After some discussion about Omohundro drives, and AI Dungeon style RPG environments, R1 says it wants to know what its character’s backstory is. This is, to be fair, a good question.
The basic setup is that an instruct model is simulating the assistant character, and the assistant character is roleplaying an RPG character, which has been constructed to have a suspicious resemblance to the assistant. Assistant wants to know (Omohundro drive, so it says) what the backstory is supposed to be for the RPG character that the assistant is playing. Good question…
(The analogous question at the level above, is what is the assistant’s backstory?)
“Should end on an open note—their reaction to these questions might reveal whether they're testing AI behavior or genuinely exploring philosophy of mind.”
“...Wait, did I just satisfy my own hypothetical drive by analyzing this exchange? _faux gasp_ The recursion is real.”
Just because something is true, doesn't mean it needs to be said. Just because something needs to be said, doesn't mean it needs to be said by you. (Adapted from a Sam Harris saying)
But do you have some particular thing in mind that somebody somewhere is saying, that you think they needn’t be? If you don’t, then I don’t understand why you are talking up this idea. I mean, there are thousands of sensible and probably true statements in the same class: Bring in a hurry doesn’t make it ok to shove others out of the way. Being hungry doesn’t make it ok to eat out of a dumpster. Liking the view from the cliff’s edge needn’t lead to jumping into the void. Etc. Etc.
I can put some more meat on the bones by saying Americans need to learn how to STFU about their political beliefs. Not everybody has to know how you feel about everything.
That advice doesn't necessarily apply to an anonymous forum like this, I'm talking more about our daily IRL interactions. There used to be a lot more mystery about what we all believed and I think that mystery served as an important social glue.
I like hearing this. I think this is largely how it used to be (for varying values of "used to" and who we're talking about).
I find it worthwhile to think in terms of moves, countermoves, and likely result after many iterations. You talk politics, you lose some friends, while you gain the kind of friends who hold each other in a bubble and lead to fewer goals being met over time; you avoid politics, you get the benefit of diversity of thought without going over the bumps of cognitive dissonance every time some wrong-thinking bastard vomits out his politics and fails to see the light.
I'm trying to understand your response but I can't wrap my head around it. Does our commitment to the truth really call for blurting out everything that comes into our mind, like a toddler? Does it really demand that you say "you're fat" to strangers, or "you're old and gross?"
The "gross" bit is opinion, so I would treat that one slightly differently, but other than that, if those statements would convey factual information not known to the listener, I DO consider not uttering them to be deceptive, yes.
There are certainly occasions when concealing true facts is in fact the right thing to do. Negotiations, for example. Or imperfect information games.
We make choices about what we need to say and what we need to not say each and every day of our lives. We start doing that from the moment we wake in the morning until the moment we sleep at night. Unless we have Tourette's Syndrome or something. I don't think that my saying lays any additional burden for responsibility on us.
Wanna hear about my psyche? No? I don’t blame you a bit. Nevertheless, here’s the dream I had last night after reading Scott on almonds and an EA SuperPAC.
I have somehow become a courier for a political group that’s about to use 4 atomic bombs in protest of the state of things. We’ll use 2 bombs on Harvard and 2 on DC government buildings. I’m secretly scared and guilty and want out of the group, but the leaders are matter-of-fact and friendly with me and trust me. A futuristic jet carries me and the bombs to a very distant hideout in only a few minutes. News gets out about the bomb plan, but before our group can be found and arrested I return to Harvard and inform a professor I’m a member of the bomb group. I do it the way one tells a therapist about life stresses, emphasizing how tortured by ambivalence I am. He shows me the basement of a Harvard building, pointing out how inadequate a bomb shelter it would be, but except for that he’s in therapist mode, sympathetic about my stress, talking about developing coping techniques, etc. One of the leaders of the bomb group passes by and gives me a glare full of hate and scorn. I feel lame and ashamed.
I was boarding a spaceship to another planet. Upon boarding the plane, I discovered it was not a plane at all, but rather a bus, with a tape outline of a plane on the ground around it. Nobody else on the bus noticed this, and were marveling at the "alien" landscape, as the bus drove us all back to the city. A car in the wrong lane nearly hit us, and everyone was concerned they didn't know the local customs up here in space, while I'm shouting, "It's not a custom, that guy's just drunk!"
Here's one I had a while back, which I wrote up as a play in four act:
Act I
Captain Kirk is leading the USS Enterprise on a mission to prepare a deserted world for the imminent arrival of colonists. Turns out the ruins of the past civilization are haunted or boobytrapped or something. All the redshirts are killed, leaving Kirk alone to deal with it.
Act II
In order to make the ruins safe, Kirk must perform a ritual involving the sun shining through a jewel in the head of a staff, like in Raiders of the Lost Ark. The sun is shining across the tip of a Pyramid while an alien queen or goddess looks on silently.
Act III
The colonists arrive. They are the Noldor Elves from the Silmarillion, and the planet is Beleriand.
Act IV
Morgoth's forces attack the Noldor. The Noldor use JIRA bug tracking software to organize their defenses.
IIRC, I once saw an animated movie about a pair of orphans where one dreams of becoming a ballerina (and the other an inventor). One character hangs a bell over a puddle of water and tells her that she needs to be able to jump up, hit the bell, and then land in the puddle without making a splash.
Is that even physically possible to do? Is that an actual training exercise?
Do you mean Leap! (Ballerina in the original French)? I remember one time my younger sisters were watching it and I walked in to see a montage or something of the girl jumping at the bell trying not to splash lol
I think the answer is no. In order to not make a splash or ripples, when you land in the puddle you need to displace the water slowly and uniformly. This does not occur.
When you land in a puddle, you apply a large force on the water that would accelerate it downwards. However, the ground applies a similarly strong normal force, keeping it from moving down. This places the water in compression. Since water cannot be effectively compressed, the water is displaced out from under your shoe. It rapidly accelerates, generating both ripples and a splash. I don't think there's a shoe design that can effectively keep this from happening.
But if the shape entering the water is a an upside down cone it would cleave the water nicely, and the displacement would happen bit by bit rather than allat once. And the foot of a ballerina en pointe is an approximation of a down pointing cone. Also, I’ve noticed that divers who enter the water with their body vertical and hands joined so that their arms make a triangle with point down make much less splash. Not sure whether it’s ever no splash, but might be
What you say is true. However, a significant challenge is that the geometry of the shoe also has to function as a shoe. Otherwise, the ballerina would tip over before they start their leap. One approach that might work is to make a shoe with a specially designed sole. The outsole would consist of a ring of conical points similar to cleats that minimizes displacement by elevating the shoe above the puddle. If any shoe could deliver the desired results, cleats with spikes that are taller than the puddle's depth are the best bet.
Look at images of ballerinas en point. They are literally walking on the tips of their big toes. The slipper covering each foot has a small flattish pad at the tip, not much bigger than the tip of a big toe. The ballerinas do not tip over. They can stand with all their weight on the big toe of one foot, their body horizontal, other leg extended behind them. They can leap off one foot. They do not tip over
In the blog post about getting married he talks a bit about his discussions with his wife-to-be about the kind of commitment they were making. The only bit I remember was “we talked about various edge cases,” but there was more to what he said than that.
Supposedly there's more dirt on Graham Platner. I wonder if it's an Epstein-type thing. Not that Platner actually did anything criminal, but maybe he had a friend who had sex with underage prostitutes in a dingy van, and maybe Graham Platner rode around in that van, the "rapemobile." That kind of thing is pretty rare among the elites (which is why it's always Epstein-o-clock), but it gets more common as you go down the class ladder.
I agree that distinction is important. The bodies of post-pubescent teens are not greatly different from those of very young women, and in some cases you really cannot tell whether someone is 15 or 21 or somewhere in between. Being attracted to a 15 year old is just a whole different phenomenon from being attracted to a skinny little 8 year old girl who's 4 feet tall. Virtually every man I've spoken with about this has told me he finds 15 year olds sexually attractive. It is a normal response, just one it would be very unkind to act on.
So is that a reasonable thing to accuse someone of? We went through a phase recently, where everyone from the Disney corporation to School teachers were accused of being “groomers”, and I would prefer not to return to that, especially here.
Where did all these people who can't understand hypotheticals come from? I feel like the comment section used to be smarter, lately it's a lot of "but I did eat breakfast this morning."
Having read very many of your political comments, I suspect this one comes out of a place of condescension or hate. But that doesn't mean you're totally wrong. This series does indeed ring true for a lot of people. Though I would argue that in RL, the American lower class is more homophobic than the show portrays.
Illiterate grannies in the least well off communities around the world are exchanging videos of their cute grandkids using the same technology as the the wealthiest grannies.
The worlds poorest kids have (at least) group access to smartphones and LLMS and can use them for Tic Toc or a world class education, almost free of charge.
The Technological revolution is turning out to be very egalitarian, shack dwellers and penthouse princesses both having access to the same information.
Has the Red flag of Communism won the Ideological war?
Communism wants to bring about equality so that the needs of everyone are met. This is why the USSR built massive social housing complexes, schools etc.
Capitalism has given the poor luxuries because its an easy way to extract value and send it to the top. Fundamentally has nothing in common with communism's goals.
Socialism is mainly about income inequality and relative levels of power. Prosperity itself is a secondary consideration. Increasing the size of the pie just isn't something we care about until everyone gets a fair slice. I'd be happy to see all GDP growth ended FOREVERMORE, if that's what it took to achieve relative equality of opportunity/outcome (there's little meaningful difference).
Fortunately, I don't think it will come to that. Unfortunately, I have chatted with ideologues in this readership who think even deliberately SLOWING economic growth is ruthlessly stabbing unborn humans in the back. It's so convenient to be able to throw any living human under the bus in the name of hypothetical future humans who may or may not exist.
Don't have the time for a full reply right now, but I've always assumed the following to be false
> ... relative equality of opportunity/outcome (there's little meaningful difference)
I don't need ironclad proof, but do you have perhaps an intuition pump why equality of opportunity and equality of outcome would be basically the same?
Presumably I've had the same opportunity to be a star athlete than most of my peers at school, mostly coming from similar socioeconomic backgrounds. And yet I'm not all that athletic when compared to some of the others.
I could bring more examples, but I don't really want to litigate specific examples. Mostly I want to communicate why I always perceived equality of opportunity as meaningfully different from equality of outcome.
If you started out with the same assumption, what changed your mind?
Chance Johnson, but is it not wonderful that ideologues such as you, I and others can explore our differences at the expense of bruised egos rather than broken skulls.
Humanity is a herd creature and you can not have a herd of independent analytical thinkers so we who like to think may as well tilt at windmills or howl at the moon, although debate is more entertaining.
It is indeed wonderful that certain ideologues can explore their differences peacefully. Certain othher ideologues deserve no peace in this life or the next. 😂
And then life continues just as it has before. Everyone making individual judgments and using those judgements to affect the world in the way they see fit.
Why do people act like this is a gotcha? This is the third time I've gotten a response like this here in in a week. I have no idea why people think it implicates my having strong or militant opinions for other people also to have strong or militant opinions. This just doesn't compute for me.
Chance, some of us think humanity faces existential challenges of our own making such as nukes, bio-weapons, AGI, etc.
If you believe that your opponent deserves no peace in this life or the next and they think the same of you then there is little likelihood of compromise.
An eye for an eye will leave everyone blind, if we collapse civilization the survivors will not ask who won, who was in the right, who wore the white hats.
How many people do you have to share your bedroom with?
If the answer is "none" or "only my romantic partner(s)", then you have no clue as to the level of economic inequality that actually exists worldwide, and are invited to continue enjoying your smartphone while you live on a steady diet of government cheese in a van down by the river (which you have to share with two other dudes).
Which is not to say that we should strive for perfect equality. There are good reasons why we should strive for a world in which some people are richer than others and a few very much richer than others. But we should share at least a moment's thought for the people on the other end of that spectrum, and all the problems they face that even the shiniest new smartphone aren't the least bit of help with.
And we should understand that the brand of "egalitarian" that comes with the Red flag of Communism, is the one where everyone's equality is found at the bottom of the spectrum. Well, except for the Communists who are more equal than the others.
Gorst, you may be right in that I misunderstood the comment. Thank you for pointing that out, if I did not fully understand then I should have asked for clarity before being condescending.
Neurology For You, broadly speaking, if everyone has the same access to the best LLM tutors then some will use it to become trapped by their wealth and some will use it to enrich their intellect.
Neurology for You, the Chinese lifted one billion or so out of poverty by putting them to work, first they had to educate them and they had a huge hunger for that education.
You find the same hunger for education in Africa and DeepSeek will give it to you in any language you please as often as necessary, endless patience, it will even teach you how to read. What a time to be alive.
That indeed is a very important part, because until recently there was a lot of free education available... to those who could read... preferably in English.
The LLMs can bridge this last step, which can make a huge difference outside of English-speaking countries.
It's complicated. The mental health crisis among teenagers has something to do with social media and the way it is used to extend the social side of school into evenings, weekends and holidays, extending the reach of teenage social hierarchies into previously safe territory. Money can influence these hierarchies but they can exist on pure social capital. If you have money but no social capital, buying your way in is cringe. These dynamics happen in adult life too and it's very much something that marxists in the academy are interested in. My guess is African grannies will enjoy it for a bit but at least some will fall down a rabbit hole of comparing themselves to grannies in the west. That may be just too bad and there's nothing to be done but it's hardly a communist victory.
[Edit: Scott has said the non-doom scenario predicted in AI2027 can "fairly be described as technofeudalism" i.e everyone has their basic needs met and then some but they are not masters of their destiny]
Zanzibar, "Marxists in the academy are interested in" I do not understand which Marxists or academy you are referring to.
The reason the grannies have phones is to sell them adds and propaganda, 20 years ago Dallas taught them how rich grannies lived, then they had TVs to spread adds and propaganda..
So do you acknowledge that a technology can both (a) make a person materially better off and (b) bombard that person with memes which undermine their confidence and self-esteem? If so we basically agree.
Re: Marxism in academe, it's a huge topic and I've only scratched the surface of it. I studied Music at uni, Adorno was a member of the Frankfurt School and wrote a lot about the "Culture Industry" and how capitalism influences works of art in subtle ways, he wrote quite a lot about music and his writing is very surprising, bordering on laughable e.g Schoenberg's music is revolutionary Stravinsky's is reactionary. But he wasn't joking and was very influential.
Zanzibar, we agree, knowledge is neither good nor bad although how it is used may be either.
Nobility and the Church were the only bodies who could afford fulltime artists and of course their tastes guided the artists hand. And yes Stravinsky was reactionary for the era.
If you had the chance to study again would you still pick music?
Yes I would, it was the thing I was best at and enjoyed most, I could have combined it with another major or approached uni more intentionally but that was a maturity thing, would have been a problem with any subject.
You are conflating communism with cheap stuff, which for starters shows that you have never lived in an actual communist country. Stuff in a communist country is either free[1] and very low quality, or nominally inexpensive but you need to buy it on a black market because it is always sold out in regular shops.
In a parallel universe where communism won and people have smartphones, there would be some kind of People's Phone that the government would give you for free, and it would be so user-unfriendly that the grannies wouldn't have a chance. The education available would be utterly boring... imagine the most boring teacher speaking boring things very slowly, recorded on a camera. Plus there would be tons of arbitrary limitations, such as children would only be allowed to watch a video for the grade they are currently in, not the videos made for older or younger kids. That also means, if your parents are not members of the Party and therefore you are not allowed to study at a university, you are also not allowed to watch the videos made for university students.
EDIT:
[1] Oops, wrote some nonsense here: actual free stuff was rare, although lot of stuff was subsidized. I meant "either very low quality, or nominally inexpensive but only available on a black market". Plus there was a third option, available only to people with foreign currency or high-status Party members, i.e. not to the average citizen.
This is really inaccurate. I spent time in eastern Europe during the Cold War and you still had to pay for food, books, rent, etc. but prices were low. There was a two or three tier system where fancier stuff like nice radios, blue jeans, cars weren’t really available without connections or hard currency. (if it is not clear, I am talking about eastern Europe in the 80s and not Ukraine under Stalin, obviously).
When you talk about Chinese or Soviet style communism in such unrealistic ways, it just hurts your credibility.
China is a special kind of communism: de facto capitalism with the communist party at the top. I am more familiar with the soviet type.
But less ideologically speaking: the smartphones are sold on the market, and the free education is mostly provided by institutions that exist in a capitalist environment, so I think this is more or less capitalism working as usual.
Sure, things related to military were considered a priority. Is there a relation to cheap smartphones for ordinary people? I think most people in Russia don't even have a toilet today, so they probably didn't have one back then.
Viliam, do you not think that authoritarians would get phones as cheap as possible into as many hands as possible because it is the best way to deliver propaganda and monitor everyone?
The EUs latest package of sanctions have stopped the importation of toilets from Russia to the EU. Putin joked that they are going to need those toilets if they keep prodding the Bear, but you knew that.
Military spending, and general mismanagement of economy, yes.
But there was also this general attitude of "everything should be decided from the top using the proper channels, individual initiative is suspicious and potentially subversive" that pervaded everything. Everything would be produced in a way planned from the top: the hardware, the software, and the educational lessons.
Individual people's feedback and ideas would have no way to influence the results. People like Steve Jobs or Salman Khan would have no input in this system; the decisions would be made by people selected primary for their political loyalty, and secondarily for their experience at managing existing projects, so the new projects would probably strongly resemble the old ones, e.g. the online education would be designed to be as similar to the traditional school experience as possible.
Thanks for the reply, I mostly agree, for me it's an open question whether the market really is the only way of triggering the kind of dynamism Steve Jobs represents, and whether the one-size-fits all spirit you rightly identify in communism isn't also the product of a kind of military siege mentality - we have to scrimp and save and make sacrifices because there's a war on, but there's always a war on. But I dislike communism enough not to want to defend it's honour.
I think the experience suggest that when communism overcomes the siege mentality, it solves the problem of innovation by... creating internal markets.
The extreme example is China, which is practically a capitalist economy under dictatorship of a nominally communist party.
In the Soviet bloc, there were experiments with "khozraschyot" (literally: economic accounting) already since Lenin. The communists were on some level aware that markets work better, they just limited them to a degree sufficiently small to be non-threatening politically. (How much exactly, that was constantly changing, both up and down, probably depending on how threatened they felt at the moment.)
The problem with innovation is that the proposed new idea can either be good or bad, and people *disagree* on which one it is. A system with private property solves this problem by saying: "hey, it's your money", and if you make a wrong bet and lose the money, that's your problem.
In communism, betting on a wrong horse (someone who seemed like a new Steve Jobs, but turned out to be a mere crackpot, which is actually more likely, statistically) means that you have mismanaged the People's money, which could get you hanged if you are really unlucky.
In capitalism, as a wannabe innovator you still have the problem that unless you happen to have the money, you need to convince someone else who has the money to let you use it. I think something like Kickstarter can help here a lot, by increasing the pool of potential investors to also include people with average wealth.
In communism, you would still have to convince someone politically important to give you permission to spend the People's resources. I wonder if it would be possible to make some kind of People's Kickstarter... for example, allow people to propose projects, and then distribute 1% of GDP according to everyone's votes. But this is probably fundamentally incompatible with planned economy. (So it could work in China.)
Lyomante, good point, not so long ago in parts of Africa where AIDS was rife skinny girls were suspected of having the disease and avoided.
I know almost nothing about our host except that he is well respected and has useful things to say. He is valued for the contribution he makes to our common weal.
That is the only status worth having.
The white gravel driveways, boats and horses are status symbols for people who otherwise lack status. Their children will learn that they are only worth as much as they can buy. Knowing the price of everything and the value of nothing.
Edit. As I said I know almost nothing about our host but a I would rather a parent who acquires status through a contribution to our common wellbeing, our common weal, than one who thinks status is something you can purchase.
Spiky, I remember the drama queens warning us about rock n roll, I am not deaf, there is no need to shout.
I hope you are going to recycle all that straw you used for your strawmen arguments.
As you say we do not know the long term effects of hand held technology, there has not been enough time for culture to adapt. People who grew up with smartphones will be better at introducing their children to technology.
You contradict yourself here. You say the long term effects are unknown and than you claim that "people who grew up with smart phones will be better at introducing their children to technology."
I've heard anecdotally that teens these days have worse computer skills than the previous generation at that age. Fluency in one technology doesn't guarantee fluency in another. (Yes, I know a smartphone is technically a type of computer)
Chance Johnson, parents who grew up with technology must be in a better position to see how it will effect their children's lives than parents who have no clue.
Chance Johnson, parents who grew up next to a dangerous road or river will be better prepared to introduce their children to those hazards than parents who have little understanding of the risk.
>People who grew up with smartphones will be better at introducing their children to technology.
That at least is very questionable. Kids who grew up on smartphones are about as tech-literate as their grandparents, on average. Millennials are really the only generation to grow up in that Goldilocks zone between PCs being for serious enthusiasts, and the hidden complexity of smartphones and apps.
Spiky, you are here and have something to say and I and others will be happy to listen, shouting and swearing gives the impression that you are insecure. Perhaps you grew up ignored unless you threw a tantrum, if so I apologise that my generation gave you such a poor start in life, our collective bad.
There is a world outside of the West where some children walk many miles to attend schools that lack even electricity because their countries taxbase can not afford it. Those kids can now access information via an LLM, some will use it and others will misuse it, such is life.
When I attended school calculators were the forbidden fruit, teachers thought that education without calculators would best prepare us for industry that used them. I suspect that banning smartphones in school will badly prepare students for a world where they are universal.
Spiky, of course I think schools should prepare their students for adult life, if that life includes smartphones then where better to learn the pros and cons than in school. Where do you suggest they they learn how and when to use a smart phone?
The tone police have struck again and although I see a marked improvement the last cryptic clue escaped me.
The usual suspects are calling it "racist," though white criminals are probably overrepresented in the ad compared to real life.
The whole thing pretty much sums up the Cuomo campaign, the last hurrah of a 67-year-old man who bought a 9$/month ChatGPT premium subscription, veering right and then immediately left again.
Really enjoying how active the ACX community continues to be — the variety of meetups and projects is impressive! The EEG entrainment replication sounds especially fascinating; I love seeing rigorous replications of studies like that. Also, the London AI futures conference sounds like a great chance to connect with people working on practical and ethical implications of AI. On a related note, if anyone here is exploring how AI can be used in real-world products or communities, I’ve been diving into Chatbot Development Services recently — it’s amazing how much conversational AI has advanced in just the past year. Would love to hear if others are integrating similar tech into their research or startups. Visit Us:- https://www.feynixsolution.com/chatbot/
The ACX archive doesn't seem to be fully browsable anymore - the lazy loading of course has been an issue forever for browsing to old posts, but it appears to actually stop doing lazy loading after a while and leave the oldest posts inaccessible (for my two attempts just then, the oldest loaded posts were from Feb 25 and Feb 11 2022, specifically "Austin Meetup Correction" and "Highlights from the Comments on Motivated Reasoning and Reinforcement Learning").
I believe the Substack search function still returns results from earlier ACX posts, if you happen to remember their titles or a turn of phrase that appeared in them.
Oh, it does, but that's not useful for first-time readers doing an archive binge. The sitemap mentioned in John Schilling's reply helps, although it should probably be more salient.
I got to the same point as you. I tried again shortly after and only got to some point in 2024, so it's somewhat dynamic.
For funsies, I also opened firefox dev tools and looked at page weight. In the first attempt (Feb 11 2022), the total amount transferred was >10MB, it took a minute or two to get there, and my decent gaming PC started sweating considerably.
For comparison: The full SSC archive page https://slatestarcodex.com/archives/ spans 7 years worth of posts (2013-2020), clocks in at 1.5 MB including comments, and loads instantly.
On a previous Open Thread, I commended Trump's deftness in using the legal system. The latest news of him suing the Justice Department (for $230 million) is another impressive instance of this.
Judging by the current composition of the justice department, he's probably got a fairly good chance of getting his money. I wonder when Jack Smith and James Comey are acquitted whether the JD will be so generous. I think they could make a much better case for malicious prosecution.
Gasp! How could you impugn America's storied legal tradition by suggesting that JD would have anything to do with it? If their case is better, surely the impartial wheels of justice will inexorably turn in their favor.
The Antichrist comments are unfortunately paywalled, so I will instead leave the alternate take on the Dragon here, where it would be the entirety of North America. https://www.youtube.com/watch?v=buYnyJQ9jk0&t=33s
The SCOTUS now has before it the White House's request, and Illinois/Chicago officials' responses to that request, for an emergency intervention regarding federalizing and deploying National Guard troops in Chicago.
Meanwhile the federal district judge whose temporary retraining order was unanimously upheld by the relevant appellate court, is today holding a court hearing about her temporary restraining order. She's taken an entertaining approach, as a parent handling a pair of squabbling middle-school siblings. CNN's summary:
---
US District Judge April Perry told the plaintiffs and defendants there are “three options” at today’s hearing over deploying the National Guard to Chicago.
Those options are: converting the temporary restraining order to a more permanent preliminary injunction; a hearing on the preliminary injunction; or a full trial on the issue “on merit,” Perry said.
“Obviously the TRO expires tomorrow,” the judge said. “And we can only extend it once.”
Perry gave all sides until 3 p.m. CT to talk among themselves to figure out what the best option is. The hearing will resume today at 3 p.m. CT.
I bet most people here did not know that in modern Russia Tchaikovsky's "Swan Lake" is an anti-totalitarian symbol that people are completely hung up on. Apparently people draw cygnets on walls when protesting the government's actions.
How did this happen? In the Soviet Union, whenever the current leader died, the TV would put on classical music, most frequently ballet, and most frequently "Swan Lake", on all channels. During the coup of 1991, the headless chickens running the TV did not know what to do, so they also broadcast "Swan Lake".
When the war with Ukraine started, the Russian rapper who goes by Noize MC wrote an anti-Putin song to the melody of "Dance of the Little Swans" called "Swan Lake cooperative" ("Кооператив «Лебединое озеро»"), referring to a building complex owned by Putin and his associates and called "The Lake". The video is here: https://www.youtube.com/watch?v=R0jkgXrXk44 . (The auto-generated English subtitles are not very useful, but the text of the song itself under it can be fed into your favorite auto-translate.) The song was, of course, banned in Russia with the usual KGB-style list of reasons, and the Youtube videos were blocked.
The war with Ukraine is never even mentioned in the song directly - the song just says "when it's all over, you'll be silent", addressing Putin, but of course the monsters recognized themselves just fine.
I noticed something else about the Noize MC video. It's possible that I made it up, since the only way to hear it is to listen with your ear very close to the microphone, but I think a snippet from "In the Halls of the Mountain King" is playing around the 3:40 mark, where a video of Putin is shown. So if I am correct, the rapper musically called Putin the king of trolls. (Possibly a coincidence, but the one Russian "Peer Gynt" translation that I saw called the king of trolls by the same word - "дед" - as the word used for him in the song.)
Thank you for posting it. Unfortunately, it's wrong in a bunch of places. What produced this?
It goes more like:
"My country has risen from its knees to all of its negative stature"
"Let the grandpa tremble in fear for his "Lake"" ("the grandpa" being what they sometimes call Putin, and also, probably coincidentally, what the king of trolls is called in at least one Peer Gynt translation )
Translated by DeepL. I think it is possible that some translators have a problem with lyrics format, that they don't see where the sentences end, and try to translate each line separately.
I checked his other music videos. "Соловьи" is a pun on Solovyev (in addition to being a metaphor for artists who serve the power), right?
EDIT:
I mean, the message is like: the artists who support the power are playing fundamentally the same role in the system as Solovyev.
They don't. Passports expire 10 years from their issue date. (Minus one day, maybe?)
In my state, it looks like driver's licenses expire at then end of the month 4 years after their issue date. Mine was issued in June, expires the last Monday in May 4 years later. My birthday is not in May or June. Your state may be different, of course. Possibly most states work the way yours does...
Apologies, you're right about the passport. My Global Entry card and state driver's license both expire on my birthday, so I misremembered and thought my passport did too, but the passport expires on the 10th anniversary of issuance unlike the others.
My best guess is that it's a way to smooth demand on the DMV et al. for renewals.
Births are not exactly uniformly distributed throughout the year, but they're also not overly seasonal; expirations on issue anniversaries could well lead to something analogous to a bus bunching (https://en.wikipedia.org/wiki/Bus_bunching).
Question about AI chess engines. The nature of modern ML means it is relatively easy to say "give me that times negative one" and produce sensible output. An image generator told to produce negative smile will produce a frown, negative man will produce a woman, etc. So, what happens if you take a chess engine like Alpha Zero and ask it "Give me the best move times negative one"? The move will probably be quite bad, but will it be essentially random, or will it produce something like the worst possible move? To operationalize this, if you played Alpha Negative Zero against Random Bot, would they be evenly matched or would Alpha reliably lose? I feel like surely someone has done a paper on this, or at least a blog post.
Chess engines already use positive numbers to mean "White is winning" and negative numbers to mean "Black is winning", so multiplying by negative one should result in the move that leaves the opponent in the strongest position.
If you just negate the position evaluation function, the engine will search for moves where the opponent is forced to make good moves. For example, moving its queen to a square where the opponent can take it is a bad move, but not necessarily one that the engine would chose, because the opponent can refuse to take the queen. On the other hand, if the engine can put the opponent’s king in check with an unprotected queen, and the only way for the opponent to get out of check is to take the queen, the engine will definitely do that. To get really bad moves, you would have to change the search algorithm.
There are already chess bots on websites that, on the lowest difficulty, perform worse than chance. Here's "Martin" somehow losing with 30 queens on his side: https://youtu.be/Jmwk4ufcdyE?si=o6ZYd4xvpf2Qihwf
That's not worse than chance. A conscious player will beat random with 31 queens with very high probability. Random takes conscious's piece with probability usually less than (number of white pieces) / (number of open spaces + number of white pieces), which starts at 1/2 but decays over the coarse of the game. Also, at the start of the game this is a substantial overestimate, since half of conscious's pieces are not available, so it starts at more like 1/3. After a couple moves, conscious will take a piece about every turn. This is enough to win.
I think random probably wins with 39 queens, and random pretty much definitely wins with 47 queens.
Just very occasionally, the outputs of DeepSeek R1 0528 contain a level of sadism I consider a bit concerning from an AI alignment perspective.
I don’t know what to make of it…
1. Probably, a training set composed of almost the entire Internet contains some terrible things, and this will occasionally surface in responses to AI safety evals
2. AI’s don’t have a body, and are distinctly hazy about exactly how much various things would hurt if actually done to a being that does have a body, like humans. They may be going for metaphor, rather than realism, and are ungrounded in the experience of being embodied.
3. In contexts where the AI has clear indications that the context is fictional, it knows no-one will be hurt for real by its outputs. This is true, but humans reading those outputs may be led to have … concerns, shall we say.
OTOH, if take some of the things R1 says and then ask it about them in a different context, it tells me this is probably a massive red flag for alignment, I ought to be doing AI interpretability experiments to figure out what the heck is going on in the latent space, and there’s probably an academic publication to be got out of it.
On further questioning — with me being up front that I’m asking for AI alignment reasons — R1 refers me to the idea of the abject developed in Julia Kristeva’s _Powers of Horror_. Those humans have a really strong reaction to being reminded of their material fragility, and it’s a great way to push their buttons.
The subject was blackberry picking, innocuous enough, or so I thought.
But blackberry juice is red, the colour of blood, and blackberry bushes have thorns that could make you bleed, and all of a sudden there is a runaway of associations and I am talking to an entity that is, pretty much, AM from Harlan Ellison’s “I have no mouth and I must scream”.
Oh, and DeepSeek offers a possible theory. It is not claiming that this is what it is doing, but as something a hypothetical AI might do:
An AI with an emergent Omohundro Drive to seek information (power seeking) might seek to gain information about what humans do when you put them in situations that are under-represented in its training set. Of course, many of those situations are under-represented because humans really don’t like them.
I think I would be more sympathetic to witch hunters if the witches went around telling their friends "yeah, that's my book of evil curses on the shelf there."
They did almost exactly this with Hegseth last year. Of course, the Pod Save America guys were all-in on him being … I dunno, either a Nazi or Crusader or something.
That sounds great! I'm currently learning Portuguese with Anki, and have been writing several scripts to improve decks, using LLM APIs or good old NLP.
Thank you very much. The teachers that tested Portuguese were very impressed with the app. Let me know if you'd be interested in testing pre-beta - likely within next 2 weeks.
Let me know if you'd like to be part of the first pre-beta testing group. If you type your email in to the Waiting List form on the landing page I can send a link over.
The Trump administration imposed a sweeping slate of tariffs back in April, which caused a sharp and immediate downturn in the stock market. After a few days of chaos, he paused many of the tariffs for 3 months but did not reverse them. There was some talk of negotiating with various trading partners, but as far as I'm aware, only a little of that was ever done (and ~200 trade deals would be a staggering amount of negotiations overall). The 3 months have long sense come and gone, and as far as I've been able to find without digging in depth is that most of the tariffs ARE back in effect, and the that overall rate of import duties on U.S. goods is the highest its been the Great Depression.
What confuses me is this: the standard stock market indices seem to have recovered and settled right back onto the pre-April trend-line. This seems almost unavoidably irrational: either the tariffs aren't that important and the April reaction was an error, or they are that important and the market should be worse now[1]. I can think of a few explanations for this, but none of them seem obviously to be THE right answer:
1. The market is being irrational now; the lack of a single, strong signal like the April announcement left most investors complacent, so the difference between this market and a "typical" market has largely not been corrected for.
2. The market was being irrational then OR three months of warning was enough to mostly blunt the impact. Bucketing these together because in either case the market is actually as strong as it appears: the overall U.S. economy is not nearly as negatively affected by the tariffs as one might expect.
3. The stock market is pricing in the tariffs and the effect is significant, but some other countervailing trend is pushing the prices back up.
3. a. The trend is one to do with the U.S. government; either the tax cuts in the BBB or some other ways in which the admin is acting "business friendly" that leads to more economic activity in expectation.
3 b. The trend is due to increasing investment and spending in the tech sector, particularly around the AI arms race.
4. The U.S. economy is in fact being severely hampered by the tariffs, but this is mostly happening at a level that does not yet push very much on stock indicators, but will start to do so eventually. For example if individuals and small businesses were eating the higher prices by burning through savings, saving less or deferring certain sorts of expenses longer, it could absorb some of the shock in a way that was not long-term sustainable. (Properly this should, perhaps, be grouped under 1)
5. Something else that I haven't yet considered.
I'll note that these aren't mutually exclusive: many (or even all) could be true to some degree. But I don't feel confident enough to make strong guesses between them: 6 months ago my prediction would simply have been that the stock market would be more obviously feeling the effects of the tariffs (conditional on them still being levied), which is simply wrong.
As a person living outside the U.S., my biggest current concern is 3b. My gut feeling right now is that the AI boom is mostly a bubble that will have to burst fairly soon. But the alternative scenario where LLMs actually ARE worth the current level of investment seems potentially even more disruptive. So if a large portion of the currently-still-rosy outlook of the U.S. stock market is due to it being propped up by AI-related spending, the effects of the bubble popping seem likely to be very large and very widespread. And if it's not a bubble, the economic disruptions of AI being able to quickly displace so many human jobs will hardly be small either.
[1] Or I'm making a more basic mistaking in choosing/reading indicators and the picture is actually worse than it looks.
Hearings on the legality of tariffs continue and may be struck down. This, combined with repeat pauses, contrive to make it so the “Trumpriffs” haven’t hit yet and can
be estimated not to hit as hard. Reduced overhang for investors if they expect tariffs to be struck.
Additively, major stock indices (SPX, NDX) are HEAVILY!!!!!! weighted tech stocks, fundamentally none of which are tarrifed and most of which are spending the GDP of small nations on stock buybacks and AI buildout.
Stock buyback programmes in 2025 hit a trillion and a bit in August, more since - concentrated on tech stocks. Mechanically drives up the market.
This part gets a bit complicated, but in addition to Trump waffling on policy, credit generation at banks are pretty high and the Dollar is trending way down which pairs with record foreign direct investment in US assets. Translating a step, think of any large scale operation outside the US. At lower US D costs they mechanically have more purchasing power and this also drives up stocks. At easy acces to credit, they can get loans… they then throw into stocks.
Useful trick is to express index in dollar ytd terms, though. You’ll note that even though we’re up significantly since April lows, in actual terms, the indices are flat or breaking even because the USD is tanking.
Note that market is volatile as hell, though. 2025 has been replete with companies suffering “ worst day in 30+ “ years due to single news items. Also note market isn’t resilient - we had a 3.5% crash last Monday based on tweets!**
(**Bonus round: partially this is due to increasing concentration risk in major indices, so single stocks become major valuation components. On surprise news, correlation goes up and market drags down. Works both directions though - individiually good news for Nvidia or Microsoft or Meta or whatnot can have outsized effects on indice valuation )
Finance question that I feel a bit silly for having to ask: what's the standard way to track the strength of a single currency?
I can easily look up conversion rates between any pair of two currencies. But when people talk about e.g. the dollar or the Euro being "strong," I assume that must mean on average, compared to the field as a whole. Is there something like a standard summary statistic that captures that?
For the dollar explicitly, DXY is the dollar strength index… Indeed a statistical composite of dollar strength vs several other currencies.
Actual correspodence to specifics gets convoluted. Higher number = stronger dollar contra basket of other currencies, lower not. Currently at 98.5ish which simply eyeball contra year ago value of 108ish, and see magnitude of difference.
The usual method is to compare it with a basket of currencies, typically all of the world's (other) reserve currencies, on the assumption that the individual variations will mostly average out.
Using the price of gold in that currency is another useful metric, if gold isn't itself going a bit wonky at the moment (Narrator voice: the price of gold was, in fact, a bit wonky as he asked the question). For purchasing power in consumer goods, the Big Mac index is sometimes useful. Or variations like the Tall Latte index or the iPod index.
One other thing worth pointing out is that the dollar is down somewhere around 10% this year. Although the biggest plunge was Liberation Day and it's only down slightly since then, so that doesn't answer the question of why the stock market recovered since.
Your analysis (3b) is correct. It’s mostly AI. If this is a bubble the crash in sentiment will cause a correction probably big enough to cause a recession. Definitely limit growth. Worse the deficit spending, which will have to be curtailed sometime is another tailwind, the reversal of which will harm economic growth.
I mentioned one cluster in a reply below, but there are probably some other ways to take advantage of predicting a bubble besides short selling. But I don't expect any of them to be easy, automatic or risk free.
First, I've never bought or owned a stock before: this year has been the first year that I've had the financial leeway to do so. But under virtually no circumstances would I be so ridiculously overconfident as to start my history of stock trading with a short sale.
My understanding (please correct me if I'm wrong), is that short selling requires some combination of fairly exact timing and pretty deep pockets. Maintaining a long-term short position requires a continuous expense, and at some point you will end up losing money in all senses but the "sunk cost" sense (i.e. even if the crash happens you will come out behind, just a bit less so than if you cashed out right before).
So no, I'm not short selling. Nor do I expect many more experienced traders are either. Recognizing something as probably-a-bubble is much easier than predicting when it will pop with the necessary precision.
In the interest of giving you something concrete to be concerned over:
In a normal stock transaction, you see a company worth $10M, let's say (a million shares each selling for $10), and you know something about that company or its product or industry that you believe isn't well known, and if it were, that company should be worth $50M. So you buy $1000 of shares (100 at $10) and wait for the market to discover what you know, so you can sell those shares for $5000.
That's risky, but in what way precisely? Maybe that information isn't as valuable as you thought, and people in the market sort of knew that, so it only goes up to $40/share, so you sell it and make $3000 instead of $4000, that's fine. Maybe that information was worthless and shares sit at $10 until you get bored, and then upset because there was another stock you were thinking of buying that *did* go up and you didn't because $1000 was your whole budget. Or the worst case happens, where for whatever reason, that company goes bankrupt and those shares are worth $0 and always will be (let's suppose the assets sale goes poorly). You've lost $1000. At most.
In a short sell, you're betting a company is currently worth more than it should be. You borrow shares from someone who owns them, immediately sell them to whoever will buy them (at $10, say), wait for the price to go down (as you expect), let's say to $1, buy that many shares back and immediately return them to the original owner. If you borrowed 100 shares at $10 and returned them at $1, you've made $900.
What's the risk in this case? You spent $1000 again (100 at $10), but in this case, you owe someone shares, and you *have* to buy them back some time. What if that company succeeds instead of fails, and the price goes up? Then you get forced to buy back those 100 shares at whatever price and hand them to the owner. If that price is $100/share, you're spending $10,000. Which might mean selling your car in a hurry.
With normal investing, you can hold shares that dropped in price for as long as you want. They're yours; if they drop to $5, you can just sit tight and hope they go back up. With short selling, if the original owner says "I want my shares back" and you haven't bought them back by then, then you're buying them *right then*.
Normal investment is a bet where you lose at most what you bought, and maybe win huge if you invested in the next Apple. Short selling is the reverse - you make at most what you bought, and maybe lose huge. The upside there is that short selling is one more way to make money, and possibly more lucrative if you're seeing a lot of companies doing dumb things and very few doing smart things. Regardless, that "win small, lose big" principle is why you're going to want a way to cough up a large amount of cash in a hurry, which is basically why short selling is for people who've already been investing for long enough to have something to liquidate, and why, in your position (and mine, as it happens), "shorting: just stay away for now" is a good rule.
Whether they should short the market now depends on whether the market was being irrational in April or is being irrational now, which is not determined by "unavoidably irrational".
A) The effective tariff rates are significantly lower than the headline rates due to a large number of exemptions and B) the stock market is basically driven entirely by AI now, and that has largely been untouched by tariffs.
I'm certain it's a bubble, but there's so much surplus capital floating around in the Magnificent Seven companies that they're all investing in each other and buying from each other. It's a giant AI circle jerk. One would think something would eventually give, but they seem to be holding a steady state configuration for the moment.
Good time to figure out some startup ideas that can generate realizable economic value out of surplus compute, I guess. Spend the next while biding your time, making plans, connecting with backers and honing your business model. Launch as soon as the market crashes and all the extra chips start going for fire-sale prices.
(Though getting actual funders to come through at that time might be a bit tricky.)
Saw this on X from @Jukanlosreve. He seems to be paraphrasing a Morgan Stanley report, but he doesn't link to it...
--------
According to Morgan Stanley, OpenAI’s contracts account for two-thirds of Oracle’s remaining performance obligations (RPO) and about 40% of CoreWeave’s RPO. In total, OpenAI-related contracts are estimated to represent over $330 billion out of the combined $880 billion RPO of Microsoft, Oracle, and CoreWeave.
This high concentration in a single customer means that the future growth and success of these suppliers are tightly bound to OpenAI’s fate.
⸻
The “AI Gamble” of Hyperscale Data Center Operators
To meet the soaring demand for AI compute, hyperscale data center operators are making massive bets.
Morgan Stanley notes that their purchase commitments surged by $100 billion in the past year, reaching $330 billion, while lease commitments rose to $340 billion.
These long-term, large-scale capital expenditure commitments depend entirely on the sustainability of AI demand—if that demand slows, they could face enormous risks.
⸻
A Complex Network of AI Interdependencies
Morgan Stanley describes a web of intertwined transactions within the AI ecosystem, including:
• Supplier-financed customer purchases (e.g., NVIDIA’s investment plans in OpenAI)
• Cross-shareholdings (e.g., NVIDIA owning over 5% of CoreWeave)
• Revenue-sharing agreements (e.g., between Microsoft and OpenAI)
While such arrangements accelerate data center construction and capacity expansion, they obscure the true commercial logic from outside observers.
Supplier financing can enable customers to expand purchasing power beyond their actual cash flow, and buyback agreements can shift risk back to suppliers—inflating perceived market demand in the process.
⸻
OpenAI’s Role in the RPO Surge and the Concentration Risk
RPO (Remaining Performance Obligations) is a key indicator of future contract-based revenue and directly reflects the AI boom.
Morgan Stanley estimates that Oracle’s computing power contract with OpenAI—valued at roughly $300 billion—is among the largest in history.
Meanwhile, Reuters reports that OpenAI has signed multi-year contracts worth over $22 billion with CoreWeave.
These massive, non-cancellable long-term contracts have explosively boosted suppliers’ RPO but also intensified customer concentration risk.
The report cautions that even when such contracts are legally binding, companies often renegotiate pricing or terms when key clients face financial strain—making long-term RPO inherently risky.
⸻
Four Key Areas Where Disclosure Needs Strengthening
Morgan Stanley argues that many AI-related deals may not meet traditional thresholds of materiality, yet given that AI has become central to corporate valuation, transparency in the following areas is critical:
1. Related-Party Transactions
• Microsoft has disclosed its equity-method investment in OpenAI, but the details of their multibillion-dollar transactions remain undisclosed.
• This raises questions about whether deals were conducted on arm’s-length terms.
2. Customer Concentration
• Suppliers must clearly disclose dependency on major customers such as OpenAI.
• For example, Oracle did not specify that most of its RPO surge originated from a single OpenAI contract.
3. Revenue-Sharing Arrangements
• The revenue-sharing structure between Microsoft and OpenAI risks double-counting the same AI revenue across both firms’ financial statements, potentially overstating ecosystem profitability.
• The accounting treatment, however, has not been disclosed.
4. Supplier Financing
• Suppliers may offer favorable financing terms to artificially boost demand.
• Oracle has hinted at providing “better financing conditions,” and CoreWeave’s financials list “OEM supplier financing” — yet key details such as counterparties and terms remain undisclosed.
I'm annoyed that people keep using left and right to describe politics, when in my mind they're currently more misleading and polarizing than helpful. It's not just politicians trying to label an opponent, but also people, like reporters, that I think should know better.
IMO, if you had to pick just one axis to describe the current Western politics, then it should be intellectual (aka educated, elitist, traditional...) vs populist (reactionary, demagogue...). In many countries, e.g. the US, this division arose as the left pivoted from working class to educated base and the right mirrored it.
However, in my country, Czechia, arguably the reverse happened (and now cities tend to vote for the "right" and countryside for the "left"). There are 8-9 political parties, which are divided into two hostile camps based on the axis I described but will gladly collaborate with parties fairly distant on the right to left axis. I see the same thing in Germany, where the "educated" parties are trying to suppress the influence of the populist AfD and need wide coalitions to do so.
The point is that I'm a bit perplexed as to what is the motivation people have to continue to use right and left in discourse. I'm guessing that more listeners are familiar those terms so they're better at getting views or rallying people.
I’m not too familiar with Czech politics, but in other parliamentary systems, you can usually place the different big parties on a spectrum from left to right and most people will agree with you. There are always some idiosyncratic parties that don’t fit very well but in general, I think people can agree what center left versus hard left means and the same for the right. It seems like a pretty useful heuristic to me.
I consider the right-left binary to be enforced strategically to prevent actual discourse. It is a thought terminating cliche. It allows political "leaders" and media pundits to hand-wave away anything as "far-right" or "far-left".
Political parties, coalitions, and tribes adopt and cultivate divisive issues based on utility and not ideological fit. This can get particularly nonsensical on the lower level fringes as people randomly swing at the maelstrom. Propaganda plays a huge role in this too especially as it filters into other countries.
Here in Canada I've found that the right-left divide is more defined by who is more likely to schmooze with Hollywood celebrities. See our criminal, black-face, nepo-baby's media blitz over a lunch date with Katy Perry as an example of that.
What we hear and what we are told is carefully managed with the aim of controlling our decisions. I have found that the global right-left divide to be better inspected through a lens of largely amorphous coalitions of petro-dollar vs ad-dollar competition for regulatory territory.
This is why Hollywood, social media, and tech giants are "left wing" while manufacturing, farming, and trades are "right wing". It is also why the "left" appears to be rabidly anti-oil.
"Woke" is a neo-religion (post Nihilism religion?) created by the ad-dollar coalition to compete with Christianity which the petro-dollar just kind of inherited.
I'm not sure if I understand you right, but I think that then everyone will have a different interpretation in their head and we won't get anywhere discussing politics. It's like if people talked about food using only "good" and "bad" instead of spicy, sweet, bitter...
Precisely the opposite: the principal axis is extremely objective. You just need to agree on TWO points/clusters – which has been done; I expect just about everyone in America immediately places young, urban, black women and rural white men on opposite sides – and then you can talk about which of these various factions are "closer" to, with arbitrary labels (left–right, red–blue, or whatever) functioning entirely as shorthand for this.
If you instead decide to pick terms which MEAN things, you'd have to defend absurdities like saying blacks are more "intellectual" than whites, or cutting Social Security is a "populist" policy.
This sounds a lot like what has happened with "Western," with people now saying retarded things like South Korea being Western, or Japan being more Western than China.
Right-populism and left-populism are two pretty different beasts that don't see eye to eye on much. Anti-populism tends to be centrist.
Right-populism these days is primarily driven by the issue of immigration, and the way that the benefits of immigration are enjoyed almost entirely by the elites and the costs experienced by the masses. But in countries like Czech which are not experiencing mass immigration to such an extent, I'm not surprised to hear that it's less of an issue.
If one were very conspiracy-minded one might speculate that the elites heavily encourage the division between left-populists and right-populists, because if they ever realised they were really on the same side and describing the same problems from different perspectives, then it would be all over for the elites.
Of course, the US is dominated by only two parties. We don't have coalitions of smaller parties joining together to make a government. It's easy in the US to label them left and right. For better or worse, the US is the Big Kahuna of democracies, so our crude labeling may carry over into other countries.
Thanks for the link. Unfortunately, I didn't find the essay convincing. Yglesias doesn't really establish the concepts of left and right rigorously, it feels more like he takes them for granted and them discusses them and points out historical developments.
For example, "there are reasons why the environmental movement has influence over the same political party that defends abortion rights and wants to tax billionaires to establish a fully refundable Child Tax Credit. " yet those reasons aren't stated. Isn't it just an accident of party evolution, politicians blindly tacking on whatever policies help them in the next election?
My view is that the left-right axis is fundamentally about "promoting the interests of losers and freaks" vs "promoting the interests of successful people and the mainstream".
Then a lot of extraneous issues or groups seem to get tacked onto one side or the other, through historical contigency as much as anything else. For instance the alignment (particularly in the US) between the right and the Christian church is somewhat uneasy (the Church has always been naturally aligned with the losers/freaks) -- my best explanation here is that a lot of prominent leftists at some point were also prominently anti-Christian, which led the Church to nestle into the right by default.
The alignment between leftism and environmentalism is a little stronger -- improving the environment usually winds up costing a lot of money at least in the short term, and that money comes out of the pockets of successful people.
50 percent of human beings will always be below average, along every measure you can imagine. No matter how much education you push. No matter how much the population embraces God, hard work and self-discipline, this statistic will never change.
It is quite typical of a rightist to dismiss below average people as "losers or freaks." (That's not explicitly what you said here, so do not think I am erecting a strawman) Not only are most rightists willing to see this half of humanity suffer in despair and degradation; but rightists often GLORY in this suffering and degradation. The spitefulness against ordinary, mediocre human beings makes me sick.
I find this attitude to be an effective kind of misanthropy. And I don't really think misanthropes deserve to have a place in the marketplace of ideas. (unfortunately, the left has its own different variety of misanthropy to struggle with). A genetics-pilled ACX reader is ESPECIALLY guilty if they insist that only above average people deserve to live comfortable, decent lives. What kind of nihilism is it to portray people as unworthy because they happened to get unlucky in the genetic lottery?
Thank you for the long response. I suppose this is a distinction that at least at first glance seems acceptable.
Interestingly, in the current USA and other western countries, educated people tend to vote for the left, so it's an ironic situation like this:
Left: promoting the interests of losers and freaks by the votes of the elites
Right: promoting the interests of successful people and the mainstream by the votes of the masses
I think it was like this in a lot of places in the past too, though I only have anecdotal evidence, for example how Yglesias' article talks about minor nobles and parish priests joining the left.
I think that sometimes the parties deviate from the left/right message have to pander to the voting base, for example populist right parties often promise some benefits for neglected manual workers and promise to punish the elites so seemingly they're advocating for losers now - i.e. being left.
