like a genuine step in the right direction.)
Then neither humans nor computers count as General Intelligence. Humans cannot learn to multiply 100-digit numbers in 2 milliseconds no matter how much they train, nor can they beat Stockfish at chess or AlphaZero at Go. On the other hand, as Mr. Doolittle said, AI *can* be trained to count the number of characters in words, but undoubtedly they'll still be unable to do things a human can do easily. Insisting on judging AI by the skill they lack the most while not doing the same for humans makes no sense.
2 miliseconds is an arbitrary restraint that has nothing to do with the concept of generality.
Jack of all trades needs not be a master of any of them, again, it just needs to be capable of figuring each one out. Simple tasks like counting letters in a single word are particularly telling because if you can't get even them out, then, well, you just can't do it at all.
It has a lot to do with trust. I would trust an AI chess program to give good advice on moves, but not infinitely. I heard someone came up with a really silly way to beat AlphaZero that would beat approximately zero human opponents. I don't know Go, but I understand it had to do with initial board setup and doing something that would never have happened in any real game for the training run. AlphaZero would repeatedly fall for the same setup and lose due to the weird technicality. This could be fixed for future versions, but it's an example of how a "superhuman" system can be incredibly limited - not just to the fact that it only plays Go, but also to not being able to work with many simple variations or unexpected changes.
Real life is far more complicated, so even a system the repeatedly did well at a common function (like self driving cars in SF) isn't trustworthy outside of where it's strong. In a city that isn't mapped out precisely (or if something big changed like an earthquake or a fire) the cars would not be able to function like they do in a fully mapped city. I don't expect to see many self-driving cars in any cities that experience winter, at least during the cold months.
I think it's important to remember that in 1980, people said computers would never be able to play Chess, and they were phenomenally wrong by any metric you want to use. In the long run, there's no reason to think there's anything a human can do that they can't also automate.
I'm very skeptical of AIs being some kind of fundamental shift. I happen to agree that computers can be programmed to automate almost, maybe all, human tasks. But that's far less interesting than it sounds. If you had the best and brightest team on the planet trying to automate a regular function you could have done amazing things in the 1980s as well. We in fact did a lot of that! But none of it was even considered AI. A CNC machine can be programmed to cut very precise tools regularly at superhuman levels, but no one argues that makes them AI.
Can a chess bot help you run a factory or fold my laundry? It can't even try. It's nonsense to think about. Billions of dollars can definitely allow the brightest people on the planet to automate a factory. That's neither new nor particularly useful. It's certainly possible without AI, and at this point trying to get any of the AIs we have to do that kind of work would be a disaster.
The goal of AI companies is to produce a general intelligence, not individually automate thousands of functions through a non-ideal interface. I happen to think that LLMs do a great job with natural language. Spending lots of money and time on getting them to do okay at math, okay at programming, etc. is probably bad ROI. The fact that they are better at programming is very likely an artifact of the people who work on them being programmers and making that a high priority. We've seen much less progress on much simpler functions (still waiting on that laundry!).
...great, now I'm disappointed that my dryer doesn't fold my laundry.
A chess bot and a factory bot can both be put under a supervisor bot, who would be trained to select whether the problem is related to chess or factories. Put all three in a box and it looks like a general AI from the outside.
But more to the point, everyone's endgoal is automating the scientific method, from hypothesis to conclusion, and everyone's fanciful futures stem from their vision of what happens after computers can pose novel questions to themselves and then answer them. And every time someone cajoles a computer into producing a better poem, it's an indication they're getting better at the overall cajoling process, that might at some point lead to automating research.
Quite possibly! That's the aggravating thing about these conversations. As you say, on one level an AGI *might* be as simple as putting many different bots together with some kind of supervising system that identifies the kind of problem in front of it and then choosing the correct superhuman response.
But in your example system with three bots, what happens when instead of running a factory or playing chess, the problem involves playing checkers? If we keep stapling on new bots to solve every individual problem, we're not talking about AGI, we're talking about the brightest people in our society individually automating thousands of discreet functions.
And this is important to your stated goal. If AIs are just referencing an existing automated bot, then they'll probably never be creating a new hypothesis (unless humans already automated science, in which case "AI" wasn't needed) and not answering novel questions.
There's three other issues I see.
One is related to the novel questions. If an AI is referencing hundreds or thousands of other bots for specific tasks, it's very likely to be bad at any new task. If I invent a brand new game, does the system need millions of examples of playing the game to figure it out? A whole new sub-bot? That might be useful in limited senses, but terrible for regular use.
Another issue with that approach is that the supervisor program might misidentify which problem it's dealing with and use the wrong subsystem. In a simple sense, that might be the system thinking there's a biology problem in front of it, but it turns out it's actually chemistry or whatever. This gets way way more common and way worse when the system is doing something truly complex (like running a national economy or taking over the world, but even something much smaller like running a factory).
Thirdly, what happens when something changes? Chess doesn't change, so the chess bot is probably going to continue being fine for most uses. But people like to vary things, so maybe someone comes out with a chess variant they want to play. Is that AI bot going to play the game at a superhuman level anymore? Is it even going to be able to play at all? Something more complex like a factory involves hundreds or thousands of individually automatable points, but also regularly involves significant changes. Even getting a new order for more/less of a widget will result in massive changes throughout the factory. Is the bot running the factory also automated to change itself to a new bot that runs the new setup of the factory? Each step is fraught with difficulties.
In the long run, I completely agree. However, that's the long run. Perhaps it should've been possible to predict Stockfish in the 1980s, perhaps not; but if you went back to 1820s and predicted that no mechanical difference engine (built out of gears) would ever beat a human at chess, you would've been *correct*. It took several radical paradigm shifts to go from arithmometers to Stockfish, and if you were to predict that such paradigm shifts are expected sometime in our future... then again, I might provisionally agree with you. But that's not the same as saying that present day LLMs would ever become superintelligent quasi-omniscient human-squashing oracles.
I'm not sure this demonstrates what you think it does. We didn't build Turing machines out of gears (or pencil and paper, for that matter)...because we found something *better* than gears. I'm really not sure what your point is supposed to be.
My point is that the following two predictions are perfectly compatible and can both be true:
1). Present-day LLMs will likely never become "superintelligent" in any non-trivial sense of the word (just as mechanical calculators could never beat humans at chess).
2). Perhaps one day we will discover some completely new and hitherto unheard-of technology that will lead to the creation of a superintelligent AI (just as electronic computers backed by novel algorithms, both of which were unimaginable in 1820s, managed to beat humans at chess).
But note that statement #2 is not compatible with the following:
3). ...And I can predict exactly when this completely new and hitherto unheard-of technology will arise and how it will function.
> just as mechanical calculators could never beat humans at chess
But this simply isn't true. You *could*, in principle, run Stockfish on a mechanical Turing machine. There's just no reason to, because electronic ones are better!
No, you couldn't -- not unless the machine was made of frictionless spherical cows.
Alice: "AI will never produce a story that will win an award"
Bob: "An AI story has won a prestigious award"
Also Bob, but quietly, so Alice doesn't get the full story (sorry, intentional bad pun): "An AI story selected out of N attempts, after a painstakingly engineered series of prompts, by an actual human".
That ought to count. We don't penalize humans for asking their friends for feedback on their stories, or for making changes an editor requests. Why should we penalize AI?
If the prompt is multiple pages long and walks out all the important points, that doesn't feel like the AI is doing the important parts. It also seems silly to say an AI wrote the story when you ask it for thousands of examples and then pick the best. That's not too far from the monkey typists "writing" Shakespeare.
"An AI story has won a prestigious award" implies that AIs can generally write good stories, and at least occasionally write very good stories. Similarly, no one would be impressed if the award was "best AI written story" which could only be won by AI submissions.
It ought to count if the author AI asked other AIs for feedback, not humans.
https://www.astralcodexten.com/p/my-bet-ai-size-solves-flubs
For reference, this is Scott actually taking the view of Bob against Gary Marcus's Alice. I think it's illustrative to see how close to Bob and Alice they end up being.
My view is that this isn't even close to what has happened. That Alice does not get to wriggle out of their claims just because there's a cosmic sense in which they're still right. In addition, in what sense does Alice have actually good predictive ability if they *can't* articulate what they're looking for? Do we also implicitly trust Alice to keep track of all the reasoning tasks that she can't come up with concrete examples for but AI has implicitly accomplished? I don't think so, because we also don't trust Bob when he says "I've been using AI and there's an ineffable sense in which it's qualitatively better"!
I agree with the debate between Scott and Gary Marcus. But when you look at all the various benchmarks that supposedly measure intelligence, it’s clear that they don’t measure what we want them to measure. GPT-4 (not even 4o) was already maxing out various LSATs and bar exams and such, but you definitely don’t want even the modern LLMs practicing unsupervised as a lawyer!
This isn’t moving goalposts - this is the difficulty of coming up with good goalposts to begin with. Precisifications usually miss some significant features of what we actually cared about when we came up with the vague idea.
I think tests designed to evaluate whether a human student learned what he was supposed to learn in a course of study are just the wrong tool for the job of evaluating the performance of a computer. Ultimately, the test comes down to how well it performs at an actual job.
Consider IQ tests. The reason we care about them is that they predict performance in school and at work. That correlation is why they matter--a test that was similarly heritable and consistent and all but didn't help us predict anything would be pretty pointless. But that's only for humans. If Claude can get a perfect score on an IQ test, it won't tell us much--we don't know the correlation between LLM IQ scores and performance.
That works in both directions. It's quite possible to have many things AIs can't do nearly as well as humans, but also to have AIs be more capable than humans in the right set of domains to let an AGI take over the world or something. There's nothing that says that an AGI that paperclips us all has to also be able to write consistently funny jokes, or make beautiful art, or whatever.
We have a little experience with nonhuman "intelligence" in things like evolutionary processes, markets, bureaucracies, and eusocial colonies. It's entirely possible that you can write poetry and develop philosophy, but your species gets wiped out by unthinking evolutionary processes that lead to the rise of some virus that kills everyone off. In the same way, it's possible for AI to have a very different kind of intelligence than humans, but still be very dangerous. In some sense, the potential danger from LLMs is easier to see because they can talk to us like humans
I agree with all the points in the last paragraph!
But I think the people who believe in the concept of AGI really *do* think that it requires *general* intelligence, which they think is what is behind all
human cognitive abilities, and thus would allow these things.
I don’t believe in general intelligence - I think it’s important to note that things can be dangerously capable in many ways, even while lacking many major capabilities average humans have, because there is no one thing that is general intelligence.
"...in what sense does Alice have actually good predictive ability if they *can't* articulate what they're looking for?"
This may sound backwards, but Alice would actually have *better* predictive ability the less likely she is able to articulate an advanced concept. It suggests that she is able to think of things at a higher level of abstraction by integrating information that is not easily available to Bob. The more pieces of information you work with, the harder it becomes to condense them into a sentence.
For me at least, Bob appears much dumber here because he should know perfectly well what Alice means when she talks about a story. He should have read enough books to know, too.
> This may sound backwards, but Alice would actually have *better* predictive ability the less likely she is able to articulate an advanced concept. It suggests that she is able to think of things at a higher level of abstraction by integrating information that is not easily available to Bob. The more pieces of information you work with, the harder it becomes to condense them into a sentence.
The operational word here is predictive and not smart. What is Alice predicting? Do you think Alice would even grant that the AI has crossed any invisible threshold? Do you think Alice would be any good at predicting capabilities the AI does end up with? Because your response would indicate no, she just has an undefinable and unaccountable ocean of invisible knowledge you can attribute to her.
Reverse the situation. If Bob confidentially says that he sees AI getting smarter in some ineffable way, Alice makes a precise prediction that that capability doesn't emerge, so we then say that no, actually Bob is more predictive than we thought?
> For me at least, Bob appears much dumber here because he should know perfectly well what Alice means when she talks about a story.
Yes, because ascend wrote it so that Bob looks dumb, and then doesn't mention that quietly, 10 months later the AI actually does end up writing a passable short story. But of course, this in fact doesn't come up on any news story, *BECAUSE* it is subjective and because Alice is exactly the genre of person to not earnestly check their claims against reality (oh yeah, says Alice, I used the free model 10 months out of date without any prompting and it failed. I'm definitely still right.) In what world do you see *any* Alices have a mea culpa? Please, if you believe this is true, list them.
The thing is, once the AI is general that's past the point anyone cares to contest the issue, so of course, Alice gets a free pass to say whatever she wants while not actually thinking, only vibing, and everyone attributes deep 10 dimensional chess cosmically and philosophically correct insight to her.
"If Bob confidentially says that he sees AI getting smarter in some ineffable way, Alice makes a precise prediction that that capability doesn't emerge, so we then say that no, actually Bob is more predictive than we thought?"
I think we try to understand what is the true claim being made. Bob doesn't have to lay out every step for how he thinks AI will manufacture nanobots, but he does have to rescind his claim when a bottleneck is suggested and he is unable to pull a single credible counter from his ineffable pool of knowledge. The current method being employed in this space is to trust whoever has the more ostensibly "precise" claim as a type of meta-credibility, without actually evaluating properly the facts of the claim. Whoever can make their claim *signal* more precision by using specific cognitive techniques and figures ends up being trusted more regardless of actual truth, whereas I would urge for true understanding of what is said. This might mean you cannot make or trust some predictions at all until more information is consolidated, and that's almost certainly a good thing.
"Alice is exactly the genre of person to not earnestly check their claims against reality (oh yeah, says Alice, I used the free model 10 months out of date without any prompting and it failed. I'm definitely still right.) In what world do you see *any* Alices have a mea culpa? Please, if you believe this is true, list them."
I don't think you want to go band for band over who rejects reality more, and I don't think it's productive, but at least some Alices *can* have a mea culpa. It is literally impossible for Bob to have a mea culpa because he is predicting fiction and can endlessly shift goalposts (Alice can shift them too, but she can also actually be proven wrong with empirical events).
> I would urge for true understanding of what is said. This might mean you cannot make or trust some predictions at all until more information is consolidated, and that's almost certainly a good thing.
You have also deflected and not answered the question, in the reversed situation, would we say that Bob is more predictive? Or is it really some other generically good sounding word?
Because, my impression is that the type of person who claims the existence of "higher order unspoken concepts" having validity are the type of person who are most eager to excuse the word and status association going on in their minds. Unfortunately, the margins of a deeply nested ACX thread are TOO SMALL to express the complicated social model in my head, and I can only ask forbearance that you do all the intellectual heavy lifting to make what I'm saying consistent. Please and thank you.
I agree that concepts can be too complicated to express, but what I heavily disagree on is that we should consider this a merit instead of a demerit. Adding degrees of freedom in how you express statements goes *exactly* against what good predictors do, as expressed directly by Philip Tetlock and indirectly by Kahneman and co. And I think it's illuminating that the instant we reverse the hypothetical, epicycles get added to make your point hold anyway.
> The current method being employed in this space is to trust whoever has the more ostensibly "precise" claim as a type of meta-credibility, without actually evaluating properly the facts of the claim.
And anyway, if you believe this is what is being done, please link to actual examples of this that have been falsified, in your view. Several times I've asked this of people, and it turns out they were doing word association and not reading the actual statements made. (See: claiming that data centers need to be unilaterally nuked, instead of as a clarifying statement on what constitutes a real vs fake treaty signed between two countries)
> I don't think you want to go band for band over who rejects reality more, and I don't think it's productive, but at least some Alices *can* have a mea culpa.
Between Gary Marcus and Scott Alexander, only one has a mistakes page and has confessed to changing their minds substantially. If you have concrete examples, please link to them instead of hinting at them. I think both in terms of "fake" metrics of things like having mistake pages and "real" metrics like actually changing their mind, Scott is the superior here, and so are people like e.g. Daniel Kokotajlo, who do scrupulously keep track of things. Is there a skeptic who also keeps track of what they're saying and compare it to reality *at all*? I think the closest is someone like nintil or Matthew Barnett, but they hardly strike me as Alices.
I would change my mind if you find such an example that I can agree on, and I'm registering that right now. I will also change my mind if you find an example that I don't agree is reasonable but I can at least see an attempt was made.
So, since you seem confident that mea culpas can be had, where does your confidence come from?
"You have also deflected and not answered the question, in the reversed situation, would we say that Bob is more predictive?"
I mistakenly thought that the answer was implied. Bob *is* more predictive, just as anyone who is capable of higher level reasoning and knowledge will naturally be more predictive. But you asked what should we do to resolve the dispute if both parties have complex ideas in their head which signal a high degree of complex knowledge, and I said that a better framework is to work through the true object-level of the claims than to just believe whoever has written down more information.
"Unfortunately, the margins of a deeply nested ACX thread are TOO SMALL to express the complicated social model in my head, and I can only ask forbearance that you do all the intellectual heavy lifting to make what I'm saying consistent. Please and thank you."
I think this person is not who I am talking about. I am talking about people who:
- claim to know something you don't (ie: Hypothetical AI is not possible, trust me bro im an engineer)
- put down 1 or 2 points of reason which explain a part of this knowledge (ie: Hypothetical AI will suffer bottlenecks 1 and 2 which I know of because of engineering experience)
- claim that the rest of their knowledge is too complex to express, but invite you to dialogue so you can scrutinise or understand it better. (ie: These bottlenecks might not look like enough information to you, and it certainly seems unfair that I have countered your 50000 word blog post with a 1 sentence bottleneck, but I have all the information in my head so we can go over it in detail as best we can)
"I think it's illuminating that the instant we reverse the hypothetical, epicycles get added to make your point hold anyway."
I addressed this one already, but I applied the same logic in the reverse case.
"And anyway, if you believe this is what is being done, please link to actual examples of this that have been falsified, in your view."
You can just look at Scott's AI 2027 predictions a couple of weeks ago. Several people in the comments noted physical bottlenecks or information-theoretic limits associated with the arguments around AI-based algorithm improvement. Multiple times now Scott has dismissed or ignored them as basically just "We expect AI will solve this". That's not reasoning, but if Scott was willing to put forward a small counter-claim and then invite readers to a greater well of knowledge then sure. He can't really though because he is not an expert in anything related to AI and his lack of technical knowledge shows.
"Between Gary Marcus and Scott Alexander, only one has a mistakes page and has confessed to changing their minds substantially. If you have concrete examples, please link to them instead of hinting at them. I think both in terms of "fake" metrics of things like having mistake pages and "real" metrics like actually changing their mind, Scott is the superior here, and so are people like e.g. Daniel Kokotajlo..."
Big problem with these discussions is that the only true/correct way to evaluate which side accepts reality more is to take a large sample of people in a controlled way and determine it from there (it would be rather difficult). You have only listed 3 guys which is obviously not a fair comparison. That's why I said it's not productive. People should be analysing claims independently of who made them anyway.
"I would change my mind if you find such an example that I can agree on, and I'm registering that right now. I will also change my mind if you find an example that I don't agree is reasonable but I can at least see an attempt was made."
"So, since you seem confident that mea culpas can be had, where does your confidence come from?"
Notice I never said I had examples of Alice having a mea culpa, I said it was theoretically possible because their claims may resolve as either true or false when an example is clearly presented in reality. With Bob, the claim is that one day, AI will kill everyone or something outlandish. When it doesn't, Bob can argue it might still be true. I have seen this happen multiple times with people like Yudkowsky. It's very similar to the classic "You can't prove god isn't real" problem.
I was and still am a moderate AI skeptic. I confess to having moved my mental goalposts by a nontrivial amount. LLMs and image-generative AIs have gotten a lot more capable much faster than I expected early on, and I've had to update based on this.
I am a more radical AI skeptic, and yes, I do admit that AI had developed faster than I thought it would -- but in a different direction from the one doomers are constantly predicting. In terms of "intelligence" (whatever this means) AI is still comparable to a powerful search engine; but its usability had skyrocketed, to the point where you can reliably use it for repetitive low-impact tasks (as you would use a search engine).
Indeed, I see AI proponents moving their goalposts. The term "superintelligence" used to refer to some kind of a quasi-omniscient oracle who can answer any question with perfect accuracy even as it converts the Earth to computronium using the power of molecular nanotechnology. Now, it refers to e.g. an engine that can outperform most (though not all) human geoguessers (some of the time) at figuring out where a photo was taken.
My worry is not that AI will awaken and kill us all, but that we will downgrade our notion of "intelligence" to the point where even modern LLMs fit the bill. They would indeed become "superintelligent", but only compared to our new level of natural stupidity.