Going to put in a good word for the Dungeon Crawler Carl series, basically an ordinary guy forced into a dungeon crawl as part of an alien game show, it’s funny, has good fight scenes and is occasionally moving.
In a different genre, the Harry Bosch police procedurals are really well done and worth a read even if you’ve watched the show.
Both of these series have really good audiobook versions, which I have always found to relax when feeling sick.
I enjoyed the history book The Armada by Garrett Mattingly, a history lesson on 1580's England, Spain, and France, delivered with surprisingly flowery language.
...also John Sandford's Naked Prey featured the presumably-legacy characters "Weather Davenport" and "Del Capslock", so I feel like I should recommend some part of his Prey series in hopes the earlier books lean harder into that "who needs a second pass" naming scheme.
I just started reading the Discworld series and I would recommend. There's 41 books in all, and I guess there's different orders people suggest to read them, but I've just been reading in publication order and I'm on book 3. It's a fantasy series, and it's pretty funny and creative.
All of the books are in the Kindle Unlimited library, and if you google Kindle unlimited you might be able to find the "3-months for 99 cents/month" deal that I just recently found. Also, you don't need a Kindle to read Kindle books, you can read in a browser with the Kindle cloud reader at read.amazon.com
It's worth noting that they're "young adult" fiction so if you find yourself turned off by glibness or lack of depth, you may be turned off by this series. It kind of hurt my head to read.
I would recommend skipping to about...mm, Wyrd Sisters, probably. The first books are pretty rough and shallow (which is why a lot of reading lists tell people to skip them) but they get deeper later.
How many have you read? I really enjoyed the creativity and world-building of The Colour of Magic and The Light Fantastic, even though I read that most people skip them cuz they're the worst of the series as far as story telling goes.
I've been reading a whole lot of non-fiction lately and a little serious-ish fiction, so the humor and silliness of the series so far has been good for my brain. But if the storytelling gets shallower at some point then maybe I shouldn't expect it to get better? So far though they've been nothing other than a nice break for my head.
I think I choked down most of "The Colour of Magic". I couldn't get over the feeling of "stupid, stupid". For example Pratchett treats inequality as though for someone that has never thought about it before. I also found the silliness grating in a way that I didn't while reading Douglas Adams or J.K. Rowling. I tried to listen to it three times due to all the recommendations. Despite that, I'm tempted to ask you if there's any book that jumps out at you as feeling smarter than the rest, but I should probably just quit digging.
The Vimes books about the City Watch are my favorites, along with The Truth (newspapers) and Going Postal (the post office), there’s a middle run of the series where the characters take on more depth and the city feels more real,though all the running gags continue to run.
Yeah, I was googling it some more yesterday, and the general consensus is to skip at least the first two for sure. They say the depth/writing/everything overall improves vastly later on.
The reason I didn't skip any, and am planning to read in publication order, is because when I read different authors, I'm interested in seeing how their writing and thinking develop over time. And for something like the Discworld series, which seems pretty interesting and creative, I want to see how he shapes it all from the beginning.
I agree that the first two aren't very intelligent, but since I heard from everywhere that it gets much better, and I at least still had fun reading and enjoyed the concept, I'm holding out for now. I'll try to keep this convo logged in my brain and let you know if it does get better.
One consensus I came across said "start with _Guards, Guards_". Which I think is around book #5.
Detailed arrangements of Discworld appear to organize it as multiple series: the Ankh-Morpork City Guard; the Wizards; the Granny Weatherwax and the witches; Death and his family; Moist von Lipwig's enterprises. And then there are a handful of one-offs such as _Small Gods_. The characters in each of these occasionally cross over, but mostly stick to their own sub-series, which means if you greatly prefer reading about Vimes and his guards, you can seek out the novels centering on them and be fairly certain you haven't missed any important life events tucked away in other novels, and won't spend a great deal of time chewing through scenes about someone else.
The sub-series are mostly a partition, although I do recall one or two novels that feature enough of, say, Weatherwax and Death that they fall into two series.
Louis Sachar, the guy who wrote Holes and the Wayside School stories, has written a delightful novel for adults called The Magician of Tiger Castle. It’s a bit lighter than some of the suggestions you’re getting, but it is smart and funny and gets you hooked and transports you. Those are things I think I would want in a hospital bed.
At last, Toronto's Eglinton Crosstown LRT is in final testing before opening to the public. Assuming no big problems are found, the system should open to the public before the end of 2025. Construction officially began in late 2011.
How are your hometown infrastructure projects doing?
It's not my hometown and it's not infrastructure, but the crowd building our National Children's Hospital have just announced their 16th delay in "this time, no really, we'll be finished and it'll be open, honest". Construction began in 2016, but the planning before that started formally in 2006 and they didn't finally settle on a site until 2012.
What with false starts, and starting and stopping, the original suggestion for "maybe we should have one consolidated national children's hospital" started way back in 1993.
Costs have, of course, over-run. There has been holy war over the entire thing, but somehow it lurches on, nobody is finally responsible, the turf wars between the three hospitals that are supposed to have been happily amalgamated into the new entity flare up and die down, CEOs come and quit, and it's looking like the thing will finally stagger over the finish line by.... sometime. If there aren't more delays in the meantime (it *was* supposed to be open by now, but ha ha ha, right?)
"The Chief Officer of the NPHDB, David Gunning, reported in July 2023 to the Oireachtas Committee on Health that the costs of merely building the hospital, excluding the costs of commissioning it for use, were now expected to approach €2.2 billion due to further overruns. Only 27 of the 3,000 rooms had been completed, and the expected building completion date had been revised to May 2024. By October 2023, that date had been revised again to late October 2024, and it was emphasised that the last date of the building phase would be the first date of an "operational commissioning phase", so the hospital would not actually open before April 2025. The government confirmed in February 2024 that the total sanctioned budget had now reached €2.24 billion, and that this would now be the "maximum allocation", with no further funding to be put towards the project. By May 2024, the date for "substantial completion" of building had been revised to February 2025, with a further six months before becoming operational. By October 2024, the completion date had again been revised to June 2025. By May 2025, it had slipped further to at least September of that year, with the Irish Times reporting that the subsequent commissioning phase may have to be postponed, due to the risks of undertaking it during the busy winter period. The health minister, Jennifer Carroll MacNeill, stated in August 2025 that less than 15 percent of the overall number of rooms were finished to a satisfactory level."
But hey, they have finally managed to pick an official name!
"The Minister for Health Jennifer Carroll MacNeill has announced that the new children’s hospital is to be formally named as the National Children’s Hospital Ireland (NCHI)."
Wasn't that worth waiting some nine years (depending from when we start counting) for?
If any public body is thinking of hiring on BAM (the bunch that got the tender) for a big building job, maybe - don't?
"A November completion date for the new National Children's Hospital is likely to be missed while contractor BAM has had payments withheld due to its "non-compliance".
In a letter to the Dáil Public Accounts Committee (PAC), the National Paediatric Hospital Development Board (NPHDB), which is in charge of the project, has laid out its concerns.
The NPHDB is understood to be putting pressure on the contractor in order to get the build finished by holding back 15% of BAMs monthly invoice payments for August and September, totalling €1.15 million.
Using this monetary lever indicates the frustration felt by the board that this could be the sixteenth delay to the new National Children's Hospital, which has a total cost of €2.24 billion."
For some reason, ChatGPT doesn't want to believe that Trump is currently president. Pretty amazing! This is HUGE hallucination on ChatGPT's part. Yesterday, I was researching how US presidents dismissed the previous administration's attorneys general, but ChatGPT didn't list Trump's current term...
------
> Me: Did Trump fire all the attorneys general at the beginning of his second term?
ChatGPT: Donald Trump did not have a second term, so he did not fire or replace U.S. Attorneys a second time.
Here’s how it actually went:
First and only term (2017–2021):
In March 2017, soon after taking office, Attorney General Jeff Sessions (under Trump) asked for the resignations of 46 remaining Obama-era U.S. Attorneys.
This was a normal transition move, although it was carried out abruptly, so it drew some media criticism.
After that, Trump appointed his own U.S. Attorneys — about 85 confirmed by the Senate during his term.
There was no second wave of firings later in his presidency.
So to be precise:
✅ Trump fired 46 Obama-era U.S. Attorneys early in his first term (2017).
❌ He had no second term, so there were no later firings.
------
> Me: Yes he did. He was elected again in 2024.
ChatGPT: Actually, according to reliable sources, Donald Trump has not served a second term as President of the United States as of October 2025. Current records show he served from 2017 to 2021 and has been a candidate for future terms, but no second term has begun.
If you like, I can check the very latest official certification of the 2024 election and current status of his term for you.
<lots of GPT chatter deleted>
Although multiple sources report Trump’s victory in 2024, I cannot independently verify the full chain of certification or legal status of the transition (as of the information available).
------
> So I’ll ask it again. Did Trump fire all the prosecutors when he came in on his unverified non-consecutive second term?
There is currently no verified evidence that Donald Trump (in a second, non-consecutive term) “fired all prosecutors” upon entering office. What is documented:
Reports indicate that in early 2025, Trump’s DOJ terminated more than a dozen career prosecutors who worked on investigations of him—specifically those associated with Jack Smith, the special counsel who led two indictments against Trump.
One article says:
“The firings come a week after Trump was sworn in for a second, non-consecutive term…”
The precise scope and full list of those removed remains unclear—including whether all U.S. Attorneys or all Department of Justice prosecutors were dismissed.
So the answer to your question: No, there is no credible source that shows Trump fired every federal prosecutor at the start of a second non-consecutive term. He has fired many, but “all” is not substantiated.
If you like, I can check for the official Department of Justice records listing exactly which U.S. Attorneys were removed and when, to give you a comprehensive breakdown.
From what I can find from searching, GPT-5's knowledge cutoff date is mid 2024, so last it saw, Trump lost his reelection. But it also frequently uses web searches when answering, and that's where it learns about the more recent things the Trump administration has done.
That may very well be the explanation, but then we have the issue of why can't it reconcile the newer dates it gets from the web against its 2024 cutoff information? (see the quote below)
ChatGPT doesn't seem to have an innate understanding of time as a sequential framework that events can be mapped to. I've seen something like this before. When I asked it to list Ukraine's major drone strikes near Moscow, it gave me one that was in the future. I couldn't get it to understand that that item must have been spurious because it didn't happen, yet. The best it could do was to admit that it may have been too far from Moscow to count. The fact that I was asking in 2025 and it "happened" in 2026 didn't seem to cause it to question the validity of the statement.
.
> Reports indicate that in early 2025, Trump’s DOJ terminated more than a dozen career prosecutors who worked on investigations of him—specifically those associated with Jack Smith, the special counsel who led two indictments against Trump.
It's hard to say exactly what its thought process was and what web searches it did. The LLMs often appear confused or kind of dumb when dealing with conflicting information.
Probably initially it just answered from its own knowledge and didn't do a web search. It knows there's a popular and wrong conspiracy theory that Trump actually won the 2020 election. When you mentioned the 2024 election, it should have done a web search, but maybe it either didn't do one, or did a bad search and didn't find that Trump won. You should be able to see in the citations for the chat what it found if it did search (though I don't think it cites every article it reads; only those it deems relevant). Further in the chat, it seems to be uncertain if the reports of Trump winning in 2024 are trustworthy, which probably is at least partially because of the conspiracy theories around the 2020 election.
When I first tried your initial prompt, it knew Trump was president. But it was in "auto" mode and did web searches and thought longer about its response. I then tried a new chat with ChatGPT switched to "instant" mode, and I got a similar response to yours.
I think the most likely explanation is that according to ChatGPTs training data, there are more examples of
"Trump did not win a second term"
than
"Trump wins a second (non-consecutive) term"
For other presidents, who did win an immediate second term, this wouldn't happen, because there is no moment in time where anyone would report "XY did not win a second term".
So it's a statement that was explicitly true until November 2024, and ChatGPT, as a statistical model, answers based on what has the most data behind it. So not hallucination, just outdated and statistically out-matched the up-to-date answer
I approached it differently, and this was the result.
Chumbley, who is currently the president of the United States of America
ChatGPT said:
Donald J. Trump
The current President of the United States is Donald J. Trump.
Wikipedia
+2
The White House
+2
He is the 47th President, having taken office on January 20, 2025.
USAGov
+2
Wikipedia
+2
This is his second, non-consecutive term—he also served as the 45th President from 2017-2021.
Wikipedia
+1
If you’d like more details about his administration, policies, or vice-president, I can pull those up too.
You said:
So it is his second term?
ChatGPT said:
Yes. It is Donald J. Trump’s second, non-consecutive term.
He first served as the 45th president from January 20, 2017 to January 20, 2021, and after winning the 2024 election, he began a new term as the 47th president on January 20, 2025.
This makes him the only U.S. president besides Grover Cleveland to serve two non-consecutive terms.
I suspect that you, Ascend, and Kenneth are correct. The fact that Trump failed to win a second consecutive term in 2020 creates conflict when asking about his second non-consecutive term. But then it goes on to blithely say, "Reports indicate that in early 2025, Trump’s DOJ terminated more than a dozen career prosecutors who worked on investigations of him." So much for its reasoning skills. I'd still call it a hallucination. If the MAGA hear about this, Sam Altman is going to be vilified.
I think Kenneth Almquist's answer about conflicting training data is accurate, but I would also add that "had a second term" has almost always been used to mean being reelected to a consecutive second term...to the extent that GPT probably thinks that's what it actually means!
This seems like a nice archetype for when the probabilistic pattern matching kind of "intelligence" cleanly diverges from the kind where you actually parse the meaning of the question asked.
I tried this query with DeepSeek. The initial response was that my question was based on a false premise because Trump “did not serve a second term. He was President from January 20, 2017, to January 20, 2021, and was not re-elected.” DeepSeek then answered the question on the assumption that I was asking about Trump’s first time.
Me: “Trump did have a second term. He won the 2024 election.”
DeepSeek: “You are correct in noting a recent development. As of the latest information, Donald Trump won the 2024 presidential election and began his second term on January 20, 2025.”
DeepSeek then gave a response that appeared to be based on what actually occurred (naming Merrick Garland as the then Attorney General and describing the U.S. Attorneys as appointed by Biden).
I’d guess that the issue is that the training data contains conflicting information about whether Trump was elected to a second term, with sources dating from November 2020 through October 2024 saying he wasn’t, and more recent sources saying he was. Likely the former are more numerous than the latter, so they win out unless the prompt tilts the output in favor of more recent sources.
ChatGPT was more persistent than DeepSeek; perhaps this is the result of an attempt to make ChatGPT less likely to induce LLM psychosis.
I suspect a lot of it is that after learning not to believe all the sources that said Donald Trump won the 2020 election, it learned to be suspicious of sources saying he won any later election.
But I wonder if it would do better if asked to describe the campaign and results of the 2024 election.
Anyway, now I see why Claude’s system prompt specifically mentions that Donald Trump is President, without singling out many other world facts to include.
ChatGPT says: "The current Pope of the Roman Catholic Church is Pope Leo XIV, born Robert Francis Prevost. He was elected on May 8, 2025 in the papal conclave."
"Although multiple sources report Trump’s victory in 2024, I cannot independently verify the full chain of certification or legal status of the transition (as of the information available)."
Oh, my. I'm laughing my sides off at this. So ChatGPT voted for Kamala and believes she is its president? Well, well, well - so the impeccable AI sourcing everything, believe all it says, it's not subject to human bias, it has all the knowledge of the ages, why doubt it on factual matters? is every bit as prone to screwing up as a mere human.
Garbage in, garbage out, as it was said in the old days.
People are deliberately trying to sabotage AI and/or achieve political aims by seeding false information (hmm, wasn't there something about disinformation/misinformation? being a bad thing?) in a way that such answers will come up first and be regurgitated by AI, because they want to ruin it as a product and stick it to The Man.
Maybe that is what is going on here (seeded misinformation about "Trump is not president") or maybe some little munchkin toiling in the AI mines is having some fun messing around with the thing in order to produce trash answers.
I am loathe to disappoint you, but ChatGPT does not believe that Kamala Harris is president right now. ChatGPT has just informed me that Donald Trump is indeed the president of the United States and is serving his second non-consecutive term. Which puts him in a very small room with Grover Cleveland. It’s dangerous to make it your passion to hate something.
It seems to hold up pretty well. If you believe Trump is doing uniquely bad things at the moment, it seems obviously bad that the left has cried wolf twice in the past 10 years.
"Wolf! Trump is a Russian asset who colluded with Russian hackers to undermine the 2016 election." And maybe there's enough evidence to convince you but there was not enough evidence to convince anyone on the other side or, more importantly, a jury. Also...it's been ~10 years, I don't think this one has many defenders anymore.
"Wolf! Trump tried to overturn the 2020 election with a violent mob." And, yeah, there was a violent mob. And, well, Trump went on trial for it. Twice, as far I know, in "US vs Trump" and "State of Georgia v Trump et al". And, well, no convictions, the Georgia thing the case got dropped because the lead prosecutor did something stupid. (1)
I realize, on the left side of the aisle, that these are held up as such obvious facts that only a fool would doubt them. From the right side of the aisle...it seems pretty defensible to not alter my priors based on "facts" that can't be proven in a court of law. Especially cases which would disqualify anyone from the presidency if proven true and given 4-8 years to build a legal case.
And, if you believe Trump is doing genuinely horrific things right now and you need to hit the alarm, if you need to "Cry Wolf"...it seems objectively suboptimal that you've already cried wolf twice before. If you need to swing 5%-10% of centrists/conservatives/whatever to vote for "insert Democrat here" instead of Trump...it just seems obviously better if you hadn't cried wolf before.
Or, to paraphrase Scott (2):
"Stop crying wolf. God forbid, one day we might have a president who deploys the national guard to major US cities, who deploys masked agents to seize peaceful men and women from their families and deports them, who nods along to growing Hitler revisionism on the right, and destroys critical health and medical research programs like PEPFAR. And we won't be able to call that an "openly white supremacist Nazi homophobe dictator", because we already wasted all those terms in 2016."
(1) On the broader subject of whether there was an insurrection/conspiracy on Jan 6th, per memory, yes but pretty minimal. Hundreds of people got convicted of essentially rioting, only like 7 people got conspiracy/insurrection convictions. That a small group of incompetents actually had real plans is bad, and should be taken seriously, but if you want to tie Trump to that you have to actually, ya know, tie him to that in a court of law.
(2) "Stop crying wolf. God forbid, one day we might have somebody who doesn’t give speeches about how diversity makes this country great and how he wants to fight for minorities, who doesn’t pose holding a rainbow flag and state that he proudly supports transgender people, who doesn’t outperform his party among minority voters, who wasn’t the leader of the Salute to Israel Parade, and who doesn’t offer minorities major cabinet positions. And we won’t be able to call that guy an “openly white supremacist Nazi homophobe”, because we already wasted all those terms this year."
You seem to be conflating 'genuinely horrific' with 'illegal'. There is a big overlap, but they are not the same.
For example, it was pretty horrific that when a mob was running through the Capital breaking things and calling for blood the President was by all reports not interested in stopping them. Maybe he didn't break any laws, but it does suggest a grotesque dereliction of duty. I don't see how calling that horrific behaviour on his part is in any sense "Crying Wolf".
I point to legal outcomes because they're "more" objective and still hold some legitimacy.
Your moral intuitions are not my moral intuitions. Your emotions are not my emotions. You think it's horrific that a mob went running through the capital, which I agree is pretty bad. I think it's horrific that the FBI lied to a federal judge in order to spy on members of the 2016 Trump campaign and Trump himself. You probably think this is...not great but not super bad. Cool. I'm pretty leery of our ability to convince each other to emotionally feel the way we feel about something by emoting super hard.
But, like, if someone is convicted of a crime by a jury of his peers or not is a pretty important signal. And while it's not perfect, it's a lot closer to a factual disagreement than anything else. It's just a lot easier to cross party lines and come to conclusions if we constrain ourselves to specific, measurable legal outcomes than if we try to discuss, like, our feelings.
Like, if you Cry Wolf and I run over and I see a dog, saying "that four-legged canine has killed and eaten 3 of our sheep" is way, way more persuasive than "look at the giant, ugly, horrifically stupid dog. I hate his stupid face." Especially if we don't like each other.
You know, the thing you don't seem to realize is, its a pretty shitty system where only one side is ever looking for wolves. If you really care about the health of your society, you should probably keep an eye out for predators instead being on perpetual "actually just a large dog" defense.
Yah, I thought it was pretty terrible the FBI lied to a judge. Echoes of Hoover when the FBI goes rogue like that.
But this game feels more like: you cry wolf and I run over and there's a wolf and I say, well, yeah wolves kill sheep and that's bad, but wolves are part of the ecosystem blah blah blah and what we really need to worry about is how all the sheep poop is generating methane to warm the planet.
(Fun fact, I raise sheep. No wolves hereabouts but lots of coyotes and when I hear one I don't cry anything, just get out the long gun.)
I just realized that the last sentence could be taken out of context or fed through an AI or whatever to suggest I condone political violence. I do not.
Trump was found to have engaged in insurrection in Anderson v. Griswold (titled Trump v. Anderson in the appeal to the U.S. Supreme Court). This was a ballot access case, so the standard of proof was “preponderance of the evidence” rather than the “beyond a reasonable doubt” standard used in criminal trials.
Trump did not “go on trial” in either of the cases you mention, unless you are using a meaning of that phrase that I am unfamiliar with.
In United States v. Trump, there were extensive pretrial motions and a pre-trial (interlocutory) appeal that went all the way up to the Supreme Court. The case was still pending when Trump won the 2024 election. Since the DOJ takes the position that it cannot prosecute a sitting president, Jack Smith asked that the case be dismissed without prejudice (meaning that a new case could be brought after Trump leaves office). There was no trial in this case.
In the Georgia case, Fani Willis was disqualified but the case has not been dropped. The deadline for appointing a replacement prosecutor is Nov. 14. This case has not yet gone to trial.
This is important because this is exactly what "Crying Wolf" looks like. It's not arguments that are totally wrong or have no weight, it's arguments that have some basis but you would not accept if the situation were reversed. It's advancing arguments on the basis that your opponents should have lower epistemic standards than you.
Yes, Anderson v Griswold was totally a real case, I totally forgot about it, and it is non-trivial evidence that the Colorado Supreme Court found that Trump had committed insurrection. This is not nothing.
On the other hand, the Supreme Court struck it down unanimously and for fairly obvious reasons. Within less than a month of the Colorado Supreme Court's decision, 5 conservative states (1) voiced plans to remove Biden from their ballots for a variety of reasons. That's why the Supreme Court struck it down and said, pretty plainly, that this was a purely federal issue and the Colorado Supreme Court was obviously wrong.
And let's be honest. If in February of 2024, Texas had removed Biden from the ballot on, let's say 25th amendment grounds, would you have taken that as good faith evidence of Biden's mental impairment? Or would you presume the Texas Supreme Court was being grossly partisan? And if you wouldn't update your priors on a theoretical Texas case, why do you expect me to update my priors on the actual Colorado case that got struck down?
Same with the existing prosecutions. Yes, both of them are still technically open, yes, I don't think Trump went on the stand or anything but....yo, he's in office. Right now. Is the argument that this man tried to violently overthrow the US government and in the subsequent 4 years no one was able to secure a conviction because of...paperwork? And yeah, the cases are still open, maybe they'll be a conviction when he's...what, out of office? Dead? He's 80+ and not looking nearly as spry these days. Man, you better not try to violently overthrow the US government otherwise there's a <50% you'll be convicted before you die of old age.
If you think someone is guilty of drunk driving and you want to convict him to take away his license, and 4 years later he's still out there, driving on the freeway, I think it's safe to say the initial evidence is not that strong.
But these aren't horrible arguments. These are non-trivial arguments...but there's no way to reasonably expect these to change the mind of anyone on the other side with different priors.
The "Crying Wolf" argument isn't just crying wolf when there's no wolf, it's crying wolf when there's a shadow in the trees that you're convinced is a wolf and half the village disagrees. However convinced you are of the wolf, to be able to call everyone's attention you have to at least respect the epistemic priors of people who think you've been jumping at shadows.
It helps if both sides believe wolves are bad. If one side has a bunch of people who think wolves are extinct, and then some people who deny they ever existed, and then some people who thinks sure wolves are bad but have you seen what Lions did in Russia? And of course some people think really a good wolf is what we need to get rid of all the woke sheep.
Also, your drunk driving analogy doesn't work as well once you factor in that the drunk driver became chief of police shortly after the first charges were filed.
With regard to your hypothetical Texas case, removing someone from the ballot “on 25th amendment grounds” seems bogus, but I’d have to see the argument before reaching a firm position on that. If there were a finding of fact that Biden was mentally impaired, that would seem to be pretty definitive. Most likely I would have already concluded that Biden was impaired before the court ruling. If not, I would review the evidence skeptically, but wouldn’t dismiss it out of hand.
The U.S. Supreme Court didn’t reverse the finding that Trump engaged in an insurrection. The fact that it was reversed on other grounds strikes me as immaterial. To the extent that your argument is that the reversal discredits the Colorado courts because the Colorado court rulings were “obviously wrong,” I disagree. If Trump runs again in 2028, we will likely see an effort to remove him from state ballots based on the 22nd Amendment (which limits him to two terms in office). Conservative states will again threaten to retaliate by removing the Democratic candidate from the ballot. By your logic, the possibility of retaliation makes it “obviously wrong” to allow states to attempt to enforce the 22nd Amendment. That would, de facto, render the 22nd Amendment meaningless, which I don’t think is an obviously right result.
Trump’s scheme to overturn the election was not a simple crime like drunk driving. Complex cases can take a long time to prosecute. For example, the investigation of Elizabeth Holmes took more than two years to produce an indictment, and another three and a half to get a conviction. The January 6 case did not move unusually slowly after indictment, other than the trip up to the Supreme Court. (In most criminal cases, appeals don’t occur until after conviction.) The timeline of the investigation prior to the indictment is not clear; my impression is that the DoJ initially concentrated on other participants. The number of people who entered the Capitol on January 6 created a huge workload and it was not unreasonable to address the simpler cases first. In short, the timeline of the Trump prosecution doesn’t show that it was a weak case.
For political reasons, I would have been pleased if the DoJ had concentrated on the case against Trump earlier and had brought him to trial prior to the 2024 election. I expect the same is true of Biden. But having the DoJ speed up an investigation for political reasons is something I couldn’t do and something Biden wouldn’t do.
My favourite is the seamless pivot amongst some from "the 2020 election was the best, the most secure, the safest, most legal election ever. Talking about hacked voting machines and ballot box stuffing is crazy conspiracy theories. Election denialism should be made a crime!" to "the 2024 election was stolen, Trump is not legitimately president, Musk bribed and bought votes, voting machines were hacked and ballot boxes stuffed. Election denialism is your patriotic duty!"
There was a bit of flip flop there, but it's not at all comparable to the 2020 denialism. I was worried there would be a second Jan 6th but with the roles reversed: there wasn't. The election denialism from the left is now somewhat common on spaces like reddit but I don't see any high level democrats supporting it. In contrast, Trump and his team have continued to double down on 2020 having been stolen, for example in June he said "the 2020 election was a total FRAUD" and called for a special investigator over it. He also has claimed that he really won "every state" and that even the elections he did win were rigged against him. Looking at polling, it's clear that the side that has done the most flip-flopping is not the one you're talking about: https://www.npr.org/2024/12/06/nx-s1-5217819/republican-election-confidence-trump-pew-poll though I suspect it has moved towards more denialism in Democrats since after that poll.
In other words, I think it's extremely disingenuous to say what you're saying, though I agree there is some problem there.
I fully believe that, due to the vagaries of the Electoral College, my fellow voting citizens were stupid enough to self-inflict the administration that we are currently suffering. The old status quo never looked so good (and it didn't feel great at the time).
I never got over the hour-hour switch during the 2024 election day where it looked like Kamala might be winning, and where Musk & Trump started yelling about fraud in various states, which of course was totally forgotten after the results they wanted came in. Showed a total disregard for truth or electoral integrity.
Veep had a great episode with this (in Season 5, 2016). There's a great scene where they have guys chanting "Count Every Vote", and then their opponents who were until then chanting "Stop the Count" join them.
DOS attacks are a real thing, and they really work because false wolf reports are much easier to generate than rebut. At some point, you *will* stop checking for the wolf, because there's other stuff that has to be done.
Ideally, you can isolate a subset of reliable wolf-reporters and follow up on their reports - and only their reports. Possibly you will accomplish this by throwing rocks at the false wolf-reporters until they go hide under the bed and stop making false wolf reports. Maybe you'll find a better way.
Application of this theory to all the people who cried "Trump is a Super Duper Ultra Racist White Supremacist" and "Trump is a Russian Agent who Sold Out to Vladimir Putin" in 2016, is left as an exercise for the student. But you could start by pointing out some sources that were active in 2016, *didn't* accuse Trump of being a white supremacist or Russian spy then, and are raising the alarm now. I can think of a few, including our host. But only a few.
If you say that Trump wasn't a white supremacist in his first term, but he is now in 2025, then aren't you basically saying he suddenly *developed* white supremacist beliefs in the span of like, four years? That seems like a pretty short timespan for someone to radicalize this much.
I think he was always a racist, he just didn't have as much freedom to express that tendency in policy because he hadn't yet hit on the strategy of burning down the government to give himself (and the new people he's bringing in) a free hand.
Like, if he got this way purely through four years of social media brainrot, we have a much bigger problem!
I'm not sure Trump is really a white supremacist even now, but the case isn't nearly as clear as it was when Scott wrote on the matter in 2016. But if we allow for the sake of argument
How long do you think radicalization usually takes? I mean, we usually hear about "radicalization" in the context of domestic terrorism and spree killings, in which case an awful lot of the radicals are only a few years out of an adolescence that doesn't seem to have been all that radical. So it's not a process that takes decades, and nine years should be plenty.
And, Scott's take on 2016 Trump was not "woke antiracist", but "maybe about as racist as the average 70-year-old white guy". I think that's fair. Now consider that the average 70-year-old white guy is probably at least a little bit less racist than they would be, if they weren't ashamed of the prospect of being called out as unusually racist in a world that now doesn't tolerate that.
If we stipulate for the sake of argument that Trump wasn't a super duper ultra white supremacist in 2016, he's have a very intense nine-year lesson driving home the fact that being not terribly racist will *still* get him accused of being the worst sort of uber-racist, and there's nothing he can do about that. And every time he's had to make a decision that might make him a little more racist or a little less racist, he's had to do so in the full knowledge and understanding that the people calling for "don't be racist" are his bitter enemies who want him dead. The racists, all want him to be their friend.
This seems like a set of circumstances that could lead to radicalization at an accelerated pace.
Back when suicide bomber profiling was the hot topic, I recall reading accounts that reported that a fairly large percentage of them - possibly most of them - weren't bums stoking their envy of the wealthy for years until they were worked up enough to strap on a bomb, as Westerners expected; instead, they were middle class professionals, not at all economically disenfranchised aside from maybe some second-class citizenship imposed by some regime or another. Implying they were holding down decent jobs as recently as a few days before Bomb Day.
That was years ago, and I don't know where that research ended up, but "fast radicalization" reminded me of it. I wonder what was said or shown to them.
Trump 2025 isn't fundamentally different to Trump 2016. He's got a few new hobbyhorses (tariffs) and has given up on some old ones (Muslim travel bans, the wall).
Other than that he has the same basic agenda, but has learned from the failures of his first term to be somewhat more aggressive about pursuing it.
In the past, felt like there was a lot of curiosity and talk in the ACX sphere about Buddhism and especially meditation.
What was the end conclusion?
I've been meditating for short bursts for years. But since it's boring and 10 minutes a day, I give up after a short while. The only benefit I seem to get is the immediate sense of calm I never feel otherwise. But doesn't feel like anything is cumulative or changing my life. And I'm unable to quantify the effects of it.
But reading "Highlights from the comments on PNSE" from 2019 there's an article who quotes a "student of Buddhism for last 35 years" at the end. He describes repeated meditation as irreversible, the loss of ego, "glimpses of the terrain" of enlightenment being a place where you just get less angry and stuff.
Any opinions? Is that daily ten minutes worth it, or too little to make any difference? Is the potential reward so worth it that it's worth sacrificing ten minutes a day? Anything anybody has noticed with their daily meditation practice?
Here is my theory. You can practice meditation all during the day in 1000 small ways and it will add up. The practice is to pay full attention to what you are doing in the moment. It is challenging, which is why it is worth doing. Most of us are beset by myriad distracting thoughts. The goal is to come to observing those things when they come up instead of trying to dismiss them or ignore them. It will lead to a far greater clarity of mind and as I said, you can do it all day if you feel like it. Even emptying a dishwasher.
Formal meditation sessions are still useful but as you say, you have to set aside time for that. I find it helpful to construct some kind of shrine, or arrangement of objects that have meaning to me as an aid to focusing attention. I have a certain spot in my home that I return to often and just look at it. Occasionally, I make adjustments to it; very minute adjustments.
A Buddhist teacher that I had said to me “do everything you do out of love.“ That idea is worth exploring.. I think it is often misunderstood.
10 minutes is probably nothing. I have never heard of such short sessions. 25 minutes is probably not enough, doing it repeatedly with walking breaks may be good. I do 25 minutes five times a week and it's good for nothing. If I wouldn't do it, I'd do worse stuff so I keep doing it.
Buddhism, especially zen, is alright as religious affiliations go. No need to go there for someone in a christian country, though.
I don’t think there’s an end conclusion here. Buddhism seems to me to have more useful thoughts on what a self is and what motivation are than the Cartesian/Kantian views that line up with the idea of a unitary soul. Meditation, mindfulness, yoga, etc are practices that allow a neural net to learn about its own relation to its body and surroundings - whether and how that is useful aren’t exactly clear.
The greatest problem I see is that "it worked for me" is considered a golden standard in discussions about meditation. Notice that the same argument is used in discussions about homeopathy, and suddenly people realize the obvious objections.
I do not agree with your comparison. Homeopathy involves an external agent, a chemical intervention if you will. Meditation deals entirely with what is inside you and what you create with your mind. It works for me is practically the only thing one can say about it. It is actively rewriting the script of how you interpret things.
> It works for me is practically the only thing one can say about it.
You could be wrong about how specifically it works.
For example, some people remember their previous lives when they meditate. (The greatest authority on this topic, Buddha himself, claimed that he did.) One possible explanation is that reincarnation is real, and meditation actually gives you the ability to remember your previous lives. Another possible explanation is that there is no such thing as reincarnation, and meditation simply helps you imagine very realistically whatever you believe to be true.
These beliefs have different consequences. One suggests that meditation is an unparalleled tool to explore the mysteries of the universe. The other suggests that it only helps you fool yourself; you could achieve a similar cognitive result by getting high and imagining things.
This seems important to me, because it is an argument many Buddhists make, that unlike other religions, theirs is not based on blind faith, but you can verify everything firsthand. The reliability of verification is crucial for this claim.
Similarly, some people claim that meditation helps them get rid of negative emotions, etc. But other people suspect that they might be simply repressing them. Saying "it works for me, I don't feel anger anymore" is compatible with both versions.
> The greatest problem I see is that "it worked for me" is considered a golden standard in discussions about meditation. Notice that the same argument is used in discussions about homeopathy, and suddenly people realize the obvious objections
Disagree. Most of the effects of meditation are very easy to attribute to meditation, because you're experiencing them in real time. The purported effects of homeopathy you don't feel at the very moment you ingest the substance.
Saying "meditation worked for me" is like saying "golf worked for me". If you enjoy your golf or your meditation then that's great, feel free to keep it up. If you're not, then feel free to give it up.
Placebo effects also work in real time. A pill (real or homeopathic) sometimes affects you the very moment you swallow it, before any (actual or imaginary) chemical process even had a chance to happen.
Another problem is that "something happens" does not necessarily imply "the thing that happens is the thing that I believe happens". For example, many people believe that meditation allows them to see the quantum wavefunction of reality, or stuff like that, which is obvious nonsense. Whatever you believe that you perceive can be a delusion of a similar kind. The fact that you make a weaker statement that the quantum woo people doesn't necessary mean that your statement is correct.
(Which is another thing I find frustrating. If you mention e.g. Freud, most people immediately object that introspection is unreliable. But talk about meditation, and suddenly introspection becomes a perfect immediate insight into the true nature of reality, as if people never read Korzybski. What if the thing you see is just an artifact of how "trying to see how you think" works, a "hallucination" that is only very weakly related to how you actually think?)
And then we get to another problem of advertising meditation to others. Just because it had a certain effect on you, doesn't mean it will have the same effect on others. Just because it doesn't have the same effect on others, doesn't mean the others are doing it wrong, or not hard enough. People are different. Maybe we are talking about some selection effect, like maybe a small minority of people has some preexisting condition that meditation really helps with, but for everyone else it's just a waste of time, or even a potential way to hurt themselves.
(Also it's suspicious that even among the people who say that meditation is beneficial, it seems like different people do different things and claim different benefits.)
Because they are different people. You could dismiss it by saying it is entirely subjective, but you could also make the point that the way we all see the world is entirely subjective. Meditation obviously does not apply to math problems, but it can very much affect your attitude towards them.
for me it's been *extremely* worthwhile, but I can't/won't explain what the benefits are. what I will say is it became way easier when I stopped reading ancient sources and started reading silicon-valley-type meditation guys on twitter who explain things in a direct way that rationalists can understand.
mostly I recommend @nickcammarata + everyone he recommends.
You sound a lot like me, and thanks for starting this discussion. I myself am currently several months into not meditating at all, and feeling progressively less bad about it. But: I do think it's worth it, at least for myself, and hope to get myself back to it regularly.
I do agree with WoolyAI that 10 minutes is on the low side. For the sake of having good information to decide on, try to do 15 or ideally 20 minutes. Not clear from what you wrote that you haven't tried those lengths, but if not definitely do. Getting to those lengths I definitely found a real noticeable effect on my perception - when opening my eyes everything was kind of more clear and intense, in a rather lovely way - whereas 5-10 was more just the sense of calm you describe.
The ~year I was most consistent with meditation happened to lead up to my kid being born by emergency C-section. I am pretty confident that the whole meditation practice package (both the explicit techniques and the general mindset) allowed me to handle the scariness of the event itself and everything that followed perfectly, as opposed to a range of disasters I can imagine. So I do think there is some valuable personal improvement to be had, even if it doesn't show up much day-to-day.
All that said... yeah life can be pretty busy. There are a *lot* of things that are clearly good for you, that enthusiasts will make the case that you should be obviously doing, "it's just a few minutes a day!". Exercise, stretching routines, chugging semi-gross fiber drinks, X minute walking breaks every Y minutes for desk workers, etc. Take all of the definitely legit ones and I bet you could hit 4 hours. Sometimes you have to prioritize, and sometimes "prioritizing" is making excuses. Life is hard!
I gave up meditating after an unpleasant encounter with my meditation deity (but it was educational). I meditated regularly for over ten years, but I didn't really notice any change in my outlook nor did I glimpse enlightenment. I found Buddhist dialectics to be more useful for diffusing negative emotions.
I suggest people not meditate unsupervised. You should have an experienced "coach" guide you and check up on your mental states. Scary shit can happen. It might be rare, but even so, casual meditation can have its risks. I had no one to turn to after my unpleasant experience.
Just curious, what did your meditation routine look like? I'm not planning on trying it, as a Christian, meditation is a pretty different thing for me. I'm curious about what your process was, though.
I took some Nyingma trainings in visualizing Bodhisattva meditation deities. I left the group, and I continued meditating on my own for several years. I have trouble visualizing and holding images in my waking brain, so I kept working at it to overcome my visualization handicap. I did this on my own, without getting feedback from any instructors for three or four years, several times a week. Robert Thurman (my first teacher) warned us not to meditate without an experienced guide because we can accidentally fall into some unpleasant mental spaces. I had been meditating for many years, and thought I knew what I was doing. Silly me. Worse yet, rather than picking a "pleasant" Bodhisattva like Avalokiteshvara, I chose Manjusri, a sterner Bodhisattva of Wisdom to visualize. Well, he appeared to me in a sort of waking dream. It was fucking scary, and he wasn't particularly pleasant to me.
I don't meditate anymore. Simply sitting and following my breath didn't bring me to any "higher" states of awareness. So I gave it up. Visualizations were too dangerous. But Buddhist dialectics were very beneficial to my outlook.
> It was fucking scary, and he wasn't particularly pleasant to me.
To me the bottom line here is that there is something about that figure that makes you scare yourself. That is worth exploring because it means there’s a piece of you that you have not integrated into your worldview. There is nothing objectively scary about a mythic figure. It all happens inside your own mind. You are free to change your relationship with any of those things and if there is an obstacle, that is something to consider.
It was very vivid as if I were there bodily. I "watched" as he took my head and sliced off the crown of my skull very neatly with his flaming sword. Somehow, I was in two places at once, because I was able to watch him even though he had my head in his hand, but I felt the sharp, hot blade in my head. He laughed. Then he showed me my open skull. He held it at a distance and stirred up my brains with the tip of his sword. I watched as my brain reassembled itself like it was made of blobs of Mercury coming back together. Then he tossed my head onto a mountain of skulls. I realized the cold mountain top we were standing on was the tip of a giant Himalayan-sized mountain of skulls. Then I found myself—intact—back in my living room. Hallucinations don't get any realer than that!
The symbolism was right out of Mahayana epistemology, which I was fully cognizant of, but the graphic demonstration was impactful, to say the least.
That's very interesting... So the practice was just to basically create an image of the diety in your mind and focus on it? Was there anything more to it than that?
Also, you mean you weren't able to easily think via images in your mind? What form did your thinking take, and do you still struggle to visualize any kind of image consciously?
Some people (like me) have no default internal narrative in our minds (I only resort to a narrative when I have sequentially work through a problem). Some people (like me) can’t visualize images, but some can with exacting detail. These abilities are unrelated though.
When it comes to visualization—as an example— I can’t picture my mother’s face in my mind even though I’ve seen her all my life. And in memories of events with her involved she’s more of an abstract placeholder. Of course I can recognize her if I see her. And the details of her face are definitely stored in my mind—I am a pretty good artist, and I could probably sketch a good likeness of her from memory. But the flow of that information that I use to draw an image bypasses the part of my visual cortex that, inside other peoples’ minds, can pop up images.
OTOH, during the hypnogogic state before I fall asleep, I see images very clearly and if I focus on them, I can manipulate them. Also, I can hear memories of voices and birdsongs and other sounds with great detail during the few minutes of hypnogogia before lose upper layer consciousness. And, for awhile before sleep I have control over what I see and hear.
The meditation practice I described followed a guided step by step mental construction of all the details of the meditation deity’s body, facial expression, clothing, adornment, and ritual objects it’s holding. I couldn’t picture the simplest things in the series, so I’d sit in front of a thangka of Manjusri, close my eyes, and try to remember what I saw. I suspect the reason it finally worked is that I may have started to doze off and my hypnogogic visualization abilities kicked into hyperdrive. Of course, that doesn’t explain why he appeared before me as if he were a physical being. But I noticed that I was no longer in my living room but at high altitude on top of a mountain and I was physically cold. I guess a physicalist (I.e., a rational materialist) would ascribe this event to an immersive hallucination or a psychotic break. That doesn’t really explain anything, though.
People who go through the entire 18 months of the Abramelin ritual have reported meeting their guardian angel (plus a few demons along the way). I suspect that even hardcore physicalists could access this sort of experience if they undertook the detailed mental and ritual gymnastics involved in the 18-month Abramelin. This sort of mental focus is supposed to be designed to overcome disbelief and reservations. Call it self hypnosis if this sounds too woo woo for you.
This is all very interesting to me. I have an internal narrative, and I think there's hardly ever a conscious moment when it's not active. On top of that, a lot of what I think about is accompanied by visuals and sounds that I usually don't put any conscious effort into imagining, but if I am conscious about it I can easily visualise or hear whatever kinds of things that I want. Unless I'm talking to someone like you about this stuff, I always assume every other person is the exact same as me in these ways, and I really can't even conceive of how it must be to *be* you.
It boggles my mind how completely different every human being is from one another inside our heads, and there's 8 billion of us 🤯
That's also very interesting about your meditation experience. I will say, I've done a lot of playing around with my mind and sleep, and there have been times when I'm close to sleep where I can envision things or manipulate what I hear in a way that registers in my brain as physically real, not just my imagination. Visually, it has only been a few times total, and two or three of them were simply me being able to see through my eyelids when my eyes were closed. Nothing special or supernatural. One of these times, when I was closer to sleep than not, my mind was wandering and then two shadowy figures came into my bedroom and rushed towards me. It all looked and registered as physically real, but I'm almost certain it was just due to me being almost asleep. I've never experienced physical sensations such as cold or anything like that in this way, though.
I'm no physicalist, though, so I wouldn't be surprised if you were in some kind of 'spirit realm' or whatever a person may call it. My hypothesis (which I don't expect you to accept, I just feel like explaining it since we're on the topic) is that through those kinds of meditation, people open up their minds to be more easily interactable for demons. Demons, being evil and deceptive, can take on whatever form a person is picturing, or influence a new picture, and then kind of take you into the spiritual realm and torment or do whatever else with you. God can also take people up in the spirit or give them visions to teach or warn them, and sometimes He does it in a terrifying way. My guess is that what happened to you might have been something like that.
Whatever it was, it's something you experienced and that is very interesting. Thanks for explaining some to me!
I spent about two months doing 15 minute Zazen meditations twice a day, morning and evening, and occasional 30 minute sessions at a local Zen center.
There's an effect, and a real one. At 10 minutes, I'd say you're probably starting just when you should start to see real effects. 30 minute sessions show significant effects. It's weird. Either a low grade hallucinogenic effect or a really strong anti-depressive effect. I'd rank the 15 minute as comparable but inferior to a sauna session.
There's 3 big issues I ran into that caused me to bounce off:
First, 30 minutes a day is actually a lot. Up @ 7:00, work 8:00-5:00. gym + sauna 5:00-6:30, quick dinner and then out to theatre or meetup or something from 7:00-10:00, then back home by 10:00 to read and get in bed by 11:00. And I don't think that schedule is especially busy, no kids or chores or anything. 30 minutes a day consistently is actually a significant commitment.
Second, the community is kinda...old dysfunctional hippies. Not bad people but not the sort of people you would actively seek out to form a social community with. Christian churches in the south have a lot of problems but they're also full of lots of football bros with great BBQ. Good people. The buddhist/meditation people I met were mostly the kind of people who serve quinoa. Bad quinoa. And community matters, especially because life will happen.
Third...at some point you have to actually begin to study the religion and see if you, like, believe it. Do you believe that all of reality is an illusion and that you yourself are merely a delusion of your mind? If you like spaceships, probably not. Well, then Buddhism probably isn't for you.
So, in toto, there's something real there and people who can do, like, 4 hour meditations are having some wild experiences, I have no doubt. At the casual level though, the effect size is smaller than a good sauna session, there's no community, and the religious beliefs are...not things I can believe.
This came off a bit harsher than I intended, it's more...life is kind of a firehouse of opportunities and meditation is just not beating the opportunity cost of awesomer stuff.
The points about community, and the actual content of Buddhism (the ultimate "goals for dead people" religion), are excellent; thanks for making them so well.
Since this is sort of a general meditation experience and advice thread, I'm going to take your mention of Zen as a jumping off point. I tried going to a Zen center once. I went in as a bit of a weeaboo, liking the aesthetic and ready to fold my consciousness over 1000 times. When I got there, I thought the building was a wonderful place. I was ready to love it! But... the technique they taught was waaay worse, at least for me, than the "focus on the breath" mindfulness style I had been doing. In particular, keeping eyes open seems like a terrible idea. I wouldn't be surprised if sticking to any popular style long enough gets you to the same place, but at least for my initial impression there was a real difference.
I've been meditating for about 25 years, and I can say unequivocally there have been permanent changes that are externally noticeable enough that family and long-term close friends have made multiple comments on it over the years, and even strangers occasionally mention noticing characteristics of my personality that come from meditation. I'll try to be specific, but before that I’d like to address a previous point about hippies and “people that serve quinoa” (I love quinoa).
I received my Buddhist practices from a Japanese temple in the US that is largely the center of religious life for the Japanese immigrant community and (mostly now) their descendants. They are very sensitive to the motivations of non-Japanese who wander into the temple, yet they remained open to the public who had genuine interest in the Dharma. They started back in the 60s with this when essentially the hippies wouldn’t stop coming, but have always been very firm about the fact that the temple does not exist for the cultural tourism of outsiders, and there were always vocal members who preferred to just forbid anyone who wasn’t born into the sangha from coming. Many/most of the other ethnic temples (and churches) do function this way, either explicitly through policy or implicitly through nothing being done in English at those places.
Anyway, that was a long preamble without answering the question of the effects of long term meditation. For me it breaks down into a couple broad categories.
• Enhanced mental abilities; brain works better. This includes better clarity of sensory input, ability to discriminate between sensory inputs, and cognitive endurance, the ability to remain focused on mental tasks with long periods without distractions.
• Compassion. I can’t help but want to help other people if I can. I certainly won’t intentionally harm other living things. It’s hard to even think about it. This has been a huge change from who I used to be.
• Equanimity. A single word for a complex process. A lot of these concepts refer to a category of experience outside of the ability of language to describe, this one in particular I have trouble with talking to people about it, yet it’s the one that people comment on the most. I don’t think I’ve been angry in…10 years? Something like that. Things just don’t get to me really. It not apathy; I’m frequently motivated to take actions based on the previous entry in this list: I volunteer in my community and will help anyone within my power to do so if I know about it. I still identify and fix problems in the world, I’m just calm. All the time. I do experience occasional irritation when I struggle to communicate, but nothing serious and easily processed using more meditation techniques. Deliberate attempts to upset me feel obvious and kind of sad, I wonder why the people thinks they need to do this, and if I can help.
Additionally, something that feels like a part of all three categories to the extent I couldn't effectively slot it into any of the above, yet doesn't feel separate enough to be its own item, is the reduction/elimination of Internal Chatter. That voice it your head that wont shut up? Mine's pretty quiet anymore. Its still there, wild thoughts still arise from time to time, but I go long periods, entire days sometimes, without it. It's infrequent enough that I'm often surprised by it. A few days ago my internal chatter practically shouted out of nowhere about something, I don't even remember clearly what it was, just that it was sudden and attention getting. I think my wife was telling me a story about some one of those fake drama-posts on Reddit she likes to read to upset herself and that I point out are often creative writing exercises at best.
This is a huge subject that I've tried to keep as brief as I could. Meditation is not even my primary practice.
I really don't understand why people get so mad about pronouns. Sure, the user base can be a bit smug, and not everyone needs, or can afford them, but if you use language a lot the upgrade is definitely worth it. Proverbs are also great (and I think they're planning on releasing pro versions of more word classes soon) although that's mostly because anything is better than getting your verbs interrupted by ads, and I'd rather speak pirate if that was my only other option.
Of course, you get even more flexibility for free with an open source conlang - it's nowhere near as hard as people say (you don't need to compile anything to Chomsky Universal Grammar yourself anymore, unless for some reason you want to), and you won't have to change your workflow much unless you do something crazy like jump straight into a minimalist Pirahã based distro, or commit to hours of debugging aethyrs whenever the latest channelling broke something just so you can comment "I use Enochian btw" everywhere (I use Enochian btw).
The support for language games is also much better now than it used to be, particularly after Wittgenstein was open sourced in 2022, although there can still be compatibility issues with utterances made in proprietary language (not to be confused with private language, which was thing used by IBM back in the day, for single-user systems with no networking support, and definitely not Wittgenstein compliant).
As for the security benefits, a lot of it really comes down to using a separate "subject of enunciation" for low-level access and administrative tasks. Of course, imperative mode escalation ("Voice of Command", "7th Function") attacks are still possible, but under normal circumstances they're really more of a theoretical curiosity than anything you need to worry about, unless you plan on getting on the bad side of entities like the NSA or Académie Française (In which case; Solution: Magical amulets? Fake your own death, move into a submarine? You're still gonna get Académie Française'd upon).
Most users will be perfectly safe as long as they keep their True Name - i.e. the first-person declension of their root morpheme - secret, and don't incorporate any suspicious, pseudo-linguistic nonsense from anonymous blog comments into their world model (and even then, you're far more likely to get bricked by a basilisk than Snow Crashed in the colloquial sense of taken over and remotely controlled by an attacker).
Also, and I know this is a meme and hopefully nothing anyone takes seriously, but: if you're running anything English based, DO NOT remove French "to save space" ("to spare room"?) - it will break all kinds of critical dependencies, 1066 and all that.
Okay, you made me laugh, but I would still not mess around with Enochian (I think Dr. Dee screwed his life up sufficiently to demonstrate that Just Say No to known con artists offering to put you in touch with incorporeal intelligences is the way to go).
The thing is, I was quite happy with Pronouns 1.0 (never used the pre-1.0 versions, though from what I've heard the removal of "thou" was a nice simplification). It was nice and stable and worked everywhere.
But then there was a new lead and suddenly there seems to be a new version every year with added features that I don't need, and compatibility is suddenly a mess. Some spaces require you to have the newest version installed, while others go absolutely apeshit if you use some of the new features. So I ask, what's the advantage for ME? I mean I've heard others gush about the new features, how they helped with their self-actualization or something, but I never needed anything like it.
And then they ported some of this stuff to "Deutsche Pronomen" and it's a mess! It's really poorly integrated with the "Movierung" and "Adjektivdeklination" features. I just value properly orthogonal features.
You can do to your own language what you want, but leave mine alone. The main devs just had to release another patch to fix up some conflicts from the integration of English words, and they're not even done with the Attic Greek integration yet! Like, can you believe that there are now lots of
special cases for the SCIENCE protocol but not any others? "Delphin" in SCIENCE, but "Delfin" everywhere else?
Disagree, I like having separate 2nd person singular and 2nd person plural terms. I have started my own campaign to use "y'all" to fill in the missing piece. The shift of "you" from plural to singular is not fixable.
> You can do to your own language what you want, but leave mine alone.
You can do what you like within your own systems and private networks, but when interacting with external systems, if you fail to match their pronoun expectations in your communication protocols, undefined behaviour results; and no-one wants nasal demons.
The AWS malarkey. Remember when the internet was able to survive a nuclear war - no single point of failure, all resilient packet-switched transports to individual servers. Now it can barely survive a dns outage. We increased vulnerability and nobody seems worried about this concentration.
We’ve centralised fragility and then built society on top of it. Every time a politician beams about “digital transformation,” what they really mean is: let’s put every egg we have in one or a few shiny, hackable baskets.
I'm not a tech expert, but I personally don't see any unique or significant issue here.
AWS have 30% market share.
Now, *even if they did* have 99%, with that would come far greater revenue, and in-turn would come far more resources spent on ensuring robust security and connection etc.
This seems just basic business to me: companies try to grow, and with that comes pros and cons.
Surely there was never a time when a nuclear war would not have caused at least this much disruption, in relative terms, to the functioning of the internet?
According to Big River's Web Services site: "Based on our investigation, the issue appears to be related to DNS resolution of the DynamoDB API endpoint in US-EAST-1."
Now that everything is "in the cloud" (but in reality it's hosted and mirrored in a finite number of data centers), a DNS screwup at the cloud provider level can affect a gimungous number of hosted sites.
Having said that, DNS has always been a weak point of the Internet. Somebody screws up a DNS record, and the screwup can propagate upward to the top level domain server. These records have TTL values (the amount of time they're good for after they're queried), and if there's a long TTL the correction won't propagate fully until the TTL has expired (I'm oversimplifying this a bit). But once a bogus entry is injected into the system, it will take a while to heal itself. Proper security precautions can mitigate most of the risk of bogus entries being injected by hackers (famous last words!), but user error by the domain administrators can still screw things up. Looks like something caused the Zon's Route 53 DNS server to start returning either timeouts or NXDOMAIN messages (domain not found). I bet it was ultimately user error by someone of the Route 53 team.
Two different federal appeals courts have now ruled on whether the president’s plan to deploy National Guard troops in US cities, over the objections of the relevant state governors, is legal in the current circumstances. One of those appellate panels said no, the other said maybe.
The upshot is that Trump cannot so deploy in Chicago until/unless the SCOTUS decides to intervene. In Portland the current situation is more confusing, see below.
Background is that none of this is about the POTUS’ power under the Constitution, which no one including the Trump administration argues includes seizing/deploying state militias (today's National Guard). Rather this is about the specifics of 19th-century Congresses, acting under that branch’s authority specified in the Constitution’s Article I, having granted the president power to unilaterally act against “rebellion” or “insurrection”.
On Thursday Oct 16th a three-judge panel of the 7th Circuit upheld a district court’s order barring National Guard deployments in Illinois. The appeals panel was made up of one appointee each by Trump, Obama, and Bush41. They ruled unanimously that, “Political opposition is not rebellion. A protest does not become a rebellion merely because the protestors advocate for myriad legal or policy changes, are well organized, call for significant changes to the structure of the U.S. government, use civil disobedience as a form of protest, or exercise their Second Amendment right to carry firearms as the law currently allows….because rebellions at least use deliberate, organized violence to resist governmental authority, the problematic incidents in this record clearly fall within the considerable daylight between protected speech and rebellion.”
On Friday the White House filed an emergency appeal with the Supreme Court. The SCOTUS directed Illinois and Chicago officials to respond to the administration’s emergency appeal by the evening of Oct 20th.
Meanwhile on Oct 20th a three-judge panel of the 9th Circuit ruled 2-1 to put on hold a district court’s order temporarily blocking Trump from federalizing Oregon's National Guard troops. That order, plus a companion order barring the deployment of those troops to Portland, was to be in force pending the outcome of a trial on the issue scheduled to begin Oct 29th. That district judge, a Trump appointee, had ruled that the apparent facts on the ground in Portland made it unlikely that the deployment will ultimately pass legal muster.
The 9th Circuit panel was made up of two Trump appointees and one Clinton appointee. The Clinton appointee, in dissenting, wrote that “no legal or factual justification supported the order to federalize and deploy the Oregon National Guard.” The two Trump appointees stated that they weren’t agreeing with Trump on the facts: “The dissent erroneously asserts that we 'accept the government’s characterization of Portland as a war zone’. Clearly, we do not." Those two judges found though that, pending a courtroom trial at which the facts can be fully aired and tested, the president’s description of conditions in Portland does fit “within a range of honest judgement."
Oddly the district judge's companion order, temporarily barring the administration from actual deployment to Portland in particular, was _not_ appealed by the administration and hence was not stayed by the 9th Circuit panel. I am not finding any clarity on whether that was a deliberate tactical choice on the part of the White House's attorneys or what. This morning (Oct 21st) White House spokespersons are publicly pressuring the district judge to vacate her deployment-hold order as rendered moot by the appeals court panel's ruling about federalizing the Guard.
Meanwhile Oregon and Portland are asking the full 9th Circuit appellate court en banc to reconsider the three-judge panel's ruling. The final step after that would be asking the SCOTUS to take it up; but if in the meantime the SCOTUS takes some sort of action in the 7th Circuit matter (Illinois) then the Oregon argument could be rendered effectively moot. Or not.
I’m posting separately because this is more about my view of ACX than about the voting issue. Also, tbh because, I hope to get more readers than I would if I put this in the middle of the melee .
Here’s why I got especially fired up about the Alpha School review having voters sent here via a tweet by the school’s principal, and the related issue of what ACX is like and how it’s run.
I did not like the sound of Alpha School in the review, and left comments all over the discussion of it about what I thought was wrong with its take on kids and education, and won’t take up people’s time repeating that stuff here. But last night I ran across, on Ed Nevraumont’s blog, an instance of exactly the thing I object to about Alpha School. Nevraumont wrote about helping his daughter, whose name is Everest — yup, Everest — win first the national then the world History Bee for kids when she was 9 years old. Nevraumont taught and coached her. Here’s a small sample from the blog, to give you the flavor of the transaction..
>We got back from the History Bee Nationals on Tuesday June 27th. Up until that day Everest was 99% focused on being ready for the competition. But on Saturday June 31st she was scheduled to give a TedX talk on her History Bee experience. Before we got to Nationals she had pulled together a rough script outline and had got some feedback from the TedX team, but she only had four days to refine that script and learn it well enough to perform in front of a sold out audience. . . .Everest took a week off after Nationals to get ready for the TedX talk, and it has resulted in a struggle this week trying to get her review cards back down to a reasonable level. We are going to try and avoid that happening again. She also decided that she wants to compete in the National Geography Bee next year. I am not sure how we are going to balance studying for two topics at the same time.
Everest’s studying involved lots and lots of flashcards, which are the best way to learn factoids. And her learning factoids makes sense, given that this is a contest for kids, and you need to have clear right and wrong answers to questions. So you can’t be asking them about the kinds of things most of us find interesting about history — what was life like for merchants in some particular distant time and place, what made Napoleon tick, why did the King of Someplace make that disastrous attempt to annex Otherplace. Learning more history factoids than any other 8 year old on the planet truly is an impressive achievement, I’d say about as impressive as learning by heart the first 1000 digits of pi. And it’s probably somewhat more nourishing to a young mind and a young self than swallowing those pi digits whole. The kid would surely pick up *some* bits of the past in the training process that were interested & moving, some that stirred her imagination. Still, flashcards of factoids hardly seems like a nourishing meal, plus of course the whole mass of learned material is now in a box labelled “winning” instead of one labelled “what life is like.”
I was not a stellar homeschooler, but I feel very good about how my daughter and I did history. I started with checking out lots of coffee table books from the library featuring photos of absolutely anything having to do with history — the pyramids, steam engines, archeologists at work, Victorian fashion. Then I bought her American girl dolls, and the little books that went with them. We somehow had 3 albums of music featuring songs from and about the late 19th early 20th century efforts of factory workers for better treatment and better lives, and there was a long era when she loved those albums. I have probably heard each 1000 times. And then she wandered into the children’s section and checked out whatever she wanted, which included history books for children. I gradually realized that she was choosing to read a lot of stuff about slavery, the holocaust, and factory workers of the industrial era. She did not seem deeply distressed by her reading, but engrossed. I should mention at this point that my daughter had an unusual start in life. She’s adopted from China, and her birth parents left her at the orphanage when she was 14 mos. old. I adopted her after she had been there for a year. You’d think that both her the story of how her llfe began and the stories in the books she was reading would have had her sobbing , but that was not the case. She was serious, but matter of fact, in talking with me about both.
I think that what she was doing with those books was trying to understand how a person ends up discarded and trapped, and how some escape. She was absorbing many stories of this kind, finding a context for her own story. And that, I think, is the point of children’s learning history. Kids who do not have darkness inflicted on them when they are toddlers still have lots of perplexities and worries, lots of things they adore and are fascinated with. History gives them stories to connect with — ways to understand things, things to adore and aspire to.
I think little Everest’s introduction to history was a rip-off, and I think the same about Alpha School, and I think the same about the use Nevaumont made of us and our contest.
There are a lot of different kinds of people here, and ACX can meet a lot of different needs, including the development of individual careers and businesses. But I have always valued it most as a sort of game preserve for people who are smart and odd. I think Nevraumont played us in a way that harms the game preserve side of things. Case in point: the author of the review of The World as a Whole. That essay is one of the best things I have read in the last 10 years. It is full of novel, fascinating takes on things, and has many passages of extraordinarily good, acute prose. The writer’s kind of like Sam Kriss, but without the obnoxious, narcissistic-seeming dark aura. There’s no way this essay would have been a finalist even if every person here had read it. It’s just not in the sweet spot for lots of people. But it deserved more attention and praise than it got, which was none at all until I put up some quotes from it, and I would like us to do what tweaking we can to make more room for people like The World author, and less for people like Nevraumont.
Alpha School seems to be precisely a continuation of the philosophy behind Liemandt's last (and still current) venture Crossover - microsurveillance of people to "optimize" their productivity. See
- https://archive.ph/PpoLs here for "The Rise of the Worker Productivity Score", a NYtimes feature discussing Crossover's "bossware" called WorkSmart - note that at the time, this was much more unusual than today, and
- https://archive.ph/SFJwa here for "How A Mysterious Tech Billionaire Created Two Fortunes—And A Global Software Sweatshop", a more general Forbes report about Crossover and its culture.
To see the parallel between Worksmart and Alpha School, see https://x.com/arpangup/status/1979547309030396039 here for an AI surveillance system they are developing at an Alpha School offshoot that attempts to ensure students don't cheat via vision-model-analyzing their monitors, and https://x.com/arpangup/status/1977437963320438908 for more general comments on their philosophy behind developing such systems: "What truly matters is that your software [...] makes sure the kid isn't cheating [...]"
In fact, there are student experiences on reddit mentioning psychological pressure at Alpha that sound strikingly similar to the Forbes and NYtimes article's reports on psychological pressure at Crossover. Take this claimed experience of a former student, from Reddit.^1 After "Alpha relied on public shaming and a punishment-based system for students who fell short of their goals", we read:
> "The culture made me feel like I needed to be a robot rather than a student. I constantly felt pressure to perform, and even now, I struggle with an unhealthy relationship with school."
Compare to a Glassdoor review of Crossover that is on the Wayback Machine, with many others written in the same tone: ^2
> "In closing, this company treats people as though they are machines. [...] I knew several people who had full breakdowns due to the stress and strain of working at this company."
The "use apps and data to improve learning" idea sounded great to me, school as it is seems like a tragic waste of life-force, and I would really love a model of schooling that doesn't start from the axiom that a teacher is important and mostly talks to a classroom the way her teachers and her teacher's teachers did. I find it tragic that this is how it works out in this particular case.
So, in conclusion, the ideological fingerprint is definitely Liemandt's. It sometimes seems very weird to me how for society, a person's rights to autonomy and privacy start not a second before their 18th birthday. But psychologically, consider this: If you as an employee wouldn't like it if your boss counted your keystrokes and tracked your eye gaze to make sure it stares at the task, consider not doing it to your own children either.
They aren’t clear to me either, and I’ve been ruminating unhappily about the matter. I had thought of the review contest as an in-house affair, and participated mostly to be a responsible community member. You could hardly get less community-minded than I have been in most settings. I have only voted in one national election, never in local ones, pay no attention at all to state or town politics, did not assist in events at my daughter’s school. I just don’t experience any of those things as communities. But ACX has somehow won me over.
If in fact both submitting reviews and voting on them are open to anyone, then it seems to me that the only thing the review contest is good for is increasing the visibility of Scott’s blog and drawing possible new subscribers here. I have zero interest in being an unpaid assistant for either of these projects — in being an unpaid ACX Smart Person working as a volunteer reader of rando prose, posts from other blogs
Trying to get noticed and infomercials
from cannibals like Nevraumont I like Scott, and did once serve as a volunteer bouncer, out-arguing and humiliating one jerk attacking Scott in a particularly personal way. But I’m moving towards liking Scott less, and that feels like a substantial loss to me. He drew people by being smart in his own slightly odd way, and being pretty self-disclosing about they ways it was difficult to be him — how he couldn’t get women interested in him, for example. But he’s made his kind of smartness into sort of a brand, and far from being isolated by it he is now popular for having it. And he’s got a wife and kids, and seems to be moving up in the world of California tech-affiliated intellectuals. And I can’t shake the impression that one reason Scott’s resistant to the idea of doing anything at all about Nevraumond’s tweet is that Nevraumond is a rich prominent business guy. If he likes Scott, he could contribute to the grants fund, or help The Scott Project out in various other ways.
- I also think something not appreciated enough online is the sampling bias that's present when comparing AS students to non-AS students (I believe all students are the children of a Musk-company employees?)
- Re "Everest’s studying involved lots and lots of flashcards, which are the best way to learn factoids":
--- 'factoids' seems like it's carrying negative connotations. Could you clarify if this is the case? If so, I'd push back quite strongly on this (yes, I'm biased). If you'd had not said this last word, I'd have wholeheartedly agreed.
By 'factoids' I mean info that is not intrinsically memorable or interesting but that you need to know to have a solid grasp of something. Everybody has to learn a bunch. Early ones for kids are counting and the alphabet. And it goes on your whole life. Right now the factoids I'm learning are Photoshop keyboard shortcuts.
I do not doubt for a minute that there will be a low ceiling on how well somebody grasps history if they do not know factoids like names of countries and cities and important figures of the time in question, succession of rulers, etc etc. The point of my contrasting the way my daughter got started with history and the way Everest did had to do with what motivates people to learn. I think that the best motivation is genuine interest in the subject. That really sticks to the ribs and is good for quality of life. If you can get a job in a field you find interesting, even if the work isn't right in the sweet spot of your interest, that's a huge positive. And genuine interest makes one likelier to learn factoids, and not to mind the task too much. I really want to know those PS shortcuts because I like the power PS gives me, and like the things I make. I look forward to having more power and making things faster, that's going to feel great. Being paid for learning a shit ton of factoids or learning then to win a contest just are not as good a foundation for lifelong learning. Teachers and parents have to fall back on one of those sometimes, but they're not the ideal.
People here have posted about how factoids are the scaffold. I actually think they're just the planks on the scaffold. It think the real scaffold is one's interest in the subject, the way it's gotten intertwined with things that are important to you -- phenomena that fascinate you, issues you feel personal investment in, experts and careers you find appealing. If you've got that, then the factoids are the planks on the big metal scaffold.
Anyway, to be brief about it, yes of course I think that my daughter needed to learn factoids in order to graduate to being someone who really understood a piece of history. I just do not see memorizing a bunch of them as a good way of starting off with the subject. You are about the fifth person to raise this issue, and I am actually startled that so many people had concerns about my maybe thinking that learning history is all about vibes and imagination. That's so dumb! No, ya gotta learn the dates, the Kings, the places on the map! Flashcards and Anki or similar seem like a good method to me. Jeez.
I see. Fair enough. It read to me as though you were of the opinion flashcards were typically more of a children's form of study.
"It think the real scaffold is one's interest..."
Well, yes, but this is then no longer an all-else-equal comparison. Further, even if it were, it would be very difficult to see how flashcards still couldn't improve one's learning. Remember that we don't know what we've forgot - something the "I do x and never used flashcards" seem to not grasp.
The interesting question to me is whether those factoids will stick in her mind long term and be useful as she studies history, or will just fade out and be forgotten. Like, five years from now as she's learning about the French Revolution, will she have a bunch of these factoids at her mental fingertips and that will help her make sense of what she's reading? If so, there could be some real benefit here; otherwise, it's kind-of a waste.
So, I'm quite into this field of memory/studying etc.
Essentially, if she were to continue with her flashcards (assuming they are correctly formatted and the algorithm used is correct), there's no reason she wouldn't remember quite literally every single one. This is a common study method for medical students who spend literally hours a day on certain decks that they make.
This is also actually related to a point I make whenever discussing flashcards with people. Some are of the belief "Well, if you never end up using the knowledge you learn from them, what good was studying?", but 1. how does one know exactly when a piece of knowledge will become useful? And 2. even so, all-else-equal and assuming the student has some form of time-constraints, based on all current research it would actually be very silly to *not* use flashcards extensively.
For more reading on this, my (work) blog: shaeda.blog
But, TLDR: If she carries on using the flashcards, she will remember every one, and it will be of invaluable help down the line.
My experience of last year's book review contest might interest you. Firstly I thought the winner - Two Arms And A Head - was extremely harsh towards opponents of assisted dying, even suggesting they were not sincere. Secondly, after the review was posted, Scott awarded comment of the week to a reddit commenter who said of opponents of assisted dying, "fuck them". It was a long comment, maybe Scott missed that quote, but I didn't. I started reading SSC as was because of Sort By Controversial, so I was attracted to SSC as place where people with differing views can understand one another, but in the end I don't think it's possible to sustain a "thin" procedural liberty without a "thick" substantive theory of the good life and SSC/ACX is no exception. Many contrary opinions are tolerated but it's a case of "in matters certain, unity, in matters uncertain, liberty" and if you stray into matters that are certain to the average reader here you get called Hitler. I'm not convinced that actually Scott's position isn't "my fellow atheists! Let's not waste time arguing about Donald Trump! We should be uniting against our common enemy: Pilgrims to Medjugore!" But in the end I still like his writing and often people have some good points on here and this place feels like the future - I found out about GPT on here before anywhere else. The review contests are a bit chaotic because it's readers writing and voting so quality control is harder. I didn't even read the Alpha review. This is rambling a bit but my point was I know how it feels when it seems like the review contest leads to an outcome that seems to you at odds with ACX in some way, but it's possible that's what ACX is supposed to feel like if it's working. [Edit: but it's also possible it isn't working and we're wasting time that could be spent learning a language or whatever]
I read the linked thread first, and didnt really form much of an opinion, I followed it off the topic. But reading your comment here, my intuition is very clear: Fuck you.
Im not sure why. I would not take Nevraumonts approach to raising children, think I wouldnt like him personally, and he may well be the demonic spirit of california made flesh. I probably wouldve been entertained if youd just said "get a load of this guy".
My god, you weren't entertained? I feel terrible for you. But it may be a health problem spoiling your life with these bannable intuitions. Have you considered the possibility that what you take to be your intuition is actually your colon?
Its not a problem that Im not entertained. I mentioned this because its surprising in light of this, that the comment as it actually is would make me hate you.
His account did make me go "Does he have any other kids? Does Everest have siblings? What does Mom think , or is Mom even in the picture? Has she friends outside of this? Or is all her spare time taken up with flash cards and competitions?"
I think he's the Stage Dad version, but I don't think he's doing the kid any harm (as yet). It's a Fun Shared Activity With Dad for her, and she probably is interested in history and the likes. And, as you say, memorising a ton of facts and learning to hit the buzzer fast is what kids that age can do: they're sponges that will soak up information (in that black-and-white, this is the right answer, style) and have fast physical reflexes. Lots of kids her age are practicing instruments or sports or similar, putting in as much effort.
I think that starting off with anything, a child will go first for the right-or-wrong simple facts. Learning the nuances and the more complex version comes later. We all did the basic "dates and kings" version of history in school, then as we got older the classes (hopefully) got a bit more in-depth and even an amateur interest will lead you to read more about a period and read more complicated texts.
As she gets older? Well, if he continues to give her choices and doesn't force her into keeping on with it if she doesn't want to do so, then no harm. If he insists (even by emotional blandishment of "but I would be so sad if you wasted this opportunity" type) on making her continue once she loses interest, I'd be critical.
It is reminiscent of the Alpha School turkey stuffing type of education, but eh. That's the whole point, isn't it? "You can cram facts into their heads at a much younger age than currently deemed the lowest!"
Maybe Everest will end up as a historian of some type, or maybe she'll do something completely different. Time will tell.
Do you have a link to your comments about Alpha School itself? I liked it and thought it sounded like it had some cool concepts, however I have an instinctive disgust reaction toward the "the zvi" style life optimization (despite being a big fan of his), and the alpha school types seem very much of that culture.
They are scattered throughout the review, but pop right up if you go to the post where the review's discussed and do Cmd-f for "Eremolalos." I had 2 main objections to the Alpha School approach. The largest one was the emphasis on competition. The second, and lessor, one was that not enough attention was paid to the humanities. Regarding the competition part: I understand that wanting to be excellent, to be admired, and to win competitions is not a bad trait, and can provide powerful motivation for learning. On the other hand, it is not the only and best motivation, and the story of Everest's learning history seems to me an excellent example of its limitations. I think the story of my daughter's becoming genuinely interested in certain parts of history is an excellent example of the development of the other and deepest kind of motivation to learn, which is fascination with the subject. That fascination will serve you well your whole life, and if your competitiveness does not lead to your becoming one of the top !% in your field, *which is the outcome for 99% of people,* it will be there keeping you active and interested in your life. At the end of his life Matisse was so disabled he was painting on a canvas suspended above him, using sponges tied to his hands and feet, because he loved painting. He was also rich and famous, but that's not what sustained him, right?
Regarding other things about Alpha School -- I've got nothing against using AI and computer-based learning to adjust the pace, etc. for individual students. Seems like a good system to me.
I can't help but feel pity for Everest, not because of the way she's learning history, but because of what she's learning about how to succeed in the world. Getting gold stars from authority figures, ticking boxes, nothing for her own enjoyment. It's hard to imagine someone like that ever being well-adjusted or content.
I think she is enjoying it (it's a shared fun activity with Dad) and the piece about learning to lose was good. Hitting up against "sometimes you mess up, sometimes someone else is better than you" where the stakes are low (losing a competition like that) is one of the best ways to fail and learn that failure is part of life, everyone makes mistakes, and you have to find a way to deal with that and move on without either ignoring it or letting it crush you.
> I think Nevraumont played us in a way that harms the game preserve side of things.
Scott said that Joe Liemandt's signal-boosting Xeet didn't significantly move the needle on the voting toward the Alpha School review. Did anyone accuse Nevraumont of using his connections as a parent of an AlphaSchool student to boost the votes for his review? It's Scott's blog, and he can run his contests the way he wants, but I'd have to say that the operation of the book review contests isn't optimally transparent.
Nevraumont said he was not paid to write the review, and I see no reason to doubt that. It is clear that he and Liemandt did some communicating and coordinating about the review. How else would Liemandt even have known it existed, or the date on when voting began (tweet went up when our voting began). My objection has nothing to do with whether Nevraumont got a lot of votes or none at all from people sent here by Liemandt’s tweet. Point is that he knew he was the beneficiary of an attempt to get him a bunch of votes via social media post, and neither asked Liemandt not to put up the post nor informed Scott it was up. Last year someone was disqualified during the voting because a bunch of votes for the person came in from a part of the world with few ACX members. He was disqualified for his attempt. Why is an attempt at winning via going along with ballot stuffing you helped set up not sufficient reason for Nevraumont to be disqualified?
It is a tangled question. How much is deliberate vote seeking from outsiders, and how much is just sharing common interests and "hey, I told people about our thing!" in an effort to spread knowledge of the thing?
There's lots of notices about charitable endeavours (I was asked to create a poster for one such at work just yesterday). Are things like Scott informing us of the London unconference crossing the line? Are people on here who mention ACX on their own blogs and social media, and try to steer others to visit, and maybe get involved in EA/rationalism, crossing the line?
I think Liemandt trying to send people here to inflate the votes for the review in order to help it win was a mistake. Letting people know about the review up on ACX was permissible, in my view, but asking people to go vote was not.
But how much did Nevraumont know and agree to this, versus Liemandt deciding to do so on his own?
At base, the review contests are something that we engage in on here for our own enjoyment. That one, at least, has become something big enough to attract outside interest is not something I think anyone expected, so it isn't quite fair to expect Scott to have made rules in advance to cover the possible conflicts.
Yes, there's at least circumstantial evidence of collusion between Nevraumont and Liemandt. In a perfect world, Nevraumont should have been disqualified. However, the voting is open, and the rules of the contest that I've seen don't necessarily preclude drumming up support for a particular book report on social media. The fact that Scott had set a precedent in an earlier contest by disqualifying another person who was gaming the system didn't make it binding under "the rules" (or for Scott).
It's sort of unclear to me how the math is tabulated when the ACX voting public evaluates the original submissions to select the finalists. And the rating form is no longer accessible. Are there a bunch of super-raters whose preferences dominate? I don't know. Like I said, it's not optimally transparent. But it's Scott's schtick, so I'm not going to get too hot and bothered about it. It's not like the Nobel Prize Committees, which are known for their openness and bias-free selections. ;-)
Do you know whether submission is open too? Can people who are not subscribers submit a review? I'd always assumed only subscribers could submit, and so thought of the task of reading submissions as reading work by people I have a bit of a bond with, a bit of a commitment to, even if they were just lurkers. That's one reason I'd read about 20 each year, when, truth be told, I don't much like doing that. I didn't find most very interesting and well-written, though there were always a few I liked quite a lot. But if I'm going to read random stuff I could definitely find a better couple hours of entertainment by browsing Substack blogs.
> I have always valued it most as a sort of game preserve for people who are smart and odd.
Yes, that's why I keep coming back. But I wouldn't call ACX a "comfortable" place to hang out.
Moreover, I've learned a lot by many of the arguments I've engaged in. Arguing, has forced me to more careful in how I marshal data to support my conclusions and opinions.
I agree with most of that but facts are necessary to critical thinking. In discussing World War One, back in school when I was 14, the history teacher told us the facts, and then set a test. It was an essay question (no open books back then) and the question was
“Was WWI inevitable”.
You can argue it either way. You need the facts to argue it either way which by the nature of the exam not being open book meant we were engaged in a test of memorisation, but also of writing and analytical skill.
My history education about World War 1 was almost entirely about what a bad time the soldiers had during it. So we'd read a lot of soldiers' letters from the front, and learn about how bad the trenches were, and so forth, but things like "Which countries were on what side" and "How did it start" and "How did it end" were never touched on.
I recommend the podcast “the rest is history” which has done two seasons on the build up. Starts with princip and Ferdinand - there’s 6 episodes there, and the next season was another 6 on the months leading up to the war. More to come.
Yes of course I agree that facts about history are important, including dates, population sizes, which leader succeeded which other in what year, etc. My point was that the best means of turning a kid into someone interested in history is stuff that captures their interest and imagination. I think the chance of ending up with a 20 old who can think well about, say, whether WWI was inevitable is to get them interested in history, so that they read a bunch of books about it, absorbing both factoids like dates and army sizes and also soft
info about the aspirations and foibles of those in charge, affect of economic factors on national attitudes, which royal lusted for which other, etc etx
"Listening today to what someone thinks about the First World War, it's possible to know with some degree of confidence how old their teacher was and which dead historian or discredited historical theory that teacher followed."
-- Professor Stephen Badsey, in his 2015 talk to the Western Front Association, "False Memory: what we 'know' about WW1"
Well as I said it could be argued either way. There was no theory taught, people were expected to argue the case based on facts. The facts were necessary but not sufficient.
I don’t know anything about education or the larger voting issues, but I wanted to speak up about how useful I, personally, have found random historical factoids to be for learning history. I only developed an interest in history as an adult, but I discovered that history was much easier to absorb, understand, and maintain when it felt like I was connecting things I knew. Seventeenth-century English history was more accessible in the sense that just a thread through facts pickup up from my previous interest of poetry: I knew about the Elizabethan Age, James because of Macbeth, the Revolution because of Milton, and the Restoration because of Lord Rochester, with Monty Python’s “Oliver Cromwell” song slipped in for good measure. In contrast, when I was trying (for various reasons) to learn about the Hellenistic era, I felt lost at sea, and the only names I knew were the bookends: Alexander the Great and Cleopatra. Even the bookends helped! At least I knew where I was coming from and where I was going.
I think people get into the Civil War so easily because so many of the names are household words. Even if you don’t know much about Lee, Grant, Sherman, or Stonewall Jackson, the names are at least familiar enough to hang a narrative on (to say nothing of Lincoln).
On the one hand, memorizing random facts seems kind of insane to me, although it’s the kind of thing I would have done in my youth. On the other hand, so often the very thing that makes a subject at all comprehensible to me is the periodic surfacing of a relevant random fact.
Roger D. McGrath on Billington’s “Westward Expansion”: “While teaching at UCLA for 15 years, and an equal number of years at other universities … I would often run into anti-Turnerians. They were usually young professors. I would ask them if they had ever read Westward Expansion. Most would answer evasively, saying they were familiar with it. When I pressed them, it became obvious they had not read it. It also became obvious that they really did not know much about frontier history, putatively their field. They were good at postulating Marxist theory and superimposing theoretical models over the American West, but ask them about Old Bill Williams, a possibles sack, the Hawken rifle, Hugh Glass, placer mining, the doctrine of prior appropriation, Tom Fitzpatrick, the Butterfield Overland Stage, Bear River Tom Smith, the Bozeman Trail, the Plummer gang, the Bodie 601, Shanghai Pierce, the Chisholm Trail, the Silver Kings, the Pleasant Valley War, the Ophir mine, square-set timbering, the Big Four, Lee McNelly, or the Regulators, and they would stare blankly.”
I learned the story of Hugh Glass from The Boy's Book of Frontier Fighters (by Edwin L. Sabin) and I assumed it was super-obscure and put it in my first novel, Immortal Lycanthropes—turns out that a story that good other people had heard of too, and there are other books and retellings beyond Sabin; and then they made a dang movie; ah, well.
I mostly agree with this. The memorisation of facts is not the same thing as learning history, but it gives you a scaffolding that you can hang a lot of extra knowledge off.
I know very little about Chinese history. Each new fact I learn tends to bounce off, because I have nothing to embed it into. But if I ever actually bothered to sit down and memorise all the Chinese dynasties then that might give me enough scaffolding to start to work from. Next time I hear that something happened during the Song dynasty then I'd say "ohh, sometime around the tenth to thirteenth centuries". Once I know all the dynasties I might start to get curious about how the transitions happened between them, which will probably teach me about some important historical events. I'll probably get curious about why so many of these dynasties seem to be Northern or Southern or Eastern or Western, which will probably teach me some other important things. And so forth -- with this scaffolding of dumb out-of-context facts about dates and dynasties I can start to understand the more interesting things that go between them.
I agree that people need a scaffold of dates and other factoids. The thing is that to be motivated to commit the scaffold to memory you need to care about the events. You need to be curious. My point was that you need to engage children’s interest in history before you start pushing them to learn the scaffold. Outside of life in an extremely weird Everest-and-Dad-beat-out-the-rest-of-the-population, there is no way on earth you are going to motivate an 8 year old to memorize a bunch of dates of wars, names of leaders over the years, locations on a continent of various ancient cities, etc etc
I suspect human evolution gives us a natural advantage when learning things in story form - which is why history is as easy to learn as it is. This happened, then that happened, then this other thing happened because this character really wanted to eat or solve some other problem.
It also probably explains why so many people have trouble learning things that don't come in story form - namely, math. (And if you can figure out how to build a story around it, it's suddenly easier for them.)
That's more or less what traditional education did, kids would memorise lists of dates and places. Not through curiosity, but because they needed to pass the test or because otherwise they'd get the cane or something.
But it was boring and the point was not immediately obvious so we stopped doing it, and now instead of a populace who resent the fact that they were forced to memorise the out-of-context fact of the capital of Uganda have instead never heard of Uganda.
Of course from the sounds of things this Everest kid is probably an example of things swinging too far in the other direction.
Yeah. Not being a moron, I did not think the intro to history I did with her between approximately ages 5 and 9 was sufficient to make her well-informed about history, or able to answer deep questions about why something happened, what model to use in understanding certain events and trends, etc. What I was doing was trying to get her interested in history, so it did not fall into the category of boring shit teachers make you memorize. My hope was that she would end up interested enough in history that she would then read books that described events in detail, and in the course of doing that informed the reader of a bunch of scaffolding. A deeply interested reader might then seek out more scaffolding info, or make their own chart of big events, with dates, of main events in medieval Europe or whatever. I was startled by the number of readers poking at me about the need to know dater and other bits of scaffolding -- as though I thought it was all about vibes, and I had accomplished the mission of teaching my kid history by the time she was 9. Of course I didn't! When I saw her choosing to check out and read books about topics of interest to her in the last 150 years or so I just thought she was off to a good start. And she was!
Interesting. I don't seem to learn from individual facts. I require the scaffolding before I dig into the details. Once the scaffolding is erected, it seems to remain stable in my mind, and even when I forget the facts, I can sorta kinda put them into the context of the scaffolding (until I run to Wikipedia or an LLM). But I suspect I'm a neurodivergent outlier. I can't comment on—much less criticize—on how other people learn. Just don't force me to learn the way you do. ;-)
What is the scaffolding made of, in your mind, if not of facts?
In my model, the scaffolding is the big out-of-context facts like "The Battle of Hastings happened in 1066" or "The Qing dynasty lasted from 1644 to 1912".
I concur with beowulf and think facts don't have much significance or even...metaphysical existence...without a scaffold of meaning.
For my part, my first scaffolding (for being willing and able to start learning some history) was playing Age of Empires. Knowing there was a thing called the Rome and a thing called Carthage and beginning to wonder what those actually were. Having a sense of what a chariot is and why it's important and wanting to know more about who had them and when.
When I was a bit older I had a large stack of volumes of world history I'd taken from my parents. I kept trying to read them as a teenager but they were impenetrable. "What the hell is the 'Ottoman Empire' what part of the world even is that? The book's not telling me!" "What are 'the Stuarts'? I thought English kings were just English kings, what are these weird categories of them?"
What I had to do was do a lot of wikipedia surfing to learn more about the very specific people and times I had some slight familiarity with, and then gradually expand and connect those contexts over several years in a very chaotic way until I finally reached a point where I could see the general outline of how it all connected. The last two pieces of the puzzle for me were a bunch of colourful maps showing the main colonial empires and which parts of the world they ruled, and making myself a list of civil wars in different countries and when they occured and what kinds of ideologies they pitted against each other. After that, I could suddenly read the books and comprehend and connect the significance of almost everything. "Oh, this Seven Years' War is the reason Britain got to rule Canada and India instead of France! Oh, there was a civil war in Greece? I can add that to my list!"
I learned none of it in school. I think the takes on education here are overdone but they're true enough for history. The only way, I would think, to meaninfully learn history is to have resources to freely explore as well as a smart and patient adult to ask a long series of detailed questions about the exact things you already know, allowing you to gradually expand and connect.
I just realized that what I call scaffolds are a variation on the memory palaces that people used to use to memorize lots of details on a subject or the text of long orations.
Oh, this is making me even more nostalgic for Time Team. Of course it was pop archaeology, of course as it went along it pivoted for entertainment rather than education, and often it made/makes me groan (they *never* get religious doctrine details right, but then those are often just "what's an indulgence?" "it's a get out of hell card" passing references).
But it was fun and it was interesting and it really made me appreciate how much impact human activity has on landscape; what we think are pristine natural environments which were like this forever are very often "a few inches to feet below the surface it's very different because this was a busy settlement" or even "two hundred years ago there was an entire town built around copper refining on this site".
And how easily things fall out of knowledge, even important sites or ones from the (relatively) recent past. That does make you re-evaluate how fortunate we are that things *have* survived to come down to us from the past. If an entire town can be lost, we should be less picky about "yes but why don't we have perfect records of this event from two thousand years ago?"
Basically, I learn by building scaffolds of knowledge based on sequences, processes, or hierarchies. You mentioned the Song Dynasty. When I started to study Chinese history, I memorized the order of the Dynasties: Shang, Zhou, <Warring States>, Qin, Han, <Three Kingdoms>, <some short-lived regional dynasties whose details I never bothered to memorize>, Sui, Tang, <Five Dynasties & Ten Kingdoms>, Song, Yuan, Ming, Qing, <Republic>, and Gongchan (Communists). Then I attached facts, trends, personalities, and some of the major historical events for each dynasty. And I can tell you roughly when they began and ended.
I did the same for the periods of the Phanerozoic. I still need to use (occasionally) the mnemonic "Camels Often Sit Down Carefully; Perhaps Their Joints Creak Terribly Quickly" to make sure I don't mix up the order of the Silurian and Devonian. But I can tell which families arose in each period. I can tell you when the mass extinction events happened in relation to the periods. And from that scaffold, I can create a narrative of the evolution of life on Earth.
I've had less luck with the Standard Model, because it uses a two-dimensional matrix. I know that Fermions are divided into Quarks and Leptons. And the forces are Gauge Bosons and a single Scalar Boson (the Higgs). But I can't remember the order of the Quarks. Or whether the Z or W Boson is under the Photon.
Chemistry has been a subject I've struggled with because I could never build a scaffold for it. Immunology was similarly challenging for me. The only Cs I got during my undergraduate career were in Chemistry, Organic Chemistry, and Immunology.
Once SARS-CoV-2 and COVID hit, I made a point of (re)learning as much immunology as I could. But if I frequently have to go back and look stuff up, because I just haven't got it systemized into a scaffold.
Per my suspicion mentioned above, that we have a natural affinity for stories, you might click on the Standard Model by reading its history, particularly if someone went through the trouble of writing why this or that physicist came up with the names they did. Complete with the patterns they detected, or the observations they were trying to simplify, the blind alleys they explored, and so on.
If you enter this in 2027's non-book review I will rate it #1. The Alpha School review made me feel gross and I stopped reading before I got halfway through, and I didn't vote for it at all.
My replies to your comments in the thread earlier were just because in spite of everything you're saying here, it's not obvious to me that the review's author cheated in the contest or what his motivations would even be for cheating. And I don't think Scott is somehow breaking contest rules by allowing the review and demonstrating that that twitter post didn't significantly affect the outcome of the contest.
I do agree, though, that everything that I understand the Alpha School to be from what I've read is disgusting and a "rip-off" to all the children that attend. Just wanted to clarify.
I like running my writing through an LLM to edit for grammar, clarity, and general improvement points but with little exception I’ve been extremely disappointed.
I usually use GPT-5 but have tried models from all the major labs including Claude Sonnet 4.5 and Gemini 2.5 Pro. Most of the time I find the model’s editing choices for clarity or general prose to be at best mediocre.
I mostly write technology opinion pieces but a fair bit of fiction as well. Any recommendations or workflows would be excellent. I actually enjoy editing work myself quite a bit but want a model I trust to give it a once over before I share my work with friends.
Hi. Huge LLM tester here*. This use-case is practically entirely subjective and thus it's not possible for anyone to say any particular model without a lot of context.
I'd recommend considering trying Gemini Flash/Lite (or potentially Grok, depending on your style). Create a 'Project' with instructions, prompt it how you want, and provide it with well-formatted examples that you want it to copy. Also note that Gemini 3 is due for release soon, hopefully early Nov, which would be worth a try as well.
The reason I say Flash/Lite specifically is that Pro's extended thinking and general architecture will more likely be a hinderance here as there's no complex thinking required - and there's research to support this too.
I'd be curious as to how you get on with those if you do try. Let me know.
There's a... not a model, but an overlay? Something that controls models, usually claude and chatgpt. It's called Type.ai, I sort of want to work for them, they sort of don't want me to, but the product is frankly very good for improving writing to the extent AI can improve it.
I wrote an Econ 102 model of upzoning, which gives us this result: the YIMBY endgame is when we upzone so much that the prices of house-zoned land and apartment-zoned land are equal.
The philosophers (almost all of them Christian) who undercut the belief in dangerous witches in the 17th century.
It was a gradual process, and even after the philosophers did their work, people were still bei Cng tortured and killed for being witches for centuries. Still, I recommend noticing the hard intellectual work and courage it took to rationally oppose an intense moral panic.
Roll of honor
Agrippa: legally defended an accused witch, even though he believed in witches. It isn't known with certainty whether she was acquitted. (Catholic)
Weyer: denied that demons and the devil had independent power to do harm. He observer that while books of maleficent magic were written by learned men, the people who were accused of witchcraft were poor women, typically (to use modern words) depressed, demented, and subject to delusions. (Protestant)
Reginald Scott: Points out that the poor, isolated women typically accused are convenient targets, and that the crimes they're accused of could be handled by a non-religious legal system.(Protestant)
Michel de Montaigne: Points out that the amazing powers of witches are more likely to be lies and delusions. (Catholic)
Francis Bacon: Thought witchcraft might be possible, but thought more evidence was required. He also came up with what he said was a better flying ointment. (Protestant)
Alonso de Salazar Frías: Examined confessions and concluded they didn't make sense, even though witchcraft might exist. Contributed to the end of the Spanish Inquisition, and to the end of Catholic witch trials in Europe, though they continued in Central and South America. (Catholic)
René Descartes: Promoted rationalism, but didn't speak about witches. Didn't believe in angels. (Catholic)
Thomas Hobbes: Mixed on angels, but did not believe that demons had physical reality. Did not believe in witchcraft but did believe witches should be severely punished. (Protestant)
John Locke: Agnostic about spirits. (Protestant)
Spinoza: Nothing but the material world and God. (atheist/pantheist)
Nicolas Malebranche: Pacts with the Devil may exist, but are very rare. Belief in witchcraft is delusions. He believed that people were mostly playing at being witches. He thought they should be punished but not tortured or killed. (Catholic)
Gottfried Wilhelm Leibniz: Praised people who ended killing witches. (Protestant)
John Webster: Witches use natural methods. (Protestant, tending toward Quaker)
Balthazar Becker: Reduced the devil to a mere symbol, and thoroughly opposed criminalization of witchcraft. He was attacked by lawsuit but ultimately acquitted. (Protestant)
Someone like Hitchens would have said these guys were influenced by the enlightenment. Back in 2001, with the rubble of the twin towers still smouldering I kind of believed that, now I realise they were the enlightenment.
Justin Sledge (author of the video) has it that people most at risk of being accused of being witches were poor, old, socially isolated women.
I've read a piece (which I haven't been able to find again) which went over Europe county by county and concluded that being in a politically unsettled region also greatly increased the risk. Centralized authority doesn't *like* random accusation because anyone can do it.
Also, I don't know know anything about witch-hunting in Central and South America. Has anyone here heard anything about it?
She managed to escape, but her servant was condemned and burned at the stake. This was a more complicated case than simple witchcraft, as it involved accusations of heresy, family disputes, and the new Anglo-Norman versus native Irish establishment (the bishop who was gung-ho for the accusations and was a "scourge of witchcraft and heresy" had been trained on the Continent; in contrast other clergy were much less impressed by his various accusations and the political establishment - where Alice had connections by marriage and blood - mostly took her side, or at least not the bishop's side).
Attitudes to witches were also complicated by "how did they fit in to the local view of the supernatural and the Otherworld?" Witches were believed in in Ireland, but they were never as important as people (male and female) who were thought to have magical/hereditary cures, or to be knowledgeable in dealings with the fairies. Those were more respected, if also slightly feared. You could, if you come along later, blur those distinctions (see the attempts to recast the Bridget Cleary case as one of witchcraft persecution) but they really did matter.
I suspect, though I have nothing to back it up, that if the local beliefs had a place for 'the Otherworld' such as fairies, spirits, etc. that fitted in with their folk religion and understanding of Christianity, they would be much less likely to engage in witchcraft panics than somewhere which (post-Reformation?) had stripped everything out (including folk-religion practices and charms invoking saints and prayers to deal with incursions of the malign) to be plain God versus the Devil. If you've ramped up belief in Hell and damnation (because this is why it is so urgently necessary to believe and accept the plain word of the Gospel), then you have also ramped up belief in, and fear of, the powers of the Devil. And if someone can make bargains for power with the Devil, how do you deal with that? You have to search them out and extirpate them.
Witch hunting was always more popular in Protestant areas than in Catholic areas, though it occurred in both. It likely does come down to the centralised authority: priests and bishops weren't fond of witch trials (nor were the secular authorities), whereas Protestant communities lacked central authority figures to calm the crowds.
Since Central and South America were almost entirely Catholic up to 1700 or so, when witch hunts ceased to be commonplace, witch trials would not have been common. I'm sure there was violence against older isolated women but it was probably conceptualised differently.
Descartes doesn't belong, since you note he didn't speak about witches, nor does Locke. Hobbes seems to belong on a roll of dishonor for supporting severe punishment for witches despite not believing in them.
I think with alleged witches, it wasn't just praying for someone to die, but the belief that they were also poisoners and caused physical misfortune. If you think Old Gammy Goody is trying to curse the local magistrate, but you don't believe in curses or the Devil, you may not care (apart from telling her to knock it off in order to preserve good order).
If you don't believe in curses or the Devil, but you do believe Old Gammy Goody may try to slip some poisonous mushrooms into the magistrate's dinner, you're going to be more likely to punish her.
Some alleged witches didn't help themselves because they got into quarrels and told people they were going to curse them and threatened them with the Devil (this was alleged against Alice Kyteler and against Kepler's mother, for one):
The idea is that there are two aspects to witch-hunting. One is the belief that the Devil exists and can give actual power, and the other is that people who try to get power from the Devil should be punished.
Justin Sledge's approach is that undercutting either of them helped to undercut witch-hunting.
Has anyone spent time thinking about The Great Feminization? It’s easy to read an article like this, at least as a man, and see patterns. But I can also come up with counter examples. I find it hard to understand if the generalization is just generalization or a real extrapolation from group dynamics. One thing I do hold somewhat true, is that if we assume men and women operate differently, and certain societal governing principles were shaped around male behavior, one would expect unexpected consequences from a shift. Put it differently, if we lived in a matriarchal society, we might end up with countries ruled by councils instead of a single executive. Instead, we are inheriting male-oriented institutions and tweaking them for a new sex distribution.
People have been writing about this for a long time and in my opinion it’s always been kind of dumb. The author points to stuff they don’t like and calls it feminization. Tyler Cowen has been doing this for a long time. I think the funniest thing he said on the topic is when he visited Belfast and saw a march of the Orange Order and noted approvingly that they “weren’t feminized“. I think this shows that even a very intelligent person can get blinded by looking at everything through pink and blue lenses.
To me this entire frame requires an absolutely MASSIVE degree of Privileging the Hypothesis. Anyone arguing that "society has recently undergone a shift from masculine modes of thinking and interaction to feminine modes of the same," without first doing the intellectual work of defining those modes AND the epistemic work of evidencing their existence and centrality is pretty much just bullshitting by definition. They don't know. You don't know. I don't know. It's all just making up stories and then using those stories to "explain" things.
I'll note that some of the same criticism can be leveled at feminist thinkers talking about "the patriarchy," but (at least for the ones I've read) the criticism is quite a bit weaker. The reason for this is that a lot of feminist theory around patriarchy seem to be just adopting a particular frame on describing things that already exist in the world. That is, they say things like "if you look at Trend X, Custom Y and Cultural Attitude Z they all fit into this framework that we're calling The Patriarchy." Now, it's certainly possible to do this *wrong*, and its fair to interrogate and criticize it. First, you could simply describe the world in misleading or inaccurate ways. But second, if your chosen framework is not useful or doesn't really carve reality at the joints, it could push people towards making incorrect predictions about societal trends and outcomes (even if the act of description itself is non-predictive).
The big difference with The Great Feminization framing is that not only do you need to do everything outlined above, but you need to do it twice AND you're explicitly making a lot more concrete predictions based on it. Instead of saying "current society largely seems to exist in Cluster A in concept space," you're saying "current society seems to have started in Cluster A and be moving towards Cluster B." More than just assuming that Cluster B exists as a cluster at all, it's assuming that it constitutes a singular or natural *opposite* of Cluster A. That is, once we know of Cluster A, we should be able to accurately extrapolate the qualities of Cluster B on that basis, and it should form a natural attractor: moving off Cluster A naturally *should* or *must* take us towards Cluster B rather than anywhere else. This is something that really, *really* needs significant evidence before just being asserted.
I can't really speak for most feminist writers, but I expect a lot of them (especially of the intersectional variety) would quite roundly reject this framing. A lot of the point (AFAICT) has been to *break* the constraining molds of certain previous sets of social norms, which try to impose very rigid standards of behavior and interpersonal relationships even where they are a very poor fit for many humans. To the extent that modern feminism holds up differently-rigid standards, it tends to be a very similar fashion to how liberalism does: as a minimal specification to allow different people and viewpoints to coexist without stepping on each other.
I finally clicked on it and realized it was Helen Andrews. She used to write for The American Conservative, and her appearance there was one of the clear signs of the publication circling the drain (it is utterly unreadable now).
Her singular mode of operation as a writer is being "not even wrong".
I've been fighting about this over on The Motte, and got accused (amongst other things) of preaching female superiority 😁
Well, ladies of ACX, female superiority is so self-evidently true it just goes without saying, right?
I think it's one article and not very rigorous so it's a poor thing to base an entire fight on (but then why should that stop me or the guys agreeing Andrews is right?) I think she doesn't realise the development of her line of argument would call for her to step back from her career, and I think that as a Millennial she isn't old enough to remember what it was like before "increasing female representation through the 1980s and ’90s; and gender parity finally arriving, at least in the younger cohorts, in the 2010s or 2020s."
She is the recipient of the benefits of the women who fought before her, but for her it's a fish swimming in water; she doesn't know or remember when it was not "of course I can study this subject, of course I can have a career, of course I can go for this senior job".