Thank you, this is an excellent AGI summary:
- a quasi-omniscient oracle who can answer any question with perfect accuracy: yes, sure, why not.
- even as it converts the Earth to computronium using the power of molecular nanotechnology - nope.
The fact that so many smart people think the first one inevitably and "almost instantly" leads to the second one tells us a lot about the ability of smart people to create fantastic tales.
I agree, though personally I don't think even the first one is possible. At least, not any time soon, on the astronomical timescale.
I guess it really hinges on our definition of the word "any": if "any" refers to the "currently-available knowledge", i.e., the thing has read every published paper and reasoned which ones are actually "true"... as I'm typing this I begin to see how even that is not anywhere near within the capabilities of the current LLMs...
If it means answering "how to cure a patient with a stage 4 malignant melanoma" with a perfectly designed treatment proposal, no, that will not be happening anytime within this century.
I would argue that "reading every published paper and reasoning which ones are actually true" is, for the most part, beyound the capabilities of *any* LLM, be it present or future. The reason people write scientific papers to begin with is because they actually went out into the world and ran experiments to determine whether their hypothesis is true; the "replication crisis" in science refers to the lack of papers written by people who -- again -- go out into the world and attempt to replicate another team's findings. The key step in this entire process is going out into the world and running experiments: growing plants, launching stratospheric balloons, building supercolliders, running brain scans, whatever. You can't discover very many truths about the physical world merely by thinking about it really hard in your head -- not even if you can think 1000x faster than any human.
Thanks for writing this, I hadn't really fully appreciated how Alices feel. That said, I think that one of the tenets of rationality is that shifting our claims *should* make us anxious, even if it's not "our" fault. We need to try our absolute best to precommit to criteria that make sense, and if for whatever reason we have to abandon those precommitments, we should "feel guilty" about doing so. In other words, because we can't tell Alices and Davids apart, Alices should be extra careful. Actually, we can't even trust our *own* internal feelings to be consistent over time. Maybe someone who feels like an Alice is actually being a David and their values are shifting over time. Sometimes you need to abandon the inside view of yourself and act according to the outside view.
There are some particular failure modes that are prevented by precommitting to criteria, but in am not convinced that these failure modes are so frequent or important that it makes sense to say “We need to try our absolute best to precommit to criteria that make sense, and if for whatever reason we have to abandon those precommitments, we should "feel guilty" about doing so.”
Precommitment involves your less informed self promising something without the benefit of all the distinctions you’ve learned to make after seeing things in reality.
Early consumers of canned vegetables said they thought they tasted like they were fresh from the garden - and they are in fact much more like fresh vegetables than they are like pickled or dried or smoked or salted vegetables, but once you’ve tasted them a couple times it’s easy to tell the difference. Thomas Edison ran a series of “tone tests” (worth googling) where he sent a singer and a phonograph, and wowed audiences when he put down the curtain while the singer was singing, and lifted the curtain but it was just the phonograph, and the audiences hadn’t noticed the shift. (Modern audiences can easily tell.)
Precommitment is good for something you are going to test statistically, and good for certain kinds of adversarial environments, but not good for getting real understanding.
Good point! I agree it doesn't make sense to *never* reassess things, but we should be painfully aware of our ability to deceive ourselves without realizing, and feel duly incentivized to try to anticipate how things might change. This situation is a little different in that the people guessing at the future have at least some sense of the technology they'll be up against, so they are better equipped to carefully craft criteria. I agree it shouldn't be disqualifying to have to readjust our criteria, but it'd be nice to make that as unnecessary as possible since it damages a really good defense we have.
Yeah, I think that's all correct.
One thing going on in the dialogue above is that Alice's metrics are corrupted after she issues them by motivated opponents. When Alice says "AI can't write a story that wins an award", and then a group forms to give an award to a story written by AI for the purpose of invalidating Alice's claim, it is obvious that this "doesn't count" - we may assume that what Alice has in mind is for her claim to be true whether or not other people have heard her make it.
Compare "Goodhart's Law".
Dismissing Alice as "moving the goalposts" in this scenario looks very bad.
I agree bad actors complicate this, but I think this is mostly a reason to be more careful when choosing criteria, imagining how others might try to Goodhart them. In your example, though, it's overt enough that I think it's fair to say "clearly that doesn't address the thrust of her claims".
> I think this is mostly a reason to be more careful when choosing criteria, imagining how others might try to Goodhart them.
This cannot be done.
Maybe a nitpick, but you're describing things that don't happen in the story: it doesn't say that the award committee knows anything about Alice, or that the award was founded after Alice made the prediction. It's not clear anyone is Goodharting.
A better way to describe the dynamic is that Alice had correct beliefs about AI (it wouldn't produce a story to her satisfaction) and incorrect beliefs about the judges of literary awards (they would have a similar judgment to hers). She placed the goalposts based on those two beliefs, but only meant to be making a strong claim about AI.
She's definitely moving the goalposts now: she chose a clear criterion that was unambiguously met, and only after that wants to revise it. The question is what we should make of her doing so. In this case, I think she has a very good argument that the details of the case didn't prove her wrong about AI, and if she'd thought about this from the start she would have placed the goalposts differently.
But if we didn't have the ability to read Alice's mind, or it were more ambiguous, it would be reasonable to see "Alice's prediction was falsified" as stronger evidence against her beliefs.
No, you've overlooked the part of the dialogue where the committee makes its award "to take a stand against human supremacy". That is a reason unrelated to the story; Alice was not mistaken about whether the judges would have judgment similar to hers - she was mistaken about whether the judges would consider the quality of the story when giving the award.
I didn't overlook that, I took it to be included in "different judgment than Alice" -- they judged based on very different criteria than what Alice was predicting, and which she doesn't see as legitimate. That's why we know she's justified in not changing her mind. I was pointing out that you framed this as an adversarial play against Alice's prediction, which is not how it's described above.
Quick edit: by "judgment" I meant "judgment about what should win an award." Alice's mistake was to think this meant "judgment about literary quality," but that's not the criterion she specified.
I think it's not unreasonable for Alice to think this criterion is implied. Otherwise the issue is reduced to a triviality: make the only criterion the fact that the story must be written by AI, AI writes "Once upon a time, the end", award goes to AI, boom, done.
Generally, I think "skeptics said AI wouldn't be able to do X, and then it did X" is still a pretty good argument against someone saying "AI will never do X." At least, if Alice doesn't have a decent response to this, I feel justified in not giving her position much credence.
AI systems keep improving. If you claim to have some insight into the kinds of tasks AI systems fundamentally can and can't do, you should have been able to apply this insight at any point in the past and been correct. You can't personally do this. The best available proxy is the smart people from the past who confidently and incorrectly said "AI will never do X."
You need to be able to say why you have better insight into the fundamental limits of AI than they did. There are some good reasons you might: you have access to more evidence than they did! But if you're not somewhat humbled by the failures of people to make this exact same prediction in the past, and don't have an explanation for why you're not making the same mistake, then I'd think you're the appropriate target for this argument.
While I agree that hypothetical "AI systems" in general can do anything you can imagine, present-day LLMs seem to be quite limited. For example, they are incapable of learning on the fly and synthesizing their discoveries into a continuously changing worldview, the way humans can (even, or perhaps especially, in childhood). They are also significantly more prone to hallucinations than humans (even humans who have imbibed a lot of alcohol, weirdly enough). Also, all of the most spectacular feats of AI intelligence that I've ever seen require a huge amount of cherry-picking (by humans).
In addition, most of those smart people from the past who said "AI will never do X" turned out to be... correct ! You are significantly underestimating the sheer number of AI claims that have been made over the years, and focusing on the few that panned out; and even those have not yet been fully realized. For example, people claimed that AI had achieved intelligence since ELIZA, but while modern LLMs can hold a conversation much better than ELIZA could, talking to them is still kind of like... well... talking to a really sophisticated ELIZA. So, thus far, I remain unconvinced.
"They are also significantly more prone to hallucinations than humans (even humans who have imbibed a lot of alcohol"
This sounds like a specific claim. Is it based on an experiment?
anecdata, part of a conversation with Claude yesterday:
Claude:
For the metals in the list (Ga, In, Sn, Zn, Al, Ag, Au, Cu), the freezing points are used because these materials have very high triple point pressures that would be impractical for calibration purposes. These freezing points are realized at standard atmospheric pressure.
Is there a particular fixed point you're interested in learning more about, or would you like to explore another aspect of thermodynamic fixed points?
me:
"For the metals in the list (Ga, In, Sn, Zn, Al, Ag, Au, Cu), the freezing points are used because these materials have very high triple point pressures that would be impractical for calibration purposes." This sounds strange. What is the vapor pressure of Ga at its melting point?
Claude:
You raise an excellent point, and I need to correct my previous statement. You're absolutely right to question this.
The vapor pressure of gallium at its melting point (29.7646°C) is extremely low, on the order of 10^-35 Pa, which is essentially negligible. This is true for most metals at their melting points
( I don't have data on the _rate_ of hallucinations, nor of the analogous rate for humans, sober or drunk. )
I asked the Tencent LLM about Steeleye Span a while ago, attempting to lead up to the question about The Victory that I've mentioned in the past.
The conversation went something like this (the LLM responses are quite long in reality, but I'm providing only the parts that I consider of interest. All of this takes place in Chinese except that English names appear in English):
[me] Do you know England's Steeleye Span?
[LLM] Yes!
[me] I like their song The Victory.
[LLM] There is no record of Steeleye Span performing a song by that name; maybe you're thinking of another band, or a song that doesn't appear on their studio albums?
[me] What? What songs appear on Steeleye Span's "Storm Force Ten"?
[LLM] [A tracklist, with blurbs for each song. None of the songs listed appear on Storm Force Ten. None of them are by Steeleye Span. Some of them are actual songs by bands similar to Steeleye Span. One of them is a title structurally similar to the title of an album (but not a song) by Steeleye Span. One of them is "The Storm Force Ten", noted in the description of the track as the song that gave its name to the album (no such song exists). Mostly they're just made up with no obvious characteristics of interest.]
-----
I followed this up by asking Google Gemini, which provided a correct tracklist that was noted as coming from Google search results rather than text generation. I can't fault the choice of approach.
Ouch! Many Thanks!
I kind-of wonder whether any of the current LLMs that have been put in customer service roles have advised any customer to do something fatal... Of the "helpfulness / harmlessness / honesty" triple, I really wish the AI labs would put _way_ more emphasis on honesty...
I called my credit card company a few days ago, and in the time between the automated system telling me my balance and me trying to pay my balance, the system forgot how much I owed and changed the number on me. I had to jump over to an actual person to resolve it. Which they did instantly, because that's the dumbest kind of mistake to have.
No, and I also admit that drunk humans can see a lot of weird shit. However, when humans get to that point they usually just shut down; in addition, the weird shit they see is still reasonably comprehensible, e.g. pink elephants and such. LLMs usually hallucinate things like perfectly detailed reports of court cases (complete with references) which do not exist; and they do it all the time. Although I also do admit that there exist many drugs other than alcohol, some of which can trigger much more potent hallucinations (in humans, that is).
I think the issue is that LLMs hallucinate things that are *much* more comprehensible than the things people actually hallucinate. Though some amount of human confabulation is very much like LLM hallucination. Ask a person why they did something, or ask them to fill in some explanation of why some steps work in something they do all the time but have never studied academically, and they’ll come up with something that sounds plausible but is actually totally made up.
Drunk people may not be the best example for humans hallucinating, though some drunks may see imaginary insults.
True in principle, but I think the overwhelming majority of purported Alices are actually Davidim.
I agree that Alice is right not to change her mind here. She has held a consistent position that she's just been unable to precisely articulate. I also think Bob is right to be a bit annoyed with her: she espoused some positions that turned out to be false, and now she's claiming she was secretly right in some inarticulable way the whole time. Unless Alice is able to specify her position more clearly, she shouldn't expect to convince anyone except herself. (Which is unfortunate, since she's in possession of a true belief that others would benefit from!)
So Alice shouldn't be too comforted by being technically correct here. It's easy to shift your own goalposts without noticing, and it's basically impossible to check after the fact whether they were "really the same" all along. One of the best ways to prevent this drift is by making clear predictions -- and making predictions that are later falsified is *evidence* that you had a false belief. (In this case, Alice has a good argument that the false belief was about the award committee rather than about AI, but in many cases it won't be as clear.)
"Unless Alice is able to specify her position more clearly, she shouldn't expect to convince anyone except herself."
But Alice has already convinced me. I think she can convince any reasonable person, because obviously "in the beginning there was a tree. Then it died. The end" isn't what any reasonable person means by "story", and obviously an AI winning a literary award just because the judges wanted to fight human supremacy isn't what Alice meant by winning a literary award. Yes, Alice could have been more clear about her criteria, but there are obvious things in the world that are hard to define.
Have you considered that there are some advanced concepts which are just not practically conveyable with language?
Sure -- like I said, Alice isn't doing anything wrong with respect to her own beliefs.
But if you can't communicate your beliefs, you shouldn't expect to convince anyone else or be recognized for having been right. If you try telling Bob that you were right all along but can't explain why, he will have no reason to believe you.
I don't know, what I am trying to say here is that it is not possible for Alice to convey her point because it is too advanced, so asking her to be precise is unreasonable. The onus would be on Bob to reach her level of inference, rather than expecting too much out of human language.
I don't know if we disagree?
From a personal epistemic perspective, Alice is correct and has done nothing wrong. This is what I take you to be saying and I agree.
From a social perspective: she's talking to Bob and trying to convince him of something. Once she decides to do that, she needs to be able to communicate! If she's not able to communicate her ideas to Bob, she will not be able to convince him of anything.
The issue might be with Alice, with Bob, or with the content itself. My point is that if Alice accidentally miscommunicates, Bob isn't wrong to point that out and say she's moving the goalposts. She is, and he has no way to distinguish her from a David who's doing so maliciously.
(I don't quite understand your last point: It seems like you're both trying to say the content is not transmissible, but also this is Bob's fault for not being on her level? But if he were at a high enough level to make the right inferences based on Alice's words, that's another way of saying the content *is* conveyable through language. I'm not sure assigning blame for the communication breakdown matters though.)
"She is, and he has no way to distinguish her from a David who's doing so maliciously."
Good point, I think. It's one I wrestle with a lot in online discussions. I've seen people who -might- be right, but are doing such a terrible, abuseable job of making their point that they end up doing it a disservice.
And it takes two to tango here as well. An audience can be so bad at absorbing an idea that they might claim Alice is David, not realizing their problem.
Troublesome news for a third-party observer.
I think it can be very hard for people to differentiate the internal experience of "I hold a consistent position that I can't articulate" from a bundle of aesthetic(/moral/whatever) preferences that don't actually cohere to a falsifiable test. As non-mind-readers, we should be suspicious of Alice' claim, whether it's about AI or anything else.
I think it’s problematic to assimilate “aesthetic/moral/whatever” to “don’t actually cohere” rather than to “consistent position that I can’t articulate”. Usually it’s probably going to be some mix of the two.
I didn't mean to imply that because subjective preferences inform/support a position that we should disregard that position as incoherent.
I'm not really sure what you mean by "Usually it's probably going to be some mix of the two" - surely, for deciding if any given thing belongs to a category (in this instance, whether or not the output of the AI counts as a story), either there is a consistent position for what the category encompasses (there is some explicit definition of story which Alice would agree to, even if she couldn't formulate that definition herself), or there is no consistent position, and we can't reason about the category (there is no string of text that Alice would consider a story, if it happened to be assembled by an AI, even though she would not endorse that position).
If we're making arguments about what a technology can/can't do, those have to be falsifiable in order to have predictive value. Arguments of the form "AI can never do X because things are only X if I personally determine them to be so" don't get us anywhere.
If it were up to me, I would refine this to, "If Alice can conceive of a claim that is worthy of our serious consideration, then she ought to be able to phrase it coherently."
What do you mean by "ought" here? If Alice doesn't phrase her claim coherently, is it because the claim isn't valid, or is it because Alice is just deficient in articulation?
David's comment seemed to suggest that we should be skeptical of Alice's claim for various ambiguous reasons. My point was that I am okay with rejecting claims purely on the basis of inability (or unwillingness) to phrase coherently. Part of this is that intentionally ambiguous phrasing is a way to keep moving the goalposts but really, if one has a claim that is worthy of our consideration, then it should be well thought out. And if they were capable of that, then they should either write it properly or get help from somebody who can.
Partly, I guess, it feels like we are doing engineering here and, as an engineer, nothing annoys me quite like being handed a document which shows a whole lot of engineering work was done, but completely failing to communicate what the result of that work was.
I don't disagree with you, but for the sake of sparing us all more ambiguity I'll spell out my argument a little more clearly.
Claims like Alice', that "I can't define it but I know it when I see it" can resolve in one of two ways: Either Alice eventually arrives at (or sees and endorses) an explicit definition for the thing she currently can't describe, or alternatively there is no explicit definition which Alice would ever endorse, and the thing we're talking about isn't a coherent category.
We can't come to a coherent conclusion arguing from an incoherent premise, so until we know which way Alice' implicit definition resolves, we should avoid engaging on those terms. We can have fruitful discussions to come up with an explicit/falsifiable definition of "story", but we can't argue about if AI will every write a story if there's no way to know ahead of time what counts as a story. We should take seriously the argument that "there's something meaningful here that bears on {AI}/{Other important topic}, but we can't describe that right now" but we should disregard the argument that "This thing I can't describe right now bears on {AI}/{Other important topic}, failing to address this is a hole in your argument.".
As a software developer, I can sympathize with ambiguity annoyance.
My angle here is that our goal is to get the right answer, whatever it is. We normally go about it by temporarily playing the role of an adversary, but we and Alice ought to share this goal. It might be that Alice is being deliberately ambiguous; it might be that Alice is trying her best, but is bad at communicating; it might be that Alice is trying her best, but her claim can't be defended in the limit - a master communicator couldn't defend it, either.
Part of our task is to play the adversary role, but unless there's a designated arbitrator, we should probably also have a side process trying to understand what Alice is saying, including asking her questions to try to clarify her claim. We're permitted to consider the possibility that she's ambiguous on purpose (or stalling for time, etc.), and disengage and perhaps even decline to engage again in the future.
Matching your engineer's annoyance (and mine) with ambiguity is mine with cases where I've seen people abuse the principle of concluding the other side is ambiguous, by insisting on hyper-specificity, pummeling an Alice who might be trying her best, and if she stumbles even once, they invoke the ambiguity clause in order to win. In other words, I've seen abuse in both directions. (I've even seen abuse of this observation of abuse. Some people can be very clever, and a lot of people have less incentive to get a right answer than they do to just win arguments.)
I wrote about why we should not moralize procreation, and why there should be different measures of epistemic position depending on what criminal justice goals you have. https://hiphination.substack.com/archive
The only meaningful response to that idea was best expressed by Razib:
"Should Religion Play a Role in Politics? [link to a blog post featured on the New York TImes] This is like asking 'should men have sex with women?' It's always going to happen."
Not if Gen Alpha continues the trends started by Gen Z!
Following on from AI’s success at GeoGuessr, I decided to test the free platforms on radiology. Neither was a big success. ChatGPT undercalled an acute radial head fracture, saying it was a normal elbow X-ray, and Grok overcalled it saying that there was complete dislocation in addition to fracture. Neither was able to correct with prompting by query (Chat doubled down on normal when I asked it specifically if a radial head fracture was present, Grok insisted the elbow was dislocated when I asked if it was sure.) I only got correct diagnosis from Chat when I insisted “this image shows a radial head fracture.” Oh well. Radiologists to stay employed for now.
Chatbots were not (generally) trained on medical data or images, and it likely would have been low-reliability data. There are quite a few radiology companies that are using machine learning specifically with high quality data of MRI, CT, and PET scan images. So far, it has seemed quite promising. However, a chatbot would not be the relevant usage of ai in this case. Take a look at companies like Therapanacea, deepc, Aidoc, or Siemens (partnering with Britain's NHS). For example, if you can take a look at the white paper for Rad AI (Siemans), so far results seem to be very good, particularly for expediting radiologists' workflow. There are white papers available with relevant studies and sourcing.
Sure, but what does that say about some "ChatGPT ver. X" becoming literally superhuman in every respect and taking over? Having a bunch of narrowly specialized expert systems sounds useful, but is a far cry from your typical doomsday scenario where that radiology AI becomes sentient, finds an Internet-connected device, and takes over the planet.