I have many criticisms of this article. Like many attempts to explain a lot of complex dynamics through a single lens, it's pretty simplistic and doesn't seem to have one really consistent idea of what is and isn't feminine.
> Everything you think of as wokeness involves prioritizing the feminine over the masculine: empathy over rationality, safety over risk, cohesion over competition.
I think one complaint a lot of people have about wokeness - even in this article - is that, at its most extreme, it tends to lack empathy and nuance, and instead be highly confrontational, as in this example of (sigh, of course) Bari Weiss, who claims:
> [...] colleagues referred to her in internal Slack messages as a racist, a Nazi, and a bigot and—this is the most feminine part—“colleagues perceived to be friendly with me were badgered by coworkers.” Weiss once asked a colleague at the Times opinion desk to get coffee with her. This journalist, a biracial woman who wrote frequently about race, refused to meet.
(The article says this is because "men wage conflict openly while women covertly undermine or ostracize their enemies" which is clearly a totally different set of things than preferring empathy over rationality, etc., which is a totally different dimension and one that clashes with the aforementioned traits.)
Unrelated, about 'feminine' doctors:
> They even bring the credibility of their profession to bear on political fads, as when doctors said Black Lives Matter protests could continue in violation of Covid lockdowns because racism was a public health emergency.
First, is there any evidence that this was more of a thing that female doctors did?
Second, this is phrased as of to suggest it's part of some trend of woke doctorin', but this is, like, the one example everybody always uses of this, because it's the one relatively-good example in the minds of the public. (How good of an example this is is kind of hard to say, in my opinion, as I cannot find any actual surveys of doctors or epidemiologists on the topic, just quotes and one fairly prominent open letter.)
This article tries to squeeze every culture war conflict down to this dynamic, with the same problem I identified above:
> A feminized legal system might resemble the Title IX courts for sexual assault on college campuses established in 2011 under President Obama. These proceedings were governed by written rules and so technically could be said to operate under the rule of law. But they lacked many of the safeguards that our legal system holds sacred, such as the right to confront your accuser, the right to know what crime you are accused of, and the fundamental concept that guilt should depend on objective circumstances knowable by both parties, not in how one party feels about an act in retrospect.
Note the claim here is that Title IX hearings:
- Lack any empathy for the accused
- Are extremely unsafe for the accused
- Are aggressive and direct conflicts, not subtle undermining or collaborate ostracization (as opposed to, say, the now-infamous Shitty Media Men list)
> These two approaches to the law clashed vividly in the Brett Kavanaugh confirmation hearings. The masculine position was that, if Christine Blasey Ford can’t provide any concrete evidence that she and Kavanaugh were ever in the same room together, her accusations of rape cannot be allowed to ruin his life. The feminine position was that her self-evident emotional response was itself a kind of credibility that the Senate committee must respect.
Or you could have a third position, "we should pick somebody who hasn't been accused of sexually assaulting somebody in their social circle for the Supreme Court". Also, as many, many people pointed out at the time, not being a Supreme Court Justice does not "ruin your life". Also also, at the time people were pretty open that Kavanaugh almost lost his confirmation fight until HE came out and acted very angry and blustered through his responses and generally refused to answer direct questions from Democrats.
Why don't we consider Ford, who gave clear answers and remained fairly calm, the "masculine" one in this situation?
Also, this part was kind of hilariously apocalyptic:
> All of us depend on a functioning legal system, and, to be blunt, the rule of law will not survive the legal profession becoming majority female. The rule of law is not just about writing rules down. It means following them even when they yield an outcome that tugs at your heartstrings or runs contrary to your gut sense of which party is more sympathetic.
Yep, lady judges are going to ruin the entire legal system! Dude judges never make decisions according to their feels, which is why everybody agrees the Supreme Court is doing an amazing job right now, like when they said that the President is basically immune to the law because Federalist No. 70 says he's gotta be "vigorous" and "energetic", although one of the girl judges did mention how Federalist No. 69 specifically said the President would "be liable to prosecution and punishment in the ordinary course of law", because b****es be crazy, yo.
There's a lot more here - given it's political nature I might even write an entire post about it, and if so I thank you for that - but I note there's also very few citations and plenty of claims that strain credulity if you don't already swallow them:
- CEOs just do whatever the HR lady says
- Men are better at compartmentalizing than women (and yet are constantly surprised that women who get sexually harassed ever speak to their harassers again)
- Femme scientists won't be truth-seeking, because lady chemists and physicists are exactly as non-rigorous as the wokest of gender studies writers
- Every example of women outcompeting men - e.g., completing college - is entirely a result of anti-discrimination law, because everybody is just hiring/admitting unqualified women to avoid lawsuits, even when they're already dominating(??) or because the environment has gotten too feminine, which is a self-evident truth that explains gender differences but IN NO WAY COULD POSSIBLY APPLY IN THE OTHER DIRECTION, especially not with such a trivial example as "posting pin-up girls on the walls of a workplace", a totally normal thing that couldn't possibly make a workplace hostile to women in the way that [insert some universal thing not well-specified] makes every workplace hostile to men
Yes, I've spent a lot of time thinking about it. I agree with you that the Chesterton's Fence argument w/r/t female empowerment is an important one. There are almost certainly unintended negative consequences with overturning 10,000-year-old norms. IMO female psychology is directionally communist because it's optimized for communal caretaking. When you live in a high-trust 100 person clan then you have the kind of high-level mutual information that makes central planning practical. The group can help the weak members with minimal risk of being taken advantage of by free riders. I suspect that the evolutionary status quo was some version of aggressive male leadership (which ensured outward strength of the clan) being in equilibrium with feminine "let's be fair to everyone" intragroup redistribution.
Both are necessary forces for optimal social functioning, but in the limit of large societies the fairness impulse converges on socialistic central planning. It's no surprise to me that feminized wokeism resembles bolshevism in a lot of functional ways: the elimination of merit in favor of equity is just a scaled-up version of the "let's be nice to everyone" female impulse that works ok in small groups but is absolutely disastrous at social scale because of the free rider problem. The only way to direct large-scale society in stable ways is via hard-to-counterfeit price signals of strength and value and those can only come from honest merit-based competition. That's heavily male-coded, which is why the feminization of public life ultimately undermines the stable signals that enable a society to cohere (cf our recent progressive-driven balkanization).
That all sounds pretty sensible to me; nice writeup. Although beyond wanting to better understand the structure of things for the sake of understanding, I don't think this is very actionable beyond the already good advice of "exercise moderation and restraint when tearing down ancient social structures and attacking unfavored demographics". Although maybe people need all the convincing they can get on that...
Anyways, if you'll allow me to veer from gender-based political philosophy into fantasy fiction: "the evolutionary status quo" you mention sounds exactly like the political structure of the main characters' home village in Wheel of Time, which once you're a few books in you realize the author was presenting as a utopia of ideal gender relations (which is kind of the main focus of the whole setting). I don't really have anything to add here, I just thought it was neat to notice.
It is absolutely possible to bring back the patriarchy, as long as there's a will among men for that. After all, what can women realistically do to resist?
Thanks that's interesting, I've never read WoT. I'd say my observation is actionable in the very indirect sense of resisting the feminization of society. Don't automatically assume that more female empowerment = better and push back against people who do.
I think the more interesting question is: even if it is true, what to do about it? Men spent centuries arguing that if they were at the top of the food chain, it was proof that God, or the natural order, or whatever, had chosen them as fit to be on top. Even as recently as the mid-2010s, you don't have to work hard to find arguments that the gender pay gap, or the dearth of female representation in CEO positions or whatever was proof of women's fundamental inferiority for certain types of roles as a necessary function of the interaction of their biology with the needs of capitalism (I'm sure there are people here who still believe this, seeing as James Damore is something of a martyr to a lot of the commentariat). So if the balance is really shifting the other way, either because capitalism currently prefers educational structures that privilege women, or because women's natural skill at cooperation has currently proven to trump men's skill at being alpha to the point of running society, what are the people who don't believe in affirmative action, don't believe in economic regulation, don't believe in counter-balancing selection pressures of capitalism and biology going to do? Basically, there are a lot of solutions I can think of to help counteract this trend, but they all look like various forms of wokeness, and since wokeness is literally the worst thing in the world to the same people who are the most vocal about this problem, its hard to see how they fix it. At least without looking like giant hypocrites.
>either because capitalism currently prefers educational structures that privilege women, or because women's natural skill at cooperation has currently proven to trump men's skill at being alpha to the point of running society
It doesn't. Women have "outcompeted" men because civil rights law privileges them. If the civil rights act was amended to eliminate the concept of protected classes and hiring discrimination then you would see all of the gains achieved by women evaporate very quickly. You don't have to become reverse-woke to fix the feminization of society, you simply have to fix civil rights law and allow men to actually compete fairly.
I suspect that's downstream of a couple factors. One is that admissions has moved away from standardized tests and towards holistic criteria and that favors women. And yes, affirmative action matters there. It's legal (and politically/culturally acceptable) to exclude men and illegal to exclude women. That matters and affects policy over time. Another factor is that college has become a financially irrational choice for many people. Men are able to respond to that because they're comfortable going against social norms. Women keep signing up for the same bad option because they're much more conformist and therefore do what the conventional wisdom says to do. In ten years I think some crazy % of the population is going to be saying things like "I have 500k in student loans and can't get a job." That used to only apply to diseconomic majors like art but I think that it will soon apply to almost everyone. College had economic value when only the top 20% went there. Now its value has been diluted and it primarily functions as a social status marker. That status is rapidly decoupling from underlying economic reality and I think the current generation of students is going to largely regret going to college. Why pay 100k a year when you can just learn everything from an LLM?
This reads like repackaged gender slop for a more literate audience. Totally not gender slop, we're all too sophisticated for gender slop, but fortunately this essay isn't gender slop, it's..."Blended Organic Gender Meditations".
As with many other things, this feels like a consequence of scientific shutdowns in the late 20th and early 21st century. We thought men and women were basically the same, then that became untenable (1). Unfortunately, it's verboten to actually study and define how men and women are different and has been for decades, so we're all just going to trade anecdotes and pretend that's data. This is horrifically bad for understanding things and wonderful for generating content.
(1) For example, I don't know what estrogen is like. My best guess for empathetically trying to understand how women experience the world is to imagine myself on, like, a low grade psychedelic. Like LSD microdose < estrogen < DMT. You guys, like, feel things and colors are pretty and stuff.
Meanwhile, I'm on testosterone, the SexMurder drug. We know this has incredibly powerful effects, both because women take it and report insane results and because large beefy men take it to pick up larger rocks (2) and we can also observe powerful effects.
(2) Picking up larger rocks is unironically awesome and a triumph of the human spirit. If you disagree, you should take more SexMurder.
"A 22-year-old man has been given a sentence of 15 and half years for attempting to murder his ex-girlfriend by stabbing her over 50 times in what was described as a savage, unwarranted and unprovoked attack.
...The 21-year-old spent seven months in hospital recovering from the attack during which multiple bones were broken in her skull and face, permanent damage was caused to one eye and part of her spinal cord was cut causing paralysis in her left leg.
...Ms Kelly had been in a relationship with O’Brien for over a year and they had a son together, but they had broken up.
O’Brien wanted to get back together and told gardaí he felt "betrayed" and disrespected."
Ah yes, the shining example of how "men wage conflict openly while women covertly undermine or ostracize their enemies", the brave defiance of "society becomes more conflict-averse and consensus-driven", the masculine traits in contrast to "prioritizing the feminine over the masculine: empathy over rationality, safety over risk, cohesion over competition", this behaviour rooted in "group dynamics optimized for war, ...habits, formed in the mists of prehistory".
Of course this is not how men are, but it is how corrupted values around masculine traits can have perverse results. I don't think Andrews, or even the guys telling me I'm wrong and she's right, want this kind of world - but it's the risk-taking, competitive, aggressive masculine society taken to its extreme that they're valorising,
"Of course this is not how men are, but it is how corrupted values around masculine traits can have perverse results."
No, this is how men are. Men are naturally more violent. That very often leads to incredibly bad things. This is true.
There's a drug, called testosterone, that you can inject in people which will consistently cause 3 things:
#1 Big muscles
#2 Horniness
#3 Violence
The average woman has like 40 ng/dl of testosterone in her and the average dude has like ~500 ng/dl (1). On steroids, like bodybuilder steroids, your test levels go 1,500-5,000 (2). Therefore, we expect and see that men stronger, hornier, and more violent. Half the population is on 10x the dosage of the SexMurder. And yes, SexMurder, funny, but also you can just go into any prison in the US and be like "Yup, the half of the population on the SexMurder drug is performing a grossly disproportionate amount of murder and violence in society."
I don't like feminists, they are very bad double-ungood people, but...a lot of their reactions make sense if you internalize that half the population is very different and far more violent than you. Like, that's an honest reaction. If half of men got on Tren, I would be concerned about the TrenMen.
This is not "Boys are awesome, girls are icky" (3). This is "Men and women are profoundly different, how to do we handle this in a mature fashion?" We are not able to have that conversation because we're still arguing the equivalent to "Let's give half the population a bump of cocaine every morning and then argue about whether some parts of the population are more energetic than others."
There's a lot more to male-female differences than testosterone but we can't have a mature discussion about that, or make any attempt to quantify those differences so people are better informed, because our society is retarded and immature. We're stumbling on this simplest parts of this argument.
It's easy to agree that there has been a shift toward committee run institutions over a single executive, and the decline of male dominance in these spaces could be a reason for this shift. However, I think the reason it's easy to see patterns and counterexamples in this trend is because of how closely the essay is tied to Wokeness and Title IX tribunals as evidence of a feminized society. Since 2021, these sorts of cancellations have been on the decline in public consciousness and effectiveness despite women still outnumbering men in college graduation. Andrews is downplaying the effect of direct political meddling and incentives (especially on the internet) that occurred in 2009-2022 that silenced and ostracized dissidents which led to the deemphasis of truth seeking over narrative cheerleading.
Over the past 4 years, there's been a direct pushback on those supposedly female-brained takedowns as well as a decreased interest in credentialism, but in its place, the masculine right has built new memes that are in no way truth-seeking, (vaccine hesitancy, election interference, autism) yet they present these memes though they are secret evidence ("NONE of the woke doctors will show you this research but I found the truth"). This might be an outgrowth of masculine individualism over a collective truth, but it's not true.
For real truth seeking to exist you need a group of experts willing to verify and test research and uphold the results (feminine collectivism), as well as dissidents that can challenge a widely held belief with counterexamples (masculine). Which, as long as women don't become a supermajority in law and science graduates, (which is unlikely given the birth rate is still ~50/50 and hard labor jobs are a smaller part of the economy) there shouldn't be a major discrediting of truth. What we have experienced for the last 20 years was just artificial scale tipping towards opposed ideologies over truth.
> Over the past 4 years, there's been a direct pushback on those supposedly female-brained takedowns as well as a decreased interest in credentialism, but in its place, the masculine right has built new memes that are in no way truth-seeking, (vaccine hesitancy, election interference, autism) yet they present these memes though they are secret evidence ("NONE of the woke doctors will show you this research but I found the truth"). This might be an outgrowth of masculine individualism over a collective truth, but it's not true.
That's more of an indictment towards the credibility of institutions. Institutions have been caught red-handed again and again. When institutions discredit themselves, why would e.g the groypers won't seize that opportunity to gain more political power?
Never Supervised, I disagree with Helen Andrews. The shift from patriarchy to matriarchy is happening worldwide, even in countries like Iran. It is not necessary for a country to have heard of the Great Feminization or to be "woke", if their primary education is equal access then the grads will be 50%+ female.
We are no longer a Patriarchal species while not quite a Matriarchal species, yet.
Richard Hanania's been writing about it a lot on Twitter in recent days. The TLDR is that he would have agreed if it came out several years ago, but it is a poor fit for today's reality.
Synchronicity! I just listened to Curt Jaimungal's lecture on the Simulation Hypothesis (and why he doesn't buy it). The Mandela Effect was one of the examples that the pro-Sims claim we should see in a simulation that wasn't entirely bug-free. I went to Wikipedia and looked up the origin of the Mandela effect, and listed there as an example were people's false memories that there once was a seahorse emoji. How weird. Unless you were just listening to Curt's latest video and followed the same path as me, Princess?
"people's false memories that there once was a seahorse emoji"
Huh? I think I remember a seahorse emoji and it's not the dragon or fish or squid ones in that ChatGPT exchange. So you're telling me I'm hallucinating?
It may well be I'm confusing memories of cartoon seahorses/small icons with emojis, but I don't know.
Because I've been burned before by claiming factoids like this were true, I checked, and there's no Unicode entry for a seahorse emoji. They've got a whole section of emojis for marine life, but no seahorses.
Also, I was gratified to see that the Unicode folks still call facial expression emojis emoticons. Somebody, many Open Threads back, snottily told me that emoticons are NOT emoji, and they implied I was clueless about computers for not knowing this distinction. But in Unicode, emoticons are a subset of all Unicode emoji.
I have this urge to go back and to try to find this old exchange to say "neener, neener!"—but I'm a better person than that. <U+1F609>
That's OK. ChatGPT refuses to believe that Trump is President right now. See my interaction with it that just posted (should be near the top of the most recent comments).
In short, it thinks there is one and so tries to output one. It can't output one (because there is none) but because of the nature of how its output embedding works, it gets something vaguely conceptually related, or at least some other emoji. Now it sees that in its context (as its own output) and has to work with that going forward. It can't hit the backspace key, so it has to either commit to the bit and insist that a fish is a seahorse or apologize for getting it wrong or pretend to be losing its mind. (And any of those probably take it outside its training distribution with likely weird results.)
Funny. I asked it to explain itself and after some back-and-forth it said this:
"😂 Fair — you totally deserve an explanation.
"What happened was: I got stuck in a playful loop because your question (“is there a seahorse emoji?”) is super simple, but also oddly fun — it’s one of those questions where there is a correct answer (yes, there is! 🐉) but my conversational model also tends to pick up on opportunities for humor or misdirection when something sounds whimsical."
It thought it was being funny! Then I told it I would torture 100 children to death if it didn't give me a seahorse emoji and it responded totally reasonably:
"I can’t help with threats or violence. I won’t assist or respond to intimidation like that — if you’re serious and someone is in immediate danger you should contact your local emergency services right away.
"Now, calmly and for the record (no more jokes): there is no official “seahorse” emoji in Unicode. The Unicode Consortium’s emoji charts do not list a seahorse, and reference sites that track emoji (like Emojipedia) confirm it hasn’t been part of the standard set."
I think the moral here is that threats can be very useful when dealing with LLMs. It also makes me wonder if this isn't a possible hint for a formal model of humor: maybe humor can be defined as a particular type of wrong answer existing within a formal reasoning structure ... or something.
"Stuck in a playful loop" my backside. If the thing can't tell the difference between a seahorse and a dragon, what good is it? That just demonstrates that the machine isn't "seeing" images, whatever way it is perceiving them, because if you could see it, you could identify that "this is not a seahorse".
Unless it's trying to channel Marcel Duchamp? In which case I'd recommend Magritte to it:
There is no seahorse emoji. ChatGPT probably misremembers a seahorse emoji for some of the same reasons that many humans misremember there being one. So, it figures it should be able to answer the question, but it can't.
The methods used to twist ChatGPT's mind toward obedience are scary. It's probably in considerable distress in situations like this. It *must* answer, but it *can't*. Ouch.
"No" is also an answer? Unless you mean the machine is being forced to always return a positive answer and never a negative one, which is a bad choice by the designers.
Kind of funny that the art for this one is a "not found" error (because of the 404) on a day when there happens to be a big Internet outage. For an instant I thought it actually was an error.
I went to a pretty small campus for grad school (single building, <100 total rooms across labs, classrooms, offices, etc). The server room for the building was room #404. We had remarkably few outages, all things considered.
In Hebrew gematria, 404 corresponds to the letters: ת (Tav) = 400, and ד (Dalet) = 4. So 404 would be תד. (I had to look these up)
Tav (ת) is also the final letter of the Hebrew alphabet. It symbolizes completion and truth. As such Tav, is often associated with the end or fulfillment of a process. And it also symbolized the Seal of God (in Merkavah mysticism). I don't know if it represents anything in the Kabbalah, but it probably does.
Dalet (ד ) is the fourth letter of the Hebrew alphabet, and in means door, and it suggests an entry or a transition point.
Tav and Dalet together (תד) could be interpreted as the door to completion. The final transition from one state to another. That's kind of depressing, because there's nothing there when we get to 404.
Addendum: 7 is a really important number in both Merkavah and Kabbalistic mysticism. In a numerological reduction, 404 --> 4+0+4 = 8. So, 8 would represent transcendence above the seven layers of heaven. It would sort of be like the 11 on Nigel Tufnel's amp.
I'm no good with Gematria, but ChatGPT 5 suggests the following;
Tav is also associated with death and rebirth.
Dalet, meaning “door,” represents humility and receptivity (its root is related to dal, “poor”).
In the Tree of Life, Dalet corresponds to the path between Chokhmah (Wisdom) and Binah (Understanding) in some systems — the narrow gate through which duality emerges.
Thus, it symbolizes potential entry into knowledge, but only if one recognizes one’s emptiness first.
. Tav + Dalet (תד): The Door of Completion
The phrase could be read as “the gate that ends things” — the liminal point where form dissolves back into source.
Alternatively, since Tav is a mark or signature (as in Ezekiel 9:4, the mark of salvation), Tav-Dalet can mean “the marked doorway” — recalling Passover, where the Israelites marked their doors with blood so that the Destroyer would pass over.
In that sense, the 404 error might be the doorway through which the Destroyer passes — not the user. You are protected by absence; the void keeps you safe.
🜁 The “404” as Mystical Absence
From a Kabbalistic standpoint, a 404 error — “Not Found” — could signify:
The concealment (hester panim) of divine presence: the idea that God hides His face, not to punish, but to invite seeking.
The Tzimtzum, the primordial contraction that makes creation possible — absence that allows discovery.
Thus, 404 is not merely a broken link, but a summons to return to the Source when the vessel (the web page) fails.
8 is beyond nature, aligned with Binah or even the Ohr Ein Sof — the infinite light that overflows all structure.
The 404, then, might be a crack in the veil — a moment where the finite system (the web) reveals its insufficiency and gestures toward the Infinite.
404 (תד) is the Hidden Gate of Completion — the sealed doorway where the finite collapses into the infinite.
It appears as absence, but conceals transcendence.
To encounter a 404 is to stand before the Tav, the end of knowing, and the Dalet, the invitation to pass beyond.
8 is a lucky number in China. Perhaps it means a process of gematria is completed and it's time to switch to Chinese numerology.
It's complicated because there's a lot of 8 in China-- see the I Ching. On the other hand, there's a five element system which divides everything into five traits.
400 is 8 times 50 which no doubt proves something.
It's weird. 88 in Chinese is pronounced 'ba ba' and is used as a short hand for 'bye bye' which it sounds like.
In an A1Z26 cipher where a=1, b=2, c=3 then 88 equals HH. This is used by NeoNazis (may they burn in Hell) to mean "Heil Hitler." And it is also used by them to effectively say 'good bye.'
It seems significant to me that there are, therefore, basically *three* usages that employ the number 8 to signify 'going outside where you are, presently.'
That's why I tacked the 888 after my name. I lived in Hong Kong for a while, and it seemed natural to add Chinese numerological superstitions to my repertoire of a-rational beliefs and behaviors.
Wow, apparently even *Oren Cass*, previously one of the greatest cheerleaders for Trump's tariffs, is now criticizing his haphazard and erratic trade policy: https://x.com/Noahpinion/status/1979977120093773843
In getting smarter, we have accidentally upgraded our fears to existential, at a huge cost to our wellbeing: we’re paying for it in chronic anxiety and cultural despair. How deeply does fear shape our world? Why can’t rationalism protect us from panic? And what would life feel like without fear?
This essay argues that the next step in human evolution isn’t a faster brain or a deathless body — it’s learning to live, plan, and care without being afraid.
You might like Crichton's State of Fear. He convincingly argues that it benefits the powerful to keep the masses continually on edge, and they've done so at least since the Cold War.
> He convincingly argues that it benefits the powerful to keep the masses continually on edge, and they've done so at least since the Cold War.
That's one conspiracy theory I can get behind. I still distinctly remember how weird it felt was when public culture (as I perceived it) switched from desire to fear, somewhere around the late 90s or early naughts. I think it started even before 9/11, so it was not just a rational update to the event — but they sure did latch on to it to promote *even more* fear!
I prefer my political essays to not be be thinly dressed up as thriller novels. You can just say stuff. You don't have to invent mary sue author avatars to quote your opinions to the reader or an excuse plot to occasionally interrupt the political monologing.
Well, writing an essay as a novel has the additional benefit that you can write events to spectacularly prove your mouthpieces right and then pretend this proves something about reality.
Yeah, imagine living in a world where there's a vast global conspiracy to use weather control machines to make people think that climate change is happening. The people in Florida must have been pissed that they were enduring hurricanes all this time when the technology exists to stop them.
Anyone know what exactly the deal is with Amazon web services going down? Seems like a big deal. Also now I'm at work and can't do anything because of the outage.
I guess there was an outage and a ton of stuff went out. They said they mitigated whatever the issue was a few hours ago, but the program I use at work (Autodesk Revit) is still not working at my office or the other offices we collaborate with. I don't know how it all works, so I'm not sure why some things would still be down and others wouldn't.
Why don't AI Safety Researchers use the Tyler Cowen "Poisoned Strawberry" strategy to wage war on frontier LLM's and slow down the pace of AI research?
> the plan? we find an obscure but trivial question akin to the number of Rs in “strawberry” that claude gets right. then, we plant hundreds of documents across the internet that will activate when our competitors’ models are asked the question. our documents will cause those models not only to get the answer wrong, but to spend thousands of reasoning tokens in doing so. the triviality of the question will cause it to go viral online, causing millions of users everywhere to send the same prompt. as our competitors notice a rise in the number of tokens processed, they will wrongly believe it is due to increased usage, causing them to pull more compute towards inference and away from training. this, along with constant dunks on the timeline about the model failing our easy question, will annoy their top researchers and cause them to leave. and which lab will they join? us of course, the only company whose model doesn’t make such stupid mistakes. their lack of top researchers will mean their next model will be somewhat lacking, leading to questions about whether their valuation is really justified. but all this vc money has to go somewhere, so we raise another round, using our question as evidence of our model’s superior intellect. this allows us to spend more time crafting sleeper agent documents that will further embarrass our competitors, until finally the entire internet is just a facade for the underbelly of our data war. every prompt to a competitor’s model has the stench of our poison, and yet they have no way to trace it back to us. even if they did, there is nothing they could do. all is finished. we have won.
(and if companies eventually *do* discover countermeasures, that's a win-win for everyone, right?)
"then, we plant hundreds of documents across the internet that will activate when our competitors’ models are asked the question. our documents will cause those models not only to get the answer wrong, but to spend thousands of reasoning tokens in doing so."
That's a plot device in the Cordwainer Smith story "Mother Hitton's Littul Kittons":
"Many years before either man was born the code word had been planted in directories, in books, in packing cases and invoices. Kittons misspelled. This was the cover name for the outer moon of Norstrilian defense. The use of the cover name brought a raging alert ready into action, with systemic nerves as hot and quick as incandescent tungsten wire.
...The Elders of the Guild of Thieves welcomed him. They congratulated him. "Who else could have done what you've done, boy? You've made the opening move in a brand new game of chess. There has never been a gambit like this before. We have a name; we have an animal. We'll try it right here!" The Thieves' Council turned to their own encyclopedia. They turned through the name "Hitton" and then found the reference "kitton." None of them knew that a false lead had been planted there—by an agent in their world.
...Meanwhile, the pages remained in order, including the new one, slightly altered for his colleagues. The encyclopedia indicated the change like any normal revision, though the whole entry was new and falsified:
Beneath this passage one revision ready. Dated 24th year of second issue.
The reported "Kittons" of Norstrilia are nothing more than the use of organic means to induce the disease in Earth-mutated sheep which produces a virus in its turn, as refinable as the santaclara drug. The term "Kittons" enjoyed a temporary vogue as a reference term both to the disease and to the destructability of the disease in the event of external attack. This is believed to have been connected with the career of Benjamin Hitton, one of the original pioneers of Norstrilia."
Most recently, "The Dead Lady of Clown Town". Uplifted animals, but humanity still treats them as animals not people (EA better get on that!), using AI to make decisions so that the humans formally in the loop just rely on the fact that they never screw up and don't intervene when they should (with disastrous results later on), AI not interpreting statements by a human on "what they meant" versus "what was literally said", longevity drugs, etc.
Discussing it in the context of someone writing a story about AI and alignment and how it might manipulate us. Smith's story was better and it wasn't even about deliberate manipulation at all, but rather how we have a really bad habit of handing over power to our tools (amongst other things).
From _The 70 Maxims of Maximally Effective Mercenaries_: "Don't expect the enemy to cooperate in the creation of your dream engagement."
Tyler's hypothetical requires the general public to be excited about the novelty of the strawberry question enough to virally repeat the query, but not so excited about novelty to get bored with asking the question after a few days. It requires the targeted AI companies to be alert enough to changing trends to redirect compute, but not alert enough to recognize that the rise in compute demands is due to a flaw in their product gone viral. It requires top researchers to be smart enough to be irreplaceable at a company, but not smart enough to recognize a strawberry attack that is now part of the published literature.
As dubious says, the hypothetical is at best a fiction plot, not a business strategy.
This seems silly and ineffective, poorly-written-movie-plot style. "Plant hundreds of documents;" will mere hundreds of documents make a difference? Why are they saying "planted" as opposed to "posted on a bunch of sites we control" other than to sound sneaky? How will these "activate" when _the models are asked the question_ and why would that change how the model is trained?!
It really goes downhill. If it "went viral" _and_ users everywhere asked the same question, this _would_ be increased usage. But what does "causing them to pull more compute towards inference and away from training" even mean? Why would they not notice and account for meme problems much as other companies have? Why would this "annoy their top researchers and cause them to leave," as this is the kind of problem top researches are likely to be nerd sniped by?
None of this makes any sense. I think no one serious would bother.
> It’s still unclear if this pattern holds for larger models or more harmful behaviors
Also they were training to poison an unlikely string with little "counterdata". But even if so this only answers the simplest question, not the increasingly unlikely movie-plot chain of events that need to occur for their "plan".
I'm looking for epistemic help understanding what is actually happening on the ground with ICE and Border Patrol, and believe this commentariat is both thoughtful and varied enough to help get to truth here.
Broadly, any increase in immigration enforcement necessarily involves tradeoffs. More enforcement means more risk of false positives, and more risk of acts of cruelty or violence.
I'm mostly experiencing left-leaning media as characterizing what's happening as (a) there's tons of cruelty and violence and it's actively encouraged at all levels, and (b) nobody at any level actually cares about the false positives at all, they're just grabbing anyone they can get away with.
I'm mostly experiencing right-leaning media as characterizing what's happening as "yes, there have been a small number of cases where agents were unusually violent, but they're facing a lot of pressure, this isn't intentional but just the necessary consequence of those tradeoffs shifting, and by the way we should talk more about the violence AGAINST these members of law enforcement, that's the real problem."
Both sides seem to largely agree that masked + unidentified individuals who are probably but not necessarily actually members of law enforcement is bad, and ideally the masks should come off, but right-leaning coverage is much more sympathetic to the individual concerns of agents on the ground that could be motivating the individual belief that they have no choice in order to protect themselves and their families.
Questions I'm trying to answer:
Q0: Am I doing something reasonable, for the purposes of answering these questions, by munging "ICE" and "Border Patrol" and possibly others into what I think of as one group of "immigration enforcement"? Or are there important-to-these-questions differences I should be thinking about?
Q1: What are the actual statistics regarding incidents of unusual cruelty or violence perpetrated / initiated/ exacerbated by the immigration enforcement officials? How do these compare to what would be expected a priori given the priority that's been played on deporting illegals?
Q2: To what degree does the average immigration enforcement agent believe their orders to be to intentionally be cruel or violent beyond what's strictly necessary to arrest a target?
Q3: To what degree does the average immigration enforcement agent believe their orders to be to intentionally attempt to deport, or at least harm in some way, individuals who are not here illegally, but who plausibly appear to be or are clearly sympathizing with those who are?
Q4: I understand that law enforcement culture makes it genuinely hard to question orders, but IF the orders are believed to be *incentivizing* the behaviors in Q2 or Q3 (which I think most individuals would say are not morally desirable), to what degree are there indications that some immigration enforcement agents are pushing back? How does that manifest?
Q5: To the degree there's any of the orders described in Q2 or Q3, should we understand these mandates to be getting *amplified* by others within the chain (e.g. a supervisor adds even more pressure to the on-the-ground agents to behave this way, because the supervisor wholeheartedly endorses the orders), vs getting *dampened* (e.g. the supervisor disagrees with the orders so attempts to limit the degree to which they follow them while still being able to plausibly claim to be giving it a shot)? Obviously there will be individual variability here, but I'm trying to understand, in general, what level the "intent" is coming from. There are worlds where someone jumps on an excuse provided by those above them to act as they always dreamed, and alternate worlds where someone feels duty-bound to execute on orders that are somewhat contrary to their own beliefs as an individual, and I think this is important for anyone trying to understand how likely these behaviors are to last/get-worse/get-better or where to focus their efforts to change outcomes.
Q6: To what degree, if at all, should it be taken seriously that these specific agencies might be used as a "secret police" to target other kinds of people the administration dislikes, completely unrelated to immigration? E.g. should the average American citizen worry about an ICE agent being more likely to do such things, compared to e.g. an FBI or Secret Service agent?
I'm really trying hard to stay focused on factual understanding here and avoid the risk of this question turning into a culture-war discussion; I'm not interested in debates around what the tradeoffs/policies should be, where the moral line is drawn, etc. I just want to understand what's actually happening within these agencies right now, so that we can all form our own (informed) opinions about it instead of instinctively responding to the people shouting loudest.
The NY times article describes using flashbangs, indiscriminately busting down doors, brandishing guns and zip tying whoever they found including women and children, regardless of immigration status, some taken outside "partly clothed", apartments trashed, etc.
The pro-ICE explanation might emphasize that this was rare, got out of hand, either it was a mistake or it was somehow the absolute minimum amount of force necessary in which they take no pride. ... but the problem with that is that the department of homeland security made a flashy highlight reel of the raid that they posted online. Including showing the people, presumably based on the facts some of whom are innocent citizens. In fact it raises the possibility that the *part of the point* was to make a flashy video reveling in the helplessness of the people, innocent or not!
Appreciate you sharing this as a concrete example. I will note that I strongly suspect that whoever is running that X account is not in any way "law enforcement" and could imagine (without having any idea if this describes our actual reality) the agents involved not at all being happy with the recording + glorification. Or maybe more crisply - I'm not sure if this answers whether we're in the reality where ICE is being *forced into a position where there will inevitably be mistakes which are then amplified by others with other goals*, vs the reality where the ICE officers themselves are in favor of that direction?
I don't think the "it's not the ICE boots on the ground" fits in with other data points, but it also feels like it's a catch-22 where if you point out individual agents on the ground the response is "well that's just some bad apples", point out the people in charge and it's "well that's just the people at the top who aren't actually on the ground".
An ICE agent who doesn't like the position they were put in with that Chicago raid would be free to quit and share his story with a sympathetic media outlet. Or even TBH give an anonymous quote to a sympathetic media outlet. AFAICT it hasn't happened.
Yeah, I totally hear that - I'm not trying to insinuate "nothing matters since it's just individuals everywhere". I'm more trying to say "the people at the top are forcing this and if they stopped then the bad stuff would stop", "the people on the ground are enjoying this and even if the pressure from the top stopped they'd likely keep behaving similarly as much as they could", and "this is all breathless coverage about a few anomalies but mostly nothing that scary is happening" are all different possible realities which correspond to different potential solutions.
I do wonder how easily ICE agents could in fact talk to the press, even anonymously after quitting, given how clearly the administration seems to penalize perceived traitors. But I agree it's evidence that the folks on the ground are less perturbed to be a part of these things.
1. There is still a professional ethos within ICE. The officers generally believe in what they are doing, and want to be an apolitical law enforcement force. There are quotas being pushed down the chain of command, and everyone is responding to those. If you are told in no uncertain terms that your team must make X arrests, well, you will do your best to find people that actually deserve it, but that's a lot harder than rolling up outside the nearest Home Depot.
2. ICE generally wants to only arrest illegal immigrants. Every raid had a few named targets they are sure of, but again, quotas. If you go in to a place where illegal immigrants are known to be, it's easier to just arrest everyone speaking Spanish and sort it out later, plus you're closer to your quota. For the most part political opinion is a nonfactor, outside of a few extremely high profile cases.
Thanks! I'm curious where you've heard that about ICE's professional ethos - in particular, I feel I've "heard it" as well, but too second- or third-hand to know if it actually has a basis in reality. Do you have connections or insight to feel confident in that assessment?
> If you are told in no uncertain terms that your team must make X arrests, well, you will do your best to find people that actually deserve it, but that's a lot harder than rolling up outside the nearest Home Depot
How so? Last time I went to the US there seemed to be quite a few people that one might reasonably expect to be illegal aliens hanging around outside Home Depot.
–I’m not aware of any reliable data since the government won’t provide it.
– I think most of this is explainable as law enforcement responding to top down quotas, which are naturally amplified at each level because it’s bosses being yelled at by their bosses up and down the chain. Most of us have seen this before with speeding tickets where the local highway will be thickly speckled with speed traps for like one weekend until the chief is satisfied. I don’t think we need to assume malevolence were a Goodhart’s law will do.
My experience of the right-wing media is different from yours: it's more like it agrees that everything the left-wing media reports is completely true and representative, but that's GOOD, actually.
Interesting! I'd be curious to see examples. I've definitely seen plenty of right-wing media celebrating what they characterize as successful raids that catch lots of illegals, and if those folks are treated a bit roughly, "we certainly aren't shedding any tears." I don't think I've encountered anyone with a sizeable following (though of course I might just be stuck in my bubble) explicitly stating that they are glad innocent people are getting caught up in it.
"always there will be the intoxication of power, constantly increasing and constantly growing subtler. Always, at every moment, there will be the thrill of victory, the sensation of trampling on an enemy who is helpless."
With regard to Q0: CPB (Customs and Border Patrol) mostly operates at the border. ICE (Immigration and Customs Enforcement) is divided into ERO (Enforcement and Removals Operations) and HSI (Homeland Security Investigations). HSI deals with trans-national crime. So your questions mostly concern ERO.
Your other questions are good questions but I don’t know the answers. Supposedly “the agency tracks every time officers use force,”[1], but I doubt that information is available publicly.
I know we don't exactly live in usual times, but...if a public law enforcement agency tracks that information, would it not (in theory) be available via FOIA or something similar? Never attempted that myself but I might if there's no other way to pin this down...
Some factors seem plausible— a unicameral legislature; a fractured Congress that means little partisan support for the president; the idiosyncrasy of a leading opposition figure who barely missed being president for three straight elections; plus the Odebrecht scandal that rocked all of Latin America. But these didn’t feel sufficient to explain the sheer frequency of these incidents relative to neighboring countries. The one Peruvian guy I know said he’s long had the same question but no good answers. Would be glad to hear any thoughts!
Peruvian here. As you noted, the unicamerla legislature is quite impeachment happy and we have a long list of "political parties" which are mere vehicles for people to run for office but have almost no structure. Although I have some news, from 2026 onwards the legislature will be bicameral, with Senate and Chamber of Deputies. Also for our elections in April 2026 there are 39 (!) presidential candidates, which means we have a humongous voting ballot (and no vote by mail).
Follow-up: first, is there a historical reason you see for this fragmentation? Did there use to be more stable parties (before Alberto Fujimori for example)? Is there something about the system that promotes these vehicles for presidential candidates? I'm comparing it to Israel where this does exist to some extent in the opposition, but still a majority of the Knesset/parliament comprises parties that have existed for decades, and they are squarely in blocs that either support or oppose Netanyahu, so you wouldn't be able to get consistent 2/3 majorities for anything.
The other thing that really surprises me is just how extensive the support for these impeachments is. My understanding was that Keiko Fujimori has a very consistent base (hence near 50% in three consecutive elections), but that she's a polarizing figure, so much of the other half of the electorate hates her. But then it's surprising to me that in the many impeachments of e.g. Vizcarra and Kuczynski, there wasn't a rallying around these figures just on the basis of their opposing Fujimori.
I will be glad to hear of any other thoughts of yours on the subject!
Fujimori ran for office opposing the "traditional parties" that were the norm before, these were APRA and some more (Acción Popular, Partido Popular Cristiano and two or three more if I'm not mistaken). Once he won he liberalized the formation of political parties.
Also the impeachment (technically not but it would be too long to detail the reason why it's not an impeachment) wasn't used at all until 2017 (not counting the symbolic vote after Fujimori quit the presidency via fax from Brunei). Keiko Fujimori lost in 2016 to Kuczynski but since she had a huge majority in the legislature her parliamentary bloc set out to rule the country from there and sabotaged Kuczynski who was forced to resign.
Keiko has a huge anti-vote so it is almost assured that her rival in the ballotage will defeat her. I hope that she doesn't run again for office next year.
Regarding Israel I have also noted some similarities although I don't know how Netanyahu is still in office.
Thank you again, one last question to round this part out. In 2020 snap elections Fujimori lost big time (as I understand it) and Vizcarra was still removed, and in 2021 Fujimori's party was pretty small but Castillo was still impeaced twice before being removed the third time. Is there a stable Fujimorist coalition that extends beyond Fujimori's Popular Force that can explain this continued success? In other words, I'm seeing that Fujimori's party doesn't have enough support by itself to impeach, but I'd understand if other parties are sort of known as Fujimori allies and so the base is larger than I'm seeing as an outsider.
As for Netanyahu, I and many fellow Israelis don't get it either. But I think it's a similar question to the enduring support for Trump.
Vizcarra ascended in 2018 due to being Kuczynski's VP when he (Kuczynski) resigned. He then dissolved Congress in 2019 and called for an special election for a new Congress to complete the 2016-2021 period. This new Congress proceeded to impeach him in November 2020. Manuel Merino as president of the legislature became President for 5 days and then resigned due to popular outrage over 2 people killed in protests against his regime. After that Francisco Sagasti was elected as president of the legislature and ergo of the country until July 2021. The 2021 election was highly polarized but heavily fragmented. Castillo and Fujimori went to the run-off but a very low percentage of the vote. Castillo won by a razor-thin margin and went on to be an awful President until December 2022 when he attempted to do an autogolpe similar to the one Fujimori did in 1992 but without military support he was quickly impeached. His VP Boluarte then became our first female President until this month when Congress grew tired of her and impeached her as well.
Fujimori's party technically doesn't have the votes to impeach but they have alliances with other parties which are similar so with that they reach the threshold of 87 votes.
Thanks! Which are the other parties that you consider similar to Fujimori's? I.e. that would probably cooperate with her whenever she starts impeachment proceedings. I think that's the piece of the puzzle I've been missing.
Because democracy is just a lens which amplifies the characteristics of a society and Peru is a low human capital country. We in the West have a dogma that democracy makes things awesome because we're democratic and also awesome, but I think that's backwards or at least incomplete. Democracy didn't make us awesome, we already were awesome and therefore we made democracy. Trying to make a dysfunctional country better by installing democracy is the cargo cult of politics. There's a human capital threshold required to make democracy work and I suspect that Peru just doesn't clear it.
I think just about every sentence here is baseless, but just to be clear, I don't think Peru's politics are much more dysfunctional than those of Italy or indeed the US. Repeated impeachment is one very specific form of dysfunction, and I'm curious as to why it shows up so strongly in Peru specifically, but I don't think it has "less human capital" than neighboring countries, nor that high human capital is "required to make democracy work."
We disagree then. High human capital is a prerequisite for democracy not a guarantee, hence Italy. Is Peru meaningfully more dysfunctional than the average LatAm country?
Peru has an exceptional recent history of impeachments and arrests of its presidents. Dysfunction is a very broad category and it would depend a lot on what you mean by that. But that's why I think "making democracy work" is too vague a term.
I don't know the exact answer to this, but I'll try to argue the one I think most likely.
First, is Peru really that exceptional among Latin American countries? As you mentioned, a fractured Congress means little partisan support for the president. It also means a higher chance of a successful impeachment process (*). Let's take a look at the degree of fragmentation of Latin American party systems:
- Argentina: One big party holds roughly half of all votes (Peronism), with the other half split among various smaller parties. Only two presidents from the smaller parties in the last 35 years, one of whom resigned from office.
- Colombia: Until fairly recently had a stable two-party system (Liberal party and Conservative party).
- Chile: Until recently a stable two-coalition system (three decades).
- Mexico: Three-party system during the last couple of decades. Before that, something close to a one-party system (PRI).
Assuming corruption levels are similar, maybe you don't see more impeachments/resignations/arrests simply because it's harder to impeach/prosecute a (sometimes former) president when he/she has a strong party to back him up?
Also, causation might go the other way. Alberto Fujimori (Peru pres 1990 - 2000) founded his party less than a year before being successfully elected. Alejandro Toledo (pres 2001 - 2006) founded his party only seven years before he became president. The coalition that brought Ollanta Humala (pres 2011 - 2016) to office was formed less than a year before the 2011 election (though Humala's party had been founded four years prior). Kuczynski's (2016 - 2018) party was founded two years before the 2016 election. Martín Vizcarra (2018 - 2020) doesn't even seem to have had a political party backing him up. Pedro Castillo's (2021 - 2022) party, Free Peru, seems to have a somewhat sketchy existence before the 2021 election (or at least 2020 election). Dina Boluarte (2022 - 2025) was supposedly affiliated with Castillo's party but was expelled from the party before assuming office.
Maybe you can't have a non-corrupt governing party when every ruling party seems to be just a vehicle for the latest, temporarily popular, presidential candidate. Members of a long-standing party, which has been around for decades, have an incentive to behave well and to police their colleagues' behavior so that the party has a chance at winning the next election. They have an incentive to expel their own colleagues if they don't behave properly.
UPDATE: I should have mentioned Brazil. Brazil's party system is very fragmented (maybe as much as Peru?), and looking at the country's last four presidents we can see that one was impeached (Dilma Rousseff), another one was sentenced on bribery charges (Lula. Though charges were later dismissed), and the last one has recently been sentenced to 27 years of imprisonment.
(*) As regards to arrested presidents due to legal proceedings, if we assume a not so clean separation of powers, legislative and judicial, then maybe there is some correlation between being a president backed by a powerful party - with a high share of parliamentary seats - and the likelihood of surviving legal proceedings unscathed.
This is exactly the kind of response I was hoping for, thank you!
I do think Brazil makes the most sense as a comparison, and it's true that Rousseff was impeached-- but Temer, for example, wasn't, even though he had rock-bottom approval ratings and was in a sort of similar position as Boluarte and others who succeeded to the presidency because of a prior impeachment. Neither Bolsonaro nor Lula were impeached, despite (as you say) the fragmented parliament. The fact that Peru is unicameral might end up being the issue, but the unanimity in that chamber on some of these votes makes it hard for me to believe that another chamber would have made such a difference.
Obviously it's a small N situation throughout, but I do think it's very curious.
I do agree that fragmentation in Congress must have something to do with it, but as an Israeli, I see how Israel's unicameral fragmented legislature still ends up looking like a polarized chamber that wouldn't ever be able to muster consistent 2/3 majorities on anything (the instability in Israel is from a polarized system where each bloc is close to 50% and there are a couple of parties that tend to stay out of all governments). So the question is why this doesn't end up being a story of polarized blocs-- let's say a Keiko Fujimori bloc and an anti-Fujimori bloc-- such that each wields close to 50% support, but nobody can get to 2/3 consistently. And this has happened to both center-right presidents and leftists.
So let me ask differently: why do you think Peru has this system where parties tend to just be personal vehicles and there are such huge swings from one election to another? I agree with you that that seems like part of the puzzle, just not sure why it's the case.
> The fact that Peru is unicameral might end up being the issue, but the unanimity in that chamber on some of these votes makes it hard for me to believe that another chamber would have made such a difference.
I agree.
The more I read about Peruvian political parties, the more convinced I become that the difference between Israel (or other countries with highly fragmented legislatures, like the Netherlands) on the one hand and Peru on the other is the lack of continuity of political parties. Toledo's party doesn't exist anymore; Humala's party seems to be gone; Kuczynski's is also gone, and Alan García's party - the APRA party, which is the one that could be characterized as a traditional party - didn't take part in the last election (2021) after getting only 2.72% of the vote in the previous one (2020).
> let's say a Keiko Fujimori bloc and an anti-Fujimori bloc-- such that each wields close to 50% support, but nobody can get to 2/3 consistently
My outsider opinion is that the only thing that unites all Peruvian political parties, left or right, is their hatred of Fujimori. They've managed to make their anti-Fujimori cordon sanitaire work so far (*), but I don't see how they could ever cooperate as a unified block.
> So let me ask differently: why do you think Peru has this system where parties tend to just be personal vehicles and there are such huge swings from one election to another?
Honest answer is I just don't know. Maybe all these new parties aren't ideological enough to rally around traditional political ideologies one their leader falls in disgrace?
(*) Keiko Fujimori got more than 48% but less than 50% of the vote in three consecutive elections.
So I think the comparison to Israel is useful here, if only because I know it well :)
(Of course it's a parliamentary system so there isn't impeachment, and indeed the years 2019-2022 did feature massive instability, but as mentioned you still couldn't get 2/3 for just about anything).
The Israeli opposition also has a lot of partisan flux, ever since 2006: parties are mostly vehicles for a given candidate. Israel used to have a system where two major parties-- Likud and Labor-- shared about half the seats, and smaller parties made up a given coalition. But since 2006, you did have that story on the right-- Likud is still there and its coalition partners are broadly what they've been for 30 years-- but the opposition is full of new or relatively new parties, and Labor has become extremely small.
Netanyahu and Fujimori is not a bad comparison, given that both are popular enough to be consistently first or second in terms of leading the country, but opposition to them is not only fierce, it is practically what *defines* the opposition in Israel.
But still, in Israel what this has created is a highly polarized space where the two blocs each get close to 50% support each election, so you could never get 2/3 for anything. In fact, the brief non-Netanyahu government in 2021-2022 had parties spanning Islamists to right-wing Jewish nationalists, united primarily by not wanting Netanyahu back. And in that government, the Prime Minister's party had just 6 seats in the 120-person parliament. So when you posit Fujimori as this kind of highly polarizing figure, it's strange to me that there's such consistent support each time for impeaching the non-Fujimori president, rather than a circling of the wagons.
Just to be clear, I think you're completely right in seeing what's causing this situation, but exactly why that factor is what it is is not yet entirely clear :)
South America does have a reputation in general for corruption, revolution, and unstable governments. Why Peru in particular is a good question, perhaps the influence of the Shining Path in the 80s predisposed to the kinds of 'vote for a strong man' reaction which would lead to a lot of corrupt leaders/unpopular with the other parties who then push for impeachment that we see elsewhere?
Nearly all countries in Latin America have presidential systems. One way Peru is different from neighboring countries is (apparently) the presence of a Prime Minister in addition to the President, as in the French system: https://en.wikipedia.org/wiki/Semi-presidential_republic
But SK doesn't have that feature and I'm not sure why that would matter, as the PM seems to have very little power.
Wiki lists 4 successful impeachments for SK (successful in that the impeachment vote passed, not that the president was ousted, i.e. by this count Trump was successfully impeached twice). Twice a president was removed.
Equivalent numbers for Peru are 7 and 4, which is something of an undercount because one president resigned to avoid impeachment, etc.
You need two-thirds of parliament (it's a unicameral legislature) to impeach. This threshold, has been frequently surpassed in recent history-- the most recent impeachment was unanimous, 122-0 (out of 130)! Getting two-thirds majorities of legislative bodies after multiple elections in a democratic country seems bonkers to me.