Sure, if an LLM would suddenly know, how to anaylyze Xray images out of the knowledge, how the human body is composed and how Xray works, that would be a step towards superhuman intelligence.
However, I am pretty much sure, an LLM canot be better than the training data. Without superhuman training data, no superhuman output.
If the creators of ChatGPT ver. X want it to be good at analyzing CT scans, they could train it on a high quality dataset and it would be good at radiology and still have its general knowledge. The specialized radiology ML isn't going to become general purpose; the general purpose AI will incorporate the specializations.
Yes, this is why my views on "AI" are a perfect mix of wild-eyed optimism for specialized systems and an "I can't believe smart people expecting this to happen in two years" pessimism for "AGI" or "ASI" or whatever acronym is supposed to do Magick.
Would you expect to see a NN-based specialized radiology software designed to do one thing only, read x-rays, with modules for, e.g., "Fractures", "Lung disease", "Brain Tumors", to be much better than a generic ChatGPT?
That's kind of where I see the path forward, huge open field for specialized AI applications as we already see in image processing, for example, while "AGI" remains mostly hot air and fearmongering.
This dialogue with Moldbug baffles me, mostly because there's a dialogue at all.
From the outside, it's like watching a twitching, seething mass of concentrated hatred and contempt, scream in a voice of a thousand howling voices "Coward! Tear down society! Build the God King a throne of skulls if he will tear down our rivals for us! REOPEN AUSCHWTIZ!"
Granted, 82 of the eyes if his twisted frame are winking in an exaggerated fashion when he says the last line, but other half-formed hands are sharpening knives and drafting legal memos justifying the active under unitary executive theory.
And then Scott nods sympathetically and says "hmm yes, you make some good points and nobody can doubt your lucidity, but on balance I think we should perhaps not drown the streets in blood so that our biological lessors are swallowed up before we are. I will of course update my priors if conditions change. Thank you for giving me much to consider."
This is such a confusing position to me. None of what Moldbug says sounds like it's in a uncivil/violent tone at all. His and Scott's levels of courtesy are comparable. The points being made are real, so they receive counterarguments. Why wouldn't they?
Isn't the point not so much about tone as much as coherence? Scott did a great job making as much sense of Curtis's 'argument' as he could in his response. But, it did seem like he was contorting himself to make a normal counterargument. In a lot of places it was really hard for me to understand what Yarvin was even saying. For example he refers to a 'regime' twice:
"You forget that this regime has spent the last century managing public opinion with every carrot and stick it can find, up to and including asking professors to compose their own inventive and detailed loyalty oaths to gay race communism."
"You forget that during the lifespan of this regime, whose crimes have no statute of oblivion, many once civilized places became desolate and dangerous, forcing whole populations to flee—including most of most of our cities, even the hometown of the new Pope lol. Oh yeah that"
What is his definition of regime here? There has been a regime for at least a century, but during this 'regime' people have fled 'most of most cities?' It's committed "crimes with no statute of oblivion". Is the regime the US's current governmental system, is it the political left, is it democracy itself, or classical liberalism? What existed before the regime? Has it been here forever? Presumably someone built the 'civilised places' to begin with. Who sets the regime's positions on managing public opinion, how has orthodox opinion changed so much over time if it's managed so tightly, surely the 1962 version of the 'regime' wouldn't approve of 'oaths to gay race communism'?
I realise things like this might be more evident to people more familiar with his writing. But I honestly don't get it, just comes across as rambling stream of thought and it's kind of funny comparing it to Scott's response.
To be clear, I don't give a single shit about whether Yarvin is courteous to the people he speaks with, although I think throwing around the term "libtard" disqualifies him from that, as does his threat to re-open Auschwtiz to show the libtards how much worse Trump could be. Curtis Yarvin should be cast in the Outer Darkness because his stated political goals are to overthrow democracy (not narrowly, but any form of government where the consent of the governed is a thing that exists and rulers need to worry about) in favor of some mishmash of Nazism, Imperial Russian tsardom, and worship of corporate profits.
I do not believe we as a society should legally declare Yarvin to be a wolfs-head or an enemy of all mankind who sits outside of any legal protection. That would not be a fair or just act. But Yarvin has said that he would do exactly that to all of us and every single other political opponent he has if the boot were on the other foot. Just something to keep in mind when deciding how we all voluntarily choose to interact with Curtis.
Indeed, the stakes are high. Seems like all the more reason to engage with the arguments properly. Proper debate leads to better accuracy. Or, if you want to dispose of investigation entirely and go into the "arguments as soldiers" mindset, then at least ensure that the battlefield is the one where correctness has an inherent advantage, rather than the coin-toss of conflict. https://slatestarcodex.com/2017/03/24/guided-by-the-beauty-of-our-weapons/
My guy, you're lecturing me on being insufficiently polite and charitable while Curtis is saying we all deserve to be thrown in Auschwitz as a reminder to know our place and not complain.
Do you have a link to this Auschwitz post?
It's in Yarvin's reply to Scott. It's phrased as a joke of course, some sort of elaborate demonstration that liberals need to stop comparing Trump to Hitler. Not that Yarvin really condemns Hitler anymore, except in so far as Hitler lost.
https://x.com/curtis_yarvin/status/1921541829683445936?s=19
I find the debate frustrating as well. In my opinion Yarvin is little more than a comparatively eloquent charlatan. But I also think it’s likely useful for someone, at least, to engage with charlatans matter of factly and politely.
If a Nobel prize winning virologist were to publicly debate RFK jr on vaccines, it’s not at least self evident to me that the optimal approach would be to treat the latter with the degree of contempt he deserves. Once someone’s inside the Overton window, that’s not really an option anymore.
I would fully expect this debate resulting in even more people concluding that the virologist was one of those contemptuous elites who want to chip everybody, and RFK was just the Man to resist the Cathedral.
One of the core tenets of rationalism is that you treat the argument itself, not the person making it, or what it would imply if true. Scott is the true Caliph of Rationalism.
A lot of people disagree that this is a good thing. You're also allowed to hold that position.
But you coming to his blog and complaining that he's taking Moldbug's arguments at face value is like telling the Pope that Jesus wasn't the son of god.
This is a bit problematic coming right on the heels of “the purpose of a system is what it does”, where he didn’t really make any attempt to come up with a steel man of what the position might mean, and just criticized the literal claim.
I'd be more sympathetic to your point if _anyone_ could agree to what the position might mean.
Quoting Scott:
> In this comment thread, people have claimed that the real meaning of POSIWID is:
Chesterton’s Fence
Moloch
Alienation of labor
Pournelle’s Iron Law of Bureaucracy
People follow incentive gradients
If nothing is changed, things will stay the same
If a system keeps going despite side effects, it’s okay with those side effects
If a system has side effects, those side effects are secretly the whole point
It’s about machines and was never intended to apply to social systems
These are pretty different things! So I continue to think that, if you like one of them, you should consider the possibility that this phrase isn’t a clear way to communicate the thing you like.
> ENDQUOTE
If after 2 articles and a couple of flamewars people can't agree on the definition, I'd say using the empirical way people use the phrase is valid.
I don't like the phrase myself! But there are several valuable and meaningful ideas that are adjacent to it in obvious ways, and Scott didn't even bother to identify those.
(I'm personally partial to "the purpose of a system is those aspects of what it does that explain why it persists and works the way it does".)
POSITAWIDTEWPWTWID doesn't really roll off the tongue, and that's saying something compared to the original haha
One of the core tenets of any way of thought worthy of the name "rationalism", is that a rational being must have a way of quickly and efficiently discarding many, many arguments *without* engaging with them. And frequently the behavior and/or reputation of the person making the argument, will be decisive in that. If it's a good argument, some less obnoxious person will make it sooner or later. Until then, we at least have a defense against the DDOS attack.
Unfortunately, enough "rationalists" have engaged with the bundle of obnoxiousness that is Curtis Yarvin, that someone like Scott probably can't just dismiss them out of hand. But I agree with the OP that it is sad this is so.
I haven't read Curtis Yarvin, so take the following as my outsider perception:
It seems like Curtis really was the only one making these arguments for a dictatorship. And seems like he used to be less obnoxious.
Scott just started paying attention to him before he built a reputation that would trigger someone's mental spam filter.
Though yeah I agree with you in general.
A pervasive problem with the Rationalist Movement™ and with people (like me) with personalities drawn to it, is that we don't have robust defenses against Gish galloping. We feel instinctively that anything that looks like a syllogism or a factual claim is worthy of engagement, so we spend hours contemplating it, while the galloper has distracted throngs of us with pseudofactual or pseudological claims and has since moved on to make unchallenged pronouncements to vulnerable people. You're right that any movement that prizes logic needs to reckon with this failure mode, but ACX readers have largely not figured out how to do so.
Mehti Hasan (Win Every Argument) recommends dealing with Gish gallopers by picking out one weak argument and hammering on it. I believe he's talking about interviews.
It's worse than that. Firstly, plenty of people hang out on rat forums who are themselves not rationalists. Secondly, some of them (and even some of the rationalists) *have* figured out a defense against Gish galloping - namely, declare it's Gish galloping and use that to justify not engaging.
The catch is that they'll also use it against arguments that aren't Gish galloping, but resemble it. Hence comes the strategy: Step 1 - make a succinct claim that smuggles a large number of mistaken or controversial premises. Step 2 - wait for the inevitable response that points out all the smuggled premises. Step 3 - declare "that's a Gish gallop! I must be be right after all!".
To a third party observer, it's not enough to side with the party that called out a Gish gallop, because there's often no substitute for inspecting the actual argument, especially if that argument is in a domain the third party isn't familiar with (which happens a lot).
Many such third parties are rationalists, who are well aware of Gish gallops and other debate anti-patterns, who are willing to use their presence as a heuristic to decide who's right, but who are also aware of the anti-pattern of calling out other anti-patterns in order to appeal to the rationalist!
Yarvin does not make an argument. Or if he does, he purposely uses obscurity to hide what his actual argument is, likely so that he can shift goalposts later. The most generous thing I can say about Yarvin's thread is that I legitimately don't know what he's saying, and would invite him to actually make an argument grounded in statistics instead of vibes. I suspect he would never do that because his personal identity would be quickly be annihilated by the contradiction in his views and the actual facts on the ground.
How can you take the arguments of someone who presents boatloads of sneering irony at face value?
I’m reasonably sure he is running the long con but I do kind of like the guy’s boho fashion sense.
This may be a tenet of "Rationalism" but it is not a tenet of "rationalism." It is clear that in developing a rationalist philosophy, it is necessary to tame the reflex to judge everything without listening but clearly this can be taken too far. One also needs exercise rationalism in reevaluating the consequences of blind adherence to tenets.
To live in the world, you need to filter out the crazy people, bored trolls, conmen, spammers, and the like. But IMO rationalism is about the tools you use to understand the world and engage with ideas once you've decided to pay them some attention.
You don't need Ed Jaynes' version of Bayesian probability theory as the more useful and complete version of logic to evaluate the claims of the guy ranting on the subway about how Donald Trump is the antichrist or the widow of the Nigerian oil minister who has just emailed you an amazingly generous offer. But at some point, you've got to evaluate some idea or argument or claim. Sometimes, that idea will be weird, will violate everything you grew up believing, will be covered in the stench of the outgroup, will be offensive and upsetting and low-status, etc. And that's where you need the tools of rationality most.
But is Moldbug really adhering to tenets? Or using whatever tenets and 'rationalist sounding' arguments to whitewash a personal, emotional response?
I tend to pretend that I am more objective and rational than most, but have learned that we all need to be more aware that emotion has a much stronger influence than we'd like to believe (not that we shouldn't try to control it at times).
Silverax appeared to be excusing a behavior that Scott has, on occasion, been accused of here. That is, "taking at face value the argument of individuals whose arguments may not warrant quite so much deference." He was excusing it based on, "...core tenets of rationalism is that you treat the argument itself, not the person making it." My comments were that real rationalism requires less tenets and more (continuous) self examination. A poor performing tenet should be discarded or modified.
I really agree with this, and to be more direct, Scott:
I absolutely love all of your posts on medicine, psychology, pharmacology, art, and rhetoric. I don’t agree with everything you say but it’s all fantastic writing and has contributed a huge amount to my view of the world. I’ve shared many of the ideas in these posts with my family and friends.
But despite this, I’m extremely hesitant to recommend Astral Codex Ten to anyone who isn’t super close to me (when they ask where I saw an idea, I usually say “a blog”), because of this deliberate courting of and deference to open bigots. Yes, they are free to express whatever beliefs they choose and it’s fine to engage with them seriously rather than with ridicule, but the level of focus and steelmanning they get can seem to evince an actual preference for them.
Would you treat an “extreme woke” viewpoint with the same level of consideration? If not, why not?
I think that if you don't recommend ACX to the sort of people who would get mad at this sort of thing then this is best for you, for them, and for ACX.
Open bigots run the world's most powerful country now, a country in which Scott happens to live.
The only realistic alternative at this point to engaging with and pushing back against their ideas is sealing yourself away in smaller and smaller bubbles, until the power of the outside world inevitably pops them.
If you can't clearly defeat open bigotry in free debate then perhaps you should reconsider your prejudices about what, exactly, you consider bigotry. I would point out the danger of conflating positive and normative views as an obvious failure mode there - HBD isn't bigoty by any reasonable use of the term, for example. From an Outside View perspective the attitude of "well we can't talk to them because they believe in <taboo topic>" generally doesn't come from people who have the truth on their side. Consider if creationists or evolutionists are more likely to require bans of the other perspective in an open forum. If your beliefs require special dispensations to protect them then they're likely not good beliefs.
I doubt anyone has defeated bigotry by open debate before. Therefore, bigotry doesn't exist at all, everything is permitted, nothing is taboo, it sounds fun to be sure. But the taboo exists to prevent you from getting sick, you're not a brave rebel for eating roadkill.
You can "defeat bigotry" in the sense of persuading outside observers. Bigotry is by definition a bad heuristic so it should be easy to prove it wrong if it is indeed bigotry.
You can't "defeat bigotry" in the sense of persuading those who are themselves bigots, again pretty much by definition of the way that bigotry works.
I don't believe so. I'm having a hard time imagining an outside observer who is genuinely unsure on whether we should open an Auschwitz to imprison political opponents, as Moldbug colorfully puts it, and who will be swayed one way or another by anonymous commentators bickering about statistics. This sort of stuff is probably already solidified to some extent by the time you leave childhood.
It's not so much that the observer is on the fence about torture camps. It's more often that both positions don't say "therefore, we should start torture camps" and also say "but the other side will totally lead to torture camps" and the observer can't trivially tell, especially given that each side has an enormous incentive to do that.
The worst thing about Hitler are all the "literally Hitler" arguments. (Cue Norm McDonald spinning in his grave.)
Meanwhile, Melvin's right. Especially on online forums, your audience isn't just the person* directly responding to you, but whoever else happens to be reading but not responding. If they see both sides screaming "no, *he's* the real fascist! Shoot *him*!", then one of the few heuristics they have left is noticing which side seems most willing to give up that type of rhetoric and at least attempt to argue rationally.
*assuming, of course, it's a real person
>I doubt anyone has defeated bigotry by open debate before.
Then how did segregation end?
The old guard goes away, replaced with a new generation with different prejudices, more fashionable/better adapted to their changing conditions.
I don't think that's accurate. Earl Warren wasn't exactly a kid when Brown v Board was decided. Neither was LBJ when the CRA was passed.
If "I know you're not racist but I'm concerned *what people would think* if they read your stuff" worked, it would have worked 10 years ago.
He treats extreme woke viewpoints with this level of care and consideration all the time ?!
People, who are smart, agree with Moldbug. Smart counterarguments are one of the useful vectors to persuade them.
Did you read his post about “the purpose of a system is what it does”? Here he just stuck with a few very surface level attempts at something and didn’t attempt to charitably read what truth (or at least, plausible point) it might be pointing at.
Is that an extreme woke viewpoint? On the surface, it doesn't seem to have any of the characteristics of wokeness.
Not exactly, but I associate it with the claim that police violence isn’t a malfunction of the system but just evidence that the point of policing is to oppress minorities, or the claim that the number of uninsured people in the United States is not a failing of the healthcare system but the intention. It goes along with revolutionary ideology rather than reformist ideology.
Moldbug has written many thousands of words on the topic.
Point me to an equally interesting post about the truth behind 'the purpose of a system is what it doesn't that a minority of the people using the phrase at least somewhat endorse and there will be something to talk about.
Or, would careful treatment of some other, different, woke point be sufficient evidence?
> People, who are smart, agree with Moldbug.
Some people who are smart.
I think the comma conveys the "some" part. That's how I read it.
Uh, I hope it was obvious I didn't mean 'the majority of people agree with moldbug, and crowds are wise' or 'smart people are smart, and smart people are more likely to agree with moldbug'. If unclear, I was saying 'a portion of the people who agree with moldbug are persuadable by honest, logical arguments'
Fair enough. Thanks.
Couldn't agree more. It's like his depictions in Unsong of Satan being at UN meetings! If i had to steelman it, he needs to appear reasonable *to the target audience* aka Moldbug sympathizers, rather than his audience. In politics you need converts, and he is confident any bridge will see net positive immigration... even if his current base don't want them around...
Fractals!
It really does all come down to mass immigration. If one believes in a strong version of HBD, and observes that the normal political process has completely failed to stop the problem, then building the God King a throne of skulls gets a lot more reasonable.
You’re right that these bravery debates are counterproductive, but no one has figured out how to have the real debate yet.
> then building the God King a throne of skulls gets a lot more reasonable
No, it doesn't. There is never a scenario where it is reasonable to build a throne of skulls. If you are advocating for building a throne of skulls, you have lost the plot.
Like, just step back and think about what you wrote. "Because I believe in IQ differences, it is correct to murder a bunch of people in service of a dictator."
I suggest you watch this (https://www.youtube.com/watch?v=ToKcmnrE5oY) and think about the implications.
Except modern immigration is far from the problem it's portrayed to be. It has not increased Crime, it has not brought the US or most countries for that matter to poverty and it has not systematically changed the culture and law of a single country.
Anyone who wants a dictator forgets 2 of this 3 points, and depending if you count repression and crimes against humanity even the 1, would strongly increase in most democracies where this is applied.
Yes if it's a benevolent king/CEO whatever that won't happen, and yet the cases of an illuminated dictator in modern times can be counted on your hands.
> It has not increased Crime, it has not brought the US or most countries for that matter to poverty and it has not systematically changed the culture and law of a single country.
Crime has increased in Europe.
The claim about poverty is true but neither has immigration helped gdp per capita or wage growth. Poland is catching the U.K. in GDP per capita (at least in PPP) and it’s doing that with no immigration and falling population. (So much for that population decline doomsterism).
And the best we can say is that laws haven’t changed much yet, although sharia courts do exist in civil law.
I was thinking about the US but interesting information, is there any proof that this is related to immigration?
Poland is not a good example since it's an emerging economy whose GDP increase has nothing mysterious, you would need two mature economies of similar nature and significantly different immigration flows.
The Sharia courts do you mean in Europe or US? Because i found no mention of that, would you have an example?
There are no general Sharia courts in civil law in Europe or the US.
The vague claim that they exist in Europe rests on one misunderstood anecdote plus a couple of tightly-limited examples that are outside of civil law.
The misunderstood anecdote is that in the UK in 2018 a court ruling recognized the application of Sharia law when a couple had married only in an Islamic religious ceremony not under civil law. The court said that in that circumstance they could not pursue divorce under UK civil law (which considered them to be simply cohabitating); they could however pursue divorce from a religious marriage under the particular religious law that applied. UK courts remain clear that if couples have legal (civil) marriages then they are bound by UK divorce laws without regard to religious status.
In the far-NE corner of Greece, in a single small province bordering Turkey that has long had a sizeable Muslim minority, there is a hundred-year-old treaty allowing practicing Muslims to use Islamic courts on matters of marriage, divorce, adoption, etc. That treaty currently applies to around 100,000 people.
In Germany marital matters between Muslims with no German citizens involved are handled by German courts attempting to follow Sharia law, with the somewhat-vague stipulation that the outcomes can't violate the German legal system's "broad principles". That has thus far resulted in muddled rulings making, it seems from this distance anyway, roughly nobody happy.
And that's it, that's "Sharia courts" in Europe.
There are no Sharia courts in the UK, but there are Sharia councils: https://www.theguardian.com/law/2017/mar/01/inside-britains-sharia-councils-hardline-and-anti-women-or-a-dignified-way-to-divorce
"Prompted by fears that they were discriminatory, Theresa May launched a government inquiry in May last year. Just a month later, the Home Affairs Committee announced its own. In December, the Casey Review by Dame Louise Casey into integration included claims that sharia councils “supported the values of extremists, condoned wife-beating, ignored marital rape and allowed forced marriages”."
A Sharia council doesn't need to have formal legal power to have de-facto power over Muslim women. They are dangerous for the same reason that cults are dangerous even though any member is legally free to leave at any time, and for the same reason that abused spouses often stay with the abuser despite being legally able to file for divorce.
To begin with, "Crime has increased in Europe" may or may not be true depending on what you look at.
https://ec.europa.eu/eurostat/statistics-explained/index.php?title=Crime_statistics
Non-sex crime has been on a long-term trend of decline except for a very modest rise starting in 2021. Sex crime has been significantly rising, but that is probably higher reporting rates.
The God King is bringing in African refugees now!
But there's a nuance: https://www.bbc.com/news/articles/crljn5046epo
Yes, of course there is. Another troll. This one is up there with nominating Matt Gaetz to be AG.
> "A refugee is someone who has to leave their country out of fear of political persecution, religious persecution, or economic persecution," Ramaphosa said. "And they don't fit that bill."
Considering "Kill the Boer" is a popular political slogan there, I'm not sure how Ramaphosa is saying that with a straight face.
I think saying it's a "popular slogan" is misleading: one very provocative politician is famous for singing it; he was tried though acquitted for hate speech for it (which I think the general pro-free speech vibe of this forum should at least have some sympathy for); in the most recent election his party got just under 10%--not trivial, but "popular" seems pretty strong.
I think Afrikaners face the *prospect* of persecution if the EFF were ever to come to power, which isn't out of the question, but I think currently their grievances are crime, corruption, and mismanagement that don't target Afrikaners specifically. If Afrikaners are refugees for having a government that doesn't deal with crime, then so are all other South Africans, regardless of ethnic background.
> he was tried though acquitted for hate speech for it (which I think the general pro-free speech vibe of this forum should at least have some sympathy for)
OK, but the problem there isn't hate speech. It's incitement.
The one good thing about Trump is how idiotic he makes all his "smart" supporters look. The first go-round it got most of the IDW types. Now he's come for the neo-reactionaries.
The latest illustration comes curtesy of Qatar:
Qatar (trying to manage the obviously bad look of a bribe): a Qatari spokesman described the claim that the plane would be a gift as "inaccurate".
Trump: "They're giving us a gift," Trump told reporters at the White House on Monday, adding that he would be "a stupid person" if he did not accept it.
https://www.bbc.com/news/articles/cy5ell3gkxvo
Actually, Trump does that to everyone. His enemies make fools of themselves overreacting to some random outrageous tweet or comment, and then his friends make fools of themselves by jumping in to agree with the outrageous tweet or comment.
The "Trump is all bluster" crowd looks even worse right now.
c.f. Politifact labeling Kamala Harris's claim "half true" for assuming Trump would actually implement tariffs as high as he propmised on the campaign trail. https://www.politifact.com/factchecks/2024/aug/30/kamala-harris/would-donald-trumps-proposed-tariffs-hit-typical-f/
I'm not saying Trump is all bluster, I'm saying he gets his opponents to make fools of themselves overreacting to things he says and does. That doesn't mean he's not also doing bad things (and indeed, often enough he is), just that a lot of the loudest outrage is about dumb stuff instead of the really bad stuff.
I disagree: the majority of his actions are being undone by judges, and it looks like his second term will ultimately have no more lasting consequence than his first.
So maybe less than a million people dead of novel diseases?
I'm still betting that by the end of it, we'll be Canada's 11th province. \S
As if we'd take you! (I kid, I kid, Americans are mostly great)
Unless you count all the extra dead people in Africa that is.
Looking sane is part of the argument.
See also: https://kennaway.org.uk/writings/Insanity-Wolf-Sanity-Check.html
If AGI does everything better than humans can, then it will improve the charity sector. You'll be able to know exactly which charity creates the most benefit per dollar donated, and exactly how much you'll need to give and to whom to save a human life.
And if it's *really* good it'll understand that charity is a social net-negative and stop promoting it.
Can you say more on why you think this? I can see disadvantages of charity, but large parts of the world don't seem to be able to pull themselves up by their bootstraps. Is your view that the externalities of charities are directly worse than any good they do, or is it mianly about opportunity cost ("the money would do more good here")?
Both, actually. Yes I think charities do long-term harm because they establish poor incentives and those inevitably lead to suboptimal outcomes. But I think the stronger argument is that a utilitarian economic analysis shows that it's better maximize your return on investment regardless of how you value human life. I've made this argument many times here and never gotten a robust response. Here's a more general argument:
https://open.substack.com/pub/astralcodexten/p/matt-yglesias-considered-as-the-nietzschean?r=fo2bp&utm_campaign=comment-list-share-cta&utm_medium=web&comments=true&commentId=63872063
Here's a more narrow economic analysis:
https://open.substack.com/pub/astralcodexten/p/altruism-and-vitalism-as-fellow-travelers?r=fo2bp&utm_campaign=comment-list-share-cta&utm_medium=web&comments=true&commentId=64627110
I've actually considered starting my own Substack just to collect my anti-EA arguments into a single resource. I feel that I have a fairly comprehensive philosophical and economic refutation of the movement and no one has ever provided a good counter either here or on reddit. By all means feel free to cross-examine me if you think you have a good reply.
If you're still using Google for internet search, here's a One Weird Trick to get rid of all the AI junk and other various and sundry "innovations" of the last decade or so: add ?udm=14 after the url, like this:
https://www.google.com/?udm=14
and get Yer Olde Goode Google back, like it was in the ancient times a score years ago.
UPDATE: looks like it's still pretty far from Yer Olde Goode Google, just a bit cleaner....
After doing a very cursory search, it seems like literally all it does is remove the AI summarization. That is so far from the only problem with modern Google and not even close to "Ye Olde Goode Google". Google has been (relatively speaking), garbage for years before it implemented the AI summaries.
The closest I have actually gotten to Ye Olde Goode Google (and still not quite there) is the paid search service Kagi.
Not everyone will think that paying for search is worthwhile, but I'm personally definitely not going back.
I stopped using Google for search years ago, DuckDuckGo is my default go to. But this Tedium piece came across this morning, so I posted this after doing a quick test. Looks like the rot is more systemic and this hack only removes the top layer of decayed matter.
I tried DDG a while back, and what I found was that I kept having to add the !g because searches weren't finding what I needed, and even in it's fallen state, google was better. Kagi also has the !g bang option, but I don't find myself using it.
This is not to argue against your experience at all. I believe you when you say that you find DDG better.
In discussions about these searches, I have found a lot of people that have your experience, and (on hacker news where a lot more kagi users can be found), people who share mine. My best guess is that perception of search experience across the various options must depend a lot on the kinds of things one is searching for and the way in which different people search.
For whatever reason, my combination of search targets and search technique gets better results on Kagi than on either google or DDG.
This is exactly why it's nice to have other multiple options, and I would encourage people who are using google by default to try other things. It is no longer the head and shoulders obvious best option (and, to bring it back around to the original point: AI search is not the reason for it's lack of Pareto-fronteir superiority)
I'm kind of not even claiming DDG to be "better", it's just that it doesn't track me (at least not to the extent Google does), and it gives consistent results for a given query. It was more of "given that I don't want to use Google, what's an easy alternative that works well enough?" question. DDG fit the bill.
I used that link, searched for something, and immediately got an "AI Summary"
Interesting... I'm not seeing AI summaries when I do this. I don't normally use Google for search nor plan to use it even with this hack.
Here's the original source: https://tedium.co/2024/05/17/google-web-search-make-default/
Re #3, in the comments to the response Scott says that housing prices are a stronger case for the claim that the median person is worse off than in the past, but housing does not actually seem to be all that strong a case
https://fred.stlouisfed.org/graph/?g=fXwf
https://fred.stlouisfed.org/graph/?g=18JRa
You really want to look at rent. I was in the rental market until a decade ago and even then I was getting poorer in discretionary income every year. With recent price rises everywhere it’s probably worse now, even controlling for rent.
The second link is re rent.
Consider the context. We know that in the past it was the norm for people to afford kids and their own home while in their early twenties, even on a lower middle class income. But now most people can't afford that. It's likely that something is missing in the data or the analysis. Otherwise we have to explain why everyone in society seems to have noticed a change that isn't real.
People in their early twenties don't choose to get married and have kids any more (because we've decoupled sex from marriage and kids).
If you do choose to get married in your early twenties, and you don't choose to live in a super expensive area, and you make an okay income, then home ownership in the US doesn't seem that difficult to me.
I think we need to stop listening to people who live in San Francisco or New York whe they talk about property prices. We all have a tendency to assume that our local problems are more universal than they are. I want to hear what people in Milwaukee think about house prices.
Which people can’t afford that? People who want smaller houses in poorer cities on 1950s salaries can still have them. But modern houses are bigger and people don’t want the smaller older ones any more, so they hold off to save up for the big modern ones.
Mostly we’ve made it illegal to build those smaller houses from the 50s?
>We know that in the past it was the norm for people to afford kids and their own home while in their early twenties, even on a lower middle class income. But now most people can't afford that.
Except that isn't true. Were that true, home ownership rates would have declined over the last sixty years, but they haven't. https://fred.stlouisfed.org/series/RHORUSQ156N
>we have to explain why everyone in society seems to have noticed a change that isn't real.
Because, as always, people look at the past through rose-tinted lenses.
The homeownership rate is Owner-Occupied Homes/Total Homes. If I rent out a room in a my home to someone who can't find any other housing, the rate is unchanged. If I'm forced to live with my parents because there isn't any available housing, the rate is unchanged. That's why I'm hesitant to jump to conclusions. I've seen historical data showing that the ratio of median home price to median household income has increased dramatically. If I take those numbers and run with them then the housing situation looks terrible. So clearly nuance is needed to figure out what's actually going on.
>If I rent out a room in a my home to someone who can't find any other housing, the rate is unchanged. If I'm forced to live with my parents because there isn't any available housing, the rate is unchanged
But, average household size has declined over time, so it is unlikely that that is happening to any degree, if at all. https://fred.stlouisfed.org/graph/?g=15tO7
>historical data showing that the ratio of median home price to median household income has increased dramatically
Not so much, if interest rates are taken into account. And much of the increase has been very recent (last 5 years), whereas the claim is about long-term trends.
See https://fred.stlouisfed.org/graph/?g=coAW
and note that the interest rate on a 30-yr mortgage in 1984 was 13-14 percent, so of course home prices were relatively low. https://fred.stlouisfed.org/series/MORTGAGE30US
Can you explain that first graph? I thought mortgage payments were cheap in 2006?
Well, it isn't my graph, but it is mortgage payments as pct of disposable personal income. Wasn't the whole subprime mortgage thing all about people with relatively low income getting mortgages they couldn't get previously?
Plus, IIRC, a lot of those were adjustable mortgages, and it looks like adjustable rates rose at that time. https://fred.stlouisfed.org/series/MORTGAGE5US
I think the problem with the first one is that it is total quarterly required mortgage payments divided by total quarterly disposable personal income. So it isn't median, its mean, and as you say, lots of new people on the market, and lots of investment properties. From memory houses were actually really cheap at that time for the median person. Lots of zero deposit loans.
So you'd want to find something like mortgage payments for the median house, then divide by disposable personal income of the median person (or household?). Still need some way to factor in the deposit.
>From memory houses were actually really cheap at that time for the median person. Lots of zero deposit loans.
But they weren't cheap, because the mortgage payments were high. Take a look at an amortization calculator. Someone who buys a $200k house with 20 pct down and a 5% 30-yr mortgage pays $308k over the life of the loan. With zero down he pays $386k (actually more, bc he won't get the same interest rate).
Moreover, if homes were so affordable, why did so many borrowers default?
Finally, while it would be nice to have median, since we are comparing over time, what is important is that we are comparing like with like, unless we have reason to think that the gap between median and mean has changed over time (if anything I think the gap has grown over time, so the graph probably overstates current costs).
Just like to point out that the illustrated tails are trifid not bifid.
Curious Scott's or any psych researcher's thoughts on Cheng, Y. C., Huang, W. L., Chen, W. Y., Huang, Y. C., Kuo, P. H., & Tu, Y. K. (2025). Comparative efficacy and tolerability of nutraceuticals for depressive disorder: A systematic review and network meta-analysis. Psychological Medicine, 55, e134. I read it (and comments elsewhere about it) as saying a series of nutraceutical may match or beat SSRIs for depression.
I've been considering whether it would be possible and interesting to run prediction markets in a LARP setting. Does anyone know of a system I could use to run this which I could execute with nothing more advanced than a slide rule and ledger? (Ideally I could pre-calcuate some ratios and have a reference tome to look them up in.)
(The system in question is Empire in the UK, which has ~4000 players trading such that there is actual economic activity happening, like a recent liquidity crisis caused by spending too much coin with foreign merchants.)
Further specifications that might be relevant:
* I'd want to denominate things ideally in the smallest coin, a "ring" - players get an income worth somewhere around 120 of these, though not all of that in pure coinage. Expecting people to have ~10 rings in their pocket is reasonable; basically I'd like to keep the stake sizes a) integer and b) small enough they're accessible by as many people as possible.
Robin Hanson has a prediction market board game for guessing movie outcomes. Maybe you can adapt that. I’ll see if I can drum up a link
had the same thought after seeing Eve online, you might find that setting even more interesting.
The simplest method to implement would be a parimutual betting pool. The Wikipedia article is pretty good. There are reasons why prediction markets use other methods - the expected return on you bet depends on the bets placed *after* your bet - but it should be simple, transparent and straightforward to implement. Note that payouts might result in fractional rings.
To remove the risk of fractional rings entirely you could have some sort of open outcry system. If there is a contract that pays out ten rings when the blue team wins and zero when the red ream wins, then people could meet in the "pit" and call there buy/sell offers: "I am willing to buy 1 blue-team contract for 4 rings", " I am willing to sell 2 blue-team contracts for 10 rings", "2 blue-team contracts for 10 rings? I accept your offer; here are ten rings, thank you for the two contracts", etc. If you've ever played the cardgame Pit, you should get the vibe. This method solves the fractional rings problem but can be very very chaotic.
The constant product automated market maker is actually fairly simple to implement with pen and paper - whatever liquidity the market starts with, that will be the liquidity throughout; you only need one reference table with only two columns - but there will be a huge fractional rings problem.
My recommendation: Avoid constant product market makers. Pick from Parimutuel pools or open outcry or maybe an auction based model.
Naomi Kanakia was profiled in the New Yorker https://www.newyorker.com/books/page-turner/is-the-next-great-american-novel-being-published-on-substack
A new project from neal.fun (known from things like The Password Game and Stimulation Clicker) is Internet Road Trip:
https://neal.fun/internet-roadtrip/
a collaborative Streetview roadtrip where users online at any time get to vote on which direction to go. You can also vote to honk the horn or to change the radio station.
I've been checking in every now and then. The trip started a few days ago in Boston, and right now it's covered 376 miles (it's much slower than a real road trip). The trip started heading southwest as if it was planning to head to California, but upon reaching Providence it twisted around towards the north and now there seems to be a consensus in favour of heading for Canada. At the time of writing (while everyone in the US is asleep anyway) we are heading up a coastal road in Maine.
What's interesting is the social dynamics of the voting. There's clearly people who want to stick with the plan -- we're heading north, let's just keep doing that as efficiently as possible. But then there's people who want to mess things up, taking every turn they possibly can because heading in a straight line is boring. These two teams fight it out every time an intersection or turnoff comes along. Often the chaos party will win, but then we all discover that the side road is frustrating and collaborate to get back on track, leading to a path which is overall progress marked with occasional fits of wandering. There's a comment section which will sometimes turn into name calling between these two parties as well. Oh, and occasionally there's a fit of interest in honking, which dissipates the moment we've honked. The only thing I haven't seen happen yet is a change in the radio station, which has been set to a jazz station in New Hampshire for a long while now.
Implications for real world politics are unclear but definitely present.
Without looking I can tell you the jazz station is WJSK out of Jackson, and that IRL it doesn't reach more than 15 miles into Maine, let alone the coast, but I'm happy that you are all enjoying it!
Ah well it looks like we've now switched to WBOR 91.1 FM - Bowdoin College Polar Bear Radio which is playing an eclectic mix of... I'm not even sure what but right now it's Talking Heads. It's broadcasting out of Brunswick and I have no idea whether you can hear it down here on the far side of Portland but they're at least making some effort to give us state-appropriate radio at least.
Right now it seems like the host has figured out he's currently internet famous (not _that_ famous, there's only 550 people online right now but that's a lot more than usually listen to Bowdoin College Polar Bear Radio at 10pm on a Monday) and he's having a great time and playing all his favourite songs.
I wonder how many people IRL will be killed by Twitch Drives a Car?
The divide between the "chaos party" and the "keep going party" sounds a lot like the dynamics of the original Twitch Plays Pokemon. The original mode was "anarchy mode" where the game just took the first input it saw in the chat, and then they added a "democracy mode" where people voted for the moves, which made it a lot easier to actually progress in the game, and the game shifted between the modes based on voting.
A lot of the dynamic there was people trying to actually win the game trying to get it into democracy mode (required a supermajority), while the other group tried to keep the game in anarchy (requiring just a majority) and would vote for "start9" (press the start button nine times - did nothing but waste time) while in democracy mode. The latter group was a mix of people who just wanted to make trouble for its own sake, and people who just felt like anarchy was the "truer" way to do TPP.
Maybe a similar dynamic here where some of the people voting for turns are just making chaos for its own sake, but others just generally prefer the original goal and don't want to go to Canada.
What's interesting is that I can't quite decide on a team myself. Obviously if I'm going to be watching this darn thing then I want the car to go somewhere, but if we just head inexorably north by the most efficient possible route then that's no fun at all, the game is only fun if there's a reasonable balance between collaboration and conflict.
In the thirteen hours since my original post the car has gone another 36 miles, so we're doing a brisk walk. Team Progress has managed to get the car onto I-95 heading north and Team Zen Wandering is annoyed at the lack of exits. People in the chat are naming themselves things like "#1 interstate hater" and "#12 interstate lover" and occasionally the haters will stop the car to honk the horn in protest for a few turns before getting bored and deciding that they'll never get off the interstate unless they keep going to the next exit.
Other social dynamics are interesting. The big question of where to go seems to have been resolved, and even the "just dick around" party seems to be less active, the big question is whether to take the interstate or the side roads. Given that you don't actually travel faster on the interstate and it's much less interesting, the back roads seem like the best option, but they provide more opportunities for team chaos to dick around.
I think there was original enthusiasm for heading across the country (which was probably Neal's intention when he started off in Boston) but once people realised how long that would take they chose Canada as a more achievable goal.
I mostly appreciate his website for the increasingly absurd Trolley Problems.
It seems obvious that the core implication is that jazz is the Great American Uniter.
How can one find good editions of public-domain books? A popular public-domain book will usually have a lot of different versions available on Amazon, but no way to tell which ones have decent typesetting and printing. I am not looking for expensive collector’s versions, just good-quality trade paperbacks. For example, I recently ordered a copy of The Wind in the Willows that I figured would be nice because it had the first-edition cover design, but it was actually low-resolution facsimile and poor type. Is there a good brand in this market, or good heuristics?
I happen to have a Barnes & Noble Classics sitting on my desk right now. It's pretty nice. I've also been reading some Penguin and Dover books recently. I think the Penguins are probably a bit nicer than the Dovers.
Just go to the bookshop and have a look?
Don't even have to go to a bookshop, you can preview a limited number of pages for most books on Amazon.
What kind of books? Penguin is known for paperbacks of the classics. Everyman’s Library books are nice but not so cheap, usually hardcovers I think.
Dover publishes pretty cheap out of print books (in my experience mostly nonfiction).
Has anyone else noticed that "Daniel Kokotajlo", AI Safety advocate, shares the exact same name as "Daniel Kokotajlo", Some Random British Director?
Not to disparage the director, but this feels like a Problem. And when you google the name, it's still the director who comes up first.