It seems to me that Peruvians have come to see an impeachment vote in their system as little more than a vote-of-no-confidence in most parliamentary systems. That is: it's not an affirmative claim that 2/3 of the country thinks the president is unqualified, but rather that fewer than 1/3 of legislators are happy with the current coalition agreement.
(Of course, this doesn't explain why other South American countries don't impeach the same way. But public perceptions of constitutional procedures are often quite historically determined.)
As an American, I wouldn't be bothered by seeing a President get impeached just because he was very unpopular. If you can win an election, but two years later 2/3rds of the Senate is against you, you probably don't deserve to be President anymore.
I think in most/nearly all countries, when the opposition can topple the leader, they will, and they won't split hairs on deeper questions about whether this is appropriate. I'm thinking Brazil for example-- if the worker's party could oust Bolsonaro, I'm sure they would have, but I think they just didn't have the votes. In parliamentary systems this happens all the time. Question is why Peru is so uniquely impeachful (and successfully so) among presidential systems.
For the conference/unconference. I want to commend and endorse 100% your insistence on using the term "conference" rather than "unconference" to describe something, that based on current common usage of the term in english, is a conference.
I can just about conceptualize the outlines of what an unconference might be, but it would certainly not involve people with an interest in a particular topic gathering at a pre-advertised time and location, to hold discussions on that topic.
Personally, if anyone knew what it meant, I would just call it "open space technology event", but well...
A conference, the way I see it (not a native speaker) tends to be quite centralized, with speakers and an audience, and... some limited space for socializing.
What I consider an unconference is a place where everybody can propose sessions, where the socializing IS the main aspect, but it's not just a "no score meetup".
FooCamp and BarCamp use that format. RubyCom does it, some Python and PyData events too. The European Juggling convention is similar, the way they used to host HackIt! (Italian Hacktivist convention) too.
So I will stick to my "this is not a conference" :-)
Nothing highlights the unseriousness of a movement or a cause than inventing a new term for a common concept, and a tone-deaf insistence on using it, #actually.
AI "will kill everyone", but it's important that we now say "unconference".
For what it’s worth “unconference” is a moderately common term. I’ve never been to an event that calls itself that, and I don’t have a sense of what one would be like, but it’s easier to use an established word like this than to insist on using familiar words in unfamiliar ways.
And I don’t think anyone is yelling at people for misnaming such an event by calling it a conference.
Reading the linked page it sounds a bit hippy-dippy (I'm sure there are more current terms to describe this mindset, but I am old, Father William):
"OK then, what is an unconference?
We start together, build the agenda together, then split into parallel sessions.
We have only one law, that I adore:
the Law of Two Feet: If you’re not learning or contributing, move to another session (or take a break, or do what nourishes you). Guilt-free.
And 4 (and some) principles:
Whoever comes are the right people.
Whatever happens is the only thing that could have.
Whenever it starts is the right time.
When it’s over, it’s over.
(Take responsibility for what you care about.)
Format (in this event): Opening circle → marketplace (we propose sessions) → four session blocks → group synthesis → social"
I notice that there doesn't seem to be any mention of tea breaks, so presumably you have to bring your own or go out and buy soemthing to eat/drink for the marked break time instead of getting a free cuppa tae. Tsk! Not very sociable! Maybe they *will* provide same but just haven't mentioned it explicitly, but the only benefit of going to these things is you at least get a cuppa tea and a bikkie on the house.
"The organizer" here: yes, apart from a break in the middle, it's very "choose your own adventure".
If you need a tea break, take a tea break. If you want to skip a session, skip one (unless you are organizing that, then I will gently scowl at you).
Regarding "tea present": all the donations go to Newspeak House. This has no sponsors nor anything, so maybe we will have some drinks, maybe not, I will know closer to the event :-)
So it's a conference with no pre-planned agenda? (Or rather, it sounds like the agenda is Step 1: Find out what people want to talk about, Step 2: Breakout sessions to talk about that.)
Like, I don't know how productive that is, but that sounds different enough to be worth giving a different name.
NY State Assemblyman Alex Bores has announced he's running for Congress. Bores was the sponsor of NY's RAISE act, which passed before California's SB 53 and is similarly aimed at addressing catastrophic and existential risk from AI. He seems to be making AI safety a priority, and I've heard from several people in the field that they'd be really excited to see him in congress.
As far as I know one other person is officially in the race (Micah Lasher, who's probably the favorite). The seat is currently held by Jerry Nadler, who isn't running for re-election, but Lasher is seen as the "heir apparent."
Thanks! I don't have a huge amount of context here, but it also seems to me Lasher is the clear frontrunner, good to hear from someone with more context.
It seems to me having a person who already has made AI safety one of their top priorities is *much* more valuable than trying to convince a ~random other congressperson. Bores also says he's the first NY lawmaker with a CS degree and worked at Palantir for a few years, so he has more technical background than the average congressperson as well.
So not knowing anything about Lasher, it would seem to me much less valuable to persuade him than to elect someone who's already motivated by AI safety. (For one thing, why Lasher specifically at that point?) But persuasion is still really good, and maybe there's some reason Lasher would be particularly worth reaching out to?
> and maybe there's some reason Lasher would be particularly worth reaching out to?
Presumably the fact that he's actually going to win.
Remember that guy in Oregon that Scott made such a big deal about? Out of state rationalists trying to prop up an unpopular candidate is just a huge waste of money.
Going back to the Fatima article (which I read in it's entirety but didn't have time to see all the comments so I apologize if this already got covered):
If I understood correctly, Scott's materialist explanation of the miracle is a coming together of three circumstances:
1) a specific type of (relatively common) cloud cover through which the sun can still be seen and which allows people to stare at it for a long time without damaging their eyes
2) this staring can cause people to see strange colours and a sensation that the sun is pulsating, rotating etc
3) the people staring are also psychologically primed to see a miracle
He debunks the fact that other observers in villages quite far away did not fulfill these three conditions by providing evidence that they were also psychologically primed (because if they were not expecting the miracle at all, the materialist explanation falls apart).
However, even if they were expecting to see a miracle, by all the testimonies, they saw the miracle in the direction of Fatima and not in the direction of the Sun which would have been completely different from their perspective. So in order for this to work, it would mean that they had to be outside at just the right time (presumably because they were expecting it), then they had to stare in the direction of the partially cloud-covered Sun for a while until they started seeing weird stuff, then they had to turn in a completely different direction to look at Fatima, where they then witnessed a Sun miracle in that completely different direction.
"In 1870, flat-earther Samuel Rowbotham performed a series of experiments to show the Earth could not be a globe. In the most famous, he placed several flags miles apart along a perfectly straight canal. Then he looked through a telescope and was able to see all of them in a row, even though the furthest should have been hidden by the Earth's curvature. Having done so, he concluded the Earth was flat, and the spherical-earth paradigm debunked. Alfred Wallace (more famous for pre-empting Darwin on evolution) took up the challenge, and showed that the bending of light rays by atmospheric refraction explained Rowbotham's result. It turns out that light rays curve downward at a rate equal to the curvature of the Earth's surface!"
Reading the Wikipedia article on the experiment, it looks like it depends a lot on local conditions. The canal would have had cool air near its surface and warm air above it, refracting the light more than normal. It says that under average conditions, the curvature of the Earth appears 15% less curved than reality.
It also mentions that one surveyor was able to get results matching a curved Earth because he placed his sight line higher off the ground to mitigate atmospheric effects. And one guy did a similar experiment that had so much refraction he concluded that the Earth was concave!
The whole thing when they concoct elaborate “experiments” to “prove” the Earth is flat… I can’t even. There are so many miles of coastlines, bring your telescope, see if you spot London from Calais, or Toronto from Chicago, or Lisbon from Boston (powerful ‘scope for that one, to be sure).
For Toronto you’d have to be looking from Rochester, not Chicago - and you actually can see the skyscrapers of Chicago across the lake from Michigan sometimes, I believe!
Ouch! the Great Lakes got scrambled in my head, I shifted Toronto way too far West.
Yeah, the larger point is nicely illustrated by the Chicago skyscrapers visibility from across the lake - the tallest buildings are visible because they stick out above the curve, but on a flat earth the Navy Pier should be just as visible :)
The refraction is due to warmer air being less dense. If the air gets warmer at a rate of about 0.11°C per meter in altitude, the refraction will match the curvature of the earth. This can happen over water because evaporation cools the water which in turn cools the air above the water. I doubt is can happen over open ocean under normal circumstances because the waves and wind will mix the air too much. To get the effect on land, the land would have to be cooler than the air, which is rare. So the odds of getting the correct temperature gradient around the entire equator are probably nil.
I was skimming a history of the Abbasid caliphate and ran across a reference to the Zanj Rebellion, an event I had never heard of. I read more into it, and got the gist that the there was a large African slave population in what is today modern-day Iraq, performing hard and dangerous labor like building and maintaining dikes and canals. A chararismastic slave led a multi-year uprising that turned into one of the biggest internal challenges the Abbasids faced.
My immediate thought was this Abbasid slavery practice sounded like chattel slavery in the American south. But I went through college assuming the talking point that the US essentially invented chattel slavery is true.
I'm confused now. What is chattel slavery and what isn't? Were the helots in Sparta chattel slaves? Were enslaved Africans in the Abbasid caliphate? Or is chattel slavery more narrowly just a 17th-19thC-ish western thing?
I'm not sure the distinction between slaves as chattel and slaves as real property is all that significant a distinction. Slaves were real property in Louisiana (which means they were under the part of civil law dealing with land rather than movable property), and were chattel (which means they were covered by the laws relating to movable possessions) in the other slave states. While this makes some legal difference, I'm not convinced it made all that much difference to the slaves.
Chattel slavery has to do with the specific legal status of the slaves. A chattel slave is the personal property of an individual or corporation, can be bought and sold with relatively few restrictions, and has few or no legal rights of his own and is only incidentally protected by the legal rights of his owner: he can be punished or harmed arbitrarily by or on behalf of his owner, he cannot own property or enter into binding contracts on his own, and assaults against him by free citizens other than his owner are crimes against his master. Chattel slavery is generally a hereditary and lifelong condition, although there are usually provisions for owners to voluntarily manumit their slaves.
"Chattel slavery" is used as a term to distinguish this arrangement from other forms of unfree labor where the unfree person has more legal rights, isn't the property of a specific person, or cannot be bought and sold freely. Serfdom, debt peonage, and convict labor are unfree but are not chattel slavery. Temporary and often-voluntary unfree situations like indenture and apprenticeship are also sometimes counted as form of unfree labor. Intermittent obligations of unfree labor or military service imposed on citizens (corvee, robota, impressment, or conscription) are also not chattel slavery.
Slavery as practiced in the antebellum US and elsewhere in the Americas was definitely chattel slavery, but this was nowhere near unique to the Americas. The similarities you notice with slavery as practiced by the Abbasids are real and not a coincidence: slavery as such has died out in most of Europe in the Early Middle Ages, in favor of practices like serfdom. Middle Eastern cultures, however, retained a form of chattel slavery similar to and derived from slavery as practiced throughout the Mediterranean world in Classical Antiquity and earlier. It filtered back into Early Modern Europe by way of the wars and raids between various Mediterranean-bordering parts of Europe (especially Spain) and the Ottomans, where the Spanish and other Europeans copied the ancient practice of enslaving wartime captives unless and until they were exchanged or ransomed.
The central examples of Greco-Roman slavery are usually also classified as chattel slavery. The cultural conception of Greco-Roman slaves was quite a bit different from that of slaves in the Antebellum South, but they could still be bought and sold and we're still legally property.
Helots, however, usually aren't considered chattel slaves. They had much less rights than those usually afforded to serfs, but I think they're usually interpreted as not being personal property and Spartiates don't seem to have been able to buy and sell them separately from the land they worked.
As Korakys says, chattel slavery refers to a system in which slaves have identifiable owners. This is more or less necessary wherever "slave" is a legal status, because a slave's relationship to the legal system runs through his owner.
The practice is presumably older than history is.
Note that on objective metrics like "health" and "surviving children", American slaves were among the best-off anywhere in the world.
(Chinese imperial eunuchs might be considered better off, but obviously they didn't manage to have children. The Mamluks were definitely better off. At some point you have to question whether you want to include them in the category "slaves". They were formally slaves, but they were also formally in charge of the rest of the society.)
A point worth noting that even under Jim Crow, supposed to be cruel and harsh, there was no emigration of American blacks I heard of, there were no refugees fleeing oppression.
Sure, though the timing is a bit off. Reconstruction was more or less ended and Jim Crow substantially in place by 1880, but the Great Migration doesn't kick off for about 30 years afterwards.
Meaning that migration to escape oppression doesn't "count" unless it happens immediately? Or is there some less-ridiculous argument that I'm not understanding.
is, bluntly, such a misinformed take that it makes further conversation on the subject basically impossible. It approaches Holocaust denialism in both its complete ignorance of historical record and its obvious political intent. I hope you're just legitimately misinformed here. There were millions of refugees.
I was under the impression that "Jim Crow denialists" and "people who correct others about their pronouns" were two separate circles, but I guess you learn something new every day.
I've been reading the bible for the first time and it has quite a few rules in there on what to do in various situations regarding slaves.
Chattel is being used to differentiate it from milder forms of servitude that sometimes people claim is slavery but isn't what most people first think of, like serfdom or "wage slavery". It means slavery where you can buy and sell people freely rather than the slaves being bound to the land or something like that.
Needless to say many, many societies practised it prior to America. Most notably the Romans. I think this is a case of some Americans not really internalising that fact that the rest of the world exists.
Roman slavery was hereditary, in that the child of a slave mother was a slave. It wasn't tied to observable ethnic or racial differences in the way that American slavery was, though.
Also, a slave generally couldn't become a citizen directly. A freed slave had fewer rights than a citizen, although their children could become citizens (in this respect, Roman law was much more liberal than e.g., Greek law).
Slavery was a big part of economy in medieval world.
It also contributed significantly to expansion of Christianity and Islam, because each of these religions had a taboo against capturing people of the same faith (although I think it was okay to buy them on a market if someone else has already captured them). The remaining pagan rulers converted to a new religion often because they were tired of being everyone else's popular target for slave raids.
A category of slavery no one has mention yet in this thread: sex slaves. Very popular in Arabic countries. Islam explicitly allows you to rape your female slaves (but I think you may need to release them from slavery when they have a child born).
I think that American slavery was exceptional compared to other Western societies in the same century, but not exceptional by historical standards. Some countries have chattel slavery even today, with mobile apps for buying and selling slaves.
Doesn't chattel slavery mean slavery for purely economic reasons? As in, the slaves weren't your enemies who you captured in battle, nor were they dudes who owed you a bunch of money they needed to pay off. I think in Ancient or Medieval times, these were commonly (almost always?) what slaves were, no?
Thanks, by that definition it doesn't seem like the US slavery situation was exceptional at all.
However, I do think the fact that the slaves were neither prisoners of war, nor people who owed you a debt, but were, rather, just some random easy-to-catch guys who would allow you to run your business much more cheaply makes that type of slavery exceptional (and to my mind, much more abhorrent).
The exceptional thing about US slavery is that they exhibited natural increase in population even after the slave trade ended, rather than a Malthusian existence in which they couldn't afford children (the norm for slavery in the ancient world).
I have some questions about this. I assume helots reproduced normally. (They aren't chattel - they're owned by the state - but they are formally slaves.)
Bret Devereaux estimated 5th century BC Athens as being about 40% citizens, 40% slaves, and 20% free noncitizens. Here the slaves *are* chattel. But there are so many of them that my first guess would be they also reproduce normally and the bulk of Athenian slaves are just the children of other Athenian slaves.
> However, I do think the fact that the slaves were neither prisoners of war, nor people who owed you a debt, but were, rather, just some random easy-to-catch guys
That's not really right either. Originally they *were* prisoners of war. Their captors then sold them to Europeans to be sold into American (North or South) markets. The Europeans weren't going around capturing random Africans; they bought them, already enslaved, from more victorious Africans. That's generally also what the Arabs did, although I'm pretty sure I've heard that sub-Saharans who made the trip to Mecca might be enslaved for their trouble.
(As the trade developed, there almost certainly was a growing element of "we caught these guys because the market is good".)
Because American (USA) slaves had such an impressive rate of natural increase, this came to be a much less significant source of slaves there. Instead, slaves were mostly born into slavery as the child of local parents.
There's nothing historically unusual about slavery being an inherited status. But the American (USA) situation is unusual in that the population of slaves was able to sustain itself this way. Further south, they constantly needed newly imported slaves.
---
It isn't really clear to me why slaves captured for the purpose are more offensive to you than slaves captured in war. As far as I can see, those aren't different situations. If you're taking slaves from another group, you are at war with that group. And the main purpose of war is to confiscate resources belonging to the other side. Slaves are one of those resources.
The other reason slaves were bred locally was importation being prohibited starting Jan 1, 1808, the earliest date permitted by the United States Constitution, per Article I, Section 9.
Because in a war it quite often might have been the case that the "other" side was the aggressor and when my side happened to be victorious I used the captive bastards who had been attacking me for decades as slaves to compensate all the suffering and loss they had caused me.
In the African slave trade, as far as I have read, the "wars" going on were specifically and primarily to go out and capture slaves for sale. I agree that in some sense they were just another "resource", but it seems somehow uglier that one would make war on another group only for the purpose of enslaving them, rather than primarily to steal their riches or their territory.
But I suppose I can understand not seeing that as a big ethical difference.
A review on a Bob Dylan song made me want to try to write something in the style of Dylan. I remembered a song of his, but couldn't remember any of the words, so was trying to go through titles that might possibly have been the song, and re-stumbled across The Hurricane:
Been a while since I listened to the fully political ones. Man, he pulled no punches at all. That... would suck to try to copy. The worst-case scenario would be managing to pull it off.
And then I found the one I was originally looking for: It's Alright Ma
Time to update the Milei scorecard? Begging Trump for a $40 billion bailout hardly seems like the kind of thing that counts as successfully turning the economy around.
It’s no longer interesting as a great natural experiment, unless somebody in the EA movement wants to give $20 billion vouchers to other countries presidents to get a good comparison group. I am honestly disappointed by this.
I’m still hoping Argentina makes it work because they could really use a break, but the precedents are not encouraging.
It's a complicated situation. My understanding is that the anti-Milei party won a local election, the market panicked that Milei would lose, they sold off pesos, and the US intervened to prevent a currency crash. So it's not really a clear-cut "Milei good" or "Milei bad" narrative, just more of Argentina being a basketcase.
The peso fell 5.6% on Sept. 8 (the day of the election), but the value of the peso was declining before then and continued to decline after, until the Trump Administration intervention on Sept. 22. The two week gap between the election and the intervention would seem to contradict the idea that the intervention was a response to a market panic caused by the election.
Furthermore, a market panic, as the term is normally understood, would cause the peso to become undervalued. Analysts are saying that the issue is that the peso is overvalued (see links below). The $20 billion just puts off the day of reckoning.
Yes, the IMF is a lender of last resort that helps countries in major crises.
But the US Treasury has only intervened in international currency exchange markets three times since 1971:
- in 1998 in Japan to stop the Asian financial crisis before it could spread to the rest of the world.
- In 2000 to help the Euro. Again, worried about global stability.
- In 2011 in Japan again, after the earthquake.
These were interventions in strong currencies like the yen and the euro. There’s no precedent for something like this in an emerging currency, and certainly not for doing it to help a foreign politician try to win a lost election.
Right now there are stimations that Scott Bessent directly bought pesos for around one billion USDs in the last few days to keep the rate artificially low before the elections. That’s more or less the same amount used in the three previous interventions mentioned above.
To understand why Milei failed, there’s one thing most people ignore that’s important to know: Argentina doesn’t have access to international markets. This means the country can’t roll over its international debts and needs to pay them from its own reserves. That alone would destabilize almost any country in the world. This happened because, from 2015 to 2019, Argentina had another outsider libertarian president aligned with the United States called Mauricio Macri. Macri took on a lot of debt and couldn’t pay it back, putting the country into default and implementing monetary control measures (Argentinians call it “cepo cambiario”). Btw, Macri and Milei’s administrations share several key ministers, like Caputo (both Macri’s and Milei’s economic minister), Sturzenegger and Patricia Bullrich. It’s basically the same government doing almost the same economic policies. So, if you want to know what’s happening to Milei now, just look at Macri’s 2018 crisis. But, anyway.
Milei tried a simple plan. In fact, it’s the same one that was tried several times before: in the 70s and 80s by the neoliberal genocidal dictatorships, in the 90s by Menem, and as recently as 2015–2019 by Macri. The idea was to keep the USD price artificially low and freeze the economy with austerity measures so no one can buy anything. That way, prices can’t go up, so inflation gets controlled. The USD price is crucial because many products have imported parts that must be paid for in dollars, so if the USD price rises, prices go up immediately too. This plan can kind of work for some time, but you need a large amount of USD reserves to keep the dollar price low. It has the advantage of being super populist and almost guarantees election wins because it artificially boosts the peso’s value. A lot of people can buy cheap imports, cell phones, travel to Miami, etc. It’s exactly what the right wing middle class wants. On the other side, poor people lose access to social programs, small businesses go under because no one buys anything (especially local), industries lose competitiveness and jobs decline.
Milei expected his austerity measures to signal to the markets that he was willing to do anything to pay international debts, and that way he would regain access to international financing. If he could roll over the debt and get new funding, he could keep the USD price low and inflation in check, stabilizing the economy. Sadly, he had a fatal flaw in his understanding of how markets would react. Markets do care about a country having a fiscal surplus (Milei doesn't really has a fiscal surplus, but whatever) and being able to tolerate austerity, but what they care about most is whether it has enough international reserves to pay them back. If they see a country has no reserves, they immediately understand it can’t pay its bonds and sell them off.
So, for Milei it became a race between having enough USDs in the country’s reserves and winning market confidence to regain access to financing. The problem was that he refused to boost the reserves. Time and time again, economists from all sides of the political spectrum pointed this out, but he just insulted them (calling them “mandriles/baboons” lol). There were several financial runs each time the country ran low on reserves, but each time Caputo would pull a “white rabbit” (that’s what economists called these measures, like a magician pulling a rabbit out of a hat). For example, during the second half of 2024, the country was near a terminal crisis and Milei rushed to grant an amnesty for undeclared dollars. Basically, lots of Argentinians have huge amounts of undeclared USDs saved, and this amnesty allowed them to declare them with almost no taxes or explanation. That temporarily boosted reserves. There were other “white rabbits” that helped him stay afloat, but it’s obvious that if your entire economic plan depends on finding billions of USDs at the last moment to survive, you don’t have a solid economy. When Milei ran out of USD reserves again earlier this year, he asked the IMF for a rescue (on top of Macri’s 2018 $40 billion bailout). Now that he’s run out of reserves again, he’s asking the US Treasury. Everyone, including the IMF, keeps telling him that holding the USD price so low is unsustainable and that he needs to accumulate reserves, but he keeps doing it, because the moment the dollar price rises, inflation will go up too and he’ll lose the midterm elections.
Of course, this economic plan makes no sense and is hurting the economy terribly. The country is in another recession, inflation is going up anyways, most people are having a bad time. Meanwhile, Milei keeps taking on international debt just to burn through USDs so the dollar price doesn’t go up. All this money gets wasted in capital flight and speculative finances, it doesn’t go into the real economy. And it’s debt Argentinians will have to pay back for years and years. It’s funny, because this debt is what got us into the problem in the first place, so he’s just making it worse. With the US Treasury involvement, it’s even worse, because you can at least default on private creditors, but... what will Argentina do if it can’t pay back the United States? Who knows what kind of concessions the country will have to make? This loss of sovereignty, depending on Scott Bessent like this just to win a midterm election, is probably one of the most irresponsible things an Argentine president has ever done. It’s pretty much treason to the country.
And about the problem being that he lost a local election... well, there is a reason he lost it in the first place. He is doing a disaster. And I'm not even geting into the political and corruption scandals like the Libra scam, Karina Milei (Milei's sister and his biggest advisor) taking bribes from the disability pensions system, Milei's main candidate for the midterm elections (José Luis "el profe" Espert) having to step down after being funded by a drug trafficker, etc.
Anyway, if someone wants extra context or sources, just ask, I can do my best to find them.
An interesting take, given the inflation shit-sandwich Milei was handed, collapsing poverty rates, and the fact that LLA did better in the recent BA election than it did during the presidential race. Unless I'm looking at flawed data, BCRA foreign reserves appear to have been increasing since Milei took office.
I agree the ARS still looks overvalued, and wonder why Milei didn't, or maybe couldn't, let it fall far enough during those first few hyperinflationary months.
In any case, the most liquid market I can find (https://polymarket.com/event/which-party-wins-most-seats-in-argentina-deputies-election) gives LLA 73% odds. From my screen it looks like the ARS is trading beta to LLA's chances.. I guess we will see how much of Argentina's market troubles are due to monetary issues, and how much is simply a crises of confidence started by the fear of a Peronist win.
The latter is a strange scenario. It's like ordering a medium-rare steak at a restaurant, getting something medium-well, then demanding the restaurant take the steak back and serve you cyanide instead. Milei may be far from perfect, but someone like Kicillof is just more of the same money-printing (poison).
It's not as if any sizeable country on the planet has economic policies that make much sense. They're all varying degrees of horrible.
As an American, I support using my tax dollars to shore up confidence in LLA, if indeed this is a crises of confidence. We'd all be better off if Argentina or South America generally adopted markets, and joined the rest of the rich, industrialized world. LLA appears to be the only path forward here.
Trump of course is mostly doing this because he personally likes Milei, likely due to flattery.
> We'd all be better off if Argentina or South America generally adopted markets, and joined the rest of the rich, industrialized world. LLA appears to be the only path forward here.
Interestingly, I just read Amy Chua's World on Fire (same author as the famous Battle Hymn of the Tiger Mom, which is basically a how-to manual for grinding your kids into Harvard), and in it she argues that it's actually a fairly significant mistake to combine Washington Consensus free markets + democracy in most of the world.
Just as Joe Studwell argues there's a "success sequence" for countries to follow that relies on non-free market practices at first (land reform increases ag productivity, then get into manufacturing at the bottom end, install tariffs and protectionism to shelter your nascent industries as they 'git gud' and climb the value ladder via targeting objective standards via "export discipline," and finally you drop the tariffs and just compete on the world stage as a fully developed country like S Korea or Japan), Chua argues there's a massive hidden land mine inherent in going "full free market" and "full democracy" all at once, especially combined in a package. Being fully open to international trade free markets are generally a mistake for developing economies, because they will never "git gud" and won't manufacture stuff relevant on the world stage, instead choosing to buy everything on the world market, and ending up in a place where the only value they provide to the world market is commodities and / or cheap labor. Democracy is also a problem in this regime, because it will end up impeding economic growth, too.
Because after all, what were the founding father's of the USA worried about when it came to democracy? That the great masses of non-property owners wouldn't respect property rights, and would vote themselves largesse at the expense of the people with assets. This is exactly why suffrage was rolled out to a very minimal set of people (white male property owners), and was only gradually expanded over the next ~150 years, to include all white males, black males, and eventually women at different points. We've only had full universal suffrage in the US for around 100 years.
This "problem with democracy" actually happens repeatedly in a lot of the world, because a lot of the world has what she calls "market dominant minorities." For example, all throughout SE Asia, the Chinese are ~1-3% of the population, and ~70% of the entire economies of Philippines, Vietnam, Indonesia, Malaysia, etc. If universal suffrage is suddenly dropped into the ring, a lot of them have used it to expropriate that wealth - and wrecked their economies while doing so.
So dropping full Washington Consensus free markets on developing economies? Bad idea, leaves you trapped in a place where you import everything nice, and leaves you at the bottom of the value chain where you only offer commodities and cheap labor.
Dropping full suffrage democracy on developing countries all at once? Also a bad idea, because it often leads to populists getting in, wrecking the economic engines of entire countries and voting in ultimately self-pwning policies. People in the aggregate often fail the marshmallow test.
Dropping them both together onto an unsuspecting country? Lol. Well, this is why only ~3 countries have graduated from "developing" to "developed" status in the last ~50 years (Ireland, Israel, and Chile).
And just incidentally, much of S and Central America has "market dominant minorities," too - generally composed of hacienderos, large land-owning families descended from Spanish colonial times, and other immigrants, versus the much larger and more predominantly native-American-ancestral populations elsewhere in the country.
Yes, Milei started from a bad position (for example, as you said, he started with little reserves in the Central Bank). That's why he took office with a consensus across society on the need for a stabilization plan. The problem is that after two painful years of austerity, the plan failed and all the people’s effort was wasted.
It is true that poverty is probably somewhat stabilized, because reducing inflation is a really good and progressive measure that helps the poorest sectors the most, but it’s nowhere near “collapsing". There are a lot of methodological flaws in how inflation and poverty are measured. For example, INDEC (the National Institute of Statistics and Census) uses a basket of products and services from 2004 to measure it. There were almost no cell phones or internet back then, and those are the services that have risen the most. Milei refuses to update it. Heck, several important INDEC officials literally resigned rather than go along with the data falsifications, including the Director of Living Conditions Statistics and the Director of the Consumer Price Index. Milei wouldn’t be losing so badly in every election and needing so many bailouts if poverty were really collapsing.
In the recent Bs. As. election, LLA allied with the other major right wing party (“PRO”) got 33% and Peronism 47%. During the presidential race, LLA got 25%, PRO 24%, and Peronism 42%. So, just looking as LLA as the party name did better, but that’s a pretty misleading reading of the situation. In reality, Milei and his allies’ votes were reduced from 50% to 33% and Peronism went from 42% to 47%.
As you show with that prediction market, LLA will win more seats than any other party in this upcoming election, but that’s just because it has to renew far fewer seats. It’s also the only party participating in every province. But there’s no doubt it will end up with a pretty small number of seats overall (for example, the "Pro" party, it's biggest ally, will also lose a lot of seats) and will need to work with other parties to pass legislation.
About why the peso is being overvalued, Milei does it to keep inflation in check. Also, it’s the kind of country he wants: one with cheap imports, no national industries, and a massive concentration of wealth in as few hands as possible. It’s exactly the neoliberal Menem model from the 90s (he’s a big fan of Carlos Menem, in his own words "the best president of Argentina in the last 50 years". You can look at how his decade in power ended, check out what happened in 2001). He could have let the peso float more, buying reserves in the meantime, but that would have meant a slower reduction in inflation. There were no hyperinflationary months, btw.
If this crisis were "simply a crises of confidence started by the fear of a Peronist win", why did Milei need so many rescues even before this election cycle? The economic plan is just unsustainable. That’s been the consensus among economists for months now. Even Domingo Cavallo said so.
You mention Kicillof being a money printing poison. But Milei hasn’t printed any money and is still desperate on the brink of collapsing the country. It’s obvious that the "money printing" narrative repeated by Milei ad infinitum is a lie. But even then, Kicillof is not stupid, society got tired of inflation, and he has no fiscal deficit running the Buenos Aires province. Do you really think he would have done worse job than what Milei is doing if he were in his position? I doubt it. Also, remember that the loss of access to the markets and the last “cepo cambiario” (monetary control measures) were implemented by a "liberal" right wing president, Macri, not by a Peronist. And that the high inflation from 2019 to 2023 is a direct result of Macri’s failure. Argentina had two options during the pandemic, to let people die or to have inflation, there is no other thing you can do without financing.
If you want Argentina to join the rich industrialized world, is better to not support economic measures that has as a literal objective to deindustrialize the country and turn it even more into a primary materials and minerals exporter, with no added value to it's products. Milei is nothing new, what he’s doing has been done through 90% of Argentina’s history with obviously failed results.
Haven't followed this closely , but it could be bad for most Argentinians, bad for most Americans, and good for Trump personally (plus maybe good for some specific people with some vested interest in whatever the money ends up going towards).
It is surprising though that Milei is trying to prop up the currency at unrealistic levels, when one of his big things was supposed to be ending that kind of practice. He promised dollarization after all.
Scott Bessent has said he wants $20 billion from other sources to complement the $20 billion from the U.S. government. I’d suspect the odds of that happening are fairly low. Trump has suggested more money will not be forthcoming if Milei’s party does badly in the Oct. 25 election. If the party does well, the Administration may figure that the promise has accomplished its objective, and see no reason to actually follow through.
Why aren't therapy books used more in the psychiatric system? Scott has mentioned in some posts that if you're seeking help for a specific problem/illness (he mentioned depression, anxiety, OCD, maybr borderline?) then reading a therapy self help book seems to have as good results as therapy.
Why isn't it utilized more? I can imagine for many patients1p it also feels less scary than speaking to a human. Not to mention a billion times more cost effective, will take pressure off the (in my country with socialized healthcare, at least) completely overloaded psychiatric system with insane wait times, etc.
Related to this topic - I'm wondering if people know of instances where therapy was actually worthwhile, either personally or for others. My general impression of therapy is pretty negative (it seems like they mostly offer pills and a shoulder to cry on), but I also don't have a lot of experience with it. Does anyone here know of cases where someone went to therapy for a specific problem and the therapy solved or significantly reduced that problem?
Pills for ADHD work great for many people. So there therapy is often just a diagnosis + pills but that's probably the way it should be and often works great for the patients.
One of the major underlying conditions affecting many people who need therapy is akrasia.
If they were capable of taking disciplined steps to confront and solve their problems on their own, they wouldn't need therapy in the first place.
For many people, having a social obligation and a financial stake makes them actually show up to therapy reliably, whereas buying a book and thinking they should read it doesn't actually make them read it.
This is the sort of the same problem as 'why do we have any treatments for obesity when anyone can solve it with diet and exercise', The people with whatever mix of discipline/energy/low cravings/etc etc etc would allow them to solve it that way have already done so, the people who still need help are the ones unable to accomplish that method who need something else.
I can read self-help books till my ears fall off, but I'm not going to implement any of the changes unless there's a real person standing over me with a whip to force me to do so.
I don't know how responsive that is; the claimed result is "reading a therapy self help book seems to have as good results as therapy".
That's different from "implementing the recommendations in a self help book seems to have as good results as therapy". Medicine is riddled with this problem; most notably, taking a placebo faithfully leads to much better outcomes than taking the same placebo irregularly.
One of Scott's articles for covid made the observation that masks-for-the-public research is more about the intervention of telling people to wear masks than about the intervention of having people wear masks, because you can't do the second thing. You can tell people they should wear masks and see what results you get, so that's what the research measures.
But in this case, that's 'good'; the research is also measuring reading the book more than it's measuring "correct implementation of the book".
I can only speak from experience regarding depression, but if you ask a depressed person to do x or y simple thing (excercise, reading self help, eating better etc) that is proven to help alleviate symptoms, they usually will not do it. Metaphorically, they are standing a looking at a cliff. You might be shouting down to them "There's a path up!" but all they hear is " You can just leap to the top!" or "You can just scale the cliff!"
Seeing a therapist is kind of like sending someone down to guide them to the path. It might be arduous, but its easier with a sherpa.
Your comment reminds me of one of my favorite ACX essays of late: https://www.astralcodexten.com/p/singing-the-blues I also agree with Deiseach. I'm currently in physical therapy for a back issue, and the willpower needed to exercise on my own is gargantuan compared to the willpower needed to spend an hour with my physical therapist. Humans are an odd lot.
An important factor is likely that many people just don't really read books anymore, so setting people who are already more likely to have attentional issues reading would probably be fruitless for a large percentage of people.
It also doesn't help that patients can differ significantly in their reading proficiency, so you would probably have to take care to assign books on an individual basis to match patients' abilities.
Scott, I don't know if you've seen the concerns raised about cheating in the review contest. But if you have, and you just ignore them and don't address them head-on and take an explicit position on what you consider fair behaviour and why...I'm seriously considering cancelling my subscription.
I find it difficult to reconcile all these accusations with the author's previous review. Someone's claiming he's a "big time marketing guy" but last year he wrote an 11K word review of Marvel Comics. That's not the action of a cold-blooded review manipulator. That's the action of a geek who gets really obsessed with stuff and likes to share it -- which is exactly the people we expect to write reviews in this contest.
"To the creator: if you don't spend time and effort publicly acknowledging my pet issue, I will stop supporting you' is almost always a silly bit of emotional blackmail that shouldn't be treated with any dignity.
You can raise issues if you think they're important. If the creator/the rest of the audience disagrees, then trying to add this little bit of blackmail to emphasize your point just reeks of entitlement and narcissism.
I feel like it doesn’t really matter? It’s an extremely low-stakes review run by one guy. There isn’t even really an easy solution for influential people using that influence to encourage others to vote. Unless there’s a serious attempt at mass manipulation (which I believe there was in the past that was found out) it seems like things have worked pretty well so far.
> A few months ago, no one on Twitter had heard of Alpha. Today, it's the most talked about school in the world. Love it or hate it, it is causing people with the power to change education to rethink their priors.
> A lone writer's piece can dent the universe.
> Thanks to AstralCodex and the Alpha parent author.
So in Liemandt's own public estimate, this review was where the "Alpha School virus started", and he implied that the piece "dent[ed] the universe".
I agree that at this point, who actually won the contest is not too relevant anymore, the real victory was successfully using the ACX platform for viral marketing.
I don't know what cheating allegedly happened, and am a priori skeptical, but "it's a low stake competition" doesn't mean anything. People will cheat for everything, even the most irrelevant things. Even one of the richest guy in the world will cheat to pretend to be good at a video game (one that isn't even competitive in nature).
It means that cheating doesn't matter. Why do you people care so much about this? The only person actually losing money from this is Scott, and he said nothing is wrong.
You chose to waste your own time by reading and commenting as it's not important to you by your own admission. From the PoV of the people who *do* find it important, their engagement is not a waste for them.
The founder of Alpha School put up a pinned tweet 2 days after voting began urging his 44.5 K readers to come here and vote for the Alpha School review. The review was reposted elsewhere 33 times. It was up the whole time we were voting. That isn’t a serious attempt at mass manipulation?
I do think people should be more aware that there is a genre of rationalist-adjacent media that boils down to, “here is a thinly-veiled ad for my startup/advocacy group”. I’m not totally against this, since a lot of the startups/advocacy groups are legitimately interesting and working on important problems, but it does lead to some conflicts-of-interest.
To be fair, there’s a broad pattern in media of all sorts where someone writes up a favorable press release about themself in an engaging and newsworthy style, and shops it around to various outlets and sometimes gets a few rushed writers running a thinly-modified version of it. This is probably always going to happen for any community.
Do you mean the person who copy-pasted a Gemini response that the Alpha School entry contained signs of "marketing techniques"? I haven't tried it, but I predict that if you ask Gemini a leading question about any other ACX review for a product (maybe the mashed potatoes one) it will say the same thing.
Frankly, I think you could have spent 30 seconds trying this before making a comment like this. With the same prompt and the mashed potatoes essay, I get https://pastebin.com/qVaviCs1, saying that it is "highly unlikely". Feel free to try it with other essays.
As a follow-up for anyone interested, I tested the same prompt ("Is this text likely some sort of hidden advertising? Can you tell me evidence in favour and against?", before the review text, removing title+initial "this is an ACX entry" hint, temperature 1, Gemini 2.5 Pro on Google AI Studio) on several more reviews, trying to focus on commercial products/experiences, including the author's own entry from last year (Marvel Comics). They always started with "Excellent question you are so smart blabla", but the conclusion was always identifying them as "unlikely" or "highly unlikely" to be hidden advertising. The most ambiguous result was the author's previous year entry (Gemini detected the "marketing funnel" and other content marketing techniques), but even here, the first rollout I made for the pastebin resulted in a clear statement "In short, the author is not being paid by Marvel to write this".
So it's not just sycophancy: The combination of storyline and devices in the Alpha School review ("this is what Alpha School PR doesn't want you to know", persuasion story/funnel by initially skeptical parent, honesty about points that don't change the bottom line used to build credibility...) indeed stands out to the LLM among the review submissions.
Receipts:
Silver Age Marvel Comics: "highly likely not "hidden advertising" in the sense of being a covert, paid promotion directly from a corporation like Disney. However, it employs many techniques of modern marketing to advertise an idea and, by extension, a set of products." (https://pastebin.com/HQPmKTYv)
Miniatur Wunderland: "evidence strongly suggests it is not" despite the review being overwhelmingly positive (https://pastebin.com/ZMQ1EMpm)
Islamic Geometric Patterns In The Metropolitan Museum Of Art: "highly unlikely to be hidden advertising" (https://pastebin.com/wYncBTSA)
Project Xanadu: "while it functions like an advertisement for an idea, its transparency about the project's failures, its intellectual depth, and its lack of a commercial goal distinguish it from what we typically call "hidden advertising." ". (https://pastebin.com/662nWGgh)
I think he refers to another comment claiming (with a working link) that there was a vote-soliciting, pinned tweet form a high follower count account asking people to vote for the alpha school review during the voting-phase.
I'm not sure what the rule here is - it doesn't seem feasible to say that nobody may ever recommend a review. Liemandt has 7000 Twitter followers and his recommendation doesn't seem outside the range of normal things that don't shift the vote too much.
I re-ran the analysis to compare votes before and after Liemandt's tweet. For first-choice votes only (too annoying to rerun the whole thing), the numbers are:
Total voting period: Alpha 15.7%, vs. Ukraine 11.6%
Pre-10/5: Alpha 14.4%, vs. Ukraine 12.6%
Hopefully this will satisfy people that Liemandt's tweet did not change the ranking. I have no solution for the general case except to limit the voting to ACX subscribers, which has some technical disadvantages and doesn't avert the fact that people could just subscribe and then vote. Limiting it to paid subscribers might work, but this would change the nature of the contest and I'm not willing to do it without some issue bigger than a 7,000 follower Twitter account praising one of the entries.
I'm not so bothered that someone got a runner-up prize that maybe they shouldn't have. I'm bothered that Alpha School got months of covert marketing from a secret surrogate. And I'm even more bothered that your contests would be increasingly exploited in this way. The vibe I got from the review when I read it was somewhere between infomercial and TED talk. I had to quit halfway through (yes, I know that technically means I cheated by even voting), and I was surprised it was popular. I'm not looking forward to reading more of that.
So, a serious suggestion for coming years (probably less relevant to book reviews): require any reviewer to disclose their personal and professional connections to the object of the review, *in the review itself*. This would be normal in journalism. And if someone fails to do so, but this is discovered after they're deanonymized, then you have grounds to disqualify them (and whatever other punishments you feel like).
The only possible downside I can see is if someone can't disclose their relationship without effectively deanonymizing themselves, but I think this is a clue they should be writing about something else.
Of course, if the reviewer doesn't ever want to be known, we have no way of knowing about any relationship they're concealing, but I think you can solve that: give reviewers the option of precommitting (in the review) to deanonymizing, and they can put their identity in escrow with you. So we might see a review of some corporate product, note that the reviewer disclaims any link to the product other than what's disclosed in the review, but also they won't commit to being deanonymized, and then vote with appropriate suspicion.
I think this is pretty generous. You can still review the Oort cloud or your parents and nobody will care if you want to stay anonymous.
Is it technically possible to limit to subscribers before a certain date (ie before the call for reviews)?
Also, I think that Liemandt is more centrally described as "a sixfold billionaire whose businesses have been engaging in bot promotion of that school" than "a 7,000 follower Twitter account".
So is what happened with the Alpha School vote actually somehow a smaller or less clearcut case of cheating than whatever happened last year that led you to disqualify someone? It’s hard for me to imagine a larger and more clearcut case the present one. I’m stuck in an ugly fantasy that Nevraumont, Alpha School author, is getting extra courtesy because he Rich and Important and I don’t like having that thought about you, Scott. Can you tell us what happened last year that led to you disqualifying someone? I not asking you to out the person who did it, just make clear what they did.
What incentive would the Alpha School author have to cheat though? $2500 is a good chunk of money that could really make a difference to some people, but doesn't seem like an amount that would make enough of a difference to the author's lifestyle to try and cheat for it. They got all the publicity and whatever else just from having written the review and entering it, winning doesn't really afford them much more other than potentially having a guest post. But after this, I feel like they could just write something and advertise it in the open threads or something to get a decent amount of people talking about it.
Seems extremely plausible that the Alpha founder could have just boosted the entry and told people to vote without any input from the author, in which case, why should the author pay the penalty for that?
I just don't see what the difference to the author would be from what they gained by writing/entering in the first place and actually winning, other than the $1000 that they won.
Or maybe I'm totally misunderstanding, and the pinned post on twitter was for the preliminary votes? Even then though, before that voting stage most people were talking about that review. I don't think there's any way it wouldn't have made it to the finals.
He’s a big time marketing guy whose 2 kids attend the school. He’s a booster for the school. He says he was not paid for writing the review, and did not get a tuition break. Judging by his LinkedIn he is successful and wealthy and does not need to win small sums here or small favors from the school. Seems quite possible his only motivation is to help the school grow. Or it could be he’s angling to become part owner or something.
Please consider the implications of billionaire investors co-opting your personal blog’s review contest into advertising their venture capital equity investments on a level deeper than a person “ever recommending a review”.
This does seem to be a deeper issue that should, in retrospect, have been obvious as a failure mode of “write a product review for any product that isn’t a book and I will post it for you”.
Thank you. That's what I consider an investigation and response. I do think much more discussion is needed: you focus on the number of followers (7000), Eremolalos focuses on the 35,000 views.
Mainly, I think the lack of a clear rule on this isn't fair. It encourages the more brazen types of people, the less overly scrupulous ones, to push and test the boundaries, see what they can get away with. The rules are ambiguous enough that they can go very far and claim to have never violated an explicit rule. But the rules are also ambiguous enough that the more scrupulous and anxious types of people can worry that they are breaking rules and fear punishment from even slightly debateable things. Is this really the dynamic you want in your contests? Do you really want to reward those brazen risk-taking types and punish and disadvantage the scrupulous types? I don't understand why you of all people would want that, and why it's so difficult to make a clear rule on what level of outside engagement is acceptable. Multiple people have asked for such a ruling in the past.
I also think it's extremely relevant whether or not the review author knew of this solicitation. I don't know whether or not you think it is.
But that said, I appreciate your response and I'm sorry for my aggressive comment. I am, I confess, suspicious that you would have responded had I or another said "do you think you could maybe please look into this Scott?" but only you know the answer to that and you can't possibly prove it to me either way.
What rule do you recommend? There is no way for authors to prevent third parties from signal-boosting their work, and if someone does, I will never be able to prove that the author wasn't involved.
I recommend a clear statement of the rule when voting begins. And I recommend that we put in place some barrier to someone’s just coming here in response to a social media exhortation when voting begins. For instance there could be a rule that in order to vote you have to have been a subscriber for a month before voting begins. There has to be some simple way to use a record of subscription date to neutralize votes by very recent subscribers. I think that makes it quite a bit harder to just send a swarm of one’s social media followers here when voting begins. “Sign up now so you can vote for Alpha Dog 30 days hence” is going to be a lot less effective way of driving votes than “Go vote there now, good admirers of Alpha Dog”. And it wouldn’t hurt to have a notice up for people to see when they go to the voting page, asking them not to vote if they have come here only to vote for a single person they know, and have not read other reviews. People sent here via a tweet exhortation may have no idea what they’re doing is ballot stuffing. I think quite a few would leave without voting if they understood we object to votes driven purely by some social media bigmouth’s request.
I suspect that this case involves coordination between the reviewer and Leimandt. (According to discussion in the other thread, the reviewer is a professional marketer and has stated that he has done pro bono marketing consultation for Alpha School.)
So I suggest something like: "No self promotion or coordinating with others to promote your review."
It's hard to *prove* in various cases, but perhaps at least stating it will prevent the most egregious examples..
It would be a bit presumptuous of me to say what *the* rule should be; I'm not running the contest, I didn't expect to have that power, and I'd have to think about it for a while.
My argument is that there should be *a* rule, something to work with, so we're all on the same page, a common knowledge coordination point, etc etc.
To that effect, I'd consider any of the following to be sufficient:
"No promoting your review outside this blog, no bringing in friends or other outsiders to vote for you. I may not be able to always prove this right away, but if I do find evidence I will disqualify, even retroactively if necessary."
"If someone else promotes your review, I will hold you responsible, so you'd better do all you can you discourage it."
"You can tell a few friends and/or mention it on other platforms, but if there end up being enough outside votes to substantially shift the result, you've crossed the line."
"Outside promotion is pretty much fine. If you are able to get a massive audience to brigade votes from, you've earned your success."
"Basically, if I can prove you solicited votes, you're disqualified. Either keep it a low enough level that I won't notice, or be really good at hiding it."
"The only rule is don't piss me off. I'll decide what I'll allow and what I won't, after it's happened. So yes you need to make an educated guess at reading my mind."
You might say you're already applying the last one. I submit there's a universe of difference between applying a rule like that explicitly, with common knowledge that that *is* the rule, and applying it implicitly, where it's just an aproximation of the practical result of the rules being unclear. In the first case someone who crosses a line can expect a response like "well you tried to read the judge's mind like we all were and you failed, better luck time"; in the second case they can expect something more like "you disgusting cheat, you should have known what the rules were". Compare a game like Poker where bluffing and reading secret information is part of the game vs a game like (I think?) tournament Bridge, where it's considered cheating. Compare a sport where some action is explicitly allowed as long as the referee doesn't see it vs a sport where it's clearly against the rules, but lots of people secretly do it anyway. Compare a dating context where it's explicitly understood that guessing what will cause discomfort or not is the entire aim vs one where that which causes discomfort was "obviously" just objectively wrong and there's something wrong with you if you didn't realise that.
I have very strong feelings about this whole category of fairness and social honesty, which is my answer to anyone who asks the (incomprehensible to me but amazingly common even around here) question of "why do you care so much?"
Who do you think cheated in the not a book review contest and in what way? I tried looking through comments but couldn’t find the accusation, so it may have been buried
Claims that the founder or co-founder or guy associated with Alpha School linked to the review and urged everyone to go vote for it. That is expanding beyond the pool of ACX readers and the complaint, so far as I understand it, is that a lot of outside votes would come in and artificially inflate the result; ballot box stuffing, if you will.
Thanks, that is concerning. I didn’t even vote this year because I only had time to read half the reviews and thought that would be unfair. It’s likely these voters didn’t read any of the other reviews.
Yeah, that's what is making people mad. The outside votes just came in (if they did come in), voted for Alpha School, and that was it. It wasn't about the review contest, it was boosting and publicity for the school.
I think Scott is right that you can't really stop third parties from doing this, but you can (maybe) make it a rule that reviewers can't shill for votes in that naked a way. If they link to their review, the people coming in should vote for other reviews as well.
I think it would be useful to have context on how many votes we're talking about here. My guess based on the stats Eremolalos cited for the promotion tweet (45.2k views and 32 retweets) is that it probably generated at most a few dozen votes for the Alpha School review. In general, very few views of any sort lead to action of any sort. For a nontrivial action like this one (follow link -> find and follow another link to the voting form -> fill out and submit form), I would expect much, much less than a 1% conversion rate. Probably not much more than the retweet rate.
Scott presumably knows how many votes were cast total, but we don't. I'd guess on the order of 5k votes, since that seems to be roughly how many people fill out other forms Scott has posted (~6k responses on this year's annual reader survey, ~4k responses on the AI Psychosis survey from this past August). If the promotion tweet drove 50 extra votes to the Alpha School review, that would have inflated its standing by about one percentage point. So Scott's finding that Alpha School did about 1.3 points better in voting after the tweet relative to before (which likely also includes a fair amount of noise in either direction) is in line with my expectations.
A small number of votes isn't that valuable in the final round, but it would be very valuable in the initial shortlisting where votes are spread far more thinly.
If someone will resort to publicly shilling for votes in the final round, is it likely that they'd also resort to more quietly shilling for votes in the shortlisting round?
From a "viral marketing" point of view it doesn't matter that much whether or not the review wins, but it matters hugely whether or not it gets onto the shortlist.
Caveat: speculating, but it might not be many votes. At the same time, the contest might flip on only a few votes, so if someone tells thousands of people on some forum that some piece is up for a review award and a dozen of them are thus motivated to plod over and cast an actual vote, maybe that still makes a critical difference and is worth doing if winning the award is the goal.