No, no one else has noticed that.
If people had noticed, it seems clear that there wouldn't be a problem, because noticing the phenomenon by definition requires being aware of both people.
If people haven't noticed, there could potentially be a problem in that people might treat one person as if he were the other one, but I don't think that's what you have in mind.
Do you have something in mind?
I guess the British director has better SEO.
To clarify, it's a problem if 1) you take AI Risk seriously and 2) you think Daniel is an important figure/spokesperson for AI Risk.
(though to be fair, it's probably embarrassing for both Daniel Kokotaijlotlakdltkalso's)
It doesn’t matter at all. You know the name already if you are searching it. You aren’t going to pick the wrong guy.
I think there's an implicit 3) you need to agree with and that's 3) it matters whether you're the first or second hit on Google.
I once worked for a company with a name that would produce a first hit link to a sex offender registry.
At least he doesn't have the same name as a child rapist: https://en.wikipedia.org/wiki/Ian_%22H%22_Watkins#Misidentification
I don't understand what the problem is here. Two people have the same name. One of them comes up first on Google. Yes?
I think the problem is that the other person does *not* come up first on Google.
But he comes up second, so I think this is not a big problem after all.
Funny, you prompted me to check and the AI researcher really does come first for me. Stable across different search engines, too. I just took OP at their word that the director comes up first.
Perhaps Daniel read OPs post and decided to have this Problem corrected in the meantime :D
It’s a problem with a capital P.
And that rhymes with T and that stands for Trouble!
And also with D for making it Double!
I would guess that your parent comment is making a reference to "Ya Got Trouble": https://genius.com/Robert-preston-ya-got-trouble-lyrics
I like my reference better. https://www.youtube.com/watch?v=ZBVrPWwSlRM. It's only 30 years old instead of 70.
There's great Russian literature but even the authors who were dissenters have been used by Moscow.
“Why Russian literature must lose its innocence. Germany's "Russia-complex" is linked to an uncritical adulation of Russian literature. The colonial mindset expressed in many Russian classics is too often overlooked.”
https://efdavies.substack.com/p/why-russian-literature-must-lose
https://zygaro.substack.com/p/putin-and-dostoevsky
https://www.persuasion.community/p/the-slavic-battle-of-the-books
https://daryazorka.substack.com/p/i-see-russia-everywhere-i-go
https://lesiadubenko.substack.com/p/tolstoevsky-must-be-reconceptualized
https://smalldeedsbigwar.substack.com/p/pushkin-at-the-mass-grave
Imperial Russia was a state in the business of colonising Eurasia. Novorossiya was perhaps the flagship of this enterprise, with Kherson being the first major city founded as part of this colonisation effort. It is no different in kind to saying "The colonial mindset expressed in many American classics is too often overlooked."
That description puts Russian imperialism and colonialism in the past. Which is definitely not the case.
"The bloodsoaked lexicon of Russian imperialism
Seven phrases that encapsulate Moscow's violent obsession with conquest"
https://chakhoyan.substack.com/p/the-bloodsoaked-lexicon-of-russian
"In occupied Mariupol, Russian authorities are confiscating almost 6,000 houses and apartments belonging to Ukrainians"
https://azovstal.substack.com/p/in-occupied-mariupol-russian-authorities
"Russia is Weaponizing Religion
Pastors from occupied territories report all Protestant churches are gone from cities like Melitopol and Mariupol."
https://stevenmoore.substack.com/p/russia-is-weaponizing-religion
American colonialism isn’t on the past either. The U.K. poodles along with the US and is supplying Israel with arms. I’m no fan of the British empire (although I’m a fan of Britain) but I’m going to read British literature from the past and present.
I'm in favor of this, but not because I think any of these lunatics have a point. However, these books and writers really are painfully overrated and often uncritically lauded by what one suspects are pseuds who didn't read any of them. I read quite a few, enough to satisfy myself on the subject, and can honestly say that they were all meritless dogshit. If they could get canceled out of social existence and nobody would ever again have to listen to some imbecile braying on about "Dosto" that would be great.
I love Dostoyevski. I can't deny, however, that I started liking him less and less every time an idiot started going on and about his work.
Dostoyevski attracts the kind of people that the author would despise. Most love to ignore the fact that our russian author was a raging orthodox christian.
Similar to how people love to ignore that Lord of the Rings is fundamentally a christian story. Really goes to show how even the most surface characteristics go completely unnoticed by the pseudo intellectual.
I think the orthodox Christian thing is well known. He’s a great writer, but preferably in small doses.
Well known, but very conveniently ignored.
Yeah, being dead white colonialist guys is bad enough, but Russian as well? Used by Moscow? An utter outrage! To the (long overdue) scrapheap of history with the lot!
Justifying genocidal war is bad.
If you'd read those pieces, chucking the lot out isn't what's proposed there.
Quotes like "Kherson is a Russian city" don't justify a genocidal war, which should be obvious to anybody not entirely mindkilled. So what's actually proposed here is suppressing any perspective contrary to what the western "polite society" deems acceptable. If this endeavor seems worthwhile to you, godspeed, I guess.
If that's what you take away, fair enough.
Could you summarize what's being proposed then? I read the first link and it includes this: "In Western debates, the calls from Ukraine to stop promoting Russian culture as long as this war is ongoing are often framed as a problematic call to ‘cancel’ certain authors who are still held in high regard. But, as Viktoria Amelina wrote before her death in 2023, should we in the West really be debating the supposed ‘canceling’ of dead Russian authors while Ukrainian authors are being killed right now? " which sounds a lot like proposing to chuck it all out.
> In Western debates, the calls from Ukraine to stop promoting Russian culture as long as this war is ongoing are often framed as a problematic call to ‘cancel’ certain authors who are still held in high regard.
There are bigger problems. If you try to litigate what's Russian culture and what's Ukrainian culture, you often find that there isn't a distinction. Reason had a good piece on the cancellation of a performance of Tchaikovsky ( https://reason.com/2022/03/09/cardiff-philharmonic-orchestra-tchaikovsky-is-canceled/ ):
> The latest utterly pointless sanction is the Cardiff Philharmonic Orchestra's announcement that it would remove music by Pyotr Ilyich Tchaikovsky, the Russian composer, from its all-Tchaikovsky concert, calling it "inappropriate at this time."
> the objection hardly necessitates scrubbing Tchaikovsky in his entirety solely because of where he was born. That's especially true in light of what was supposed to be the program's main course: his Symphony No. 2, which, in a sort of cosmic irony, is built around…three Ukrainian folk songs.
> For those familiar with Tchaikovsky, that likely won't come as a shock. The composer spent several months a year in Ukraine and had close family ties to the region; his paternal grandfather was born there. "I found the peace of mind here that I had unsuccessfully sought in Moscow and Petersburg," he once wrote
But "Tchaikovsky was explicit that he preferred Ukraine to Russia, it's just that Ukraine was called 'Little Russia' at the time" is probably even more offensive to the movement to censor Russian culture than Russian culture is.
I feel like focusing on the historical question of whether Ukraine was "part of" Russia or not is to give into Putin's framing of the question.
It's utterly irrelevant whether or not Ukraine was part of Russia a hundred or two hundred or five hundred years ago, all that matters is whether the Ukranians want to be part of Russia now.
Similarly it's irrelevant whether Taiwan was traditionally part of China, or Southern Ireland was part of the UK, or the United States part of the British Empire; it just matters what the people who are actually alive today think.
Sure. But given that premise, how do you censor Russian culture without simultaneously censoring all of Ukrainian culture? We know that one is good and the other is bad, but how do we apply that?
There's a big difference between actively suppressing a culture like what Russians are doing in occupied Ukraine, and in ceasing to actively promote a culture, which is what I've largely seen proposed for Russian culture while Russians pursue war.
What’s your feeling on Israeli literature (or perhaps Jewish literature) in the aftermath of Gaza. I suppose German literature is already out of the question. We are deep into the decolonisation of Shakespeare here in Blighty and I suppose that makes sense since though not a colonialist himself he probably brushed shoulders with a few people whose descendants might go on to colonise the falklands or somewhere.
The Spanish and Portuguese are already banned, one expects, and as to why any American literature survives at all is a mystery for the ages. Maybe because most of this cancellation malarkey is driven by the American Academy.
I am less familiar with the details of Israeli literature and culture. That said, to the best of my knowledge, there hasn't been a concerted effort across multiple iterations of an Israeli empire to wipe out someone else's culture.
Is themotte.org still up? I've been getting "this site can't be reached. DNS_PROBE_FINISHED_NXDOMAIN" for a while now
Yeah, so I discovered that it's my wifi! When I turn off my wifi and use mobile data (different network) it works! I can't access it when connected to Optus (Australia), but can on the Vodafone network. Weird.
I've had random websites block my home IP for unclear reasons (maybe my ISP is giving me IPs that have been used for abusive purposes? I don't know). Try using tor or a VPN if you can, and see if the problem persists.
nxdomain is unlikely to be an IP block by the website, and is more likely to be related to the user's ISP either having a misconfiguration, or not having picked up a recent change, or blocking the motte.
Works for me and in https://downforeveryoneorjustme.com/themotte.org
8.8.8.8 still resolves themotte.org
Working fine for me. Must be on your end.
I have a new working paper: Music and Tears: Phenomenology, Science, Philosophy
Download here: Academia.edu: https://www.academia.edu/129310650/Music_and_Tears4_WP
SSRN: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=5250527
ResearchGate: https://www.researchgate.net/publication/391653542_Music_and_Tears_Phenomenology_Science_Philosophy
Abstract: This paper explores the phenomenon of musicians being moved to tears during performance, interpreting such moments as portals into emotional, physiological, and spiritual transformation. Drawing on personal narrative, cognitive neuroscience, and Tantric philosophy, it argues that these experiences reflect a confluence of subcortical emotional release and the disciplined control of artistic expression. Within a Tantric framework, they represent karmic purification and chakra activation—moments where the performer becomes a conduit for Shakti, the dynamic pulse of the cosmos. Music, in this view, functions not merely as art or communication, but as sacred ritual capable of dissolving ego, transmuting karma, and restoring the self through sound.
Contents: Contents
Introduction: How Do We Understand Tears in Music? 3
From a Lump in the Throat to Nothing 6
Tears for Johnny 8
Acting in the Inner and Outer Worlds 10
Where Music Breaks the Shell: Karma, Tantra, and the Tears of the Musician 15
Introduction: How Do We Understand Tears in Music?
As I have said in my preface to Beethoven’s Anvil: Music in Mind and Culture (Basic Books 2001), my oldest memory is of a piece music: Burl Ives singing about a fly marrying a bumble bee. I played a record of that song over and over, driving visiting uncle to distraction when I did it at five in the morning. That little ditty is thus my anchor to the world, my omphalos.
I started trumpet lessons when I was 10. I was a reluctant student for two or three years until my music teacher, David Dysert, read me the riot act and got me to take music seriously. Since then music has always been with me, sometimes more so than other, but always. It has given me great joy and pleasure, and it has soothed my wounded heart.
It has also given me experiences that we do not talk about. Perhaps we don’t talk about them because they are difficult to talk about. But then all sensory experience is difficult to talk about. No, I suspect we don’t talk about them because we are apprehensive about where such talk might lead us.
In the next section of this document, “From a Lump in the Throat to Nothing,” I talk about two such experiences. One of them – a lump in the throat – is about being moved nearly to tears while playing my instrument. It has happened to other musicians – I know, because I’ve read stories – but we don’t talk about them. Most of this document is about that kind of experience. The next section, “Tears for Johnny,” is about a performance Bette Midler gave to and for Johnny Carson in his last week of television. Both were moved nearly to tears and I rather suspect that many in the audience that night were weeping as well.
The last two sections are about understanding how such things happen, first from a Western scientific point of view, “Acting in Inner and Outer Worlds,” and then from an Eastern philosophical point of view, “Where Music Breaks the Shell: Karma, Tantra, and the Tears of the Musician.” How do we reconcile those two ways of looking at the world? I don’t know. But I’m working on it.
I’m working on it, in part, because the other experience I offer to you – nothing – shook me to the core. It was a mystical one, one where the world dissolved in a blaze of light and sound. What are we to make of such experiences?
For we must make something of them, must we not? The psychologist William James wrote a famous book about them, The Varieties of Religious Experience, countless philosophers and theologians, both East and West, have pondered them, sometimes to dismiss them (all too common in the West), but often to understand them, to understand what the world must be if such experiences are possible.
It is easy to psychologize such experiences, to explain them as something that the brain does. It’s highly unusual to be sure, but it’s just some kind of brain event that we can’t yet explain. And yet those experiences are so very compelling. How do we honor that? Or even: Should we honor that? Perhaps we explain it as entering higher realms of being. That is, in effect, what Hinduism and Buddhism do. Is there a way to assert that and acknowledge scientific psychology at the same time? How do we treat those experiences as an invitation, an invitation to a larger experience of the world? That’s what I’m interested in.
I don’t pretend to provide an answer in this short document. My aim is simply to put the issue before you. What you make of it, that’s your responsibility.
Ancient Hindus believed that certain types of art jolted you out of your thinking. Like the discombulated patters I recently saw in the very very old Marundeeshwarar temple in Chennai in the Tripurasubdari section.
And they believed devotional music could be a trapdoor to experience oneness with the universe...i.e. a way to experience the dissolving of your own self and experiencing your being part of a larger universal self ("Brahman").
*Tripurasundari
You can edit posts. Click on the ellipsis on the upper right of one of your posts and one of the options will be Edit.
Can't. Not on mobile Substack app anyway.
If Moldbug genuinely wants “rule by CEO” then why support Trump, who has been a disaster in every real management role he’s ever held? Good CEOs in the real world are probably too consensus driven and results oriented for Moldbug’s taste.
He probably means more of the founder-style CEO. Jobs, Gates, Ellison and the like. I think those guys are far less consensus-driven.
Most founders are not great at being CEO. It’s a different skill set. Founders take a lot of risk and generally can’t delegate well. CEOs are usually paid to mitigate risk, grow the organization and guarantee stability and predictable growth to investors. It may be we already live in a society where traditional CEO skills are less valuable. Someone like Musk would have been forced out of management a long time ago if this were still the 1990s.
Strong disagree. Companies go into decline when they lose their founders, Jobs, Gates, Moore and many more
That’s a) not universally true - Microsoft is doing just fine, plenty of examples of companies that have thrived after the founder was kicked out, and obviously some founders, like Jobs, do have good CEO skills as well and b) kind of irrelevant - you may well simply be making the case that traditional CEO skills, which is what I understand Moldbug to be referring to, are not as important as founder skills. In which case Moldbug‘s argument also falls apart. The U.S. cannot be led by a „founder“ unless you really want to tear the whole country down and rebuild. You can argue Mao was more of a „founder“ than a CEO, and that’s why he needed to incite a Cultural Revolution to keep himself relevant but is that what most Americans want?
Strong agree, especially about Musk. In rare instances (like him) being willing to continue doubling down on bigger and newer risks pays off. In most cases it means a company makes a few big mistakes and dies. A competent manager is, on average, better than a maverick. For newer companies a competent manager can't get it off the ground, as they are too risk-averse to take the necessary jumps that make the company take off in the first place.
I think that in most fields this hasn't really changed since the 90s. It's just that tech has been a huge part of the new company growth and several high profile founders have made it big, so it feels like there are new options. Coca-Cola, Exxon, or even IBM aren't hiring a young CEO to "move fast and break things" because there's more to lose than to gain from even above-average performance from a CEO trying that strategy.
Which is why people try to bet against Musk succeeding. Whether he's unusually good or unusually lucky (my guess is strongly both), he's got a really good streak going by any metric.
Still, why Trump? He’s not a founder type CEO.
Agreed. Beats me. I guess Trump *acts* like he's one of those guys even though he's far less competent and maybe Moldbug can't tell the difference. Or maybe Moldbug just takes what he can get. I dunno. I'm not trying to defend either.
I'm not an expert on Trump's business by any means, but I understood he did get a number of buildings constructed and managed around the world. I suspect that's top 1% ability level at least - meaning 99/100 would fail to accomplish that, probably closer to 9,999/10,000 at least.
Of the people who manage multi-million dollar real estate companies, maybe he's a loser and bad at the job. But that's still a pretty exclusive club with a high bar to entry. I'm certain that if I was given control of his company at any stage from 1970 to present, I would have failed to run it. Even if the only thing he did was hire effective people, that's still a pretty significant skill. Maybe a rarer skill than being a good CEO alone.
Forbes says if he had put all of his inheritance in an S&P 500 index fund and done nothing else he'd be better off today than he is.
He had a couple good projects in the eighties but most of his projects since then have been licensing deals. The Trump Tower project was run by a woman named Barbara Res. She worked for him for 18 years and doesn't have a lot of good things to say about him.
https://www.businessinsider.com/barbara-res-donald-trump-real-estate-trump-tower-2020-11
Yes, Trump is a genius at self-promotion and brand management. He was a man ahead of his time for most of his life - his skill set is perfectly suited to the social media/influencer 21st century. He was mostly a failure back in the 20th when organizational and management skills were still in vogue. Seems to me „President“ is still a job that requires the latter skills but maybe I‘m old fashioned.
I did not deeply invest time into researching this, but I don't believe the trump organization was actually much involved in the building process for (any?) of the buildings. My brief internet research implies that they were mostly involved in securing financing and name licensing deals.
Also, most of them seem to be permanently closed or no longer be Trump affiliated. Not a great omen.
If you define "good CEO" by how much money the person makes, it's pretty clear Trump qualifies. He's a billionaire.
Didn’t he make most of his fortune (post-bankruptcy) by being an entertainer as opposed to a businessman?
Citation needed. I'm not claiming one way or the other, but the only people I've seen claiming this are not to be taking seriously.
Wait, really? Can you point to an enterprise he has made significant money on that isn’t primarily about branding and entertainment? His real estate portfolio in the 80s and 90s had several famous bankruptcies, which were his main claim to fame before his reality show, and everything since then that has been profitable has been based on this bending.
He had several businesses, including real-estate businesses, *some* of them filed for bankruptcy. Moreover, I don't understand how leveraging his public persona to make his profitable businesses even more successful is somehow an indictment to his business acumen, rather than the opposite.
It’s not indicting his business skill - it’s just noting that his major successes are all based on converting attention into money, and not based on creating a product or building that does something of its own, separate from the attention of the public.
A good CEO is defined by how much money he makes shareholders.
The shareholders of the Trump Organization have made money.
By how many orders of magnitude did he increase his family fortune after inheriting it?
Hard to be sure because he's taken such care to suppress his tax returns. He inherited a total of around $400 million from his father. Guesstimates of his net worth have bounced around in the low-billions range, though those took at face value stated valuations of real estate holdings which have been exposed in a couple different courtrooms as hilariously fictitious.
Trump was always very aggressive with financial leverage (debt) which is why business bankruptcies have been such a central part of his operations. (Stiffing the lenders as a strategy rather than a last-ditch option.) Some unknown amount of personally-guaranteed debt still exists on paper, and the guy who ghost-wrote "Art of the Deal" has been quoted as predicting that Trump's probate will reveal his actual all-in net worth to be negative.
The new wrinkle which may wash away all that though is the crypto scam -- early indications are that it is proving _extremely_ rewarding for Trump and whichever family members are in with him on the back end of it.
> The new wrinkle which may wash away all that though is the crypto scam
I'm extremely confused by assertions like these. People want to give Trump money and he gets the money, right? Who is scamming whom?
People believe they're buying into a Trump-endorsed investment vehicle, but the reality is more like a donation to his bank account in exchange for near-worthless scrip
> Guesstimates of his net worth have bounced around in the low-billions range, though those took at face value stated valuations of real estate holdings which have been exposed in a couple different courtrooms as hilariously fictitious.
While this is accurate, it should be noted that all other estimates of anyone's net worth have exactly the same problem.
They don’t. A lot of people have most of their net worth in stocks of publicly traded corporations, whose valuation is very public.
The valuation of publicly traded stock is public, but it's also obviously fictitious. If a major stockholder attempted to cash his stock out, its value would crash, and everyone knows this. Net worth calculations do not even pretend to take it into account. They are not attempts to calculate the value of someone's net worth; they are only attempts to produce a number that is tenuously related. Valuations of other property meet the same standard.
Though not to equal degrees. Depends on how much of a person's net worth is in real estate.