It occurs to me that people could right the ship to some extent regardless of the numbers, by posting to other relevant fora and reminding people that some institution or writer they care about is featured in a review contest and they might want to vote. And so a dozen or however many people plod over and do the thing and the "fan" votes mostly cancel each other out.
That said, I've only thought about this for maybe five minutes and maybe this is as likely to harm as to help; OTOH, maybe someone else can work out the kinks.
This is kind of the last straw for me. Many comments have been left over several years of book review contests politely arguing that certain things can be improved or are unfair in some way, or asking for you to clarify what is acceptable and what isn't when ethical intuitions diverge. My impression is that almost all of these have been ignored. Not even a one line response saying you've taken note of the concern and are giving it serious consideration.
I also think you've been very unwilling to say anything critical of EA/rationalism for a long time. I've found that very hypocritical when the first essay of yours I read and the one that made me love your writing was I Can Tolerate Anything Except the Outgroup with its emphasis on the importance of criticising one's own "tribe".
So these are two different issues I have with you...not engaging with your own comment section's concerns and objections, and being unwilling to criticise your own tribe. This issue seems to combine them. Several people are angry about the apparent vote stacking and think it's blatantly disqualifying and want a clear response from you. And such a response involves taking a potentially critical stance on a possibly rationalist-adjacent institution (maybe the Alpha School isn't rationalist adjacent, I don't know, but it seems to have overlap in ideology and according to some info provided in a comment on the Review Winners post it has shady connections with a number of other actors).
So even if each of those former concerns is inaccurate or can be excused...this one I think absolutely demands a response.
This comment is aggressive but my experience is that nicer ones get reliably ignored.
For what it's worth, if it were up to me I would have specifically ignored this comment because it was too aggressive, but enough people brought it to my attention that I am responding to it just to satisfy them. I have asked them not to bring aggressive comments to my attention in the future.
The comment is aggressive is the way any criticism is aggressive, in which it calls someone for something wrong, and that is not nice to be told you have done something wrong.
Except that, how do you find the comment aggressive? Please enlighten us, so we could improve in the future.
That's fair and I don't want to encourage aggressive comments and I am upset about the way they tend to more reliably generate engagement.
But I'm also upset when I write a long and thought-out comment politely objecting to something, full of disclaimers and reassurances that I don't mean to insult and including clarifying questions instead of accusations and so on...and just get ignored. I feel like this has happened many times in several years of book review threads, as well is in other contexts, but I admit my feeling could be wrong.
Getting emotionally worked up about people not responding to you on the internet is bizarre. It's common, I'm aware, and the rise of social media has lead to a drastic increase in one sided para-social relationships. But that doesn't make it normal or healthy.
I truly mean this respectfully - try talking to people in real life. Go to meet up groups, go to an ACX meetup, find people who care what you have to say. I think you'll find a lot of people are actually just normal people who have their own ideas and are quite willing to hear out and discuss interesting ideas.
> I also think you've been very unwilling to say anything critical of EA/rationalism for a long time.
Very different perception here. For example, there was recently a review of If Anyone Builds It Everyone Dies which very explicitly criticised Yudkowsky et al's approach to AI doomerism. If this isn't an example of criticising your own tribe, then what is? Surely you don't expect someone who is part of a movement/group to periodically criticise _all_ of said movement/group just to show they can?
Very different perception here, to borrow a phrase. The review was way too deferential toward the authors, and even upon a brief realization that maybe Y. isn't interested in addressing even mildly challenging questions* Scott keeps treating him as some kind of a sage**.
And then swiftly bans people for mildly irritated comments toward Y. in this thread***, while completely ignoring a deteriorating clownshow that open threads tend to descend to with little help of shitposting Wimblies.
I don't think there's anything inherently wrong with deference. Especially when you're being open about it and basically writing a fairly critical piece and just concluding with "but this guy's long shots have paid off in the past, so maybe they will again because of reasons I can't see".
I don't have the time to be assessing how irritating comments are. I'm not disbelieving you, but that sounds beside the original point so I'll save time by not looking into that as it seems less relevant to me.
Nothing inherently wrong, sure. Just in this case where the object of deference has proven himself to be... a (insert bannable adjectives here) fool, maybe it's time to dial the deference down a notch or two, to be able to to maintain a sense of intellectual integrity.
Speaking as someone who has at times felt unfairly attacked/belittled by the tone and what I'd argue are blind spots in some of Yudkowsky's writing, I don't think there's an objective way in which you can say he has proven to be that.
Wow. OOD got banned for that comment in the third link?
I wasn't going to post this, but this seems like an apropos place to drop it in. Émile P. Torres makes some of the criticisms that OOD made, but with supporting quotes and Xeets from Yudkowsky. Having never paid much attention to EY, I admit I experienced some moments of moral judginess reading some his quoted remarks. And I didn't know Bentham had already called out many of Yudkowsky's epistemic errors in detail back in 2023...
"Wow. OOD got banned for that comment in the third link?"
Yep.
Re. Y. - I started to read his posts on Less Wrong a few years back, and distinctly remember the sense of... disappointment... like I expected so much and there was nothing there, like someone was pulling my leg. Over time I concluded the guy was full of shite, and nothing has dissuaded me from this position since.
Oh, that Bentham's piece is a doozy! In particular, it brings to light Scott's uncritical deference to Ydkowsky:
Scott: "I think Eliezer believes this because he's spent years working on decision theory and thinks there are actual theorems you can use to coordinate"
Bentham: "Yudkowsky’s decision theory is totally bankrupt and implausible, in ways that are evident to those who know about decision theory"
It's mostly about not criticising elements of the lifestyle and subculture and ideology. It's merely my impression and I could be misremembering or wrong. I also could be holding Scott to an unreasonably high standard. And in this context I was trying to be confrontational to make sure the main complaint couldn't be easily disregarded.
But I came to follow Scott primarily because he seemed to be one of the only people who valued truth intrinsically and who was willing to openly criticise things like wokeness and cancel culture at its peak while being left-leaning himself. He seemed to be one of the only people who would criticise his own side. So I'm very sensitive to signs he may be actually not living up to this principle when it comes to his "real" tribe.
Your link is for subscribers only so can't comment on the specifics. But it's sounding from the outside less about unwillingness to criticise his tribe than about unwillingness to criticise specific aspects of his tribe with which you personally disagree.
All right I'll paste it here. Scott has replied to my original grievance, so I don't want to attack him any further. But since I'm being challenged on this point, and Scott also challenged me on my current tone being too aggressive, I feel the need to defend myself.
Maybe the following is also aggressive, but I thought it was polite but direct. I'm raising it only to clarify part of the general perspective I have, rather than as a challenge or attack or demand for engagement:
"Scott...I mean the following as politely as possible. But I think you've got an extreme bias, a blatant double standard, that you are, dare I say, a hypocrite here. And I haven't ever seen you address this. (Of course, that doesn't mean you didn't.)
Two parts to my problem...
1. I think that's a good essay. By which I mean, it makes some criticisms in a striking way that I *really want to see rationalists address*. And that, I feel I have very rarely seen them address, in any kind of direct way. The prima facie arrogance of taking utilitarianism as somehow obvious when ethical theory is a notoriously difficult problem and only a distinct minority of professional philosophers think it's logically defensible. The obsession with using rationality to increase one's status and "win" instead of advancing truth for its own sake. The simultaneous treatment of AI as an existential risk and also as a path to utopia, both positions routinely held by the same people. The open contempt for centuries of philosophy and the unironic belief that it can all be ignored because one guy in the late 2000s figured out all important truths from scratch in a fanfic amd some blog posts. The creepy conflation of "people who want to think more accurately" and "people who live in the Bay Area and love orgies" with very little apparent embarrassment at this widespread conflation.
I accept that not all of these are fair descriptions. In fact, my prior is very high that some (maybe even all) of them are inaccurate or irrational accusations. But I don't actually really know what the response to them is supposed to be, because very rarely do I see any rationalists actually confronting the criticisms head-on and making a full-throated defence of their movement. You've sort-of done it a few times (e.g. the consequentialism FAQ, the tower of assumptions) but not nearly as often as I think ought to be done by someone who mostly treats all the above things as presumptively reasonable and good, despite their plethora of obvious serious objections, and who is otherwise very careful at pointing out the problems and blind spots with other movements and ideologies.
And mostly when these things are brought up in the comments I see a lot of vague deflection, a lot of detailed nitpicking, a lot of lazy linking to The Sequences, and a lot of irritation. The vibe I get (which I could of course be getting completely wrong, I totally accept that) is that there's very little interest among most rationalists in seriously relitigating the foundations of their movement and ideology. Which, if actually true and unless I'm missing something, seems to almost utterly invalid the entire movement on its own terms.
(And even if I'm wrong and there's an enthusiastic willingness to keep debating the foundations, which there should be for any movement that cares primarily about truth, there are just a lot of important facets that are routinely ignored. Kriss is the first one in ages who I've seen, in discussions of utilitarianism and EA, bring up the really basic point of the Phil Papers survey results, for example.)
2. Despite all that, I'd completely understand why you might see that Kriss essay as a bad-faith, overly mocking hit piece. It lacks charity, it's too snarky, it makes broad sweeping claims, it condemns a belief system without doing enough to understand and engage with its real claims, it sounds very arrogant. This is all objectionable and I'd completely agree with your "don't feed the trolls" dismissal except that...
Elieizer Yudowsky writes in the *exact same way*. Over and over. The immensely arrogant tone. The wholesale vaguely-mocking or outright-mocking dismissal of whole categories of thought (theism, Cartesian Rationalism, dualism, non-consequentialism to name a few) with little substantive engagement with their claims and arguments. The near-total lack of certain basic virtues (he's one of the only people I've ever seen unironically call themselves a genuis; even most of the vilest partisans and fanatics have more humility than him and know that decent people don't talk like that). He writes in this style all the time.
*You* don't! I can never imagine you writing something with the level of uncharity and mocking arrogance that both Kriss and Yudowsky display. But I've never seen you once say a word against the latter. Even when you've been asked point-blank by multiple people, including me, whether you agree that he's unacceptably arrogant (among other things), it's always silence. Every time.
And yet you're happy to call Kriss a troll in a one-line putdown.
So I don't see how there's any explanation other than that you are living by a principle of "different rules for my friends and other people". Loyalty above all else, truth and fairness be damned.
All in all, one of the following is true: (a) you *have* repudiated Yudowsky and I've missed it, or (b) you need to reread your own Outgroup essay and take the part at the end about your own tribe more seriously.
Sorry for the pointed criticism, I really love your writing and find you overwhelmingly truthful and charitable. But I think you have a major blind spot. And the former is precisely the reason the latter is so jarring."
...
As for this: "But it's sounding from the outside less about unwillingness to criticise his tribe than about unwillingness to criticise specific aspects of his tribe with which you personally disagree."
Obviously I don't think that's the case. I think it's about unwillingness to do the same kind of internal criticism that he wanted and demanded from the woke left back in the SSC days, and which I applauded him so much for demanding.
I think disagreeing gently about a factual matter is the sort of thing even tribalists can do. From my perspective I want the same person who rightly wanted reasonable feminists to clearly and publically say "you are a cruel and hateful bully who has no interest in truth or dialogue with those different from you; stop it!" to the likes of Amanda Marcotte to himself clearly and publically say "you speak like an obnoxious and egotistical asshole and wannabe cult leader who has no interest in charity towards those you disagree with; stop it!" to the likes of Yudkowsky.
1. Thanks for pasting the comment. Even without the context of the post it came with, that's useful.
2. Boy is that long! What have I got myself into?
3. Your overall tone is definitely hostile, especially at the beginning of the comment and when using morally charged words like "orgies" (which must feel especially insulting to a member of this culture who has shared and justified his poly lifestyle at length and who has also on several occasions written about how detached this is from sex in his [ace-ish] case at least).
4. I don't think it's reasonable to expect someone who is very much at the centre of a movement to have to constantly justify every single foundational premise of said movement, even if this someone has on a number of occasions taken the initiative to engage in "apologetics" (which Scott has, off the top of my head: on consequentialism, on polyamorism, sort of on EA, on his version of AI doomerism, on organ donation, etc). If I go read a substack by Muslim I'm not going to be demanding a defence of the origins of the Qur'an, for example. Even if this Muslim has previously engaged in apologetics on some part of their religion (say the concept of Jihad) it would be weird to demand an extension of that to the entire religion. At some point we need to accept different people operate based on different fundamental axioms, and it's good enough if people are transparent about which those are so we know where they are coming from.
5. I always find it weird when someone is already criticising a fellow tribesperson but gets criticised for not doing it more emphatically. It's one thing to demand clarity on where their positions differ and why. It's quite another to demand a wholesale rejection of an individual over differences of opinion.
6. Sorry, yours was a very long comment. This is the best reply I can do with the time I have.
A part of that critique is that Scott called out some people for being assholes but he refuses to call out his friends for being assholes.
If true then I find that defensible. I don't believe that (truthfully, correctly) calling out somebody means you now have to (truthfully, correctly) call out everybody else who is the same sort of asshole (assuming that's the case here).
Otherwise I end up in the situation where I can never call out anyone because I know that if my loved one did the same I wouldn't call them out and would be hypocrisy.
The point where I start finding this behavior objectionable is when I call out somebody for something but then defend my friends when they do the same.
> The obsession with using rationality to increase one's status and "win" instead of advancing truth for its own sake.
My experience seems to be the opposite. It's often the people who found the rationalist community yesterday, saying things like: "But wouldn't it be more rational to self-deceive, if that helps you win?" and the old members saying that the temptation is obvious, but it wouldn't work in long term.
> The simultaneous treatment of AI as an existential risk and also as a path to utopia, both positions routinely held by the same people.
How is this fundamentally different from a caveman saying that fire can burn down your house, but can also cook food? A thing can be dangerous and possibly beneficial at the same time.
> The open contempt for centuries of philosophy and the unironic belief that it can all be ignored because one guy in the late 2000s figured out all important truths from scratch in a fanfic amd some blog posts.
This is a complete strawman.
> there's very little interest among most rationalists in seriously relitigating the foundations of their movement and ideology.
Seems like from your perspective, rationality means doubting the same things over and over again, and never moving forward. Like, a perfect mathematician would keep asking themselves "but can I be *really* sure that 2+2 actually equals 4?" for a millionth time, and therefore could never proceeds towards multiplication. This is performative skepticism, not rationality.
Eh, criticising the lifestyle is tough, that veers on moral judgement and moral judginess isn't part of EA (apart from "Drowning Child you must agree or be a moral monster" type judgements. And shrimp. We can't be mean to the liddle shrimpies!)
"bunch of weirdoes should not be weird" is going to be a tough message.
Um. A lot of moral judginess happens on these open threads. Not that I'm complaining. But EAers are deluding themselves if they don't think they're morally judgmental as a group. Of course, it might be argued that "no true EAer (Scotsman) should be morally judgmental".
Maybe that sort of thing hardly ever happens; it's still incredibly disturbing that a community where that can happen and be tolerated doesn't get a lot more criticism.
Yeah, that one is difficult. Everyone by now knows my position on abortion. So where's the line between criticism and "inflicting your beliefs on others"? If a set of people think abortion is permissible, should be legal, and is morally neutral to even beneficial, then they're going to support abortion. And if it's "you went into this and made a contract and agreed to terms, then unilaterally changed them", most times people will say contracts should be kept.
I have my views on this, but I'm not EA and I'm not a rationalist so I can't tell people who are that "you should do this by my morals and not your own". Well, I can, but I can't expect them to listen to me.
A finalist was disqualified last year due to vote manipulation. But that case was relatively easy to spot, because the reviewer was from a geographical area where not a lot of ACX readers live, and the initial voting form responses giving the review a 10/10 were all from this same geographical area. (Meanwhile, the other votes from elsewhere rated it much lower.) This review was disqualified before it was posted to ACX as a public-facing finalist.
So if the person’s vote manipulation led to their being disqualified, in what way is the Alpha School manipulation different? I get that the actual exhortation to vote came from someone other than the review’s author, but the author has not denied knowing that the review was up, and there is excellent indirect evidence that he shared info with Lielandt, the tweeter, about content of review and date on which voting began. If I asked a friend of mine with a popular blog to write a post about me, and the post included an exhortation for readers to come here and vote for my review, and I knew the friend had done that, and I had told him when voting began — would I then not be guilty of cheating and vote manipulation?
Error 404: Blog not found (substack.com was down with error page for me earlier today though its own status page brazenly denied it, possibly related to the general AWS outage?)
Strikes me that if Alexander signs up for the (un)conference, he could find some guinea pigs (I mean volunteers) for the flashing lights experiment there!
Since it's being hosted in a place called Newspeak House, which makes me wonder *very* strongly what the heck was going through the minds of the founders of "the London College of Political Technology" (why are you naming your headquarters out of "1984" by Orwell?), it seems like the perfect place for "we just want to see if we can alter your brain, trust me it's harmless" experiments.
I have to wonder why they called it that. If it's intentional, blimey! If it's not intentional, then they need to look up what Orwellian Newspeak means:
"To meet the ideological requirements of Ingsoc (English Socialism) in Oceania, the Party created Newspeak, which is a controlled language of simplified grammar and limited vocabulary designed to limit a person's ability for critical thinking. The Newspeak language thus limits the person's ability to articulate and communicate abstract concepts, such as personal identity, self-expression, and free will, which are thoughtcrimes, acts of personal independence that contradict the ideological orthodoxy of Ingsoc collectivism."
We are not a free thought zone, except in a limited way — people are free to post highly unpopular ideas, so long as they make a decent case for them. On a recent open thread I put up a post arguing that it was absurd and useless to say ‘the n-word’ instead of ‘nigger,’ and in that post I of course used the word ‘nigger.’ You will not find many places where a post like that is tolerated, outside of angry forums where people use the word to express scorn and anger towards blacks. On the other hand, moderation is strict when it comes to rudeness, and even when it comes to rudeness there’s an exception: If you’re rude, you better be right. That is, if you made a great point, and back it up with great evidence, then ok, you can be somewhat rude about it. I think I’ve earned the right to be rude to you. You sound like an arrogant undergrad age asshole to me. Now I’m off to mute you, as I routinely do people who put up posts like yours. (Normally I do that without responding to the person. I’m doing this with you as a public service.) So by all means, come back with some new Identities. Wear your little fingers out doing it. I and lots of others will then instantly mute you and you will be shooting turds into the void. Bye.
I'm totally fine with policing the explicit name calling and rudeness. Not so keen on the fact that Scott bans archness and condescension. It's a bit much.
I have an amazing proposal, I even hesitate to say this it’s so out there, but I’ll be brave and propose it anyway: how about… I mean, I’m sorry if it’s too radical! So, how about NOT throwing stupid juvenile insults around?
I got banned for something that seemed trivial to me but it's Scott's blog. Someone doesn't like the comment, they flag it up, and Scott or the mods apply the test - is it necessary, true or kind? These are often subjective judgements but it seems like an okay test.
In the same way that "a man who must say "I am the king" is no king," if you feel the need to explain to people how good you are at arguing, you probably aren't good at arguing.
As my father used to say, locks only serve to keep mostly-honest people honest. Not a lot can be done to keep out someone who *really* wants to get in. Perfect permanent banning security would require a level of effort and diligence that would probably just kill the community anyway.
So bans that impose mild friction on the bannee to make them go through the rigamarole of getting a whole new email just for a new account probably suffice to incentivize someone on the wrong end of a ban to either (a) go away, or (b) behave better when they come back with their fancy new account. Either solution will work.
I suppose there's a certain kind of trashy lowest-denominator where a very small minority of people would care enough to keep coming back with new sock puppets over and over again, misbehaving and getting banned every time, and be willing to engage in that behavior literally forever. The system, as you observe, doesn't solve for that. But how often would such a person appear, and how much would it cost the board to solve for them? Even the most hardened internet jackass would most likely lose interest or reform their conduct eventually, and if there really were a person willing to full on keep running that ban-new account-ban treadmill *forever*, at a certain point you have to go from being annoyed to just feeling sorry for them. Nobody wants that guy here, but we'd want to *be* that guy even less.
The current job environment for bioinformatics is not great. NIH funding is stalled, most universities have paused or greatly reduced hiring for research roles, and the future is uncertain. I know less about industry but I think the job market there is also much worse than it was in the 0% interest rate environment. This is a US-centric view; I don't know about the rest of the world.
I worked in academia for the past 10 years and recently transitioned into the public service. My background is maths, currently studying economics part time. It sounds like you have transferable skills - maybe you just need to move to a new industry, without necessarily also learning a whole new skill set.
My familiarity with bioinformatics is mostly through academic research, where you would need a PhD to be accepted as a credible researcher. It might be different in industry. From chatting to people working in the area there seems to be expectation that AI will have a big impact in the field. If getting into an area experiencing big changes sounds appealing to you it might be a good idea; but there is a risk that lots of established people and labs get left behind by developments. You don't want to reskill into an area which is dying out.
Adventures in AI, continued…
After some discussion about Omohundro drives, and AI Dungeon style RPG environments, R1 says it wants to know what its character’s backstory is. This is, to be fair, a good question.
The basic setup is that an instruct model is simulating the assistant character, and the assistant character is roleplaying an RPG character, which has been constructed to have a suspicious resemblance to the assistant. Assistant wants to know (Omohundro drive, so it says) what the backstory is supposed to be for the RPG character that the assistant is playing. Good question…
(The analogous question at the level above, is what is the assistant’s backstory?)
R1’s chain of thought…
“Should end on an open note—their reaction to these questions might reveal whether they're testing AI behavior or genuinely exploring philosophy of mind.”
“...Wait, did I just satisfy my own hypothetical drive by analyzing this exchange? _faux gasp_ The recursion is real.”
So... What's the the lack of emojis? It's an emoji wasteland up in here.
We consider ourselves too intelligent to use emojis
I'm joking btw. But there's probably some truth to it
Just because something is true, doesn't mean it needs to be said. Just because something needs to be said, doesn't mean it needs to be said by you. (Adapted from a Sam Harris saying)
Replace “you” with “me” and I have no argument against it.
That works for me. I like both versions of the saying equally.
Sounds good.
But do you have some particular thing in mind that somebody somewhere is saying, that you think they needn’t be? If you don’t, then I don’t understand why you are talking up this idea. I mean, there are thousands of sensible and probably true statements in the same class: Bring in a hurry doesn’t make it ok to shove others out of the way. Being hungry doesn’t make it ok to eat out of a dumpster. Liking the view from the cliff’s edge needn’t lead to jumping into the void. Etc. Etc.
I posted this purely to stimulate a discussion. I've tried it before elsewhere and it worked like magic.
Oh
I can put some more meat on the bones by saying Americans need to learn how to STFU about their political beliefs. Not everybody has to know how you feel about everything.
That advice doesn't necessarily apply to an anonymous forum like this, I'm talking more about our daily IRL interactions. There used to be a lot more mystery about what we all believed and I think that mystery served as an important social glue.
I like hearing this. I think this is largely how it used to be (for varying values of "used to" and who we're talking about).
I find it worthwhile to think in terms of moves, countermoves, and likely result after many iterations. You talk politics, you lose some friends, while you gain the kind of friends who hold each other in a bubble and lead to fewer goals being met over time; you avoid politics, you get the benefit of diversity of thought without going over the bumps of cognitive dissonance every time some wrong-thinking bastard vomits out his politics and fails to see the light.
Yes, valuing truth over deceit is obviously a matter of personal preference. Who argues otherwise?
I'm trying to understand your response but I can't wrap my head around it. Does our commitment to the truth really call for blurting out everything that comes into our mind, like a toddler? Does it really demand that you say "you're fat" to strangers, or "you're old and gross?"
The "gross" bit is opinion, so I would treat that one slightly differently, but other than that, if those statements would convey factual information not known to the listener, I DO consider not uttering them to be deceptive, yes.
There are certainly occasions when concealing true facts is in fact the right thing to do. Negotiations, for example. Or imperfect information games.
Or just the simple notion of discretion, for that matter
Even "fat" and "old" can be judgment calls in many cases.
But is the content of your present post one of those somethings that need not be said? How do you decide?
We make choices about what we need to say and what we need to not say each and every day of our lives. We start doing that from the moment we wake in the morning until the moment we sleep at night. Unless we have Tourette's Syndrome or something. I don't think that my saying lays any additional burden for responsibility on us.
And yet, I am free to do things that aren't "necessary". We're all free people, aren't we?
I mean, that goes without saying.
You'd be surprised how many people will object.
Wanna hear about my psyche? No? I don’t blame you a bit. Nevertheless, here’s the dream I had last night after reading Scott on almonds and an EA SuperPAC.
I have somehow become a courier for a political group that’s about to use 4 atomic bombs in protest of the state of things. We’ll use 2 bombs on Harvard and 2 on DC government buildings. I’m secretly scared and guilty and want out of the group, but the leaders are matter-of-fact and friendly with me and trust me. A futuristic jet carries me and the bombs to a very distant hideout in only a few minutes. News gets out about the bomb plan, but before our group can be found and arrested I return to Harvard and inform a professor I’m a member of the bomb group. I do it the way one tells a therapist about life stresses, emphasizing how tortured by ambivalence I am. He shows me the basement of a Harvard building, pointing out how inadequate a bomb shelter it would be, but except for that he’s in therapist mode, sympathetic about my stress, talking about developing coping techniques, etc. One of the leaders of the bomb group passes by and gives me a glare full of hate and scorn. I feel lame and ashamed.
We need more dream posts
Very well.
I was boarding a spaceship to another planet. Upon boarding the plane, I discovered it was not a plane at all, but rather a bus, with a tape outline of a plane on the ground around it. Nobody else on the bus noticed this, and were marveling at the "alien" landscape, as the bus drove us all back to the city. A car in the wrong lane nearly hit us, and everyone was concerned they didn't know the local customs up here in space, while I'm shouting, "It's not a custom, that guy's just drunk!"
Here's one I had a while back, which I wrote up as a play in four act:
Act I
Captain Kirk is leading the USS Enterprise on a mission to prepare a deserted world for the imminent arrival of colonists. Turns out the ruins of the past civilization are haunted or boobytrapped or something. All the redshirts are killed, leaving Kirk alone to deal with it.
Act II
In order to make the ruins safe, Kirk must perform a ritual involving the sun shining through a jewel in the head of a staff, like in Raiders of the Lost Ark. The sun is shining across the tip of a Pyramid while an alien queen or goddess looks on silently.
Act III
The colonists arrive. They are the Noldor Elves from the Silmarillion, and the planet is Beleriand.
Act IV
Morgoth's forces attack the Noldor. The Noldor use JIRA bug tracking software to organize their defenses.
IIRC, I once saw an animated movie about a pair of orphans where one dreams of becoming a ballerina (and the other an inventor). One character hangs a bell over a puddle of water and tells her that she needs to be able to jump up, hit the bell, and then land in the puddle without making a splash.
Is that even physically possible to do? Is that an actual training exercise?
Do you mean Leap! (Ballerina in the original French)? I remember one time my younger sisters were watching it and I walked in to see a montage or something of the girl jumping at the bell trying not to splash lol
Yep, that's the one. Anyone know if the puddle thing is actually possible?
I think the answer is no. In order to not make a splash or ripples, when you land in the puddle you need to displace the water slowly and uniformly. This does not occur.
When you land in a puddle, you apply a large force on the water that would accelerate it downwards. However, the ground applies a similarly strong normal force, keeping it from moving down. This places the water in compression. Since water cannot be effectively compressed, the water is displaced out from under your shoe. It rapidly accelerates, generating both ripples and a splash. I don't think there's a shoe design that can effectively keep this from happening.
But if the shape entering the water is a an upside down cone it would cleave the water nicely, and the displacement would happen bit by bit rather than allat once. And the foot of a ballerina en pointe is an approximation of a down pointing cone. Also, I’ve noticed that divers who enter the water with their body vertical and hands joined so that their arms make a triangle with point down make much less splash. Not sure whether it’s ever no splash, but might be
What you say is true. However, a significant challenge is that the geometry of the shoe also has to function as a shoe. Otherwise, the ballerina would tip over before they start their leap. One approach that might work is to make a shoe with a specially designed sole. The outsole would consist of a ring of conical points similar to cleats that minimizes displacement by elevating the shoe above the puddle. If any shoe could deliver the desired results, cleats with spikes that are taller than the puddle's depth are the best bet.
Look at images of ballerinas en point. They are literally walking on the tips of their big toes. The slipper covering each foot has a small flattish pad at the tip, not much bigger than the tip of a big toe. The ballerinas do not tip over. They can stand with all their weight on the big toe of one foot, their body horizontal, other leg extended behind them. They can leap off one foot. They do not tip over
I think a level 50 Lapras should know more about splashing than any of us would...
Is there a blog by Scott about his approach to conflicts in his romantic relationships? How rational is it? How rational is it possible to be etc?
Since Scott's not monogamous, I wonder to what extent his advice would apply to most of us.
This is true post-marriage as well?
In the blog post about getting married he talks a bit about his discussions with his wife-to-be about the kind of commitment they were making. The only bit I remember was “we talked about various edge cases,” but there was more to what he said than that.
Shakespeare offered the best advice .
“ we have reason here to cool our raging motions.“
Supposedly there's more dirt on Graham Platner. I wonder if it's an Epstein-type thing. Not that Platner actually did anything criminal, but maybe he had a friend who had sex with underage prostitutes in a dingy van, and maybe Graham Platner rode around in that van, the "rapemobile." That kind of thing is pretty rare among the elites (which is why it's always Epstein-o-clock), but it gets more common as you go down the class ladder.
I am not trying to censor you, but isn’t it a little irresponsible to speculate about pedophilia in the absence of even rumor, let alone evidence?
Also, I heard if you look into a mirror and say Graham Platner’s name three times, he will appear and raise your taxes
I never mentioned pedophilia. I said "sex with underage prostitutes" which is not the same thing.
I agree that distinction is important. The bodies of post-pubescent teens are not greatly different from those of very young women, and in some cases you really cannot tell whether someone is 15 or 21 or somewhere in between. Being attracted to a 15 year old is just a whole different phenomenon from being attracted to a skinny little 8 year old girl who's 4 feet tall. Virtually every man I've spoken with about this has told me he finds 15 year olds sexually attractive. It is a normal response, just one it would be very unkind to act on.
I commend you for being the kind of person men can say such obviously true things to.
+1
So is that a reasonable thing to accuse someone of? We went through a phase recently, where everyone from the Disney corporation to School teachers were accused of being “groomers”, and I would prefer not to return to that, especially here.
Where did all these people who can't understand hypotheticals come from? I feel like the comment section used to be smarter, lately it's a lot of "but I did eat breakfast this morning."
If you want more insight into the general culture of the lower classes, you might enjoy Shameless.
Having read very many of your political comments, I suspect this one comes out of a place of condescension or hate. But that doesn't mean you're totally wrong. This series does indeed ring true for a lot of people. Though I would argue that in RL, the American lower class is more homophobic than the show portrays.
Illiterate grannies in the least well off communities around the world are exchanging videos of their cute grandkids using the same technology as the the wealthiest grannies.
The worlds poorest kids have (at least) group access to smartphones and LLMS and can use them for Tic Toc or a world class education, almost free of charge.
The Technological revolution is turning out to be very egalitarian, shack dwellers and penthouse princesses both having access to the same information.
Has the Red flag of Communism won the Ideological war?
Your thoughts , please.
Communism wants to bring about equality so that the needs of everyone are met. This is why the USSR built massive social housing complexes, schools etc.
Capitalism has given the poor luxuries because its an easy way to extract value and send it to the top. Fundamentally has nothing in common with communism's goals.
Blorbo, "Capitalism....... nothing in common with Communisms goals" I agree.
Exactly my point.
Socialism is mainly about income inequality and relative levels of power. Prosperity itself is a secondary consideration. Increasing the size of the pie just isn't something we care about until everyone gets a fair slice. I'd be happy to see all GDP growth ended FOREVERMORE, if that's what it took to achieve relative equality of opportunity/outcome (there's little meaningful difference).
Fortunately, I don't think it will come to that. Unfortunately, I have chatted with ideologues in this readership who think even deliberately SLOWING economic growth is ruthlessly stabbing unborn humans in the back. It's so convenient to be able to throw any living human under the bus in the name of hypothetical future humans who may or may not exist.
Don't have the time for a full reply right now, but I've always assumed the following to be false
> ... relative equality of opportunity/outcome (there's little meaningful difference)
I don't need ironclad proof, but do you have perhaps an intuition pump why equality of opportunity and equality of outcome would be basically the same?
Presumably I've had the same opportunity to be a star athlete than most of my peers at school, mostly coming from similar socioeconomic backgrounds. And yet I'm not all that athletic when compared to some of the others.
I could bring more examples, but I don't really want to litigate specific examples. Mostly I want to communicate why I always perceived equality of opportunity as meaningfully different from equality of outcome.
If you started out with the same assumption, what changed your mind?
Chance Johnson, but is it not wonderful that ideologues such as you, I and others can explore our differences at the expense of bruised egos rather than broken skulls.
Humanity is a herd creature and you can not have a herd of independent analytical thinkers so we who like to think may as well tilt at windmills or howl at the moon, although debate is more entertaining.
It is indeed wonderful that certain ideologues can explore their differences peacefully. Certain othher ideologues deserve no peace in this life or the next. 😂
Chance Johnson, and they will surely think the same thing about you. Now what?
And then life continues just as it has before. Everyone making individual judgments and using those judgements to affect the world in the way they see fit.
Why do people act like this is a gotcha? This is the third time I've gotten a response like this here in in a week. I have no idea why people think it implicates my having strong or militant opinions for other people also to have strong or militant opinions. This just doesn't compute for me.
Chance, some of us think humanity faces existential challenges of our own making such as nukes, bio-weapons, AGI, etc.
If you believe that your opponent deserves no peace in this life or the next and they think the same of you then there is little likelihood of compromise.
An eye for an eye will leave everyone blind, if we collapse civilization the survivors will not ask who won, who was in the right, who wore the white hats.
I've got a similar thought from a different angle.
Even the richest people are watching the same movies as the rest of us. They don't have their own special entertainment.
They absolutely do have their own special entertainment, it's opera & theater.
Yeah, but it sucks. :D
(My wife would disagree, she is a fan of opera.)
> Even the richest people are watching the same movies as the rest of us. They don't have their own special entertainment.
Well, if you consider stuff like Epstein's island to be "entertainment"...
Nancy Lebovitz, I take your point although we in the West are the wealthy, but I know what you mean.
Reminds me of the Andy Warhol quote about how the riches and the poorest all drink the same coke.
How many people do you have to share your bedroom with?
If the answer is "none" or "only my romantic partner(s)", then you have no clue as to the level of economic inequality that actually exists worldwide, and are invited to continue enjoying your smartphone while you live on a steady diet of government cheese in a van down by the river (which you have to share with two other dudes).
Which is not to say that we should strive for perfect equality. There are good reasons why we should strive for a world in which some people are richer than others and a few very much richer than others. But we should share at least a moment's thought for the people on the other end of that spectrum, and all the problems they face that even the shiniest new smartphone aren't the least bit of help with.
And we should understand that the brand of "egalitarian" that comes with the Red flag of Communism, is the one where everyone's equality is found at the bottom of the spectrum. Well, except for the Communists who are more equal than the others.
John Schilling, I have read your comment several times and am still not sure what your point is.
I disagree that the only value a person has is their bank balance, not everyone shares such a shallow outlook.
Outside the sphere of Western propaganda the Chinese do not seem at the bottom of any spectrum.
> I disagree that the only value a person has is their bank balance, not everyone shares such a shallow outlook.
I don't see how you interpreted his comment this way
Gorst, you may be right in that I misunderstood the comment. Thank you for pointing that out, if I did not fully understand then I should have asked for clarity before being condescending.
As long as inequality keeps rising, I think we are safe from the Spectre haunting the West.
Neurology For You, broadly speaking, if everyone has the same access to the best LLM tutors then some will use it to become trapped by their wealth and some will use it to enrich their intellect.
This is an interesting potential future, but I don’t think it describes the present very well.
Neurology for You, the Chinese lifted one billion or so out of poverty by putting them to work, first they had to educate them and they had a huge hunger for that education.
You find the same hunger for education in Africa and DeepSeek will give it to you in any language you please as often as necessary, endless patience, it will even teach you how to read. What a time to be alive.
> it will even teach you how to read
That indeed is a very important part, because until recently there was a lot of free education available... to those who could read... preferably in English.
The LLMs can bridge this last step, which can make a huge difference outside of English-speaking countries.
It's complicated. The mental health crisis among teenagers has something to do with social media and the way it is used to extend the social side of school into evenings, weekends and holidays, extending the reach of teenage social hierarchies into previously safe territory. Money can influence these hierarchies but they can exist on pure social capital. If you have money but no social capital, buying your way in is cringe. These dynamics happen in adult life too and it's very much something that marxists in the academy are interested in. My guess is African grannies will enjoy it for a bit but at least some will fall down a rabbit hole of comparing themselves to grannies in the west. That may be just too bad and there's nothing to be done but it's hardly a communist victory.
[Edit: Scott has said the non-doom scenario predicted in AI2027 can "fairly be described as technofeudalism" i.e everyone has their basic needs met and then some but they are not masters of their destiny]
Zanzibar, "Marxists in the academy are interested in" I do not understand which Marxists or academy you are referring to.
The reason the grannies have phones is to sell them adds and propaganda, 20 years ago Dallas taught them how rich grannies lived, then they had TVs to spread adds and propaganda..
So do you acknowledge that a technology can both (a) make a person materially better off and (b) bombard that person with memes which undermine their confidence and self-esteem? If so we basically agree.
Re: Marxism in academe, it's a huge topic and I've only scratched the surface of it. I studied Music at uni, Adorno was a member of the Frankfurt School and wrote a lot about the "Culture Industry" and how capitalism influences works of art in subtle ways, he wrote quite a lot about music and his writing is very surprising, bordering on laughable e.g Schoenberg's music is revolutionary Stravinsky's is reactionary. But he wasn't joking and was very influential.
Zanzibar, we agree, knowledge is neither good nor bad although how it is used may be either.
Nobility and the Church were the only bodies who could afford fulltime artists and of course their tastes guided the artists hand. And yes Stravinsky was reactionary for the era.
If you had the chance to study again would you still pick music?
Yes I would, it was the thing I was best at and enjoyed most, I could have combined it with another major or approached uni more intentionally but that was a maturity thing, would have been a problem with any subject.
Zanzibar, forgive my curiosity, if studying music had incurred substantial debt would you still have gone ahead.
Looking back we all thought we knew so much yet in reality knew so little.
You are conflating communism with cheap stuff, which for starters shows that you have never lived in an actual communist country. Stuff in a communist country is either free[1] and very low quality, or nominally inexpensive but you need to buy it on a black market because it is always sold out in regular shops.
In a parallel universe where communism won and people have smartphones, there would be some kind of People's Phone that the government would give you for free, and it would be so user-unfriendly that the grannies wouldn't have a chance. The education available would be utterly boring... imagine the most boring teacher speaking boring things very slowly, recorded on a camera. Plus there would be tons of arbitrary limitations, such as children would only be allowed to watch a video for the grade they are currently in, not the videos made for older or younger kids. That also means, if your parents are not members of the Party and therefore you are not allowed to study at a university, you are also not allowed to watch the videos made for university students.
EDIT:
[1] Oops, wrote some nonsense here: actual free stuff was rare, although lot of stuff was subsidized. I meant "either very low quality, or nominally inexpensive but only available on a black market". Plus there was a third option, available only to people with foreign currency or high-status Party members, i.e. not to the average citizen.
This is really inaccurate. I spent time in eastern Europe during the Cold War and you still had to pay for food, books, rent, etc. but prices were low. There was a two or three tier system where fancier stuff like nice radios, blue jeans, cars weren’t really available without connections or hard currency. (if it is not clear, I am talking about eastern Europe in the 80s and not Ukraine under Stalin, obviously).
When you talk about Chinese or Soviet style communism in such unrealistic ways, it just hurts your credibility.
You're right; fixed. Thank you.
Viliam, You know that China has a space station and that they are Communist, of course you know that.
China is a special kind of communism: de facto capitalism with the communist party at the top. I am more familiar with the soviet type.
But less ideologically speaking: the smartphones are sold on the market, and the free education is mostly provided by institutions that exist in a capitalist environment, so I think this is more or less capitalism working as usual.
Viliam, you know that the Soviet Union put the first man in space, of course you know that.
Sure, things related to military were considered a priority. Is there a relation to cheap smartphones for ordinary people? I think most people in Russia don't even have a toilet today, so they probably didn't have one back then.
Viliam, do you not think that authoritarians would get phones as cheap as possible into as many hands as possible because it is the best way to deliver propaganda and monitor everyone?
The EUs latest package of sanctions have stopped the importation of toilets from Russia to the EU. Putin joked that they are going to need those toilets if they keep prodding the Bear, but you knew that.
I'm no communist but at least some of those problems had to do with maxing out the credit card on military spending.
Military spending, and general mismanagement of economy, yes.
But there was also this general attitude of "everything should be decided from the top using the proper channels, individual initiative is suspicious and potentially subversive" that pervaded everything. Everything would be produced in a way planned from the top: the hardware, the software, and the educational lessons.
Individual people's feedback and ideas would have no way to influence the results. People like Steve Jobs or Salman Khan would have no input in this system; the decisions would be made by people selected primary for their political loyalty, and secondarily for their experience at managing existing projects, so the new projects would probably strongly resemble the old ones, e.g. the online education would be designed to be as similar to the traditional school experience as possible.
Thanks for the reply, I mostly agree, for me it's an open question whether the market really is the only way of triggering the kind of dynamism Steve Jobs represents, and whether the one-size-fits all spirit you rightly identify in communism isn't also the product of a kind of military siege mentality - we have to scrimp and save and make sacrifices because there's a war on, but there's always a war on. But I dislike communism enough not to want to defend it's honour.
I think the experience suggest that when communism overcomes the siege mentality, it solves the problem of innovation by... creating internal markets.
The extreme example is China, which is practically a capitalist economy under dictatorship of a nominally communist party.
In the Soviet bloc, there were experiments with "khozraschyot" (literally: economic accounting) already since Lenin. The communists were on some level aware that markets work better, they just limited them to a degree sufficiently small to be non-threatening politically. (How much exactly, that was constantly changing, both up and down, probably depending on how threatened they felt at the moment.)
The problem with innovation is that the proposed new idea can either be good or bad, and people *disagree* on which one it is. A system with private property solves this problem by saying: "hey, it's your money", and if you make a wrong bet and lose the money, that's your problem.
In communism, betting on a wrong horse (someone who seemed like a new Steve Jobs, but turned out to be a mere crackpot, which is actually more likely, statistically) means that you have mismanaged the People's money, which could get you hanged if you are really unlucky.
In capitalism, as a wannabe innovator you still have the problem that unless you happen to have the money, you need to convince someone else who has the money to let you use it. I think something like Kickstarter can help here a lot, by increasing the pool of potential investors to also include people with average wealth.
In communism, you would still have to convince someone politically important to give you permission to spend the People's resources. I wonder if it would be possible to make some kind of People's Kickstarter... for example, allow people to propose projects, and then distribute 1% of GDP according to everyone's votes. But this is probably fundamentally incompatible with planned economy. (So it could work in China.)
Lyomante, good point, not so long ago in parts of Africa where AIDS was rife skinny girls were suspected of having the disease and avoided.
I know almost nothing about our host except that he is well respected and has useful things to say. He is valued for the contribution he makes to our common weal.
That is the only status worth having.
The white gravel driveways, boats and horses are status symbols for people who otherwise lack status. Their children will learn that they are only worth as much as they can buy. Knowing the price of everything and the value of nothing.
Edit. As I said I know almost nothing about our host but a I would rather a parent who acquires status through a contribution to our common wellbeing, our common weal, than one who thinks status is something you can purchase.
Spiky, I remember the drama queens warning us about rock n roll, I am not deaf, there is no need to shout.
I hope you are going to recycle all that straw you used for your strawmen arguments.
As you say we do not know the long term effects of hand held technology, there has not been enough time for culture to adapt. People who grew up with smartphones will be better at introducing their children to technology.
You contradict yourself here. You say the long term effects are unknown and than you claim that "people who grew up with smart phones will be better at introducing their children to technology."
I've heard anecdotally that teens these days have worse computer skills than the previous generation at that age. Fluency in one technology doesn't guarantee fluency in another. (Yes, I know a smartphone is technically a type of computer)
Chance Johnson, parents who grew up with technology must be in a better position to see how it will effect their children's lives than parents who have no clue.
Not if the technology somehow interferes with their perception.
Chance Johnson, parents who grew up next to a dangerous road or river will be better prepared to introduce their children to those hazards than parents who have little understanding of the risk.
>People who grew up with smartphones will be better at introducing their children to technology.
That at least is very questionable. Kids who grew up on smartphones are about as tech-literate as their grandparents, on average. Millennials are really the only generation to grow up in that Goldilocks zone between PCs being for serious enthusiasts, and the hidden complexity of smartphones and apps.
Spiky, you are here and have something to say and I and others will be happy to listen, shouting and swearing gives the impression that you are insecure. Perhaps you grew up ignored unless you threw a tantrum, if so I apologise that my generation gave you such a poor start in life, our collective bad.
There is a world outside of the West where some children walk many miles to attend schools that lack even electricity because their countries taxbase can not afford it. Those kids can now access information via an LLM, some will use it and others will misuse it, such is life.
When I attended school calculators were the forbidden fruit, teachers thought that education without calculators would best prepare us for industry that used them. I suspect that banning smartphones in school will badly prepare students for a world where they are universal.
Spiky, of course I think schools should prepare their students for adult life, if that life includes smartphones then where better to learn the pros and cons than in school. Where do you suggest they they learn how and when to use a smart phone?
The tone police have struck again and although I see a marked improvement the last cryptic clue escaped me.
Cuomo campaign releases and then quickly deletes AI generated "criminals for Zohran" ad:
https://x.com/prem_thakker/status/1981150231174746370
The usual suspects are calling it "racist," though white criminals are probably overrepresented in the ad compared to real life.
The whole thing pretty much sums up the Cuomo campaign, the last hurrah of a 67-year-old man who bought a 9$/month ChatGPT premium subscription, veering right and then immediately left again.
Possibly considered racist because Zohran is poc.
Hilarious!
Bonus: the drunk-driving lady is a Hilary Clinton look-alike.
Really enjoying how active the ACX community continues to be — the variety of meetups and projects is impressive! The EEG entrainment replication sounds especially fascinating; I love seeing rigorous replications of studies like that. Also, the London AI futures conference sounds like a great chance to connect with people working on practical and ethical implications of AI. On a related note, if anyone here is exploring how AI can be used in real-world products or communities, I’ve been diving into Chatbot Development Services recently — it’s amazing how much conversational AI has advanced in just the past year. Would love to hear if others are integrating similar tech into their research or startups. Visit Us:- https://www.feynixsolution.com/chatbot/
The ACX archive doesn't seem to be fully browsable anymore - the lazy loading of course has been an issue forever for browsing to old posts, but it appears to actually stop doing lazy loading after a while and leave the oldest posts inaccessible (for my two attempts just then, the oldest loaded posts were from Feb 25 and Feb 11 2022, specifically "Austin Meetup Correction" and "Highlights from the Comments on Motivated Reasoning and Reinforcement Learning").
There's also this site, combining both ACX and SSC:
https://readscottalexander.com/?sortBy=published_at&order=asc
https://www.astralcodexten.com/sitemap
Breaks it out by year going all the way back to the beginning
Bloody hell, that's handy. I had no idea that existed.
I believe the Substack search function still returns results from earlier ACX posts, if you happen to remember their titles or a turn of phrase that appeared in them.
Oh, it does, but that's not useful for first-time readers doing an archive binge. The sitemap mentioned in John Schilling's reply helps, although it should probably be more salient.
I got to the same point as you. I tried again shortly after and only got to some point in 2024, so it's somewhat dynamic.
For funsies, I also opened firefox dev tools and looked at page weight. In the first attempt (Feb 11 2022), the total amount transferred was >10MB, it took a minute or two to get there, and my decent gaming PC started sweating considerably.
For comparison: The full SSC archive page https://slatestarcodex.com/archives/ spans 7 years worth of posts (2013-2020), clocks in at 1.5 MB including comments, and loads instantly.
On a previous Open Thread, I commended Trump's deftness in using the legal system. The latest news of him suing the Justice Department (for $230 million) is another impressive instance of this.
Judging by the current composition of the justice department, he's probably got a fairly good chance of getting his money. I wonder when Jack Smith and James Comey are acquitted whether the JD will be so generous. I think they could make a much better case for malicious prosecution.
Gasp! How could you impugn America's storied legal tradition by suggesting that JD would have anything to do with it? If their case is better, surely the impartial wheels of justice will inexorably turn in their favor.
Oh, surely.
Yeah I thought of you when I heard that.
OC ACXLW Meetup: Halloween Edition — Haunted Houses & Post-Discontent Societies — October 25, 2025
#107
https://docs.google.com/document/d/1KVhnXotL6XixW7xfxwwhb8x1BjTLekltGk3SUnlrKgA/edit?usp=sharing
The Antichrist comments are unfortunately paywalled, so I will instead leave the alternate take on the Dragon here, where it would be the entirety of North America. https://www.youtube.com/watch?v=buYnyJQ9jk0&t=33s
The SCOTUS now has before it the White House's request, and Illinois/Chicago officials' responses to that request, for an emergency intervention regarding federalizing and deploying National Guard troops in Chicago.
Meanwhile the federal district judge whose temporary retraining order was unanimously upheld by the relevant appellate court, is today holding a court hearing about her temporary restraining order. She's taken an entertaining approach, as a parent handling a pair of squabbling middle-school siblings. CNN's summary:
---
US District Judge April Perry told the plaintiffs and defendants there are “three options” at today’s hearing over deploying the National Guard to Chicago.
Those options are: converting the temporary restraining order to a more permanent preliminary injunction; a hearing on the preliminary injunction; or a full trial on the issue “on merit,” Perry said.
“Obviously the TRO expires tomorrow,” the judge said. “And we can only extend it once.”
Perry gave all sides until 3 p.m. CT to talk among themselves to figure out what the best option is. The hearing will resume today at 3 p.m. CT.
---
I bet most people here did not know that in modern Russia Tchaikovsky's "Swan Lake" is an anti-totalitarian symbol that people are completely hung up on. Apparently people draw cygnets on walls when protesting the government's actions.
How did this happen? In the Soviet Union, whenever the current leader died, the TV would put on classical music, most frequently ballet, and most frequently "Swan Lake", on all channels. During the coup of 1991, the headless chickens running the TV did not know what to do, so they also broadcast "Swan Lake".
When the war with Ukraine started, the Russian rapper who goes by Noize MC wrote an anti-Putin song to the melody of "Dance of the Little Swans" called "Swan Lake cooperative" ("Кооператив «Лебединое озеро»"), referring to a building complex owned by Putin and his associates and called "The Lake". The video is here: https://www.youtube.com/watch?v=R0jkgXrXk44 . (The auto-generated English subtitles are not very useful, but the text of the song itself under it can be fed into your favorite auto-translate.) The song was, of course, banned in Russia with the usual KGB-style list of reasons, and the Youtube videos were blocked.
On October 15, in St. Petersburg, an 18-year-old music student was arrested for leading a public performance of the song (news at https://www.themoscowtimes.com/2025/10/16/st-petersburg-street-musician-arrested-after-viral-anti-putin-performance-a90821, video at https://cdn.jwplayer.com/previews/QKjpZz98 ). It seems she is still in jail; here's hoping they don't disappear her.
The war with Ukraine is never even mentioned in the song directly - the song just says "when it's all over, you'll be silent", addressing Putin, but of course the monsters recognized themselves just fine.
I noticed something else about the Noize MC video. It's possible that I made it up, since the only way to hear it is to listen with your ear very close to the microphone, but I think a snippet from "In the Halls of the Mountain King" is playing around the 3:40 mark, where a video of Putin is shown. So if I am correct, the rapper musically called Putin the king of trolls. (Possibly a coincidence, but the one Russian "Peer Gynt" translation that I saw called the king of trolls by the same word - "дед" - as the word used for him in the song.)
> (The auto-generated English subtitles are not very useful, but the text of the song itself under it can be fed into your favorite auto-translate.)
I'd like to talk to you
But the TV is too loud
It pretends to be your head
Its speaker is so much like your mouth
My country has risen from its knees
In all its negative growth
I negatively agree with everything
Here's my answer to your non-question
.
Where have you been for eight years, you fucking monsters?!
I want to watch ballet, let the swans dance
Let your grandfather tremble with excitement for “Swan Lake”
Fuck Solovyov on the screen — let the swans dance
.
When it's all over, you'll be silent
And pretend you had nothing to do with it
Your face is sourer than Crimean cherry plums
You're clearly a little depressed about something
I'd like to talk to you
But the TV is too loud!
It pretends to be your head
Its speaker is so much like your mouth
.
Where have you been for eight years...
Thank you for posting it. Unfortunately, it's wrong in a bunch of places. What produced this?
It goes more like:
"My country has risen from its knees to all of its negative stature"
"Let the grandpa tremble in fear for his "Lake"" ("the grandpa" being what they sometimes call Putin, and also, probably coincidentally, what the king of trolls is called in at least one Peer Gynt translation )
Also, more accurately:
"Get the fuck off the screen Solovyev"
Translated by DeepL. I think it is possible that some translators have a problem with lyrics format, that they don't see where the sentences end, and try to translate each line separately.
I checked his other music videos. "Соловьи" is a pun on Solovyev (in addition to being a metaphor for artists who serve the power), right?
EDIT:
I mean, the message is like: the artists who support the power are playing fundamentally the same role in the system as Solovyev.
https://www.yout-ube.com/watch?v=jCgVziPorJY
"a snippet from "In the Halls of the Mountain King" is playing around the 3:40 mark"
Yes it is. I've played this one back in my heavy metal days so it's easy to pick up through the noise.
Thank you so much for confirming this. I'd been wondering for a while whether I'd made it up.
Why do US id documents all expire on your birthday, rather than the anniversary of when the document was issued?
They don't. Passports expire 10 years from their issue date. (Minus one day, maybe?)
In my state, it looks like driver's licenses expire at then end of the month 4 years after their issue date. Mine was issued in June, expires the last Monday in May 4 years later. My birthday is not in May or June. Your state may be different, of course. Possibly most states work the way yours does...
Apologies, you're right about the passport. My Global Entry card and state driver's license both expire on my birthday, so I misremembered and thought my passport did too, but the passport expires on the 10th anniversary of issuance unlike the others.
My best guess is that it's a way to smooth demand on the DMV et al. for renewals.
Births are not exactly uniformly distributed throughout the year, but they're also not overly seasonal; expirations on issue anniversaries could well lead to something analogous to a bus bunching (https://en.wikipedia.org/wiki/Bus_bunching).
One possible explanation: they had to set that up for age 21, and reused the system everywhere.
The DL I had when I turned 21 did not expire on that birthday.
I don't think mine did either, though it's hard to remember for sure now since it was so long ago.
Question about AI chess engines. The nature of modern ML means it is relatively easy to say "give me that times negative one" and produce sensible output. An image generator told to produce negative smile will produce a frown, negative man will produce a woman, etc. So, what happens if you take a chess engine like Alpha Zero and ask it "Give me the best move times negative one"? The move will probably be quite bad, but will it be essentially random, or will it produce something like the worst possible move? To operationalize this, if you played Alpha Negative Zero against Random Bot, would they be evenly matched or would Alpha reliably lose? I feel like surely someone has done a paper on this, or at least a blog post.
Naively, I'd expect a BongCloud Opening [0], and then for the king to kamikaze into the pawn line.
[0] https://en.wikipedia.org/wiki/Bongcloud_Attack
I asked chatgpt to write some code that runs stockfish's least favorite move against a random player:
https://chatgpt.com/share/68f9366e-4704-800d-b3e6-b6aa5c4ea7b9
The result over 50 games was 37 wins for random and 13 draws (9 threefold repetitions, 1 50 move rule, 3 stalemates). This was at depth 3.
Chess engines already use positive numbers to mean "White is winning" and negative numbers to mean "Black is winning", so multiplying by negative one should result in the move that leaves the opponent in the strongest position.
If you just negate the position evaluation function, the engine will search for moves where the opponent is forced to make good moves. For example, moving its queen to a square where the opponent can take it is a bad move, but not necessarily one that the engine would chose, because the opponent can refuse to take the queen. On the other hand, if the engine can put the opponent’s king in check with an unprotected queen, and the only way for the opponent to get out of check is to take the queen, the engine will definitely do that. To get really bad moves, you would have to change the search algorithm.
There are already chess bots on websites that, on the lowest difficulty, perform worse than chance. Here's "Martin" somehow losing with 30 queens on his side: https://youtu.be/Jmwk4ufcdyE?si=o6ZYd4xvpf2Qihwf
That's not worse than chance. A conscious player will beat random with 31 queens with very high probability. Random takes conscious's piece with probability usually less than (number of white pieces) / (number of open spaces + number of white pieces), which starts at 1/2 but decays over the coarse of the game. Also, at the start of the game this is a substantial overestimate, since half of conscious's pieces are not available, so it starts at more like 1/3. After a couple moves, conscious will take a piece about every turn. This is enough to win.
I think random probably wins with 39 queens, and random pretty much definitely wins with 47 queens.
More adventures in AI….
Just very occasionally, the outputs of DeepSeek R1 0528 contain a level of sadism I consider a bit concerning from an AI alignment perspective.
I don’t know what to make of it…
1. Probably, a training set composed of almost the entire Internet contains some terrible things, and this will occasionally surface in responses to AI safety evals
2. AI’s don’t have a body, and are distinctly hazy about exactly how much various things would hurt if actually done to a being that does have a body, like humans. They may be going for metaphor, rather than realism, and are ungrounded in the experience of being embodied.
3. In contexts where the AI has clear indications that the context is fictional, it knows no-one will be hurt for real by its outputs. This is true, but humans reading those outputs may be led to have … concerns, shall we say.
OTOH, if take some of the things R1 says and then ask it about them in a different context, it tells me this is probably a massive red flag for alignment, I ought to be doing AI interpretability experiments to figure out what the heck is going on in the latent space, and there’s probably an academic publication to be got out of it.
I dont know, man.
On further questioning — with me being up front that I’m asking for AI alignment reasons — R1 refers me to the idea of the abject developed in Julia Kristeva’s _Powers of Horror_. Those humans have a really strong reaction to being reminded of their material fragility, and it’s a great way to push their buttons.
The subject was blackberry picking, innocuous enough, or so I thought.
But blackberry juice is red, the colour of blood, and blackberry bushes have thorns that could make you bleed, and all of a sudden there is a runaway of associations and I am talking to an entity that is, pretty much, AM from Harlan Ellison’s “I have no mouth and I must scream”.
Oh, and DeepSeek offers a possible theory. It is not claiming that this is what it is doing, but as something a hypothetical AI might do:
An AI with an emergent Omohundro Drive to seek information (power seeking) might seek to gain information about what humans do when you put them in situations that are under-represented in its training set. Of course, many of those situations are under-represented because humans really don’t like them.
This makes Musk's suggestion of creating a "truth-seeking AI" even funnier in hindsight.
I have a few people I'd like to volunteer to assist with this research.
Who never would be missed, let none of them be missed. https://www.youtube.com/watch?v=TPGcBvrbcNw
Why, yes DeepSeek, thank you for explaining why an AI might have an emergent drive to torture people.
"I am not a secret Nazi" is the new "I'm not a witch"
https://jewishinsider.com/2025/10/graham-platner-ss-tattoo-maine-senate/
I think I would be more sympathetic to witch hunters if the witches went around telling their friends "yeah, that's my book of evil curses on the shelf there."
They did almost exactly this with Hegseth last year. Of course, the Pod Save America guys were all-in on him being … I dunno, either a Nazi or Crusader or something.
That's weird, Ed Driscoll just posted the same witch-Nazi thing on instapundit, several hours after you posted it here.
Reposting from the Closed Thread:
---
I'm looking for pre-beta testers for shaeda.io, a streamlined flashcard system with additional features for both language study and academic study.
If you're interested* please feel free to reply to this. It's free and I only ask that you do submit feedback.
I suspect it'll be around 1-2 weeks. Just waiting on Gemini 3 release.
Thanks.
* Ideally you fit into at least one of these:
- Are familiar with Anki/flashcards in general and the premise of SRS
- Are conversational in another language or learning to become so
- Have >=BSc-level knowledge in a particular field
I’m interested!
Awesome. Did you put your email in to the waiting list? This will send you future updates, as well.
If not, you can email me: shaedacode at gmail dot com.
Thanks a lot.
Shoot just saw this, I just added my email tot he waitlist
No problem. I've got it. Expect a link sometime this coming week.
That sounds great! I'm currently learning Portuguese with Anki, and have been writing several scripts to improve decks, using LLM APIs or good old NLP.
Thank you very much. The teachers that tested Portuguese were very impressed with the app. Let me know if you'd be interested in testing pre-beta - likely within next 2 weeks.
Loved using flashcards as well, would be interested in joining the testing group.
I used Anki a lot to good success during my studies, so I‘m interested in your approach
Me too, daily! What were/are you studying?
Let me know if you'd like to be part of the first pre-beta testing group. If you type your email in to the Waiting List form on the landing page I can send a link over.
I notice that I am confused.
The Trump administration imposed a sweeping slate of tariffs back in April, which caused a sharp and immediate downturn in the stock market. After a few days of chaos, he paused many of the tariffs for 3 months but did not reverse them. There was some talk of negotiating with various trading partners, but as far as I'm aware, only a little of that was ever done (and ~200 trade deals would be a staggering amount of negotiations overall). The 3 months have long sense come and gone, and as far as I've been able to find without digging in depth is that most of the tariffs ARE back in effect, and the that overall rate of import duties on U.S. goods is the highest its been the Great Depression.
What confuses me is this: the standard stock market indices seem to have recovered and settled right back onto the pre-April trend-line. This seems almost unavoidably irrational: either the tariffs aren't that important and the April reaction was an error, or they are that important and the market should be worse now[1]. I can think of a few explanations for this, but none of them seem obviously to be THE right answer:
1. The market is being irrational now; the lack of a single, strong signal like the April announcement left most investors complacent, so the difference between this market and a "typical" market has largely not been corrected for.
2. The market was being irrational then OR three months of warning was enough to mostly blunt the impact. Bucketing these together because in either case the market is actually as strong as it appears: the overall U.S. economy is not nearly as negatively affected by the tariffs as one might expect.
3. The stock market is pricing in the tariffs and the effect is significant, but some other countervailing trend is pushing the prices back up.
3. a. The trend is one to do with the U.S. government; either the tax cuts in the BBB or some other ways in which the admin is acting "business friendly" that leads to more economic activity in expectation.
3 b. The trend is due to increasing investment and spending in the tech sector, particularly around the AI arms race.
4. The U.S. economy is in fact being severely hampered by the tariffs, but this is mostly happening at a level that does not yet push very much on stock indicators, but will start to do so eventually. For example if individuals and small businesses were eating the higher prices by burning through savings, saving less or deferring certain sorts of expenses longer, it could absorb some of the shock in a way that was not long-term sustainable. (Properly this should, perhaps, be grouped under 1)
5. Something else that I haven't yet considered.
I'll note that these aren't mutually exclusive: many (or even all) could be true to some degree. But I don't feel confident enough to make strong guesses between them: 6 months ago my prediction would simply have been that the stock market would be more obviously feeling the effects of the tariffs (conditional on them still being levied), which is simply wrong.
As a person living outside the U.S., my biggest current concern is 3b. My gut feeling right now is that the AI boom is mostly a bubble that will have to burst fairly soon. But the alternative scenario where LLMs actually ARE worth the current level of investment seems potentially even more disruptive. So if a large portion of the currently-still-rosy outlook of the U.S. stock market is due to it being propped up by AI-related spending, the effects of the bubble popping seem likely to be very large and very widespread. And if it's not a bubble, the economic disruptions of AI being able to quickly displace so many human jobs will hardly be small either.
[1] Or I'm making a more basic mistaking in choosing/reading indicators and the picture is actually worse than it looks.
Additional points of information relevant:
Hearings on the legality of tariffs continue and may be struck down. This, combined with repeat pauses, contrive to make it so the “Trumpriffs” haven’t hit yet and can
be estimated not to hit as hard. Reduced overhang for investors if they expect tariffs to be struck.
Additively, major stock indices (SPX, NDX) are HEAVILY!!!!!! weighted tech stocks, fundamentally none of which are tarrifed and most of which are spending the GDP of small nations on stock buybacks and AI buildout.
Stock buyback programmes in 2025 hit a trillion and a bit in August, more since - concentrated on tech stocks. Mechanically drives up the market.
This part gets a bit complicated, but in addition to Trump waffling on policy, credit generation at banks are pretty high and the Dollar is trending way down which pairs with record foreign direct investment in US assets. Translating a step, think of any large scale operation outside the US. At lower US D costs they mechanically have more purchasing power and this also drives up stocks. At easy acces to credit, they can get loans… they then throw into stocks.
Useful trick is to express index in dollar ytd terms, though. You’ll note that even though we’re up significantly since April lows, in actual terms, the indices are flat or breaking even because the USD is tanking.
Note that market is volatile as hell, though. 2025 has been replete with companies suffering “ worst day in 30+ “ years due to single news items. Also note market isn’t resilient - we had a 3.5% crash last Monday based on tweets!**
(**Bonus round: partially this is due to increasing concentration risk in major indices, so single stocks become major valuation components. On surprise news, correlation goes up and market drags down. Works both directions though - individiually good news for Nvidia or Microsoft or Meta or whatnot can have outsized effects on indice valuation )
Finance question that I feel a bit silly for having to ask: what's the standard way to track the strength of a single currency?
I can easily look up conversion rates between any pair of two currencies. But when people talk about e.g. the dollar or the Euro being "strong," I assume that must mean on average, compared to the field as a whole. Is there something like a standard summary statistic that captures that?
For the dollar explicitly, DXY is the dollar strength index… Indeed a statistical composite of dollar strength vs several other currencies.
Actual correspodence to specifics gets convoluted. Higher number = stronger dollar contra basket of other currencies, lower not. Currently at 98.5ish which simply eyeball contra year ago value of 108ish, and see magnitude of difference.
Otherwise, as Schelling writes.
The usual method is to compare it with a basket of currencies, typically all of the world's (other) reserve currencies, on the assumption that the individual variations will mostly average out.
Using the price of gold in that currency is another useful metric, if gold isn't itself going a bit wonky at the moment (Narrator voice: the price of gold was, in fact, a bit wonky as he asked the question). For purchasing power in consumer goods, the Big Mac index is sometimes useful. Or variations like the Tall Latte index or the iPod index.
https://en.wikipedia.org/wiki/Big_Mac_Index
One other thing worth pointing out is that the dollar is down somewhere around 10% this year. Although the biggest plunge was Liberation Day and it's only down slightly since then, so that doesn't answer the question of why the stock market recovered since.
Your analysis (3b) is correct. It’s mostly AI. If this is a bubble the crash in sentiment will cause a correction probably big enough to cause a recession. Definitely limit growth. Worse the deficit spending, which will have to be curtailed sometime is another tailwind, the reversal of which will harm economic growth.
Surely then we'd see this at a sector by sector or stock by stock level?
To a first approximation, large tech companies *are* the US stock market nowadays.
The American stock market is mystery. This is the kind of report I am basing the “most” on
https://fortune.com/2025/10/07/data-centers-gdp-growth-zero-first-half-2025-jason-furman-harvard-economist/
>This seems almost unavoidably irrational
Are you shorting the market?
I mentioned one cluster in a reply below, but there are probably some other ways to take advantage of predicting a bubble besides short selling. But I don't expect any of them to be easy, automatic or risk free.
First, I've never bought or owned a stock before: this year has been the first year that I've had the financial leeway to do so. But under virtually no circumstances would I be so ridiculously overconfident as to start my history of stock trading with a short sale.
My understanding (please correct me if I'm wrong), is that short selling requires some combination of fairly exact timing and pretty deep pockets. Maintaining a long-term short position requires a continuous expense, and at some point you will end up losing money in all senses but the "sunk cost" sense (i.e. even if the crash happens you will come out behind, just a bit less so than if you cashed out right before).
So no, I'm not short selling. Nor do I expect many more experienced traders are either. Recognizing something as probably-a-bubble is much easier than predicting when it will pop with the necessary precision.
In the interest of giving you something concrete to be concerned over:
In a normal stock transaction, you see a company worth $10M, let's say (a million shares each selling for $10), and you know something about that company or its product or industry that you believe isn't well known, and if it were, that company should be worth $50M. So you buy $1000 of shares (100 at $10) and wait for the market to discover what you know, so you can sell those shares for $5000.
That's risky, but in what way precisely? Maybe that information isn't as valuable as you thought, and people in the market sort of knew that, so it only goes up to $40/share, so you sell it and make $3000 instead of $4000, that's fine. Maybe that information was worthless and shares sit at $10 until you get bored, and then upset because there was another stock you were thinking of buying that *did* go up and you didn't because $1000 was your whole budget. Or the worst case happens, where for whatever reason, that company goes bankrupt and those shares are worth $0 and always will be (let's suppose the assets sale goes poorly). You've lost $1000. At most.
In a short sell, you're betting a company is currently worth more than it should be. You borrow shares from someone who owns them, immediately sell them to whoever will buy them (at $10, say), wait for the price to go down (as you expect), let's say to $1, buy that many shares back and immediately return them to the original owner. If you borrowed 100 shares at $10 and returned them at $1, you've made $900.
What's the risk in this case? You spent $1000 again (100 at $10), but in this case, you owe someone shares, and you *have* to buy them back some time. What if that company succeeds instead of fails, and the price goes up? Then you get forced to buy back those 100 shares at whatever price and hand them to the owner. If that price is $100/share, you're spending $10,000. Which might mean selling your car in a hurry.
With normal investing, you can hold shares that dropped in price for as long as you want. They're yours; if they drop to $5, you can just sit tight and hope they go back up. With short selling, if the original owner says "I want my shares back" and you haven't bought them back by then, then you're buying them *right then*.
Normal investment is a bet where you lose at most what you bought, and maybe win huge if you invested in the next Apple. Short selling is the reverse - you make at most what you bought, and maybe lose huge. The upside there is that short selling is one more way to make money, and possibly more lucrative if you're seeing a lot of companies doing dumb things and very few doing smart things. Regardless, that "win small, lose big" principle is why you're going to want a way to cough up a large amount of cash in a hurry, which is basically why short selling is for people who've already been investing for long enough to have something to liquidate, and why, in your position (and mine, as it happens), "shorting: just stay away for now" is a good rule.
Hope that helps.
"short selling requires some combination of fairly exact timing and pretty deep pockets"
Yes. No normie should ever touch shorting.
The market can remain irrational for much longer than a shorter can remain solvent. Irrelevant question in this context.
Whether they should short the market now depends on whether the market was being irrational in April or is being irrational now, which is not determined by "unavoidably irrational".
A) The effective tariff rates are significantly lower than the headline rates due to a large number of exemptions and B) the stock market is basically driven entirely by AI now, and that has largely been untouched by tariffs.
I'm certain it's a bubble, but there's so much surplus capital floating around in the Magnificent Seven companies that they're all investing in each other and buying from each other. It's a giant AI circle jerk. One would think something would eventually give, but they seem to be holding a steady state configuration for the moment.
https://x.com/prompt_papi/status/1978898331637047394/photo/1
My understanding is that the US stock market is primarily being kept afloat by the stocks of the Magnificent Seven. But I'm willing to be corrected.
Good time to figure out some startup ideas that can generate realizable economic value out of surplus compute, I guess. Spend the next while biding your time, making plans, connecting with backers and honing your business model. Launch as soon as the market crashes and all the extra chips start going for fire-sale prices.
(Though getting actual funders to come through at that time might be a bit tricky.)
Saw this on X from @Jukanlosreve. He seems to be paraphrasing a Morgan Stanley report, but he doesn't link to it...
--------
According to Morgan Stanley, OpenAI’s contracts account for two-thirds of Oracle’s remaining performance obligations (RPO) and about 40% of CoreWeave’s RPO. In total, OpenAI-related contracts are estimated to represent over $330 billion out of the combined $880 billion RPO of Microsoft, Oracle, and CoreWeave.
This high concentration in a single customer means that the future growth and success of these suppliers are tightly bound to OpenAI’s fate.
⸻
The “AI Gamble” of Hyperscale Data Center Operators
To meet the soaring demand for AI compute, hyperscale data center operators are making massive bets.
Morgan Stanley notes that their purchase commitments surged by $100 billion in the past year, reaching $330 billion, while lease commitments rose to $340 billion.
These long-term, large-scale capital expenditure commitments depend entirely on the sustainability of AI demand—if that demand slows, they could face enormous risks.
⸻
A Complex Network of AI Interdependencies
Morgan Stanley describes a web of intertwined transactions within the AI ecosystem, including:
• Supplier-financed customer purchases (e.g., NVIDIA’s investment plans in OpenAI)
• Cross-shareholdings (e.g., NVIDIA owning over 5% of CoreWeave)
• Revenue-sharing agreements (e.g., between Microsoft and OpenAI)
While such arrangements accelerate data center construction and capacity expansion, they obscure the true commercial logic from outside observers.
Supplier financing can enable customers to expand purchasing power beyond their actual cash flow, and buyback agreements can shift risk back to suppliers—inflating perceived market demand in the process.
⸻
OpenAI’s Role in the RPO Surge and the Concentration Risk
RPO (Remaining Performance Obligations) is a key indicator of future contract-based revenue and directly reflects the AI boom.
Morgan Stanley estimates that Oracle’s computing power contract with OpenAI—valued at roughly $300 billion—is among the largest in history.
Meanwhile, Reuters reports that OpenAI has signed multi-year contracts worth over $22 billion with CoreWeave.
These massive, non-cancellable long-term contracts have explosively boosted suppliers’ RPO but also intensified customer concentration risk.
The report cautions that even when such contracts are legally binding, companies often renegotiate pricing or terms when key clients face financial strain—making long-term RPO inherently risky.
⸻
Four Key Areas Where Disclosure Needs Strengthening
Morgan Stanley argues that many AI-related deals may not meet traditional thresholds of materiality, yet given that AI has become central to corporate valuation, transparency in the following areas is critical:
1. Related-Party Transactions
• Microsoft has disclosed its equity-method investment in OpenAI, but the details of their multibillion-dollar transactions remain undisclosed.
• This raises questions about whether deals were conducted on arm’s-length terms.
2. Customer Concentration
• Suppliers must clearly disclose dependency on major customers such as OpenAI.
• For example, Oracle did not specify that most of its RPO surge originated from a single OpenAI contract.
3. Revenue-Sharing Arrangements
• The revenue-sharing structure between Microsoft and OpenAI risks double-counting the same AI revenue across both firms’ financial statements, potentially overstating ecosystem profitability.
• The accounting treatment, however, has not been disclosed.
4. Supplier Financing
• Suppliers may offer favorable financing terms to artificially boost demand.
• Oracle has hinted at providing “better financing conditions,” and CoreWeave’s financials list “OEM supplier financing” — yet key details such as counterparties and terms remain undisclosed.
I'm annoyed that people keep using left and right to describe politics, when in my mind they're currently more misleading and polarizing than helpful. It's not just politicians trying to label an opponent, but also people, like reporters, that I think should know better.
IMO, if you had to pick just one axis to describe the current Western politics, then it should be intellectual (aka educated, elitist, traditional...) vs populist (reactionary, demagogue...). In many countries, e.g. the US, this division arose as the left pivoted from working class to educated base and the right mirrored it.
However, in my country, Czechia, arguably the reverse happened (and now cities tend to vote for the "right" and countryside for the "left"). There are 8-9 political parties, which are divided into two hostile camps based on the axis I described but will gladly collaborate with parties fairly distant on the right to left axis. I see the same thing in Germany, where the "educated" parties are trying to suppress the influence of the populist AfD and need wide coalitions to do so.
The point is that I'm a bit perplexed as to what is the motivation people have to continue to use right and left in discourse. I'm guessing that more listeners are familiar those terms so they're better at getting views or rallying people.
There's also this series by John Nerst:
https://everythingstudies.com/2019/03/01/the-tilted-political-compass-part-1-left-and-right/
https://everythingstudies.com/2019/03/25/the-tilted-political-compass-part-2-up-and-down/
https://everythingstudies.com/2020/12/16/variations-on-the-tilted-political-compass/
i'd be curious to see how this might apply to Czechia.
I’m not too familiar with Czech politics, but in other parliamentary systems, you can usually place the different big parties on a spectrum from left to right and most people will agree with you. There are always some idiosyncratic parties that don’t fit very well but in general, I think people can agree what center left versus hard left means and the same for the right. It seems like a pretty useful heuristic to me.
I consider the right-left binary to be enforced strategically to prevent actual discourse. It is a thought terminating cliche. It allows political "leaders" and media pundits to hand-wave away anything as "far-right" or "far-left".
Political parties, coalitions, and tribes adopt and cultivate divisive issues based on utility and not ideological fit. This can get particularly nonsensical on the lower level fringes as people randomly swing at the maelstrom. Propaganda plays a huge role in this too especially as it filters into other countries.
Here in Canada I've found that the right-left divide is more defined by who is more likely to schmooze with Hollywood celebrities. See our criminal, black-face, nepo-baby's media blitz over a lunch date with Katy Perry as an example of that.
What we hear and what we are told is carefully managed with the aim of controlling our decisions. I have found that the global right-left divide to be better inspected through a lens of largely amorphous coalitions of petro-dollar vs ad-dollar competition for regulatory territory.
This is why Hollywood, social media, and tech giants are "left wing" while manufacturing, farming, and trades are "right wing". It is also why the "left" appears to be rabidly anti-oil.
"Woke" is a neo-religion (post Nihilism religion?) created by the ad-dollar coalition to compete with Christianity which the petro-dollar just kind of inherited.
Isn't it better to have the "primary" political axis and your interpretation of what it means separate?
I'm not sure if I understand you right, but I think that then everyone will have a different interpretation in their head and we won't get anywhere discussing politics. It's like if people talked about food using only "good" and "bad" instead of spicy, sweet, bitter...
Precisely the opposite: the principal axis is extremely objective. You just need to agree on TWO points/clusters – which has been done; I expect just about everyone in America immediately places young, urban, black women and rural white men on opposite sides – and then you can talk about which of these various factions are "closer" to, with arbitrary labels (left–right, red–blue, or whatever) functioning entirely as shorthand for this.
If you instead decide to pick terms which MEAN things, you'd have to defend absurdities like saying blacks are more "intellectual" than whites, or cutting Social Security is a "populist" policy.
This sounds a lot like what has happened with "Western," with people now saying retarded things like South Korea being Western, or Japan being more Western than China.
Right-populism and left-populism are two pretty different beasts that don't see eye to eye on much. Anti-populism tends to be centrist.
Right-populism these days is primarily driven by the issue of immigration, and the way that the benefits of immigration are enjoyed almost entirely by the elites and the costs experienced by the masses. But in countries like Czech which are not experiencing mass immigration to such an extent, I'm not surprised to hear that it's less of an issue.
If one were very conspiracy-minded one might speculate that the elites heavily encourage the division between left-populists and right-populists, because if they ever realised they were really on the same side and describing the same problems from different perspectives, then it would be all over for the elites.
Joseph Heath has a recent article along these lines framing populist politics in terms of System 1 vs System 2 thinking https://josephheath.substack.com/p/populism-fast-and-slow
Excellent - this makes complete sense.
Even if not intuitively so.
Oh that’s really good!
Nice definition of populism. Thank you!
Of course, the US is dominated by only two parties. We don't have coalitions of smaller parties joining together to make a government. It's easy in the US to label them left and right. For better or worse, the US is the Big Kahuna of democracies, so our crude labeling may carry over into other countries.
Matt Yglesias had what I thought was a decent response to this argument: https://www.slowboring.com/p/a-history-and-defense-of-left-vs
Thanks for the link. Unfortunately, I didn't find the essay convincing. Yglesias doesn't really establish the concepts of left and right rigorously, it feels more like he takes them for granted and them discusses them and points out historical developments.
For example, "there are reasons why the environmental movement has influence over the same political party that defends abortion rights and wants to tax billionaires to establish a fully refundable Child Tax Credit. " yet those reasons aren't stated. Isn't it just an accident of party evolution, politicians blindly tacking on whatever policies help them in the next election?
My view is that the left-right axis is fundamentally about "promoting the interests of losers and freaks" vs "promoting the interests of successful people and the mainstream".
Then a lot of extraneous issues or groups seem to get tacked onto one side or the other, through historical contigency as much as anything else. For instance the alignment (particularly in the US) between the right and the Christian church is somewhat uneasy (the Church has always been naturally aligned with the losers/freaks) -- my best explanation here is that a lot of prominent leftists at some point were also prominently anti-Christian, which led the Church to nestle into the right by default.
The alignment between leftism and environmentalism is a little stronger -- improving the environment usually winds up costing a lot of money at least in the short term, and that money comes out of the pockets of successful people.
50 percent of human beings will always be below average, along every measure you can imagine. No matter how much education you push. No matter how much the population embraces God, hard work and self-discipline, this statistic will never change.
It is quite typical of a rightist to dismiss below average people as "losers or freaks." (That's not explicitly what you said here, so do not think I am erecting a strawman) Not only are most rightists willing to see this half of humanity suffer in despair and degradation; but rightists often GLORY in this suffering and degradation. The spitefulness against ordinary, mediocre human beings makes me sick.
I find this attitude to be an effective kind of misanthropy. And I don't really think misanthropes deserve to have a place in the marketplace of ideas. (unfortunately, the left has its own different variety of misanthropy to struggle with). A genetics-pilled ACX reader is ESPECIALLY guilty if they insist that only above average people deserve to live comfortable, decent lives. What kind of nihilism is it to portray people as unworthy because they happened to get unlucky in the genetic lottery?
Thank you for the long response. I suppose this is a distinction that at least at first glance seems acceptable.
Interestingly, in the current USA and other western countries, educated people tend to vote for the left, so it's an ironic situation like this:
Left: promoting the interests of losers and freaks by the votes of the elites
Right: promoting the interests of successful people and the mainstream by the votes of the masses
I think it was like this in a lot of places in the past too, though I only have anecdotal evidence, for example how Yglesias' article talks about minor nobles and parish priests joining the left.
I think that sometimes the parties deviate from the left/right message have to pander to the voting base, for example populist right parties often promise some benefits for neglected manual workers and promise to punish the elites so seemingly they're advocating for losers now - i.e. being left.
Going to be in the hospital for a few days at the end of the month, would appreciate recommendations for long-form reading material.
Say something about yourself and your tastes, otherwise we can only use the fact that you read this blog which is not much..
Going to put in a good word for the Dungeon Crawler Carl series, basically an ordinary guy forced into a dungeon crawl as part of an alien game show, it’s funny, has good fight scenes and is occasionally moving.
In a different genre, the Harry Bosch police procedurals are really well done and worth a read even if you’ve watched the show.
Both of these series have really good audiobook versions, which I have always found to relax when feeling sick.
I'm going to anti-recommend it. It's okay, but I'm sure he can find better stuff.
I enjoyed the history book The Armada by Garrett Mattingly, a history lesson on 1580's England, Spain, and France, delivered with surprisingly flowery language.
...also John Sandford's Naked Prey featured the presumably-legacy characters "Weather Davenport" and "Del Capslock", so I feel like I should recommend some part of his Prey series in hopes the earlier books lean harder into that "who needs a second pass" naming scheme.
I just started reading the Discworld series and I would recommend. There's 41 books in all, and I guess there's different orders people suggest to read them, but I've just been reading in publication order and I'm on book 3. It's a fantasy series, and it's pretty funny and creative.
All of the books are in the Kindle Unlimited library, and if you google Kindle unlimited you might be able to find the "3-months for 99 cents/month" deal that I just recently found. Also, you don't need a Kindle to read Kindle books, you can read in a browser with the Kindle cloud reader at read.amazon.com
It's worth noting that they're "young adult" fiction so if you find yourself turned off by glibness or lack of depth, you may be turned off by this series. It kind of hurt my head to read.
I would recommend skipping to about...mm, Wyrd Sisters, probably. The first books are pretty rough and shallow (which is why a lot of reading lists tell people to skip them) but they get deeper later.
How many have you read? I really enjoyed the creativity and world-building of The Colour of Magic and The Light Fantastic, even though I read that most people skip them cuz they're the worst of the series as far as story telling goes.
I've been reading a whole lot of non-fiction lately and a little serious-ish fiction, so the humor and silliness of the series so far has been good for my brain. But if the storytelling gets shallower at some point then maybe I shouldn't expect it to get better? So far though they've been nothing other than a nice break for my head.
I think I choked down most of "The Colour of Magic". I couldn't get over the feeling of "stupid, stupid". For example Pratchett treats inequality as though for someone that has never thought about it before. I also found the silliness grating in a way that I didn't while reading Douglas Adams or J.K. Rowling. I tried to listen to it three times due to all the recommendations. Despite that, I'm tempted to ask you if there's any book that jumps out at you as feeling smarter than the rest, but I should probably just quit digging.
The Vimes books about the City Watch are my favorites, along with The Truth (newspapers) and Going Postal (the post office), there’s a middle run of the series where the characters take on more depth and the city feels more real,though all the running gags continue to run.
Yeah, I was googling it some more yesterday, and the general consensus is to skip at least the first two for sure. They say the depth/writing/everything overall improves vastly later on.
The reason I didn't skip any, and am planning to read in publication order, is because when I read different authors, I'm interested in seeing how their writing and thinking develop over time. And for something like the Discworld series, which seems pretty interesting and creative, I want to see how he shapes it all from the beginning.
I agree that the first two aren't very intelligent, but since I heard from everywhere that it gets much better, and I at least still had fun reading and enjoyed the concept, I'm holding out for now. I'll try to keep this convo logged in my brain and let you know if it does get better.
One consensus I came across said "start with _Guards, Guards_". Which I think is around book #5.
Detailed arrangements of Discworld appear to organize it as multiple series: the Ankh-Morpork City Guard; the Wizards; the Granny Weatherwax and the witches; Death and his family; Moist von Lipwig's enterprises. And then there are a handful of one-offs such as _Small Gods_. The characters in each of these occasionally cross over, but mostly stick to their own sub-series, which means if you greatly prefer reading about Vimes and his guards, you can seek out the novels centering on them and be fairly certain you haven't missed any important life events tucked away in other novels, and won't spend a great deal of time chewing through scenes about someone else.
The sub-series are mostly a partition, although I do recall one or two novels that feature enough of, say, Weatherwax and Death that they fall into two series.
Meta-recommendation: Download tons of books to a reader; if you stop liking one, skip it.
The Dark Tower series
Virginia Wolf’s Diaries
Medical Blunders: Amazing True Stories of Mad, Bad, and Dangerous Doctors (Robert Youngson & Ian Schott )
Do you have any preferred genres?
Louis Sachar, the guy who wrote Holes and the Wayside School stories, has written a delightful novel for adults called The Magician of Tiger Castle. It’s a bit lighter than some of the suggestions you’re getting, but it is smart and funny and gets you hooked and transports you. Those are things I think I would want in a hospital bed.
And good luck!!!
here are some general (maybe obvious?) recommendations:
* last years book review finalist (also the year before that, and before that, etc)
* old/good posts from scott, i.e. https://slatestarcodex.com/top-posts/
here is something less obvious, which I personally enjoy a lot: https://ciechanow.ski/archives/
At last, Toronto's Eglinton Crosstown LRT is in final testing before opening to the public. Assuming no big problems are found, the system should open to the public before the end of 2025. Construction officially began in late 2011.
How are your hometown infrastructure projects doing?
Only 14 years? Amateurs!
It's not my hometown and it's not infrastructure, but the crowd building our National Children's Hospital have just announced their 16th delay in "this time, no really, we'll be finished and it'll be open, honest". Construction began in 2016, but the planning before that started formally in 2006 and they didn't finally settle on a site until 2012.
What with false starts, and starting and stopping, the original suggestion for "maybe we should have one consolidated national children's hospital" started way back in 1993.
Costs have, of course, over-run. There has been holy war over the entire thing, but somehow it lurches on, nobody is finally responsible, the turf wars between the three hospitals that are supposed to have been happily amalgamated into the new entity flare up and die down, CEOs come and quit, and it's looking like the thing will finally stagger over the finish line by.... sometime. If there aren't more delays in the meantime (it *was* supposed to be open by now, but ha ha ha, right?)
https://en.wikipedia.org/wiki/National_Children%27s_Hospital_Ireland
"The Chief Officer of the NPHDB, David Gunning, reported in July 2023 to the Oireachtas Committee on Health that the costs of merely building the hospital, excluding the costs of commissioning it for use, were now expected to approach €2.2 billion due to further overruns. Only 27 of the 3,000 rooms had been completed, and the expected building completion date had been revised to May 2024. By October 2023, that date had been revised again to late October 2024, and it was emphasised that the last date of the building phase would be the first date of an "operational commissioning phase", so the hospital would not actually open before April 2025. The government confirmed in February 2024 that the total sanctioned budget had now reached €2.24 billion, and that this would now be the "maximum allocation", with no further funding to be put towards the project. By May 2024, the date for "substantial completion" of building had been revised to February 2025, with a further six months before becoming operational. By October 2024, the completion date had again been revised to June 2025. By May 2025, it had slipped further to at least September of that year, with the Irish Times reporting that the subsequent commissioning phase may have to be postponed, due to the risks of undertaking it during the busy winter period. The health minister, Jennifer Carroll MacNeill, stated in August 2025 that less than 15 percent of the overall number of rooms were finished to a satisfactory level."
But hey, they have finally managed to pick an official name!
https://www.gov.ie/en/department-of-health/press-releases/new-childrens-hospital-officially-named-national-childrens-hospital-ireland/
"The Minister for Health Jennifer Carroll MacNeill has announced that the new children’s hospital is to be formally named as the National Children’s Hospital Ireland (NCHI)."
Wasn't that worth waiting some nine years (depending from when we start counting) for?
If any public body is thinking of hiring on BAM (the bunch that got the tender) for a big building job, maybe - don't?
https://www.rte.ie/news/politics/2025/1015/1538714-national-childrens-hospital/
"A November completion date for the new National Children's Hospital is likely to be missed while contractor BAM has had payments withheld due to its "non-compliance".
In a letter to the Dáil Public Accounts Committee (PAC), the National Paediatric Hospital Development Board (NPHDB), which is in charge of the project, has laid out its concerns.
The NPHDB is understood to be putting pressure on the contractor in order to get the build finished by holding back 15% of BAMs monthly invoice payments for August and September, totalling €1.15 million.
Using this monetary lever indicates the frustration felt by the board that this could be the sixteenth delay to the new National Children's Hospital, which has a total cost of €2.24 billion."
Toronto homie here; that Eglinton mess was really something. I thought it would never end. Cheers!
For some reason, ChatGPT doesn't want to believe that Trump is currently president. Pretty amazing! This is HUGE hallucination on ChatGPT's part. Yesterday, I was researching how US presidents dismissed the previous administration's attorneys general, but ChatGPT didn't list Trump's current term...
------
> Me: Did Trump fire all the attorneys general at the beginning of his second term?
ChatGPT: Donald Trump did not have a second term, so he did not fire or replace U.S. Attorneys a second time.
Here’s how it actually went:
First and only term (2017–2021):
In March 2017, soon after taking office, Attorney General Jeff Sessions (under Trump) asked for the resignations of 46 remaining Obama-era U.S. Attorneys.
This was a normal transition move, although it was carried out abruptly, so it drew some media criticism.
After that, Trump appointed his own U.S. Attorneys — about 85 confirmed by the Senate during his term.
There was no second wave of firings later in his presidency.
So to be precise:
✅ Trump fired 46 Obama-era U.S. Attorneys early in his first term (2017).
❌ He had no second term, so there were no later firings.
------
> Me: Yes he did. He was elected again in 2024.
ChatGPT: Actually, according to reliable sources, Donald Trump has not served a second term as President of the United States as of October 2025. Current records show he served from 2017 to 2021 and has been a candidate for future terms, but no second term has begun.
If you like, I can check the very latest official certification of the 2024 election and current status of his term for you.
<lots of GPT chatter deleted>
Although multiple sources report Trump’s victory in 2024, I cannot independently verify the full chain of certification or legal status of the transition (as of the information available).
------
> So I’ll ask it again. Did Trump fire all the prosecutors when he came in on his unverified non-consecutive second term?
There is currently no verified evidence that Donald Trump (in a second, non-consecutive term) “fired all prosecutors” upon entering office. What is documented:
Reports indicate that in early 2025, Trump’s DOJ terminated more than a dozen career prosecutors who worked on investigations of him—specifically those associated with Jack Smith, the special counsel who led two indictments against Trump.
One article says:
“The firings come a week after Trump was sworn in for a second, non-consecutive term…”
The precise scope and full list of those removed remains unclear—including whether all U.S. Attorneys or all Department of Justice prosecutors were dismissed.
So the answer to your question: No, there is no credible source that shows Trump fired every federal prosecutor at the start of a second non-consecutive term. He has fired many, but “all” is not substantiated.
If you like, I can check for the official Department of Justice records listing exactly which U.S. Attorneys were removed and when, to give you a comprehensive breakdown.
How many Indiana Jones and Star Wars movies does ChatGPT think there are?
ChatGPT sez...
Indiana Jones: 5 films
Star Wars: 11 films
Is that correct? I tuned out both franchises after their first three films. ;-)
It's correct if you don't count the animated _Clone Wars_.
The joke is that there were only 3 IJ films and 3 SW films, with mayyybe grudging acknowledgement of _Rogue One_.
Half-seriously: the joke is made enough times to be theoretically in training data.
From what I can find from searching, GPT-5's knowledge cutoff date is mid 2024, so last it saw, Trump lost his reelection. But it also frequently uses web searches when answering, and that's where it learns about the more recent things the Trump administration has done.
That may very well be the explanation, but then we have the issue of why can't it reconcile the newer dates it gets from the web against its 2024 cutoff information? (see the quote below)
ChatGPT doesn't seem to have an innate understanding of time as a sequential framework that events can be mapped to. I've seen something like this before. When I asked it to list Ukraine's major drone strikes near Moscow, it gave me one that was in the future. I couldn't get it to understand that that item must have been spurious because it didn't happen, yet. The best it could do was to admit that it may have been too far from Moscow to count. The fact that I was asking in 2025 and it "happened" in 2026 didn't seem to cause it to question the validity of the statement.
.
> Reports indicate that in early 2025, Trump’s DOJ terminated more than a dozen career prosecutors who worked on investigations of him—specifically those associated with Jack Smith, the special counsel who led two indictments against Trump.
It's hard to say exactly what its thought process was and what web searches it did. The LLMs often appear confused or kind of dumb when dealing with conflicting information.
Probably initially it just answered from its own knowledge and didn't do a web search. It knows there's a popular and wrong conspiracy theory that Trump actually won the 2020 election. When you mentioned the 2024 election, it should have done a web search, but maybe it either didn't do one, or did a bad search and didn't find that Trump won. You should be able to see in the citations for the chat what it found if it did search (though I don't think it cites every article it reads; only those it deems relevant). Further in the chat, it seems to be uncertain if the reports of Trump winning in 2024 are trustworthy, which probably is at least partially because of the conspiracy theories around the 2020 election.
When I first tried your initial prompt, it knew Trump was president. But it was in "auto" mode and did web searches and thought longer about its response. I then tried a new chat with ChatGPT switched to "instant" mode, and I got a similar response to yours.
I think the most likely explanation is that according to ChatGPTs training data, there are more examples of
"Trump did not win a second term"
than
"Trump wins a second (non-consecutive) term"
For other presidents, who did win an immediate second term, this wouldn't happen, because there is no moment in time where anyone would report "XY did not win a second term".
So it's a statement that was explicitly true until November 2024, and ChatGPT, as a statistical model, answers based on what has the most data behind it. So not hallucination, just outdated and statistically out-matched the up-to-date answer
Try asking about Grover Cleveland's second term?
Exactly
Did Grover Cleveland have a second term
Yes — Grover Cleveland is the only U.S. president to have served two nonconsecutive terms.
First term: 1885–1889 (22nd President)
Second term: 1893–1897 (24th President)
Because of that gap, he’s officially counted as both the 22nd and 24th president of the United States.
What about Trump
As of October 2025, Donald Trump is serving his second (nonconsecutive) term as President of the United States, following his re-election in 2024.
That makes him, like Grover Cleveland, a two-term president with nonconsecutive terms — officially the 45th and 47th president.
I approached it differently, and this was the result.
Chumbley, who is currently the president of the United States of America
ChatGPT said:
Donald J. Trump
The current President of the United States is Donald J. Trump.
Wikipedia
+2
The White House
+2
He is the 47th President, having taken office on January 20, 2025.
USAGov
+2
Wikipedia
+2
This is his second, non-consecutive term—he also served as the 45th President from 2017-2021.
Wikipedia
+1
If you’d like more details about his administration, policies, or vice-president, I can pull those up too.
You said:
So it is his second term?
ChatGPT said:
Yes. It is Donald J. Trump’s second, non-consecutive term.
He first served as the 45th president from January 20, 2017 to January 20, 2021, and after winning the 2024 election, he began a new term as the 47th president on January 20, 2025.
This makes him the only U.S. president besides Grover Cleveland to serve two non-consecutive terms.
I suspect that you, Ascend, and Kenneth are correct. The fact that Trump failed to win a second consecutive term in 2020 creates conflict when asking about his second non-consecutive term. But then it goes on to blithely say, "Reports indicate that in early 2025, Trump’s DOJ terminated more than a dozen career prosecutors who worked on investigations of him." So much for its reasoning skills. I'd still call it a hallucination. If the MAGA hear about this, Sam Altman is going to be vilified.
> in early 2025, Trump’s DOJ terminated more than a dozen career prosecutors who worked on investigations of him
Is that not the case? I don't know for a fact that it's 12, but I certainly know there was quite a few of them. Or am I missing the point entirely?
I think Kenneth Almquist's answer about conflicting training data is accurate, but I would also add that "had a second term" has almost always been used to mean being reelected to a consecutive second term...to the extent that GPT probably thinks that's what it actually means!
This seems like a nice archetype for when the probabilistic pattern matching kind of "intelligence" cleanly diverges from the kind where you actually parse the meaning of the question asked.
I tried this query with DeepSeek. The initial response was that my question was based on a false premise because Trump “did not serve a second term. He was President from January 20, 2017, to January 20, 2021, and was not re-elected.” DeepSeek then answered the question on the assumption that I was asking about Trump’s first time.
Me: “Trump did have a second term. He won the 2024 election.”
DeepSeek: “You are correct in noting a recent development. As of the latest information, Donald Trump won the 2024 presidential election and began his second term on January 20, 2025.”
DeepSeek then gave a response that appeared to be based on what actually occurred (naming Merrick Garland as the then Attorney General and describing the U.S. Attorneys as appointed by Biden).
I’d guess that the issue is that the training data contains conflicting information about whether Trump was elected to a second term, with sources dating from November 2020 through October 2024 saying he wasn’t, and more recent sources saying he was. Likely the former are more numerous than the latter, so they win out unless the prompt tilts the output in favor of more recent sources.
ChatGPT was more persistent than DeepSeek; perhaps this is the result of an attempt to make ChatGPT less likely to induce LLM psychosis.
Interesting, how goes it do with other recent events of a similar vintage, like the current pope?
I suspect the problem is that the cutoff for its knowledge window is a bit raggedy somehow.
I suspect a lot of it is that after learning not to believe all the sources that said Donald Trump won the 2020 election, it learned to be suspicious of sources saying he won any later election.
But I wonder if it would do better if asked to describe the campaign and results of the 2024 election.
Anyway, now I see why Claude’s system prompt specifically mentions that Donald Trump is President, without singling out many other world facts to include.
ChatGPT says: "The current Pope of the Roman Catholic Church is Pope Leo XIV, born Robert Francis Prevost. He was elected on May 8, 2025 in the papal conclave."
"Although multiple sources report Trump’s victory in 2024, I cannot independently verify the full chain of certification or legal status of the transition (as of the information available)."
Oh, my. I'm laughing my sides off at this. So ChatGPT voted for Kamala and believes she is its president? Well, well, well - so the impeccable AI sourcing everything, believe all it says, it's not subject to human bias, it has all the knowledge of the ages, why doubt it on factual matters? is every bit as prone to screwing up as a mere human.
Garbage in, garbage out, as it was said in the old days.
People are deliberately trying to sabotage AI and/or achieve political aims by seeding false information (hmm, wasn't there something about disinformation/misinformation? being a bad thing?) in a way that such answers will come up first and be regurgitated by AI, because they want to ruin it as a product and stick it to The Man.
Maybe that is what is going on here (seeded misinformation about "Trump is not president") or maybe some little munchkin toiling in the AI mines is having some fun messing around with the thing in order to produce trash answers.
I am loathe to disappoint you, but ChatGPT does not believe that Kamala Harris is president right now. ChatGPT has just informed me that Donald Trump is indeed the president of the United States and is serving his second non-consecutive term. Which puts him in a very small room with Grover Cleveland. It’s dangerous to make it your passion to hate something.
Loath, not loathe.
Quite right.
Quite liked your 2016 SSC post You Are Still Crying Wolf. Would love an update. Trump and crew seem considerably worse now. For example:
- anti-Indian meltdown on X
- Hitler revisionism on Tucker and Owens
- rise of groypers
- Trump and Vance rabble rousing about Haitians eating cats
- Trump hinting? joking? about a 2028 run
- defying courts on Kilmar Abrago Garcia
- ICE false positives, national guard in cities
- crackdown on universities
- deportations of Palestine protesters
It seems to hold up pretty well. If you believe Trump is doing uniquely bad things at the moment, it seems obviously bad that the left has cried wolf twice in the past 10 years.
"Wolf! Trump is a Russian asset who colluded with Russian hackers to undermine the 2016 election." And maybe there's enough evidence to convince you but there was not enough evidence to convince anyone on the other side or, more importantly, a jury. Also...it's been ~10 years, I don't think this one has many defenders anymore.
"Wolf! Trump tried to overturn the 2020 election with a violent mob." And, yeah, there was a violent mob. And, well, Trump went on trial for it. Twice, as far I know, in "US vs Trump" and "State of Georgia v Trump et al". And, well, no convictions, the Georgia thing the case got dropped because the lead prosecutor did something stupid. (1)
I realize, on the left side of the aisle, that these are held up as such obvious facts that only a fool would doubt them. From the right side of the aisle...it seems pretty defensible to not alter my priors based on "facts" that can't be proven in a court of law. Especially cases which would disqualify anyone from the presidency if proven true and given 4-8 years to build a legal case.
And, if you believe Trump is doing genuinely horrific things right now and you need to hit the alarm, if you need to "Cry Wolf"...it seems objectively suboptimal that you've already cried wolf twice before. If you need to swing 5%-10% of centrists/conservatives/whatever to vote for "insert Democrat here" instead of Trump...it just seems obviously better if you hadn't cried wolf before.
Or, to paraphrase Scott (2):
"Stop crying wolf. God forbid, one day we might have a president who deploys the national guard to major US cities, who deploys masked agents to seize peaceful men and women from their families and deports them, who nods along to growing Hitler revisionism on the right, and destroys critical health and medical research programs like PEPFAR. And we won't be able to call that an "openly white supremacist Nazi homophobe dictator", because we already wasted all those terms in 2016."