Forbes just a few weeks ago published their latest estimate of Trump's net worth. They arrived at a total of $5.7B minus $600 million to date in legal liabilities. Nearly half of the total is the highly-inflated Truth Social stock. ("From a financial standpoint, Trump’s social-media venture is one of the most absurd businesses in America, generating sales of just $3.6 million in 2024 and recording a net loss of $401 million. Revenues dropped 12 percent in a year, even while Trump dominated most news cycles. Yet Trump-loving traders still bid up shares to head-scratching prices....") If Trump were to start cashing out of that stock it would plummet in value given its underlying business fundamentals, so that's kind of a platonic example of the concept of paper wealth.
Of the remaining $3.1B of Trump net worth, Forbes puts real estate at $2.2B. They don't say where they are sourcing the valuation estimates from, so it's hard to say how much those figures depend on Trump Org's own highly-aspirational numbers.
> Depends on how much of a person's net worth is in real estate.
No, it doesn't. You follow that up by noting a problem in the estimated value of Trump's stock!
Not sure if you can answer this, but I saw a report that whoever runs $TRUMP has made multiple millions off of transaction fees.
Those fees aren't in dollars, just in $TRUMP, so they would still need a party to change the cryptos into dollars, right? And that would be trackable?
You're way outside of my personal knowledge base there.
Your cookie policy banner doesn't seem to work. At least whenever I click any of the options it says "saved" and then refreshes and the banner appears again. There doesn't seem to be any way to get rid of it.
Happened to me but whitelisting the site in uBlock fixed it
All of substack seems to be doing that, at least in my experience. Substack seems to have trouble maintaining a decent web experience...
Interestingly, the Washington Post did a piece on Curtis Yarvin last week making a rather different case on his Trump pivot, titled “Curtis Yarvin helped inspire DOGE. Now he scorns it.”
Curious what’s driving the seemingly different points of view here.
https://www.washingtonpost.com/politics/2025/05/08/curtis-yarvin-doge-musk-thiel/
Key observation of the column:
‘Yarvin’s DOGE disillusionment is somewhat surreal, almost as if Marx had lived long enough to troll the Bolsheviks for misreading “Das Kapital.”’
Which Marx would have done.
Looks like Musk didn't make proper obeisances, so he's on the shit list.
The PKK, the Kurdish terrorist organization from Turkey, has announced its dissolution: https://www.reuters.com/world/middle-east/kurdish-pkk-dissolves-after-decades-struggle-with-turkey-news-agency-close-2025-05-12/
Their leader, imprisoned since 1999, had called for the group to disband this February. I assume this is great news, not just for people in Turkey, but for those in Syria and Iraq too (there are Kurdish militias affiliated with the PKK controlling territory in those countries), but I want to know what people from the area think.
These news may also have an effect on the US role in the Middle East, given that they're allied with the YPG (Kurdish militia in Syria), and possibly the Iraqi Kurds.
I think you greatly simplified the situation. PKK is a terrorist group just as any rebel force which has armed disputes with a somewhat hostile occupying power (which Turkey, without any doubts, is). Also, the territories the Kurdish militias control in Iraq and Syria happen to have a Kurdish majority in most cases (while Turkey frequently makes alliances with islamist militias).
Great news indeed!
I don't expect this to affect Syria very much. YPG has quite different focus and goal than PKK. It's not clear that they can survive if the US should withdraw from that region, because Turkey would probably conquer the YPG territory pretty quickly if no one else intervenes. But that has little to do with their ties to PKK.
Curtis Yarvin. As a long time SSC/ACX reader, I kinda get it when Scott writes on X: "I agree ... there's still a outside chance Trump manages to come through and do great stuff. Certainly I was happy to see him take steps against affirmative action, NEPA, etc. " - But why give the "others" (NYT and armies of trolls) such a fine quote to "proof" to non ACX-readers: "See: S.A. Siskind really is a racist hating nature - we told you!" ... From a more obviously Straussian way-to-look at the whole thing: What is the aim of all that Curtis-critique - the one effect it can have is giving him more readers. (Oops: the purpose is, what it does.) I did too much history to care zilch about Curtis and his fantasy-monarchies - or his opinion about Trump. At SSC Scott did some great writing steelmanning and refuting those wet dreams on monarchists/reactionaries. Chapter closed.
I kinda feel like they all hate him at this point and he might as well be honest and fair. It's a big part of his brand.
Nothing dishonest about phrasing it: "Certainly I was happy to see him take steps towards color-blind college-admission and YIMBY." - A tweet is not a personal e-mail, but a public statement. Scott is enough of a brand to attract the haters, mind-police - well, the NYT even. And those do have influence over the collective consciousness. - Most people will never be ACX readers. - I will always be, and paid. But I see nothing worthwhile in former or present Curtis to pay attention to.
Makes sense. I am not confident enough of my media savvy to give detailed advice like that to Scott; he seems to know what he's doing, he's got ...how many subscribers now?
He is number 2 in the substack-category "science". Number one is a mainstream team of medics, very liberal with a capital D - one really wants to comment - Ay, there's the rub: to comment, you need to pay-subscribe! Free to read. https://yourlocalepidemiologist.substack.com/p/birth-rates-are-falling-but-solutions Ok, I did it: a shout-out for substack worse than ACX. All are. But much better than Curtis and his ... in German, I'd call it: Dreck.
I actually have a paid account that never comments. I seriously expect the Democrats to get back into power and start throwing people who have said politically incorrect things in jail.
He probably feels like we're way past the age where the NYT and similarly aligned people had any real power over the collective consciousness, so he feels like he can speak freely. I would agree and I think it's a great thing.
Scott, calling yourself a libtard and a coward seems like poor rhetorical innovation. Just giving the other side ammo. “Scott Alexander admits that people who oppose Trump are cowards.”
It's not giving them ammo if they're going to call him that anyways; it's leaning into being Shylock. The self-awareness is refreshing, if he means it.
Signed,
Someone who has admitted to authorial cowardice in other comments elsewhere
it's also puerile and unclear, what does "libtard" even mean, never heard it in a single context where it made a minimum of sense. At the same time the association of not wanting to risk a monarchy and/or dictatorship = being a coward that's a strange association at best
Well, actual monarchies have occasionally been decent governments. (I think the rule is 1 time out of 5. Which is perhaps better than the record for presidents.) The problem is when they aren't. It has certainly been asserted "For forms of governments let fools contest, what ere is best administered is best.", though I thing is rather depends on the goals of the administration.
The thing is, democracies, and form adjacent governments like the US, are historically less reliable at keeping deals than other forms of government. Both internally and externally. They tend to feel that they don't want to be bound by the prior administration, and the populace forgets why the deal was made in the first place, if they ever agreed it should be made.
I think a modification of the old Anglo-Saxon idea is perhaps the best. The new king must be elected from a group of people, but not a small group. Say those who are either descendants, nephews, or nieces of the current king. (Perhaps expand the group if the list of candidates is less than 7. But expand by consanguinity. If there aren't enough cousins, pull in the second-cousins.) This gives a small enough group that they can be raised responsibly, and also have their characteristics known, and a large enough group that SOMEONE in it is likely to be decent. It also gives the current ruler a stake in the succession. As to who should do the electing...that's a tough one. Most people are clearly incompetent at that. I know that given a field of 12 candidates I wouldn't have paid enough attention to them to be a reasonable decision maker. Saying "just let everyone vote" is a recipe for disaster, but every filter I've thought of is subject to corruption. "Let the governors vote on the king", for example, is a guarantee that the current power structure will be supported.
> Well, actual monarchies have occasionally been decent governments. (I think the rule is 1 time out of 5. Which is perhaps better than the record for presidents.) The problem is when they aren't.
I read somewhere that for a while Turkey's succession system was to have a civil war whenever the sultan died, with two obvious consequences:
1. Turkey was ruled by something like eleven highly competent and effective monarchs in a row, noted by the essay as a record that puts every other known state to shame.
2. Turkey spent much of its time enduring brutal civil wars, which wasn't worth it.
It isn't actually true that the problem with monarchy is when the king is bad at governing. That problem is easy to solve. We prefer not to solve it. The point of monarchy is to assign rulership regardless of what people might think of the particular ruler.
1 time out of 5 for modern monarchies seems pretty optimistic and yes as you point out the consequencces when they go badly are particularly severe.
Apart from the fact that those records were written under that king without any real possibility of criticizing the king even if that were true having one example hundreds of years ago is not remotely a proof that a monarch would be even halfway successfull today.
If we look at modern dictators or kings the ones we see are not particularly inspiring. Russia is under a dictator who brought it into the bloodiest war since ww2. Sure it's marshaling it's resources quite well for the war, except the mistake was starting it in the first place.
Since you mentioned Turkey we can look at Erdogan who brought Turkey an inflation of almost 38%. Similar stories of economic missmanagement you will find in Hungary, Belarus, most African military juntas and Venezuela and in all this the population also bears the human cost of censorship and dissent being eliminated.
Even the cases where dictators have been a lot more economically successful such as China (technically the gulf monarchies too, but their economic success is purely based on oil extraction not on wise leadership) the population has to suffer a human cost.
Among the very few cases of illuminated dictatorship i can count Singapore, an essentially tiny country, a bad example to institure a dictator in any populous country.
I mean, they're worse at keeping deals, but are they better at other things? Why would you go for a form of government that hasn't been used in a millennium and existed in a much smaller polity with a lower tech level when some form of parliamentary democracy (unlike the US) seems to work OK in, say, Europe or the non-PRC parts of East Asia?
https://www.youtube.com/watch?v=aNOu6tfmOOA&ab_channel=ESOTERICA
In Charles Williams' _Descent into Hell_ (1937), a professor creates a succubus. No overt magical methods, just obsession about a young woman who isn't interested in him. He doesn't like that she's not interested. As I recall, once this is clear to him, he avoids her, and invents a false version with her distaste for him edited out.
He spends more and more time with the false Amelia until she can even be seen by someone else. His lack of interest in truth leads to his mind disintegrating.
I was surprised to find in this discussion of Paracelsus, a major Renaissance writer about magic who put much emphasis on the power of imagination (at about 25:00), a description of making a succubus by imagination, and I'm willing to bet that Williams, who had a considerable interest in magic, had picked the idea up from there.
I thought I had just found a really cool reference, but this does rather look like concerns about AI companions.
I'll never be afraid of being unable to identify AI agents until they can act as though I don't exist like human women can! Checkmate, bots.
https://www.youtube.com/watch?v=ZXNIgHov0Nk&ab_channel=BenSyversen
The rather hectic story of a manuscript copying Archimedes' letter about his "method", a socially unacceptable way of using infinitesimals to calculate areas.
The ancient Greeks didn't like them, the counter-Reformation Church didn't like them. (Let me know if that's true.) Fortunately, Newton didn't have to please the Jesuits. I feel like there's a whole conversation about gatekeeping and Damned Things* in the topic.
The text barely survived. There's one known copy, and it was bleached out for a prayer, but some of it was barely visible in the margins. A scholar copied what he could see-- recognizably lost Archimedes-- but a lot of it wasn't visible, and then the manuscript was lost and getting moldy, what with being hidden from the holocaust.
Fortunately, it was found, and modern scanning was able to recover the text. Watch the video for details of the method and animated diagrams.
The most interesting thing on the abstract level is the possibility that there is revulsion against actually good ideas, and probably not just ideas some current rebels are defending. Is there any way to get the actually good ideas to surface when there are so many bad ideas competing with them?
*Damned Things-- Robert Anton Wilson's term for things people seriously don't want to think about
"Damned Things-- Robert Anton Wilson's term for things people seriously don't want to think about"
I thought it was Charles Fort? Though maybe Wilson was quoting him:
https://en.wikipedia.org/wiki/The_Book_of_the_Damned
"The Book of the Damned was the first published nonfiction work by American author Charles Fort (first edition 1919). Concerning various types of anomalous phenomena including UFOs, strange falls of both organic and inorganic materials from the sky, odd weather patterns, the possible existence of creatures generally believed to be mythological, disappearances of people, and many other phenomena, the book is considered to be the first of the specific topic of anomalistics.
The title of the book referred to what he termed the "damned" data – data that had been damned, or excluded, by modern science because of their not conforming to accepted belief. Fort charged that mainstream scientists are conformists who believe in what is accepted and popular, and never really search for truth that may be contrary to what they believe. He also compared the close-mindedness of many scientists to that of religious fundamentalists, implying that the supposed "battle" between science and religion is just a distraction for the fact that, science, in his opinion is in essence simply a de facto religion. This is a theme that Fort developed more in his later works, New Lands and Lo!, particularly."
I've been told Wilson was referencing Fort.
It’s years and years since I had to study history of mathematics as part of math undergrad, but…
- you have the exhaustion lemma in Euclid’s elements
- techniques used by Archimedes, considered even at the time not to be respectable mathematics
- Newton and Lebnitz get a lot of flak from, e.g. Bishop Berkeley, for not being adequately rigourous.
- Eventually (nineteenth century or so) we hav an account of calculus that looks respectable
Now, when ‘m teaching ungraduate digital signal processing, we do point out that if the student is paying attention they ought to notice that what we’re doing is not mathematically respectable, and if you care you can go read a book on measure theory.
And then, you go pick up a quantum field theory book, turn to the section on zeta summation, and wince.
I maybe misremembering my undergrad history l course here, but possibly there is a step circa Newton and Leibnitz, where people know Archimedes had a method but don’t know what it was, try to rediscover it, and end up inventig something different.
Would appear analogous to the bit where the US circulated a fake story that lasers were used in isotype refinement for nuclear weapons, the soviets believed this nonsense story, and actually got it to work.
You say Archimedes' method was socially unacceptable - I don't think that's quite accurate. The mathematicians of his time had a geometric & Platonic view of mathematics. There wasn't a proper understanding of functions, or of limits in mathematics. The 'paradox' that an arrow could never strike a rabbit because in the time the arrow moves 4 feet the rabbit moves 1, and so they never meet was regarded seriously.
Aristotle's fairly sophisticated methods for working out areas (essentially by integration) didn't have the appropriate foundations, and mathematicians of his time thus couldn't either understand these ideas or work with them.
Newton & Leibnitz developed better descriptions of calculus (though Newton, as I understand it, spent several years developing 'geometric' justifications for the rules of calculus, which were probably not at all related to how he discovered the rules originally). These took mathematicians some substantial time to appreciate, and in fact analysis was placed on proper foundations only in the second half of the nineteenth century.
It's worth pointing out that the Church of England was hostile to Newton, for mathematical reasons as well as non-mathematical (his views were likely heretical). Traditionalists everywhere dislike new ideas, I guess.
Okay, interesting link but I had to stop because I was rolling my eyes so hard when we got to THE JESUIT WHO ROSE TO POWER and TOLD THE CHURCH IT COULD USE MATHEMATICS TO ACHIEVE ITS GOALS.
Said goals being of course "to re-establish its social, cultural and political dominance". Because the one thing every peasant in the middle of Europe was fighting about was which particular mathematical theory one supported, naturally.
If Clavius - who has a crater on the Moon named after him - 'rose to power' via mathematics, it was on the grounds of calendar reform. I guess the Protestants were right and the Gregorian Calendar should be junked because it's a sinister Papist plot to - what was it again? oh yeah - "re-establish social, cultural and political dominance".
https://en.wikipedia.org/wiki/Christopher_Clavius
Though I have to admit, the idea of a Renaissance figure pondering "how do I achieve power, status, wealth and success? I know, I'll become a mathematician!" does tickle my fancy.
I'd have to read this guy's book to figure out if what he is describing is so, and how the theory of infinitesimals was received outside of Italy in Protestant nations (was there any opposition from Protestant mathematicians? if so, that puts a hole in his proposition that it was purely Catholic power-grabbing) and to be frank, nothing will make me read a book about mathematics, so some braver soul must do this for me:
Infinitesimal: How a Dangerous Mathematical Theory Shaped the Modern World by Amir Alexander
https://www.goodreads.com/book/show/18593597-infinitesimal
So far, it's sounding like good old Whig history and the pernicious Papists, but since I'm only going by a video produced by some whippersnapper, that may be unfair.
Opposition to infinitesimals wasn't just a Catholic thing. From the book:
"Yet, useful as it was, and successful as it was, the concept of the infinitely small was challenged at every turn. The Jesuits opposed it; Hobbes and his admirers opposed it; Anglican churchmen opposed it, as did many others."
"The struggle over the infinitely small in the early modern world took different forms in different places, but nowhere was it waged with more determination, or with higher stakes, than in the two poles of Western Europe: Italy in the south and England in the north."
I’m inclined to agree that infinites and infinitesimals are basically mathematical concepts that don’t really exist in reality.
I think what Andrews is trying to do (and this is only an impression from the reviews) is set up the sexy opposition model of "side A versus side B, A wins because goodness, goodness = our modern stuff" and thus wants to have Italy - central! Catholic! - versus Britain - isolated! Protestant! as the boxers in their relevant corners.
But that leaves out Leibniz, for instance, who was working away in Germany. Was that not also an isolate Protestant periphery? But that doesn't suit the narrative as well, though Hobbes versus Wallis is not the first thing that comes to mind if I'm thinking of Big Maths Fights from the past.
I think the guy was trying to write the mathematical version of The Swerve:
https://en.wikipedia.org/wiki/The_Swerve
Big Concept books that pitched Science versus Religion in a sweeping historical context where the Classical past by way of the Renaissance came to save us all from the Dark Ages were very big in the mid teens of this century; Catherine Nixey and "The Darkening Age" was another such:
https://historyforatheists.com/2017/11/review-catherine-nixey-the-darkening-age/
So I think Andrews (or his publisher) went for a similar angle though with maths this time.
The Catholic Church has an interest in philosophy and theology, and I can imagine it wanting to spend a small fraction of its resources on what it thought was pinning down the corners, especially if one Jesuit thought it was important and no one else had a strong opinion.
The Church has been both fairly maligned and unfairly maligned, so what's needed is actual history.
I nearly made a fool of myself. I've only ever learned calculus through limits, so I just assumed Newton and Liebnitz used limits. Apparently that was a later formulation.
Newton and Leibniz used limits - they just thought of limits as defined by infinitesimals (inputs infinitely close to x have outputs infinitely close to x’s output) rather than by triply nested quantifiers (for every x and every epsilon there exists a delta such that inputs within delta of x have outputs within epsilon of x’s output).
"the counter-Reformation Church didn't like them. (Let me know if that's true.)" Yep, it's true. A few years ago I was going to review this book for the contest, but life and laziness got in the way. I've forgotten the details by now but yeah, there was plenty of religious debate over infinitesimals. https://www.amazon.com/Infinitesimal-Dangerous-Mathematical-Theory-Shaped/dp/0374176817/ref=tmm_hrd_swatch_0
There are roughly two poles with regard to mathematical rigor. One camp I might call pure math, and it generally cares very much about solid foundations. (Not always enough, mind you, naive set theory was a thing, after all, and Goedel has put a hard limit on how good your foundations can be.)
The other pole I might roughly call "physicists". There, anything goes, if you can write it on a blackboard and don't get struck by lightening, that means you are allowed to do it. Using mechanical (a la Archimedes) or numerical methods is fine. Playing fast and loose with infinitesimals is fine, e.g. cancelling dx in (df/dx)*dx. If a student comes up with a counterexample where these tricks don't work, you just call it a pathological example and move on, irrespective of the fact that almost all functions might qualify as pathological. What do you mean, a plane wave in an unbounded space can not be normalized? In practice, you can take the integral of a lot of sets and never once encounter a Banach Tarski decomposition.
The difference between these poles is exemplified by Dirac's delta "function". It can be defined by a physicist in ten minutes. However, if you want a sound definition which tells you what you can actually do with it, you require distribution theory, which takes a semester or so.
Both approaches have their uses, however it is important to be aware if you are cutting corners or not.
Also, early mathematicians were a bit of a cult, see for example WP on irrational numbers:
> Hippasus, however, was not lauded for his efforts: according to one legend, he made his discovery while out at sea, and was subsequently thrown overboard by his fellow Pythagoreans 'for having produced an element in the universe which denied the... doctrine that all phenomena in the universe can be reduced to whole numbers and their ratios.'[7] Another legend states that Hippasus was merely exiled for this revelation. Whatever the consequence to Hippasus himself, his discovery posed a very serious problem to Pythagorean mathematics, since it shattered the assumption that numbers and geometry were inseparable; a foundation of their theory.
> [Dirac's delta function] can be defined by a physicist in ten minutes.
Only if you want the formality of setting it up as a limit of Gaussians. If you'll accept the piecewise function (δ(0) = +∞ and δ(x) = 0 if x ≠ 0), the integral, and a plot thrown for good measure, it shouldn't take more than one.
Differentiating this function is when you might need to watch out for lightning.
> the integral
To expand on this, there was no need to gloss over "the integral". The integral is 1 when 0 is within the bounds of integration and 0 otherwise.
Personally, I would take the ten minutes to go through the limit of gaussians. It will still leave students uneasy (can you treat something which implies "when evaluating this, take the limit for epsilon-->0 over the expression it is in" as a function in good conscience?), as it should, but at least they will kind of know where you are coming from.
The piece-wise "definition" is a function which is zero outside of a null set (which happens to be {0}), so its Lebesgue integral should be zero. And even if you do not have students who have heard of measure theory, someone will notice that δ(x) = 42 δ(x) in the piece-wise definition (unless you want to argue that 42 ∞ !=∞, which is definitely not a hill I would want to die on).
So the "and the integral shall be 1" is very much part of the definition.
Now I wonder if it would be worthwhile to use infinitesimals (1/(2eps) for |x|<eps, 0 elsewhere -- or gaussian if you like) for a more robust (but likely still unsound) definition. Not that infinitesimals are typically rigorously defined in lectures where delta will pop up.
> Now I wonder if it would be worthwhile to use infinitesimals (1/(2eps) for |x|<eps, 0 elsewhere -- or gaussian if you like) for a more robust (but likely still unsound) definition.
Why would that be "likely still unsound"? Any proof involving limits can be translated _mechanically_ into one involving infinitesimals and vice versa.
From a perspective based on infinitesimals, the delta function is nonzero over an infinitesimal halo around the real value 0, and its value over that region is one of the infinitely large values such that the real part of the area under it is 1. This is related to the fact that the limit of 0 (the real part of the width of the halo around 0) times infinity (the value of the function) is indeterminate. It is possible to choose the two values, the one with 0 real part and the one with infinite magnitude, so that their product is 1.
My mistake. I was under the impression that the theory of distributions was younger than hyperreal numbers / non-standard analysis (which offer a rigorous definition of infinitesimals), and hence concluded that if a rigorous definition of the Dirac function was trivial using infinitesimals, Schwartz would have used that instead of inventing distributions.
However, the actual timeline is:
1930: Dirac introduces delta (with some prior art)
1951: Schwartz published his book on distributions
1966: Robinson published Nonstandard Analysis
From my PoV, one of the following three must be the case:
(1) Distribution theory is contained in NSA (and Schwartz was just too early to use them)
(2) A rigorous definition of Dirac delta is trivial with NSA, but other results from distribution theory elude NSA (and Schwartz was interested in these results, and not providing a rigorous definition of physicist's syntax)
(3) NSA can not trivially provide a sound definition of delta (e.g. feeding infinitesimals to the delta-function might be problematic)
Based on what I know about these topics (which is sadly not very much), I will happily plead the fifth on which one is the case. If I had to wager a guess, I would pick (3), but don't quote me on it.
Is the Hippasus story true? In a lovely example of serendipity, I never heard of it or him till this comment, and when I went looking for something about Clavius and infinitesimals, there he pops up again:
https://pillars.taylor.edu/cgi/viewcontent.cgi?article=1008&context=acms-2017
"When there are good true stories to tell, everyone suffers when we tell the false ones. This is especially important when the stories directly impact the matters students are studying. If one tells the story of a man named Hippasus being tossed from a ship with regard to incommensurables, better to talk about what it says about worldviews that such a story seemed credible when first written down centuries later."
Anyway, the paper about infinitesimals and why they were so controversial was more educational for me. First off, unlike the framing in the video, it wasn't a tidy "religious bigots on this side, Enlightenment scientists on that side". Cavalieri, Galileo's pupil and defender of infinitesimals, was a member of a religious order himself. And the ban on teaching infinitesimals only held in *Jesuit* schools, since Clavius was creating the mathematics curriculum for these schools. If you weren't attending a Jesuit institution, you could (or could not, depending) learn about these.
"A typical analogy, most often associated with the Italian Jesuat brother [As always has to be explained regarding Cavalieri, he was a member of the Jesuat/Gesuat, not Jesuit, order. It flourished from 1361 until suppressed by papal decree in 1668] and mathematician Bonaventura Cavalieri, is of infinitely thin pages of a book.
1. The pages could be so thin that, no matter how (finitely) many you stacked, all of them together would not be as thick as any actual book; you would need infinitely many. Such pages are infinitesimal.
2. The pages could be so thin that, no matter how you tried, you could not slice them any thinner; they are indivisible. They are parts of the book, but if infinite in number, whether they comprise the whole book was open to question."
So, as ever, the story is slightly more complex than the easily-digestible Youtube video version with (bad old) Church on this side, (good new) mathematicians on the other side.
"Now, there is plenty of truth in this story. First, there is no doubt as to the remarkable success of the Jesuit order in (among other things) establishing rigorous, desirable, ‘safe’ schools for the minor nobility and nascent bourgeoisie – all in the name of the Catholic/Counter-Reformation. Their most prominent mathematician, Christopher Clavius, was a solid proponent of Euclidean geometry in these schools; under his aegis the church (and hence much of Europe) achieved the long-sought goal of calendrical
reform in 1582, where October 15th followed October 4th to make up for the solar year not being evenly divisible by standard days.
...But the overall story, alluring as it may be in our society of absolute freedom and intellectual inquiry, and as exciting as it might be to think of calculus (!) as having political importance so early on, is selling a bill of goods. The suppression of
indivisibles was surely part of a long fight over new ideas and to what extent the Church could direct or restrict many activities, but not the only one, and much of the hyperbolic verbiage used was par for the course at that time.
Despite Alexander’s often deft handling of the distance between (post) modern secular readers and the worldview landscape fifty years on either side of Galileo, much of the political discussion is speculation, or simply wrong. As just one example, while in such a political age it was possible that the Jesuits got the Pope to suppress the Jesuat order, and (barely) conceivable they would have been motivated because some of the most prominent advocates of indivisibles (notably Cavalieri) were from it, among his copious endnotes there is not one to be found about this topic.
We should teach about incommensurables, and that the Pythagoreans may have had strong feelings about this; but we should also make it clear that the death of Hippasus is probably just a story. Likewise, the example of the triangles is a good warning against setting up integrals without caution, and the very real fights waged over infinitesimals/indivisibles for some of these reasons (Galileo was an early advocate)
is a wonderful topic in a Calculus II course. But let’s not suggest it is really about preparing the West for a modern secularist worldview (as even his subtitle How a Dangerous Mathematical Theory Shaped the Modern World implies) that Alexander acknowledges no one in question was actually fighting for."
As for the second half of the book and the Hobbes-Wallis hairpulling:
"Now, Hobbes reasoned if he could achieve the long-desired (Euclidean) construction of a square with the same area as a circle using his techniques, then surely his entire philosophy would be accepted – including otherwise-distasteful-to-all outcomes such as the Leviathan totalitarian-yet-not-monarchist regime. (Here is where Alexander draws a direct comparison to the top-down regime promoted by the Jesuits, adherents of papal authority as their raison d’etre.)
...For the man primarily responsible for Hobbes’ downfall was another Oxford don, erstwhile Presbyterian-party preacher and Parliamentarian-party cryptoanalyst John Wallis.
Through all the regime change of the English Civil War era, Wallis retained favor by adroit maneuvering; unlike many similar men, he seems to have been addicted to producing (and publishing) correspondence aimed at defeating anyone he disagreed with about anything. Since Hobbes was just as stubborn, and since Wallis disliked Hobbes’ theology, views on the university, and (to him) inadequate mathematics,
their dispute lasted through over twenty years’ worth of letters, publication, and ‘transactions’. They argued primarily over mathematics, but interspersed accusations of plagiarism, bad Latin, and proper authority of a minister of the Gospel.
...Still, this was not somehow an English version of the Italian controversy. In fact, Wallis was even more cavalier about his use of infinitesimals than Hobbes was about his indivisibles (see the next section for an example). In Alexander’s telling, Wallis, not Hobbes, plays the role of Cavalieri’s school, with repeated direct references to the
Baconian ideals of experimental induction and free inquiry espoused by the Royal Society of London. All Wallis wanted, in the best experimentalist tradition, were “theorems that were sufficiently ‘true’ for the business at hand.”
Hence the Society’s (with Wallis) repudiation of Hobbes (and so his philosophy) was an example of the key to English (and, by extension, later American?) pre-eminence in science being tolerance of differing opinion, where “a land of many voices. . . discover[ed] its path to wealth and power”. This seems to be an awful lot to lay at the feet of the infinitesimals, or even the Royal Society, alone.
This is especially so since the technicalities of these questions looked so different in the dawning eighteenth century, where people abandoned Cavalieri or Wallis as dead ends to grapple with utilizing Leibniz’ tools (and Newton’s in England), with different controversies. Similarly, it is very hard to imagine the political development of Italy or England going very differently with or without infinitesimals. Even if the Royal Society’s openness to work with the sort of deficiencies Wallis’ had might perhaps be a token of a more general openness in English society to once-heretical ideas, it is not a main reason we should care about this dispute."
> Is the Hippasus story true?
There is no reason to believe that it is. Note that "Pythagoras" is known to history only as the name behind a religious cult, making him parallel to Jesus, Mohammed, etc.
It is true that infinitesimal methods were disliked by Archimedes, probably for reasons of perceived rigor. He used a "method of exhaustion" to derive his results, but once he had a result, he would prove it by other means that were more acceptable to him.
So I recently found out how extremely difficult it is to get reasonably priced ADHD meds in the US and then I got linked to this pretty moving video: https://www.youtube.com/watch?v=eRPeU1DYOWA
Now im not an American, this could all just be entirely wrong for all I know but I remember hearing American private pharmacies were all on the the verge of bankruptcy a year ago - is this their response ? Is there something else going on with the actual manufacturers ?
This definitely feels like the sort of thing our resident Adderall Merchant and True Caliph should have done a deep dive on.
A useful workaround I know of for people having trouble getting the extended release version: I've known people who just get the regular version and take a fraction of a pill every couple hours. It's not much trouble, it's cheaper, and I wouldn't be surprised to learn that it actually is a better way to keep a steady lowish level of the stuff in your system. Whatever system an extended release version of a drug uses to keep up a slow release inside the person's body can't possibly work equally well for everyone.
Now I am wondering if someone could build empty diffusion-based extended release capsules for DIY pharmaceutical upgrading.
I don't think that is easy, though. The release rate depends on the solubility of the active ingredient and how well it diffuses through the membrane. And some drugs might rely on stomach acid, which will stop affecting the drug once the XR capsule moves to the intestines. (Also, one would require a two part capsule which seals on assembly, and any commercial manufacturer would likely get sued by the pharma industry.)
Re the contest: if you start organising things only tomorrow, does that mean that changes to the docs after submission but before today are automatically taken into account?
In his response to Moldbug, Scott says
“ But you're the one who likes talking about curate's eggs”
That’s the egg that has good and bad in it which is I suppose how Moldbug uses it. I used to listen to some creaky intellectual Sunday morning review on BBC 4 where every movie or book considered as mediocre was compared to the curates egg.
Well they were wrong! The curates egg is a totally bad egg.
A lot of people think there was no humour in the 19C and when you realise that the “curates egg” has survived for more than a century (however misremembered) as their best joke, then we realise that’s probably true.
The joke is in comic book form, if I recall, the curate is visiting the bishop who has offered him an egg. The bishop asks if the egg is good, the curate replies it’s good in parts.
That means the poor curate is eating an egg that smells of sulphur and he has to eat around the bad parts, which would be greenish tinted if I recall. ( we don’t really have bad eggs any more as the hens are better educated).
Anyway that’s not a good egg, or even a mediocre eggs, it’s a rotten egg. That’s the joke!
You now have more reason to dislike Moldbug and fusty BBC 4 review shows, if you haven’t already. You’re welcome.
https://archive.org/details/in.ernet.dli.2015.207098/page/n230/mode/1up
Moldbug also likes to say if your drink has any sewage in it, it's all sewage.
The fecal transplant was pioneered by mixing material into a chocolate milkshake; I tend to think that everyone involved would take issue with Moldbug's characterization.
Part of the joke is that the curate (pretty much the most humble clerical rank in the Anglican church) is having breakfast with a bishop, and when the latter asks him how his egg is his ridiculous reply that it is good in places is a desperate attempt not to offend his exalted host.
Moldbug uses the term in the original sense, not in the radio 4 sense. See for example (and why Scott referenced it): https://graymirror.substack.com/p/scott-alexander-the-disappointed
Yes. He uses it correctly.
A quote from that link:
> As a boomer, I am not and nor will I be on any “Twitter,”
And now their argument is taking place on Twitter.
The point of the joke is that an egg can’t be good in parts.
The curate is using a common formula to mitigate criticism that doesn’t make sense applied to an egg.
The modern usage isn’t that.
The modern usage is, in my experience, to say with wry irony that something has good and bad parts (while implying that the bad parts entirely ruin the good part).
That's the same as the non-modern usage.
I'm prepared to believe that someone out there is misusing the phrase, I just haven't seen any examples provided.
In my experience, there is no modern usage of the phrase "curate's egg". That would be my first answer if anyone asked.
I see that Merriam-Webster does list the phrase ("something with both good and bad qualities"), while defining it as British.
Hmm. My experience is that the modern usage is always “good in parts” unironically, and never “rotten entirely” which ignores the original context. Otherwise why not say bad.
Anyway language usage moves on as chatGPT reminded me when I complained about this.
Edit:
It’s possible that i was misreading these BBC 4 types and they were totally condemning whatever they were reviewing as rotten while pretending to find some merit in it.
This has shattered my worldview.
Apparently, I haven’t much noticed this phrase being used. But this modern usage is illogical and also pretty pointless (what’s the purpose of making a reference to something like this if one just wants to unironically say something is partly bad and partly good?).
Vibe coding is so addictive, it should be called vicoding.
(Cue lots of folks piling in saying emacscoding is better ;) )
A much cleaner high. ;)
Has anyone personally noticed end-user effects from the tariffs yet? I was hearing that by early to mid May the economy would in massive turmoil, and I've heard stories about ports being empty, but so far no complaints about products being unavailable or overpriced. Should I be surprised?
There are definitely business impacts. A bunch of small board game companies have wound up operations in various ways - apparently it’s uneconomic to print custom cards for games or make plastic meeples in the US.
My own employer buys components from a Chinese manufacturer that is basically the only possible supplier. We have stocks, but eventually it will be a problem. Also half our bespoke manufacturers capacity is in Mexico. So our margins will be impacted. But we can still make stuff.
Stocks of the board games that have gone out of print are starting to run out. The impact of my own employers cost issues won’t be apparent to the public for a year or so. Supply chains are long and very complex
If I try to buy a suit from Spier and Mackay (Canadian store), a $270.30 tariff charge is added to the order when choosing the US as the shipping destination. I wouldn't call that massive turmoil, of course.
Most things (that I've been buying) that were initially getting the China tariff markups were simply rerouted and taking an additional week or 2 to deliver.
Hannafords has avocados at $1.50 now. It's possible I'm misremembering, but I think the price has been from $1 to $1.19 for as long as I've gone to the store - I don't think I've ever seen it that high.
I'm in some SF groupchats, and the people doing electronic engineering or computer hardware stuff have been fairly regularly posting about the overnight shortage of parts, some parts skyrocketing, orders being canceled by suppliers or changed in a 'pray I do not alter the deal further' way, for both the hobbyists and the hard-tech startups. People are definitely postponing projects or not starting them at all, in the usual Unseen way. (Just another example of how life is much easier in the software world than hardware world - you don't have to worry about your next `git pull` from Github suddenly costing you $10k extra because Trump 'truthed' something last night.)
And now the God King of Skulls has folded and slunked away into his corner with the tail between his legs as is his custom when met with a push back. Tariff pause with China for another 90 days! A "reset" (how on Earth did he use Obama's term for it?!) with China!
The stock market is lapping it up.
My strategy has been to buy whenever he announces something outrageous, under the assumption that he will walk it back to some degree and stocks will rebound
I closed my eyes and increased my stock allocations a bit, knowing that when I was terrified in the past it turned out to be a good time to invest. I wish I did more.
> how on Earth did he use Obama's term for it?!
It doesn't beat the time he introduced the Space Force as a "separate but equal" branch of the US armed forces.
Scientific equipment has had lots of tariff-related price increases. This is really annoying for my company.
My Italian Wine Club offered to release me from my club obligations because they can't guarantee that I won't be hit with tariffs when my shipments arrive in the US. But they admit they don't know how they'll be enforced at our border. I decided to forego my club shipment for the immediate future because I couldn't find any clear guidance from the CBP on the current tariffs on alcohol from the EU, if any, and how they'd be collected. 10% I think, but those may only be for major importers, and not people who have alcohol shipped for personal use.
The job market in my field finally picked up, due as far as I can tell to foreign product equivalents being a bit more uncertain long-term, and I was finally able to get a good position. Can't say I've noticed anything on the consumer end.
My favorite YouTuber's merch has gone up in price.
I’m still waiting for my $1 Temu blender to arrive.
The board game space is getting hit hard. I know of at least three companies that have already been shuttered or significantly reduced, and a few others on the larger side are dramatically changing the way they do things to not have non-American customers bear the costs of tariffs while charging Americans a surcharge.
I've heard (unofficially) on social media and message groups and such that many other smaller companies are going to just try to sell current stock and hope the situation changes. I'd be shocked if anyone's actually ordering new Chinese game manufacturing right now in the hobbyist space, and I'd guess 95%+ of hobbyist games are manufactured in China.
I'm not aware of what's happening with the actually big companies (Hasbro, Mattel, etc).
I suspect the story is the same for other industries that rely on heavily tariffed sources: they're big enough to start doing something of a pivot or they're selling out current stock and crossing their fingers. Or they were able to lobby for an exception, I suppose.
Another thing to consider is that while the de minimus exception has been removed (last I checked), I don't think anyone has any clue how it's supposed to be enforced, so many places might be getting away with direct to consumer shipments at the moment.
Yes, GMT games sent out a moving message to subscribers about how their operations are going through major changes and risk company finances going deep, deep red.
Obviously, anyone working in finance has had a hard month. Thus my spouse is more stressed.
Our resident substation engineer doesn't know yet how to get new transformers of the right size within budget. But we'll see...
Yep, I've got a kickstarter board game that's on indefinite hold from shipping to the US. Rest of the Anglosphere should have gotten theirs a week or two ago, I think?
More broadly, we're likely in a recession right now so that's pretty bad. I believe I also read a few weeks ago that there are indications that people are burning through their savings, the effects of which we'll see play out in the coming months.
More narrowly, I remembered that a few weeks ago we bought a trampoline. A couple of days after we purchased it the item nearly doubled in price.
The lab where I work is already seeing "tariff" line items on orders. I work with electronics (repairing old equipment and designing/building new tools) and I've seen prices go up. Mouser Electronics is one of the main distributors and they've got notices on their website to the effect that they're trying to minimize the impacts of tariffs on prices.
A part I ordered in March for $X got "lost in shipment" from China. I had to reorder the part. The price was $2X.
At least one board game company has closed. And another company for a game I preordered had to 1) raise prices and 2) collect payment in advance (which they normally don’t do)
Of all the things I had no idea were now all "made in China", board games never occurred to me.
We really have put all our eggs in one basket, haven't we?
They're not all made in China. In fact, this is the first I've heard of that level of manufacturing in China - based on the Kickstarters I've watched and occasionally funded, a lot of the printing happens in the Netherlands or Belgium.
I'm now curious what the relative market shares are, broken down by board game material type (card printing, box making, misc. cardboard art and punching, plastic molding, woodcutting, metal casting), but as beleester said, it's not exactly a universal core industry.
To be fair, custom-printed bits of cardboard and plastic are not exactly what you call "core manufacturing capabilities."
Like, if you say "America needs a robust auto industry," you'll get applause. If you say "America needs a robust capability to cut out giant sheets of printed cardboard," not so much.
What is striking (and to an extent, frightening) me is how much we've outsourced to China about things we don't even think about or consider.
"Cutting out sheets of cardboard" is something you'd imagine a medium-sized factory in the Midwest could trundle along doing, neither making "line go up" profits nor in danger of closing down in the morning, but no. Everything comes from someplace else. Just exactly how vulnerable are any one of our domestic markets if something like Covid comes along again, not to mention tariffs/trade wars/hot wars?
Which company is the latter? I try to keep up with board game news.
As someone who was looking at the car market specifically during this:
25% Canada/Mexico tariffs went into effect March 4th, then March 5th there was a one month exemption for automakers, then March 6th further delays on general Canada/Mexico tariffs.
April 3rd the 25% auto tariffs actually started, and then April 5th the overall import 10% tariff started.
April 29th there is a change to give 3.75% of MSRP of an auto as a tariff rebate.
Companies have made plans and taken action to reduce imports, and many companies said new prices will stay the same through May or June (see https://www.caranddriver.com/news/a64375899/automakers-trump-tariff-response/).
In early April many people said that used car prices would rise, which happened (see https://www.carscoops.com/2025/05/used-car-prices-jumped-2-7-in-april-as-demand-surges/), while they predicted new car prices would be flat through June, and then we'd see where tariffs are at (eg https://www.usatoday.com/story/money/cars/2025/04/02/trump-car-prices-tariffs/82744006007/). Some auto dealers tried to capitalize on tariff panic to drive sales (https://www.koreaherald.com/article/10479836), but people making predictions targeted June as the earliest you'd see new car price increases (March 7th prediction for a summer rise in prices https://www.cbsnews.com/news/auto-tariffs-should-you-buy-a-car-now-2025-prices/).
Well, we're in May and there are some announced price increases for July: https://www.caranddriver.com/news/a64704402/ford-bronco-sport-maverick-mustang-mach-e-price-hike-tariffs/ This is despite what looks like weakening demand for new cars (https://www.cnn.com/2025/05/12/business/car-prices-tariffs-recession-economy).
That's all to say: supply chains are long, we're not yet at the predicted price increase date of June, and the only increase I've seen confirmed so far is Ford in July. Whether others follow or not will depend on if tariffs stay in place or get more exemptions/changes. TBD, stay tuned.
For the Germans here or anyone interested in salaries in Berlin, I recently published the Salary Trends report 2025, a comprehensive overview of tech salaries in town:
https://handpickedberlin.com/salaries/2025-03/report/
I would appreciate feedback, and if you shared it with anyone who'd be interested in this type of data.
Thanks!
I don't have the time to read it right now, just skimmed the intro. The two things that stood out to me is the that it's around 1.8k respondents and the gender pay cap numbers. At 20%+ and 15% they're a lot higher than the German average at around 16% and 6%[1]
I can think of a couple of explanations, but I worry it's selection bias in who answers your survey. If so then the data might just not be representative
Could also have different reasons and it's still interesting and I'll read it. But it'll probably not change my mind much, whatever the results, due to uncertainty about the issues above
[1] https://www.tagesschau.de/wirtschaft/gender-pay-gap-lohnluecke-100.html [in German]
Edit: Now that I've had the time to read a bit further into the article I see you've addressed those points in the report :)
Edit 2: I'm actually really impressed by how clear the report is at every level on what's measured exactly, what the sample sizes are for every group and so on. Great read :)
There is a selection bias, it’s a tech survey.
I mean, sure, but is it representative of the Berlin tech sector? That is, should we expect the gender pay gap to be much wider in Berlin tech than in society at large. Or should we expect that OP's getting a skewed sample from their respondents.
Thanks for your comment. As Peter mentioned, yes, def a selection bias. Plus, it's anonymous on top. I cannot be sure it's representative, but I did a bit of checking on particular job categories, and the gap is narrower in some cases.
The data def isn't perfect, but it's still better than Glassdoor or Kununu.
I think you would expect the tech sector to have greater gender discrepancies than average. Higher paid roles for engineers
My sense is that:
1. Scott is right that Yarvin is now supporting governance solutions which he used to oppose.
2. Yarvin's main intellectual contribution was diagnosing the problem of the cathedral, and his proposed solutions were always kinda silly (more here: https://x.com/RichardMCNgo/status/1880988102371754156).
This leaves me a bit confused about how to update. In some sense he's selling out, but he's selling out from a position that he never should have had. So it's a bit tricky to tell how much it reflects a sensible retreat from untenable views, versus him becoming less principled.
One question that would help me thinking about this: lately he has been using FDR, Lincoln, and Washington as examples of the type of "CEO" that Trump should be. But these are non-central examples of dictators. Should I interpret this as him backing away from his position of supporting dictatorship?
"Yarvin's main intellectual contribution was diagnosing the problem of the cathedral"
You... seriously write this? As a serious thing?
These "diagnoses" are worthless, they are not even dime a dozen, every ranting boomer on facebook has "diagnosed" the cathedral/whatever with a great deal of detail and memes, who cares.
I disagree with this. Of all the references Scott and others have made to Moldbug over the years, the Cathedral is the only concept that I found insightful. Everything else just seemed like park bench ramblings.
Like, literally, the only time I can imagine ever referencing him is to reference that concept. He's far from the first to identify a nebulous set of elite institutions that shape public orthodoxy and constantly reinforce each other to make contrary ideas almost unthinkable, but he gave it a catchy name and a somewhat-clear description (I think; I'm going by quotes because I can't stand tryingv to read a single sentence of his).
I find the general prejudice that you're not allowed to say a person had one good idea and everything else was crap extremely annoying.
What's the difference between Yarvin's Cathedral and Eric Raymond's?
eric's metaphor emphasizes the distinction between centralization (cathedral) vs decentralization (bazaar).
moldbug's metaphor emphasizes the role of the Priest Caste (Mass Media and Academia) in shaping public opinion. In this framing, checks & balances hardly matter because The 4th Estate is the defacto head of state.
I agree that we should allow that a person can have a good idea in a sea of crappy ones. But is the whole Cathedral thing really that insightful? Or, quoting beowulf888,
"Yarvin Cathedral premises were always based on his mistaken (bordering on delusional) belief that elite institutions, such as universities and the mainstream press, were *colluding* to pull the strings of public opinion. This has been a mainstay of rightwing nuttery long before Yarvin came along."
Meditations on Moloch was insightful. "Shadowy Elites conspire against you and me, dear reader" is... like I said, every Boomer on Facebook has one of those.
For the record, I too, believe that Moldbug's main contribution was identifying The Cathedral.
> "Yarvin Cathedral premises were always based on his mistaken (bordering on delusional) belief that elite institutions, such as universities and the mainstream press, were *colluding* to pull the strings of public opinion. .
Beowulf is mistaken. I'm pretty sure there's a UR post which unambiguously says that The Cathedral is an emergent phenomenon, rather than an explicit cabal. Though I can't remember the exact terms, which makes it hard to search for, but I'll search upon request.
> "Shadowy Elites conspire against you and me, dear reader" is... like I said, every Boomer on Facebook has one of those.
Imma describe to you my own evolution in understanding Moldbug.
1. Chomksy's Manufactured Consent is absorbed by osmosis. I.e. the media lies, vietnam, watergate, yadda yadda. Like you said, boomer stuff. zzz.
2. I read moldbug's UR a bit. I don't really grok what he's rambling about.
3. Scott reviews Albion's Seed. "U.S. Politics is descended from Protestant religious fundies." Huh. Interesting.
4. The Pull Request [0] (not any specific article): "The Printing Press caused not just Protestantism, but also Nationalism. Nationalism sired Democracy, Communism, and Fascism (in that order)."
5. eureka! moldbug read a bunch of old, musty books and decided that Nationalism (and its progeny) is mind-killing, people-killing, toxoplasmic brainrot. Also, Democracy doesn't answer to *We The People*, it answers to Walter Cronkite and Harvard. In this sense, acknowledgement of The Cathedral goes back even further than the boomers. It goes back at least as far as Thomas Carlyle [1]:
> "Burke said there were Three Estates in Parliament; but, in the Reporters' Gallery yonder, there sat a Fourth Estate more important far than they all."
Understanding the nature of The Cathedral comes packaged with other implications. Such as "oh, Wokism comes from Harvard". And "oh, Protestants are oughtistic [2]". And "actually, violent revolutions are kinda degenerate". And "wait, America is actually 4 nations (at least)". And "actually, voting feeds the egregore". And "maybe literacy was a mistake". And "maybe HBD has a point". And "American Civic Religion isn't really a metaphor". And "Democracy has a lot in common with Fascism/Communism". And "wait, maybe governance by kafka-esque bureaucracy is overrated".
Boomers are still stuck on the Chomsky stage. Like, they'll share memes about that one Simpsons episode. Or maybe ask questions about the Rothschilds and the Moon Landing, if they're feeling extra spicy. But generally, they don't look at the history of the Modern Age and conclude "politics is crack, and we're all microdosing addicts".
P.S. Although gunpowder was probably important to the birth of Nationalism, too. As Carlyle once said: "the real use of gunpowder is to make all men tall." I.e. it's the great military equalizer, since it's cheap & easy compared to knighthood. And since the distribution of political power tends to mirror the distribution of military power... therefore universal suffrage.
[0] https://www.thepullrequest.com/
[1] https://en.wikipedia.org/wiki/Fourth_Estate#The_press
[2] https://jdanielsawyer.substack.com/p/i-literally-cant-even
Thank you for a deeper dive, appreciated.
But still... It's almost trivial? For example, that Democracy answers to Walter Cronkite. It's the next question, where does Walter get his power that the Democracy feels the pressure to answer to him, that gets to some interesting observations., that is, whoever speaks to masses in the language they understand has significant power. We can really dig here, but I think my point stands - even if Yarvin is an interesting thing to study to understand "Alt-Right" or some such, there's 0 reason to engage with him and to promote him (yes, engaging is promoting, especially when you're Scott Alexander).
> It's the next question, where does Walter get his power that the Democracy feels the pressure to answer to him, that gets to some interesting observations., that is, whoever speaks to masses in the language they understand has significant power.
And "The Cathedral" gestures at a fuller answer. Again, it's much more than just "the media is dishonest". It's a window into a heterodox reading of modern history and modern political institutions. Let's dig into some examples that are harder to dismiss.
Remember our debate over monetary/fiscal policy? You concluded with something like "fine, debt is kinda like drugs. But drugs can be very useful if used responsibly." LOL. Please read this post by Dmitry [0] and tell me with a straight face that the U.S. national debt is going to be managed "responsibly". I'm 99% certain that Dmitry was influenced by a UR post that goes something like "hmmm... is it weird that there's no Worker's Cooperatives in the Fortune 500? Maybe because Democratic Agonism inflicts agony on the bottom line? What could this mean for the USG?" However ethically or morally upstanding you think our politicians/bean-counters are, there's a *structural incentive* for the debt to keep increasing. Because the leviathan is myopic. This is one of the reasons why nrx types say things like "cthlulu always swims left" and "conservatives are just progressives going the speed limit" and "conservatives love being beautiful losers".
Here's another example. Recently, you made a negative comment about trump's tariffs and also referenced some sort of obscure aluminum alloy. (I think it was you, correct me if I'm wrong.) I don't agree with the spirit of the alloy clause. If the Cathedral is obvious to you, surely you can articulate why your aluminum alloy clause tilts me to the moon.
Hey, remember Scott's post about various theories on the origin of woke? If The Cathedral is so obvious, then why did such a bright, erudite, elite human-asset like Hanania point to Civil Rights Law instead of our priest caste (viz. the Boston Brahmins)?
I don't think any of this is obvious. And it fills an *indispensable* niche in my epistemology, because it's foundational for understanding a wide array of tangential interests. So I'm gonna keep promoting moldbug, thank you very much. Just as everyone continues promoting Isaac Newton, despite believing in alchemy and secret bible codes.
[0] https://thedosagemakesitso.substack.com/p/gain-dependents-gain-power
> Beowulf is mistaken. I'm pretty sure there's a UR post which unambiguously says that The Cathedral is an emergent phenomenon
You're probably right. I can't dig too deeply into his twaddle without shouting at my laptop screen. But I'd agree with you that...
> Boomers are still stuck on the Chomsky stage. Like, they'll share memes about that one Simpsons episode. Or maybe ask questions about the Rothschilds and the Moon Landing, if they're feeling extra spicy. But generally, they don't look at the history of the Modern Age and conclude "politics is crack, and we're all microdosing addicts".
I don't know that it's a *good* idea, but I think it's a comprehensive (unsurprising given Moldbug's style) articulation of a *common* idea, which can still be valuable.
Like, I think the main thing I got from Moldbug was a better understanding of the grievances that conservatives and reactionaries have; it's not that I agree with his presentation, but it made me able to understand the point of view.
What I mean is, if someone described the Cathedral as "how the world works", I wouldn't find it that insightful. But if someone told me it as a description of the world view that animates reactionaries, I *would* find that insightful (and did).
Oh, sure - as a window into certain mindset, there’s value in observing: aha, this is how such-and-such folks see the world.
But there’s no reason to engage, debate, persuade.
Worse yet, a segment of otherwise sensible Substack contributors regard him as a Very Serious Person with serious opinions. Gak!
My interpretation is similar, Moldbug was an idealist writer about the best possible solution. Yarvin is a realist that recognizes his old solutions were never feasible, and has found a realistic way to accomplish the same objective. You could call him a sellout, but honestly I think most people do the same over their life. People in their 20s are idealistic and shoot for the moon, but now he’s in his 50s, and wants to accomplish something, even if it’s not perfect.
Scott's point is that *Moldbug's own stated views on this* are that the non-ideal version of his program literally leads to Hitler.
Most idealists soften into accepting second or third best over an unachievable perfect; if Moldbug is doing that then it implies either he thinks "literal Hitler" is a second-best solution (which strikes me, as I suspect it strikes most people, as completely insane), or it means he no longer believes his old ideas in full--which is fine, but undermines the rest of his program. If the democracy in the Trump movement is worth contaminating oneself with, does that mean democracy isn't actually so bad overall?
For me, part of what makes moldbug confusing is that "democracy" is weasel-word that can mean lots of different things, depending on the context. E.g. "the abstract notion of political enfranchisement", "local decision-making among friends at Red Lobster", "formal voting schemes", "implicit voting by audience cheering", "neoliberal managerialism", etc. In the context of moldbug, i think what he really means is "modern liberal democracy as descended from *Nationalism*" (see my reply to Mr Fibonnaci [0]). I think Moldbug understands what's going on conceptually, but colloquial English isn't really precise enough to easily disentangle all the stuff going on under the hood of the concept of "democracy".
And as for moldbug's pivot, it's hard to say for sure. But my own hypothesis is also that he's decided to compromise on his ideals. But if you look at his twitter exchange with Scott, the "Miller Light" and "Jan 6" references make it clear that he thinks there's not really enough energy/edginess to actually reboot the NSDAP in America. On the contrary, he's more worried that Trump isn't powerful/edgy *enough* to overthrow the Catheral.
And for further context as to my mindset lately, I've taken a bit more of a Straussian reading of history lately. E.g. I think HBDchick [1] once hypothesized that Communism was popular in Russia/China/etc because the former Mongol Empire follows a certain kinship structure (viz Exogenous Communitarianism) which engenders a lot of resentment. And Marxism was just a convenient justification to Fuck Shit Up. Likewise, I think Scholar's Stage [2] once discussed how Jihadism comes not from Islamic Theology, but from the underlying Arab Culture. So I get this feeling that an ideology's popularity is less a function of Absolute Truth, and more a function of how well it rationalizes adherents' behavior post-hoc. And this is how I view the Vance and the Dissident Right deciding to wear UR's ideas as a skinsuit. So moldbug is just like "welp, according to my Asimovian Psychohistorical analysis [3], I can either settle for Trump, or endure 1,000 more years of Gay Race Communism".
----
{edit: I didn't explain the "democracy" thing well. there's "democracy qua nationalistic hysteria". there's "democracy qua concentration of political power" (cf Aristotle). there's "democracy qua FDR's bureaucratic leviathan".
To clarify, when moldbug says "democracy is bad", he's usually referring to nationalism and the bureaucracy. But when he says "Trumpism is democratically powered", he's referring to "diluted concentration of political energy" (neutral connotations) (for reference, Aristotle said that the 3 good types of government each has an evil twin. Democracy's evil twin is Ochlocracy, aka rule by mob).}
[0] https://www.astralcodexten.com/p/open-thread-381/comment/116775338
[1] I will post links tomorrow when I have the time/energy to look.
[2] ditto
[3] https://en.wikipedia.org/wiki/Foundation_(book_series)
Au contraire, Yarvin Cathedral premises were always based on his mistaken (bordering on delusional) belief that elite institutions, such as universities and the mainstream press, were *colluding* to pull the strings of public opinion. This has been a mainstay of rightwing nuttery long before Yarvin came along. The Cathedral used to be the Trilateral Commission, and before that, it was Freemasons, and before that, it was the Illuminati. In their ignorance, people like Yarvin revive the bogeymen of the past and repackage them in an attractive garment of contemporary memes and discontent. The Paranoid Style in American Politics by Richard Hofstadter is still relevant sixty years after he wrote it, and nothing has changed except that rightwing nuttery has finally gained power in the US. There's something to Marx's quip that "History repeats itself, first as tragedy, second as farce."
What puzzles me is that the Substack ecosystem treats Yarvin as a Very Serious Person when his ideas are and always have been manifestly screwball.
I think he's said he wants Trump to reduce Congress and the Supreme Court to "advisory" roles only, which I interpet as a polite way of saying completely ignore them if they disagree. I think this is beyond FDR/Lincoln, although of course FDR's appointed court packing was a step in this direction. I really do think he means "rule by decree".
I think all of Moldbug's critique of the Cathedral survives this pivot, but I think the objection to him was always "doesn't dictatorship have some very obvious failure modes?", and that most of his answer to that question involved the weird stuff which is now obsoleted. I don't know what his new answer to that question would be.
Lincoln suspended habeas corpus and had a complete majority in Congress because the other half seceded.
You could imagine a different history that the Rebel states don't get their votes back, at least for awhile.
How much of the contest results are driven by people asking their friends to vote?
Hopefully little. Last time I disqualified one entry for obviously being like this (unusually high score with most voters from the same unusual country as the author). I do spot checks to make sure most voters are ACX subscribers (obviously not all, and it's not cheating if a nonsubscriber votes, but if one entry has way more nonsubscribers than others, I'll be suspicious).
I have to say, I think the rules on these things should be more clearly spelt out. Last year I wasn't sure if it was acceptable to rate your own review. I didn't, because it seemed untoward, but I definitely didn't see anything suggesting it was frowned on.
Getting friends to vote triggers a much stronger negative reaction in me, but I imagine norms on this could be very culture-dependent.
I second; Scott should make it clear on what would be considered cheating
(NOT how said cheating would be detected)
It might also be worthwhile to check the vote of people who are long-term and/or paid subscribers.
But I presume that the implementation details of cheater detection are probably best kept secret.
I had something of an idea for the everything-but... contest, but was not able to implement it in time with sufficient quality due to other projects.
Perhaps there's room for two contests per year? One for books and one for everything-but... or is the intention to just return to the normal contest schedule next year and hope there's more everything-but reviews than previously?
Readers tend to get contest fatigue pretty quickly.
You can do double the contests with the same fatigue if you only allow people with even birthdates to participate and vote in one and odd birthdates to participate and vote in the other
(replace birthdates with some other random oracle you can afford to assign to people)
‘Skill issue’ is a useful meme - on agency, learned helplessness, useful beliefs and systems
wrote a short essay on the usefulness of the meme “skill issue” that some of you might enjoy. I wrote it as a way to reconcile my own belief in personal agency with the reality of supra-individual forces that constrain it. The point isn’t that everything is a skill issue, but that more things might be than we assume and that believing something is learnable can expand what’s possible.
It’s part cultural critique, part personal essay, weaving through tattoos, Peter Pan, and The Prestige to ask: what happens when belief does shape reality? And how do we keep choosing, even when the choice feels like it’s left us?
I’d love to hear what you think :)
https://velvetnoise.substack.com/p/skill-issue-is-a-useful-meme
This is just me - I like it, but it is too long for what it is trying to say.
Agreed. Take this as sincere criticism, Maja — but you lost me about halfway down your essay. At that point, I wasn't sure if you believed personal agency could overcome supra-individual constraints—or not—or whether you were sorta in-between on the issue. Instead of opening with a paragraph about Dawkins and the origin of the word meme, state your opinion succinctly and resolutely, then work in the meme theme further into your arguments.