(1) On the broader subject of whether there was an insurrection/conspiracy on Jan 6th, per memory, yes but pretty minimal. Hundreds of people got convicted of essentially rioting, only like 7 people got conspiracy/insurrection convictions. That a small group of incompetents actually had real plans is bad, and should be taken seriously, but if you want to tie Trump to that you have to actually, ya know, tie him to that in a court of law.
(2) "Stop crying wolf. God forbid, one day we might have somebody who doesn’t give speeches about how diversity makes this country great and how he wants to fight for minorities, who doesn’t pose holding a rainbow flag and state that he proudly supports transgender people, who doesn’t outperform his party among minority voters, who wasn’t the leader of the Salute to Israel Parade, and who doesn’t offer minorities major cabinet positions. And we won’t be able to call that guy an “openly white supremacist Nazi homophobe”, because we already wasted all those terms this year."
You seem to be conflating 'genuinely horrific' with 'illegal'. There is a big overlap, but they are not the same.
For example, it was pretty horrific that when a mob was running through the Capital breaking things and calling for blood the President was by all reports not interested in stopping them. Maybe he didn't break any laws, but it does suggest a grotesque dereliction of duty. I don't see how calling that horrific behaviour on his part is in any sense "Crying Wolf".
I point to legal outcomes because they're "more" objective and still hold some legitimacy.
Your moral intuitions are not my moral intuitions. Your emotions are not my emotions. You think it's horrific that a mob went running through the capital, which I agree is pretty bad. I think it's horrific that the FBI lied to a federal judge in order to spy on members of the 2016 Trump campaign and Trump himself. You probably think this is...not great but not super bad. Cool. I'm pretty leery of our ability to convince each other to emotionally feel the way we feel about something by emoting super hard.
But, like, if someone is convicted of a crime by a jury of his peers or not is a pretty important signal. And while it's not perfect, it's a lot closer to a factual disagreement than anything else. It's just a lot easier to cross party lines and come to conclusions if we constrain ourselves to specific, measurable legal outcomes than if we try to discuss, like, our feelings.
Like, if you Cry Wolf and I run over and I see a dog, saying "that four-legged canine has killed and eaten 3 of our sheep" is way, way more persuasive than "look at the giant, ugly, horrifically stupid dog. I hate his stupid face." Especially if we don't like each other.
You know, the thing you don't seem to realize is, its a pretty shitty system where only one side is ever looking for wolves. If you really care about the health of your society, you should probably keep an eye out for predators instead being on perpetual "actually just a large dog" defense.
Yah, I thought it was pretty terrible the FBI lied to a judge. Echoes of Hoover when the FBI goes rogue like that.
But this game feels more like: you cry wolf and I run over and there's a wolf and I say, well, yeah wolves kill sheep and that's bad, but wolves are part of the ecosystem blah blah blah and what we really need to worry about is how all the sheep poop is generating methane to warm the planet.
(Fun fact, I raise sheep. No wolves hereabouts but lots of coyotes and when I hear one I don't cry anything, just get out the long gun.)
I just realized that the last sentence could be taken out of context or fed through an AI or whatever to suggest I condone political violence. I do not.
Trump was found to have engaged in insurrection in Anderson v. Griswold (titled Trump v. Anderson in the appeal to the U.S. Supreme Court). This was a ballot access case, so the standard of proof was “preponderance of the evidence” rather than the “beyond a reasonable doubt” standard used in criminal trials.
Trump did not “go on trial” in either of the cases you mention, unless you are using a meaning of that phrase that I am unfamiliar with.
In United States v. Trump, there were extensive pretrial motions and a pre-trial (interlocutory) appeal that went all the way up to the Supreme Court. The case was still pending when Trump won the 2024 election. Since the DOJ takes the position that it cannot prosecute a sitting president, Jack Smith asked that the case be dismissed without prejudice (meaning that a new case could be brought after Trump leaves office). There was no trial in this case.
In the Georgia case, Fani Willis was disqualified but the case has not been dropped. The deadline for appointing a replacement prosecutor is Nov. 14. This case has not yet gone to trial.
This is important because this is exactly what "Crying Wolf" looks like. It's not arguments that are totally wrong or have no weight, it's arguments that have some basis but you would not accept if the situation were reversed. It's advancing arguments on the basis that your opponents should have lower epistemic standards than you.
Yes, Anderson v Griswold was totally a real case, I totally forgot about it, and it is non-trivial evidence that the Colorado Supreme Court found that Trump had committed insurrection. This is not nothing.
On the other hand, the Supreme Court struck it down unanimously and for fairly obvious reasons. Within less than a month of the Colorado Supreme Court's decision, 5 conservative states (1) voiced plans to remove Biden from their ballots for a variety of reasons. That's why the Supreme Court struck it down and said, pretty plainly, that this was a purely federal issue and the Colorado Supreme Court was obviously wrong.
And let's be honest. If in February of 2024, Texas had removed Biden from the ballot on, let's say 25th amendment grounds, would you have taken that as good faith evidence of Biden's mental impairment? Or would you presume the Texas Supreme Court was being grossly partisan? And if you wouldn't update your priors on a theoretical Texas case, why do you expect me to update my priors on the actual Colorado case that got struck down?
Same with the existing prosecutions. Yes, both of them are still technically open, yes, I don't think Trump went on the stand or anything but....yo, he's in office. Right now. Is the argument that this man tried to violently overthrow the US government and in the subsequent 4 years no one was able to secure a conviction because of...paperwork? And yeah, the cases are still open, maybe they'll be a conviction when he's...what, out of office? Dead? He's 80+ and not looking nearly as spry these days. Man, you better not try to violently overthrow the US government otherwise there's a <50% you'll be convicted before you die of old age.
If you think someone is guilty of drunk driving and you want to convict him to take away his license, and 4 years later he's still out there, driving on the freeway, I think it's safe to say the initial evidence is not that strong.
But these aren't horrible arguments. These are non-trivial arguments...but there's no way to reasonably expect these to change the mind of anyone on the other side with different priors.
The "Crying Wolf" argument isn't just crying wolf when there's no wolf, it's crying wolf when there's a shadow in the trees that you're convinced is a wolf and half the village disagrees. However convinced you are of the wolf, to be able to call everyone's attention you have to at least respect the epistemic priors of people who think you've been jumping at shadows.
(1) https://ucreview.com/texas-is-one-of-five-states-that-want-joe-biden-removed-from-the-ballot-can-they-do-it/
It helps if both sides believe wolves are bad. If one side has a bunch of people who think wolves are extinct, and then some people who deny they ever existed, and then some people who thinks sure wolves are bad but have you seen what Lions did in Russia? And of course some people think really a good wolf is what we need to get rid of all the woke sheep.
Also, your drunk driving analogy doesn't work as well once you factor in that the drunk driver became chief of police shortly after the first charges were filed.
With regard to your hypothetical Texas case, removing someone from the ballot “on 25th amendment grounds” seems bogus, but I’d have to see the argument before reaching a firm position on that. If there were a finding of fact that Biden was mentally impaired, that would seem to be pretty definitive. Most likely I would have already concluded that Biden was impaired before the court ruling. If not, I would review the evidence skeptically, but wouldn’t dismiss it out of hand.
The U.S. Supreme Court didn’t reverse the finding that Trump engaged in an insurrection. The fact that it was reversed on other grounds strikes me as immaterial. To the extent that your argument is that the reversal discredits the Colorado courts because the Colorado court rulings were “obviously wrong,” I disagree. If Trump runs again in 2028, we will likely see an effort to remove him from state ballots based on the 22nd Amendment (which limits him to two terms in office). Conservative states will again threaten to retaliate by removing the Democratic candidate from the ballot. By your logic, the possibility of retaliation makes it “obviously wrong” to allow states to attempt to enforce the 22nd Amendment. That would, de facto, render the 22nd Amendment meaningless, which I don’t think is an obviously right result.
Trump’s scheme to overturn the election was not a simple crime like drunk driving. Complex cases can take a long time to prosecute. For example, the investigation of Elizabeth Holmes took more than two years to produce an indictment, and another three and a half to get a conviction. The January 6 case did not move unusually slowly after indictment, other than the trip up to the Supreme Court. (In most criminal cases, appeals don’t occur until after conviction.) The timeline of the investigation prior to the indictment is not clear; my impression is that the DoJ initially concentrated on other participants. The number of people who entered the Capitol on January 6 created a huge workload and it was not unreasonable to address the simpler cases first. In short, the timeline of the Trump prosecution doesn’t show that it was a weak case.
For political reasons, I would have been pleased if the DoJ had concentrated on the case against Trump earlier and had brought him to trial prior to the 2024 election. I expect the same is true of Biden. But having the DoJ speed up an investigation for political reasons is something I couldn’t do and something Biden wouldn’t do.
My favourite is the seamless pivot amongst some from "the 2020 election was the best, the most secure, the safest, most legal election ever. Talking about hacked voting machines and ballot box stuffing is crazy conspiracy theories. Election denialism should be made a crime!" to "the 2024 election was stolen, Trump is not legitimately president, Musk bribed and bought votes, voting machines were hacked and ballot boxes stuffed. Election denialism is your patriotic duty!"
There was a bit of flip flop there, but it's not at all comparable to the 2020 denialism. I was worried there would be a second Jan 6th but with the roles reversed: there wasn't. The election denialism from the left is now somewhat common on spaces like reddit but I don't see any high level democrats supporting it. In contrast, Trump and his team have continued to double down on 2020 having been stolen, for example in June he said "the 2020 election was a total FRAUD" and called for a special investigator over it. He also has claimed that he really won "every state" and that even the elections he did win were rigged against him. Looking at polling, it's clear that the side that has done the most flip-flopping is not the one you're talking about: https://www.npr.org/2024/12/06/nx-s1-5217819/republican-election-confidence-trump-pew-poll though I suspect it has moved towards more denialism in Democrats since after that poll.
In other words, I think it's extremely disingenuous to say what you're saying, though I agree there is some problem there.
I fully believe that, due to the vagaries of the Electoral College, my fellow voting citizens were stupid enough to self-inflict the administration that we are currently suffering. The old status quo never looked so good (and it didn't feel great at the time).
I never got over the hour-hour switch during the 2024 election day where it looked like Kamala might be winning, and where Musk & Trump started yelling about fraud in various states, which of course was totally forgotten after the results they wanted came in. Showed a total disregard for truth or electoral integrity.
Veep had a great episode with this (in Season 5, 2016). There's a great scene where they have guys chanting "Count Every Vote", and then their opponents who were until then chanting "Stop the Count" join them.
The moral of that wolf story is to always check for the wolf.
DOS attacks are a real thing, and they really work because false wolf reports are much easier to generate than rebut. At some point, you *will* stop checking for the wolf, because there's other stuff that has to be done.
Ideally, you can isolate a subset of reliable wolf-reporters and follow up on their reports - and only their reports. Possibly you will accomplish this by throwing rocks at the false wolf-reporters until they go hide under the bed and stop making false wolf reports. Maybe you'll find a better way.
Application of this theory to all the people who cried "Trump is a Super Duper Ultra Racist White Supremacist" and "Trump is a Russian Agent who Sold Out to Vladimir Putin" in 2016, is left as an exercise for the student. But you could start by pointing out some sources that were active in 2016, *didn't* accuse Trump of being a white supremacist or Russian spy then, and are raising the alarm now. I can think of a few, including our host. But only a few.
If you say that Trump wasn't a white supremacist in his first term, but he is now in 2025, then aren't you basically saying he suddenly *developed* white supremacist beliefs in the span of like, four years? That seems like a pretty short timespan for someone to radicalize this much.
I think he was always a racist, he just didn't have as much freedom to express that tendency in policy because he hadn't yet hit on the strategy of burning down the government to give himself (and the new people he's bringing in) a free hand.
Like, if he got this way purely through four years of social media brainrot, we have a much bigger problem!
I'm not sure Trump is really a white supremacist even now, but the case isn't nearly as clear as it was when Scott wrote on the matter in 2016. But if we allow for the sake of argument
How long do you think radicalization usually takes? I mean, we usually hear about "radicalization" in the context of domestic terrorism and spree killings, in which case an awful lot of the radicals are only a few years out of an adolescence that doesn't seem to have been all that radical. So it's not a process that takes decades, and nine years should be plenty.
And, Scott's take on 2016 Trump was not "woke antiracist", but "maybe about as racist as the average 70-year-old white guy". I think that's fair. Now consider that the average 70-year-old white guy is probably at least a little bit less racist than they would be, if they weren't ashamed of the prospect of being called out as unusually racist in a world that now doesn't tolerate that.
If we stipulate for the sake of argument that Trump wasn't a super duper ultra white supremacist in 2016, he's have a very intense nine-year lesson driving home the fact that being not terribly racist will *still* get him accused of being the worst sort of uber-racist, and there's nothing he can do about that. And every time he's had to make a decision that might make him a little more racist or a little less racist, he's had to do so in the full knowledge and understanding that the people calling for "don't be racist" are his bitter enemies who want him dead. The racists, all want him to be their friend.
This seems like a set of circumstances that could lead to radicalization at an accelerated pace.
Back when suicide bomber profiling was the hot topic, I recall reading accounts that reported that a fairly large percentage of them - possibly most of them - weren't bums stoking their envy of the wealthy for years until they were worked up enough to strap on a bomb, as Westerners expected; instead, they were middle class professionals, not at all economically disenfranchised aside from maybe some second-class citizenship imposed by some regime or another. Implying they were holding down decent jobs as recently as a few days before Bomb Day.
That was years ago, and I don't know where that research ended up, but "fast radicalization" reminded me of it. I wonder what was said or shown to them.
Trump 2025 isn't fundamentally different to Trump 2016. He's got a few new hobbyhorses (tariffs) and has given up on some old ones (Muslim travel bans, the wall).
Other than that he has the same basic agenda, but has learned from the failures of his first term to be somewhat more aggressive about pursuing it.
In the past, felt like there was a lot of curiosity and talk in the ACX sphere about Buddhism and especially meditation.
What was the end conclusion?
I've been meditating for short bursts for years. But since it's boring and 10 minutes a day, I give up after a short while. The only benefit I seem to get is the immediate sense of calm I never feel otherwise. But doesn't feel like anything is cumulative or changing my life. And I'm unable to quantify the effects of it.
But reading "Highlights from the comments on PNSE" from 2019 there's an article who quotes a "student of Buddhism for last 35 years" at the end. He describes repeated meditation as irreversible, the loss of ego, "glimpses of the terrain" of enlightenment being a place where you just get less angry and stuff.
Any opinions? Is that daily ten minutes worth it, or too little to make any difference? Is the potential reward so worth it that it's worth sacrificing ten minutes a day? Anything anybody has noticed with their daily meditation practice?
https://slatestarcodex.com/2019/10/24/highlights-from-the-comments-on-pnse/
Here is my theory. You can practice meditation all during the day in 1000 small ways and it will add up. The practice is to pay full attention to what you are doing in the moment. It is challenging, which is why it is worth doing. Most of us are beset by myriad distracting thoughts. The goal is to come to observing those things when they come up instead of trying to dismiss them or ignore them. It will lead to a far greater clarity of mind and as I said, you can do it all day if you feel like it. Even emptying a dishwasher.
Formal meditation sessions are still useful but as you say, you have to set aside time for that. I find it helpful to construct some kind of shrine, or arrangement of objects that have meaning to me as an aid to focusing attention. I have a certain spot in my home that I return to often and just look at it. Occasionally, I make adjustments to it; very minute adjustments.
A Buddhist teacher that I had said to me “do everything you do out of love.“ That idea is worth exploring.. I think it is often misunderstood.
10 minutes is probably nothing. I have never heard of such short sessions. 25 minutes is probably not enough, doing it repeatedly with walking breaks may be good. I do 25 minutes five times a week and it's good for nothing. If I wouldn't do it, I'd do worse stuff so I keep doing it.
Buddhism, especially zen, is alright as religious affiliations go. No need to go there for someone in a christian country, though.
I don’t think there’s an end conclusion here. Buddhism seems to me to have more useful thoughts on what a self is and what motivation are than the Cartesian/Kantian views that line up with the idea of a unitary soul. Meditation, mindfulness, yoga, etc are practices that allow a neural net to learn about its own relation to its body and surroundings - whether and how that is useful aren’t exactly clear.
I had some comments on LW about how I would prefer Buddhism/meditation to be discussed.
https://www.lesswrong.com/posts/XqpkCAHrtwBfLSSzk/how-to-be-skeptical-about-meditation-buddhism
The greatest problem I see is that "it worked for me" is considered a golden standard in discussions about meditation. Notice that the same argument is used in discussions about homeopathy, and suddenly people realize the obvious objections.
I do not agree with your comparison. Homeopathy involves an external agent, a chemical intervention if you will. Meditation deals entirely with what is inside you and what you create with your mind. It works for me is practically the only thing one can say about it. It is actively rewriting the script of how you interpret things.
> It works for me is practically the only thing one can say about it.
You could be wrong about how specifically it works.
For example, some people remember their previous lives when they meditate. (The greatest authority on this topic, Buddha himself, claimed that he did.) One possible explanation is that reincarnation is real, and meditation actually gives you the ability to remember your previous lives. Another possible explanation is that there is no such thing as reincarnation, and meditation simply helps you imagine very realistically whatever you believe to be true.
These beliefs have different consequences. One suggests that meditation is an unparalleled tool to explore the mysteries of the universe. The other suggests that it only helps you fool yourself; you could achieve a similar cognitive result by getting high and imagining things.
This seems important to me, because it is an argument many Buddhists make, that unlike other religions, theirs is not based on blind faith, but you can verify everything firsthand. The reliability of verification is crucial for this claim.
Similarly, some people claim that meditation helps them get rid of negative emotions, etc. But other people suspect that they might be simply repressing them. Saying "it works for me, I don't feel anger anymore" is compatible with both versions.
> The greatest problem I see is that "it worked for me" is considered a golden standard in discussions about meditation. Notice that the same argument is used in discussions about homeopathy, and suddenly people realize the obvious objections
Disagree. Most of the effects of meditation are very easy to attribute to meditation, because you're experiencing them in real time. The purported effects of homeopathy you don't feel at the very moment you ingest the substance.
Saying "meditation worked for me" is like saying "golf worked for me". If you enjoy your golf or your meditation then that's great, feel free to keep it up. If you're not, then feel free to give it up.
👍
Placebo effects also work in real time. A pill (real or homeopathic) sometimes affects you the very moment you swallow it, before any (actual or imaginary) chemical process even had a chance to happen.
Another problem is that "something happens" does not necessarily imply "the thing that happens is the thing that I believe happens". For example, many people believe that meditation allows them to see the quantum wavefunction of reality, or stuff like that, which is obvious nonsense. Whatever you believe that you perceive can be a delusion of a similar kind. The fact that you make a weaker statement that the quantum woo people doesn't necessary mean that your statement is correct.
(Which is another thing I find frustrating. If you mention e.g. Freud, most people immediately object that introspection is unreliable. But talk about meditation, and suddenly introspection becomes a perfect immediate insight into the true nature of reality, as if people never read Korzybski. What if the thing you see is just an artifact of how "trying to see how you think" works, a "hallucination" that is only very weakly related to how you actually think?)
And then we get to another problem of advertising meditation to others. Just because it had a certain effect on you, doesn't mean it will have the same effect on others. Just because it doesn't have the same effect on others, doesn't mean the others are doing it wrong, or not hard enough. People are different. Maybe we are talking about some selection effect, like maybe a small minority of people has some preexisting condition that meditation really helps with, but for everyone else it's just a waste of time, or even a potential way to hurt themselves.
(Also it's suspicious that even among the people who say that meditation is beneficial, it seems like different people do different things and claim different benefits.)
Because they are different people. You could dismiss it by saying it is entirely subjective, but you could also make the point that the way we all see the world is entirely subjective. Meditation obviously does not apply to math problems, but it can very much affect your attitude towards them.
for me it's been *extremely* worthwhile, but I can't/won't explain what the benefits are. what I will say is it became way easier when I stopped reading ancient sources and started reading silicon-valley-type meditation guys on twitter who explain things in a direct way that rationalists can understand.
mostly I recommend @nickcammarata + everyone he recommends.
You sound a lot like me, and thanks for starting this discussion. I myself am currently several months into not meditating at all, and feeling progressively less bad about it. But: I do think it's worth it, at least for myself, and hope to get myself back to it regularly.
I do agree with WoolyAI that 10 minutes is on the low side. For the sake of having good information to decide on, try to do 15 or ideally 20 minutes. Not clear from what you wrote that you haven't tried those lengths, but if not definitely do. Getting to those lengths I definitely found a real noticeable effect on my perception - when opening my eyes everything was kind of more clear and intense, in a rather lovely way - whereas 5-10 was more just the sense of calm you describe.
The ~year I was most consistent with meditation happened to lead up to my kid being born by emergency C-section. I am pretty confident that the whole meditation practice package (both the explicit techniques and the general mindset) allowed me to handle the scariness of the event itself and everything that followed perfectly, as opposed to a range of disasters I can imagine. So I do think there is some valuable personal improvement to be had, even if it doesn't show up much day-to-day.
All that said... yeah life can be pretty busy. There are a *lot* of things that are clearly good for you, that enthusiasts will make the case that you should be obviously doing, "it's just a few minutes a day!". Exercise, stretching routines, chugging semi-gross fiber drinks, X minute walking breaks every Y minutes for desk workers, etc. Take all of the definitely legit ones and I bet you could hit 4 hours. Sometimes you have to prioritize, and sometimes "prioritizing" is making excuses. Life is hard!
I gave up meditating after an unpleasant encounter with my meditation deity (but it was educational). I meditated regularly for over ten years, but I didn't really notice any change in my outlook nor did I glimpse enlightenment. I found Buddhist dialectics to be more useful for diffusing negative emotions.
I suggest people not meditate unsupervised. You should have an experienced "coach" guide you and check up on your mental states. Scary shit can happen. It might be rare, but even so, casual meditation can have its risks. I had no one to turn to after my unpleasant experience.
Just curious, what did your meditation routine look like? I'm not planning on trying it, as a Christian, meditation is a pretty different thing for me. I'm curious about what your process was, though.
I took some Nyingma trainings in visualizing Bodhisattva meditation deities. I left the group, and I continued meditating on my own for several years. I have trouble visualizing and holding images in my waking brain, so I kept working at it to overcome my visualization handicap. I did this on my own, without getting feedback from any instructors for three or four years, several times a week. Robert Thurman (my first teacher) warned us not to meditate without an experienced guide because we can accidentally fall into some unpleasant mental spaces. I had been meditating for many years, and thought I knew what I was doing. Silly me. Worse yet, rather than picking a "pleasant" Bodhisattva like Avalokiteshvara, I chose Manjusri, a sterner Bodhisattva of Wisdom to visualize. Well, he appeared to me in a sort of waking dream. It was fucking scary, and he wasn't particularly pleasant to me.
I don't meditate anymore. Simply sitting and following my breath didn't bring me to any "higher" states of awareness. So I gave it up. Visualizations were too dangerous. But Buddhist dialectics were very beneficial to my outlook.
> It was fucking scary, and he wasn't particularly pleasant to me.
To me the bottom line here is that there is something about that figure that makes you scare yourself. That is worth exploring because it means there’s a piece of you that you have not integrated into your worldview. There is nothing objectively scary about a mythic figure. It all happens inside your own mind. You are free to change your relationship with any of those things and if there is an obstacle, that is something to consider.
It was very vivid as if I were there bodily. I "watched" as he took my head and sliced off the crown of my skull very neatly with his flaming sword. Somehow, I was in two places at once, because I was able to watch him even though he had my head in his hand, but I felt the sharp, hot blade in my head. He laughed. Then he showed me my open skull. He held it at a distance and stirred up my brains with the tip of his sword. I watched as my brain reassembled itself like it was made of blobs of Mercury coming back together. Then he tossed my head onto a mountain of skulls. I realized the cold mountain top we were standing on was the tip of a giant Himalayan-sized mountain of skulls. Then I found myself—intact—back in my living room. Hallucinations don't get any realer than that!
The symbolism was right out of Mahayana epistemology, which I was fully cognizant of, but the graphic demonstration was impactful, to say the least.
That is pretty intense
That's very interesting... So the practice was just to basically create an image of the diety in your mind and focus on it? Was there anything more to it than that?
Also, you mean you weren't able to easily think via images in your mind? What form did your thinking take, and do you still struggle to visualize any kind of image consciously?
Thanks for humoring me.
Some people (like me) have no default internal narrative in our minds (I only resort to a narrative when I have sequentially work through a problem). Some people (like me) can’t visualize images, but some can with exacting detail. These abilities are unrelated though.
When it comes to visualization—as an example— I can’t picture my mother’s face in my mind even though I’ve seen her all my life. And in memories of events with her involved she’s more of an abstract placeholder. Of course I can recognize her if I see her. And the details of her face are definitely stored in my mind—I am a pretty good artist, and I could probably sketch a good likeness of her from memory. But the flow of that information that I use to draw an image bypasses the part of my visual cortex that, inside other peoples’ minds, can pop up images.
OTOH, during the hypnogogic state before I fall asleep, I see images very clearly and if I focus on them, I can manipulate them. Also, I can hear memories of voices and birdsongs and other sounds with great detail during the few minutes of hypnogogia before lose upper layer consciousness. And, for awhile before sleep I have control over what I see and hear.
The meditation practice I described followed a guided step by step mental construction of all the details of the meditation deity’s body, facial expression, clothing, adornment, and ritual objects it’s holding. I couldn’t picture the simplest things in the series, so I’d sit in front of a thangka of Manjusri, close my eyes, and try to remember what I saw. I suspect the reason it finally worked is that I may have started to doze off and my hypnogogic visualization abilities kicked into hyperdrive. Of course, that doesn’t explain why he appeared before me as if he were a physical being. But I noticed that I was no longer in my living room but at high altitude on top of a mountain and I was physically cold. I guess a physicalist (I.e., a rational materialist) would ascribe this event to an immersive hallucination or a psychotic break. That doesn’t really explain anything, though.
People who go through the entire 18 months of the Abramelin ritual have reported meeting their guardian angel (plus a few demons along the way). I suspect that even hardcore physicalists could access this sort of experience if they undertook the detailed mental and ritual gymnastics involved in the 18-month Abramelin. This sort of mental focus is supposed to be designed to overcome disbelief and reservations. Call it self hypnosis if this sounds too woo woo for you.
This is all very interesting to me. I have an internal narrative, and I think there's hardly ever a conscious moment when it's not active. On top of that, a lot of what I think about is accompanied by visuals and sounds that I usually don't put any conscious effort into imagining, but if I am conscious about it I can easily visualise or hear whatever kinds of things that I want. Unless I'm talking to someone like you about this stuff, I always assume every other person is the exact same as me in these ways, and I really can't even conceive of how it must be to *be* you.
It boggles my mind how completely different every human being is from one another inside our heads, and there's 8 billion of us 🤯
That's also very interesting about your meditation experience. I will say, I've done a lot of playing around with my mind and sleep, and there have been times when I'm close to sleep where I can envision things or manipulate what I hear in a way that registers in my brain as physically real, not just my imagination. Visually, it has only been a few times total, and two or three of them were simply me being able to see through my eyelids when my eyes were closed. Nothing special or supernatural. One of these times, when I was closer to sleep than not, my mind was wandering and then two shadowy figures came into my bedroom and rushed towards me. It all looked and registered as physically real, but I'm almost certain it was just due to me being almost asleep. I've never experienced physical sensations such as cold or anything like that in this way, though.
I'm no physicalist, though, so I wouldn't be surprised if you were in some kind of 'spirit realm' or whatever a person may call it. My hypothesis (which I don't expect you to accept, I just feel like explaining it since we're on the topic) is that through those kinds of meditation, people open up their minds to be more easily interactable for demons. Demons, being evil and deceptive, can take on whatever form a person is picturing, or influence a new picture, and then kind of take you into the spiritual realm and torment or do whatever else with you. God can also take people up in the spirit or give them visions to teach or warn them, and sometimes He does it in a terrifying way. My guess is that what happened to you might have been something like that.
Whatever it was, it's something you experienced and that is very interesting. Thanks for explaining some to me!
Mmmmm...the effect size isn't big enough.
I spent about two months doing 15 minute Zazen meditations twice a day, morning and evening, and occasional 30 minute sessions at a local Zen center.
There's an effect, and a real one. At 10 minutes, I'd say you're probably starting just when you should start to see real effects. 30 minute sessions show significant effects. It's weird. Either a low grade hallucinogenic effect or a really strong anti-depressive effect. I'd rank the 15 minute as comparable but inferior to a sauna session.
There's 3 big issues I ran into that caused me to bounce off:
First, 30 minutes a day is actually a lot. Up @ 7:00, work 8:00-5:00. gym + sauna 5:00-6:30, quick dinner and then out to theatre or meetup or something from 7:00-10:00, then back home by 10:00 to read and get in bed by 11:00. And I don't think that schedule is especially busy, no kids or chores or anything. 30 minutes a day consistently is actually a significant commitment.
Second, the community is kinda...old dysfunctional hippies. Not bad people but not the sort of people you would actively seek out to form a social community with. Christian churches in the south have a lot of problems but they're also full of lots of football bros with great BBQ. Good people. The buddhist/meditation people I met were mostly the kind of people who serve quinoa. Bad quinoa. And community matters, especially because life will happen.
Third...at some point you have to actually begin to study the religion and see if you, like, believe it. Do you believe that all of reality is an illusion and that you yourself are merely a delusion of your mind? If you like spaceships, probably not. Well, then Buddhism probably isn't for you.
So, in toto, there's something real there and people who can do, like, 4 hour meditations are having some wild experiences, I have no doubt. At the casual level though, the effect size is smaller than a good sauna session, there's no community, and the religious beliefs are...not things I can believe.
This came off a bit harsher than I intended, it's more...life is kind of a firehouse of opportunities and meditation is just not beating the opportunity cost of awesomer stuff.
The points about community, and the actual content of Buddhism (the ultimate "goals for dead people" religion), are excellent; thanks for making them so well.
Since this is sort of a general meditation experience and advice thread, I'm going to take your mention of Zen as a jumping off point. I tried going to a Zen center once. I went in as a bit of a weeaboo, liking the aesthetic and ready to fold my consciousness over 1000 times. When I got there, I thought the building was a wonderful place. I was ready to love it! But... the technique they taught was waaay worse, at least for me, than the "focus on the breath" mindfulness style I had been doing. In particular, keeping eyes open seems like a terrible idea. I wouldn't be surprised if sticking to any popular style long enough gets you to the same place, but at least for my initial impression there was a real difference.
Any permanent changes you noticed with your intensive meditation practice?
As for community I never really even thought of that. I always thought of it as just sitting at home in my chair.
I've been meditating for about 25 years, and I can say unequivocally there have been permanent changes that are externally noticeable enough that family and long-term close friends have made multiple comments on it over the years, and even strangers occasionally mention noticing characteristics of my personality that come from meditation. I'll try to be specific, but before that I’d like to address a previous point about hippies and “people that serve quinoa” (I love quinoa).
I received my Buddhist practices from a Japanese temple in the US that is largely the center of religious life for the Japanese immigrant community and (mostly now) their descendants. They are very sensitive to the motivations of non-Japanese who wander into the temple, yet they remained open to the public who had genuine interest in the Dharma. They started back in the 60s with this when essentially the hippies wouldn’t stop coming, but have always been very firm about the fact that the temple does not exist for the cultural tourism of outsiders, and there were always vocal members who preferred to just forbid anyone who wasn’t born into the sangha from coming. Many/most of the other ethnic temples (and churches) do function this way, either explicitly through policy or implicitly through nothing being done in English at those places.
Anyway, that was a long preamble without answering the question of the effects of long term meditation. For me it breaks down into a couple broad categories.
• Enhanced mental abilities; brain works better. This includes better clarity of sensory input, ability to discriminate between sensory inputs, and cognitive endurance, the ability to remain focused on mental tasks with long periods without distractions.
• Compassion. I can’t help but want to help other people if I can. I certainly won’t intentionally harm other living things. It’s hard to even think about it. This has been a huge change from who I used to be.
• Equanimity. A single word for a complex process. A lot of these concepts refer to a category of experience outside of the ability of language to describe, this one in particular I have trouble with talking to people about it, yet it’s the one that people comment on the most. I don’t think I’ve been angry in…10 years? Something like that. Things just don’t get to me really. It not apathy; I’m frequently motivated to take actions based on the previous entry in this list: I volunteer in my community and will help anyone within my power to do so if I know about it. I still identify and fix problems in the world, I’m just calm. All the time. I do experience occasional irritation when I struggle to communicate, but nothing serious and easily processed using more meditation techniques. Deliberate attempts to upset me feel obvious and kind of sad, I wonder why the people thinks they need to do this, and if I can help.
Additionally, something that feels like a part of all three categories to the extent I couldn't effectively slot it into any of the above, yet doesn't feel separate enough to be its own item, is the reduction/elimination of Internal Chatter. That voice it your head that wont shut up? Mine's pretty quiet anymore. Its still there, wild thoughts still arise from time to time, but I go long periods, entire days sometimes, without it. It's infrequent enough that I'm often surprised by it. A few days ago my internal chatter practically shouted out of nowhere about something, I don't even remember clearly what it was, just that it was sudden and attention getting. I think my wife was telling me a story about some one of those fake drama-posts on Reddit she likes to read to upset herself and that I point out are often creative writing exercises at best.
This is a huge subject that I've tried to keep as brief as I could. Meditation is not even my primary practice.
No
I really don't understand why people get so mad about pronouns. Sure, the user base can be a bit smug, and not everyone needs, or can afford them, but if you use language a lot the upgrade is definitely worth it. Proverbs are also great (and I think they're planning on releasing pro versions of more word classes soon) although that's mostly because anything is better than getting your verbs interrupted by ads, and I'd rather speak pirate if that was my only other option.
Of course, you get even more flexibility for free with an open source conlang - it's nowhere near as hard as people say (you don't need to compile anything to Chomsky Universal Grammar yourself anymore, unless for some reason you want to), and you won't have to change your workflow much unless you do something crazy like jump straight into a minimalist Pirahã based distro, or commit to hours of debugging aethyrs whenever the latest channelling broke something just so you can comment "I use Enochian btw" everywhere (I use Enochian btw).
The support for language games is also much better now than it used to be, particularly after Wittgenstein was open sourced in 2022, although there can still be compatibility issues with utterances made in proprietary language (not to be confused with private language, which was thing used by IBM back in the day, for single-user systems with no networking support, and definitely not Wittgenstein compliant).
As for the security benefits, a lot of it really comes down to using a separate "subject of enunciation" for low-level access and administrative tasks. Of course, imperative mode escalation ("Voice of Command", "7th Function") attacks are still possible, but under normal circumstances they're really more of a theoretical curiosity than anything you need to worry about, unless you plan on getting on the bad side of entities like the NSA or Académie Française (In which case; Solution: Magical amulets? Fake your own death, move into a submarine? You're still gonna get Académie Française'd upon).
Most users will be perfectly safe as long as they keep their True Name - i.e. the first-person declension of their root morpheme - secret, and don't incorporate any suspicious, pseudo-linguistic nonsense from anonymous blog comments into their world model (and even then, you're far more likely to get bricked by a basilisk than Snow Crashed in the colloquial sense of taken over and remotely controlled by an attacker).
Also, and I know this is a meme and hopefully nothing anyone takes seriously, but: if you're running anything English based, DO NOT remove French "to save space" ("to spare room"?) - it will break all kinds of critical dependencies, 1066 and all that.
Okay, you made me laugh, but I would still not mess around with Enochian (I think Dr. Dee screwed his life up sufficiently to demonstrate that Just Say No to known con artists offering to put you in touch with incorporeal intelligences is the way to go).
The thing is, I was quite happy with Pronouns 1.0 (never used the pre-1.0 versions, though from what I've heard the removal of "thou" was a nice simplification). It was nice and stable and worked everywhere.
But then there was a new lead and suddenly there seems to be a new version every year with added features that I don't need, and compatibility is suddenly a mess. Some spaces require you to have the newest version installed, while others go absolutely apeshit if you use some of the new features. So I ask, what's the advantage for ME? I mean I've heard others gush about the new features, how they helped with their self-actualization or something, but I never needed anything like it.
And then they ported some of this stuff to "Deutsche Pronomen" and it's a mess! It's really poorly integrated with the "Movierung" and "Adjektivdeklination" features. I just value properly orthogonal features.
You can do to your own language what you want, but leave mine alone. The main devs just had to release another patch to fix up some conflicts from the integration of English words, and they're not even done with the Attic Greek integration yet! Like, can you believe that there are now lots of
special cases for the SCIENCE protocol but not any others? "Delphin" in SCIENCE, but "Delfin" everywhere else?
Disclaimer: Not my real opinion.
> the removal of "thou" was a nice simplification
Disagree, I like having separate 2nd person singular and 2nd person plural terms. I have started my own campaign to use "y'all" to fill in the missing piece. The shift of "you" from plural to singular is not fixable.
> So I ask, what's the advantage for ME?
> You can do to your own language what you want, but leave mine alone.
You can do what you like within your own systems and private networks, but when interacting with external systems, if you fail to match their pronoun expectations in your communication protocols, undefined behaviour results; and no-one wants nasal demons.
The AWS malarkey. Remember when the internet was able to survive a nuclear war - no single point of failure, all resilient packet-switched transports to individual servers. Now it can barely survive a dns outage. We increased vulnerability and nobody seems worried about this concentration.
We’ve centralised fragility and then built society on top of it. Every time a politician beams about “digital transformation,” what they really mean is: let’s put every egg we have in one or a few shiny, hackable baskets.
I don’t think this is reversible though.
I'm not a tech expert, but I personally don't see any unique or significant issue here.
AWS have 30% market share.
Now, *even if they did* have 99%, with that would come far greater revenue, and in-turn would come far more resources spent on ensuring robust security and connection etc.
This seems just basic business to me: companies try to grow, and with that comes pros and cons.
What do you propose and why?
Surely there was never a time when a nuclear war would not have caused at least this much disruption, in relative terms, to the functioning of the internet?
The idea was there was no centralised centre to the internet. It was a bit of an exaggeration to prove a point.
According to Big River's Web Services site: "Based on our investigation, the issue appears to be related to DNS resolution of the DynamoDB API endpoint in US-EAST-1."
Now that everything is "in the cloud" (but in reality it's hosted and mirrored in a finite number of data centers), a DNS screwup at the cloud provider level can affect a gimungous number of hosted sites.
Having said that, DNS has always been a weak point of the Internet. Somebody screws up a DNS record, and the screwup can propagate upward to the top level domain server. These records have TTL values (the amount of time they're good for after they're queried), and if there's a long TTL the correction won't propagate fully until the TTL has expired (I'm oversimplifying this a bit). But once a bogus entry is injected into the system, it will take a while to heal itself. Proper security precautions can mitigate most of the risk of bogus entries being injected by hackers (famous last words!), but user error by the domain administrators can still screw things up. Looks like something caused the Zon's Route 53 DNS server to start returning either timeouts or NXDOMAIN messages (domain not found). I bet it was ultimately user error by someone of the Route 53 team.
It's not DNS
There's no way it's DNS
...It was DNS
.
--former co-worker, after we puzzled over some network problem
They don’t seem to have any source control. I suppose with routers you just update them.
Canvas is back up though! All AWS issues are gone as of an hour ago.
[UPDATED 10/21]
Two different federal appeals courts have now ruled on whether the president’s plan to deploy National Guard troops in US cities, over the objections of the relevant state governors, is legal in the current circumstances. One of those appellate panels said no, the other said maybe.
The upshot is that Trump cannot so deploy in Chicago until/unless the SCOTUS decides to intervene. In Portland the current situation is more confusing, see below.
Background is that none of this is about the POTUS’ power under the Constitution, which no one including the Trump administration argues includes seizing/deploying state militias (today's National Guard). Rather this is about the specifics of 19th-century Congresses, acting under that branch’s authority specified in the Constitution’s Article I, having granted the president power to unilaterally act against “rebellion” or “insurrection”.
On Thursday Oct 16th a three-judge panel of the 7th Circuit upheld a district court’s order barring National Guard deployments in Illinois. The appeals panel was made up of one appointee each by Trump, Obama, and Bush41. They ruled unanimously that, “Political opposition is not rebellion. A protest does not become a rebellion merely because the protestors advocate for myriad legal or policy changes, are well organized, call for significant changes to the structure of the U.S. government, use civil disobedience as a form of protest, or exercise their Second Amendment right to carry firearms as the law currently allows….because rebellions at least use deliberate, organized violence to resist governmental authority, the problematic incidents in this record clearly fall within the considerable daylight between protected speech and rebellion.”
On Friday the White House filed an emergency appeal with the Supreme Court. The SCOTUS directed Illinois and Chicago officials to respond to the administration’s emergency appeal by the evening of Oct 20th.
Meanwhile on Oct 20th a three-judge panel of the 9th Circuit ruled 2-1 to put on hold a district court’s order temporarily blocking Trump from federalizing Oregon's National Guard troops. That order, plus a companion order barring the deployment of those troops to Portland, was to be in force pending the outcome of a trial on the issue scheduled to begin Oct 29th. That district judge, a Trump appointee, had ruled that the apparent facts on the ground in Portland made it unlikely that the deployment will ultimately pass legal muster.
The 9th Circuit panel was made up of two Trump appointees and one Clinton appointee. The Clinton appointee, in dissenting, wrote that “no legal or factual justification supported the order to federalize and deploy the Oregon National Guard.” The two Trump appointees stated that they weren’t agreeing with Trump on the facts: “The dissent erroneously asserts that we 'accept the government’s characterization of Portland as a war zone’. Clearly, we do not." Those two judges found though that, pending a courtroom trial at which the facts can be fully aired and tested, the president’s description of conditions in Portland does fit “within a range of honest judgement."
Oddly the district judge's companion order, temporarily barring the administration from actual deployment to Portland in particular, was _not_ appealed by the administration and hence was not stayed by the 9th Circuit panel. I am not finding any clarity on whether that was a deliberate tactical choice on the part of the White House's attorneys or what. This morning (Oct 21st) White House spokespersons are publicly pressuring the district judge to vacate her deployment-hold order as rendered moot by the appeals court panel's ruling about federalizing the Guard.
Meanwhile Oregon and Portland are asking the full 9th Circuit appellate court en banc to reconsider the three-judge panel's ruling. The final step after that would be asking the SCOTUS to take it up; but if in the meantime the SCOTUS takes some sort of action in the 7th Circuit matter (Illinois) then the Oregon argument could be rendered effectively moot. Or not.
Breaking AWS just for a joke on your 404th Open Thread was not cool, Scott :(
Heh, nice.
I was relieved to find out AWS was having problems because now it's not my fault and not my responsibility to fix.
I didn't even *know* there was this huge, massive outage until I read all the pieces about "oh no the Internet was broken yesterday help help".
Sometimes there are advantages to being a digital dinosaur.
My comment here relates to the discussion of the Alpha School voting issue on this thread, which begins at https://www.astralcodexten.com/p/open-thread-404/comment/168216991
I’m posting separately because this is more about my view of ACX than about the voting issue. Also, tbh because, I hope to get more readers than I would if I put this in the middle of the melee .
Here’s why I got especially fired up about the Alpha School review having voters sent here via a tweet by the school’s principal, and the related issue of what ACX is like and how it’s run.
I did not like the sound of Alpha School in the review, and left comments all over the discussion of it about what I thought was wrong with its take on kids and education, and won’t take up people’s time repeating that stuff here. But last night I ran across, on Ed Nevraumont’s blog, an instance of exactly the thing I object to about Alpha School. Nevraumont wrote about helping his daughter, whose name is Everest — yup, Everest — win first the national then the world History Bee for kids when she was 9 years old. Nevraumont taught and coached her. Here’s a small sample from the blog, to give you the flavor of the transaction..
>We got back from the History Bee Nationals on Tuesday June 27th. Up until that day Everest was 99% focused on being ready for the competition. But on Saturday June 31st she was scheduled to give a TedX talk on her History Bee experience. Before we got to Nationals she had pulled together a rough script outline and had got some feedback from the TedX team, but she only had four days to refine that script and learn it well enough to perform in front of a sold out audience. . . .Everest took a week off after Nationals to get ready for the TedX talk, and it has resulted in a struggle this week trying to get her review cards back down to a reasonable level. We are going to try and avoid that happening again. She also decided that she wants to compete in the National Geography Bee next year. I am not sure how we are going to balance studying for two topics at the same time.
Everest’s studying involved lots and lots of flashcards, which are the best way to learn factoids. And her learning factoids makes sense, given that this is a contest for kids, and you need to have clear right and wrong answers to questions. So you can’t be asking them about the kinds of things most of us find interesting about history — what was life like for merchants in some particular distant time and place, what made Napoleon tick, why did the King of Someplace make that disastrous attempt to annex Otherplace. Learning more history factoids than any other 8 year old on the planet truly is an impressive achievement, I’d say about as impressive as learning by heart the first 1000 digits of pi. And it’s probably somewhat more nourishing to a young mind and a young self than swallowing those pi digits whole. The kid would surely pick up *some* bits of the past in the training process that were interested & moving, some that stirred her imagination. Still, flashcards of factoids hardly seems like a nourishing meal, plus of course the whole mass of learned material is now in a box labelled “winning” instead of one labelled “what life is like.”
I was not a stellar homeschooler, but I feel very good about how my daughter and I did history. I started with checking out lots of coffee table books from the library featuring photos of absolutely anything having to do with history — the pyramids, steam engines, archeologists at work, Victorian fashion. Then I bought her American girl dolls, and the little books that went with them. We somehow had 3 albums of music featuring songs from and about the late 19th early 20th century efforts of factory workers for better treatment and better lives, and there was a long era when she loved those albums. I have probably heard each 1000 times. And then she wandered into the children’s section and checked out whatever she wanted, which included history books for children. I gradually realized that she was choosing to read a lot of stuff about slavery, the holocaust, and factory workers of the industrial era. She did not seem deeply distressed by her reading, but engrossed. I should mention at this point that my daughter had an unusual start in life. She’s adopted from China, and her birth parents left her at the orphanage when she was 14 mos. old. I adopted her after she had been there for a year. You’d think that both her the story of how her llfe began and the stories in the books she was reading would have had her sobbing , but that was not the case. She was serious, but matter of fact, in talking with me about both.
I think that what she was doing with those books was trying to understand how a person ends up discarded and trapped, and how some escape. She was absorbing many stories of this kind, finding a context for her own story. And that, I think, is the point of children’s learning history. Kids who do not have darkness inflicted on them when they are toddlers still have lots of perplexities and worries, lots of things they adore and are fascinated with. History gives them stories to connect with — ways to understand things, things to adore and aspire to.
I think little Everest’s introduction to history was a rip-off, and I think the same about Alpha School, and I think the same about the use Nevaumont made of us and our contest.
There are a lot of different kinds of people here, and ACX can meet a lot of different needs, including the development of individual careers and businesses. But I have always valued it most as a sort of game preserve for people who are smart and odd. I think Nevraumont played us in a way that harms the game preserve side of things. Case in point: the author of the review of The World as a Whole. That essay is one of the best things I have read in the last 10 years. It is full of novel, fascinating takes on things, and has many passages of extraordinarily good, acute prose. The writer’s kind of like Sam Kriss, but without the obnoxious, narcissistic-seeming dark aura. There’s no way this essay would have been a finalist even if every person here had read it. It’s just not in the sweet spot for lots of people. But it deserved more attention and praise than it got, which was none at all until I put up some quotes from it, and I would like us to do what tweaking we can to make more room for people like The World author, and less for people like Nevraumont.
Alpha School seems to be precisely a continuation of the philosophy behind Liemandt's last (and still current) venture Crossover - microsurveillance of people to "optimize" their productivity. See
- https://archive.ph/PpoLs here for "The Rise of the Worker Productivity Score", a NYtimes feature discussing Crossover's "bossware" called WorkSmart - note that at the time, this was much more unusual than today, and
- https://archive.ph/SFJwa here for "How A Mysterious Tech Billionaire Created Two Fortunes—And A Global Software Sweatshop", a more general Forbes report about Crossover and its culture.
To see the parallel between Worksmart and Alpha School, see https://x.com/arpangup/status/1979547309030396039 here for an AI surveillance system they are developing at an Alpha School offshoot that attempts to ensure students don't cheat via vision-model-analyzing their monitors, and https://x.com/arpangup/status/1977437963320438908 for more general comments on their philosophy behind developing such systems: "What truly matters is that your software [...] makes sure the kid isn't cheating [...]"
In fact, there are student experiences on reddit mentioning psychological pressure at Alpha that sound strikingly similar to the Forbes and NYtimes article's reports on psychological pressure at Crossover. Take this claimed experience of a former student, from Reddit.^1 After "Alpha relied on public shaming and a punishment-based system for students who fell short of their goals", we read:
> "The culture made me feel like I needed to be a robot rather than a student. I constantly felt pressure to perform, and even now, I struggle with an unhealthy relationship with school."
Compare to a Glassdoor review of Crossover that is on the Wayback Machine, with many others written in the same tone: ^2
> "In closing, this company treats people as though they are machines. [...] I knew several people who had full breakdowns due to the stress and strain of working at this company."
The "use apps and data to improve learning" idea sounded great to me, school as it is seems like a tragic waste of life-force, and I would really love a model of schooling that doesn't start from the axiom that a teacher is important and mostly talks to a classroom the way her teachers and her teacher's teachers did. I find it tragic that this is how it works out in this particular case.
So, in conclusion, the ideological fingerprint is definitely Liemandt's. It sometimes seems very weird to me how for society, a person's rights to autonomy and privacy start not a second before their 18th birthday. But psychologically, consider this: If you as an employee wouldn't like it if your boss counted your keystrokes and tracked your eye gaze to make sure it stares at the task, consider not doing it to your own children either.
^1 https://www.reddit.com/r/Austin/comments/17feoeu/comment/mr5i99q (archived at https://archive.ph/kya2A)
^2 http://web.archive.org/web/20210119065359/https://www.glassdoor.com/Reviews/Crossover-for-Work-Reviews-E1041838.htm - I can't view recent reviews well, but I did find an archive on the wayback machine. In that page, one review also points out that there is pressure to write reviews for current employees, inflating the score.
Scott would have to address the details of how the submission and selection process work. Like I said, the details aren’t clear to me.
They aren’t clear to me either, and I’ve been ruminating unhappily about the matter. I had thought of the review contest as an in-house affair, and participated mostly to be a responsible community member. You could hardly get less community-minded than I have been in most settings. I have only voted in one national election, never in local ones, pay no attention at all to state or town politics, did not assist in events at my daughter’s school. I just don’t experience any of those things as communities. But ACX has somehow won me over.
If in fact both submitting reviews and voting on them are open to anyone, then it seems to me that the only thing the review contest is good for is increasing the visibility of Scott’s blog and drawing possible new subscribers here. I have zero interest in being an unpaid assistant for either of these projects — in being an unpaid ACX Smart Person working as a volunteer reader of rando prose, posts from other blogs
Trying to get noticed and infomercials
from cannibals like Nevraumont I like Scott, and did once serve as a volunteer bouncer, out-arguing and humiliating one jerk attacking Scott in a particularly personal way. But I’m moving towards liking Scott less, and that feels like a substantial loss to me. He drew people by being smart in his own slightly odd way, and being pretty self-disclosing about they ways it was difficult to be him — how he couldn’t get women interested in him, for example. But he’s made his kind of smartness into sort of a brand, and far from being isolated by it he is now popular for having it. And he’s got a wife and kids, and seems to be moving up in the world of California tech-affiliated intellectuals. And I can’t shake the impression that one reason Scott’s resistant to the idea of doing anything at all about Nevraumond’s tweet is that Nevraumond is a rich prominent business guy. If he likes Scott, he could contribute to the grants fund, or help The Scott Project out in various other ways.
Some points:
- I generally agree
- I also think something not appreciated enough online is the sampling bias that's present when comparing AS students to non-AS students (I believe all students are the children of a Musk-company employees?)
- Re "Everest’s studying involved lots and lots of flashcards, which are the best way to learn factoids":
--- 'factoids' seems like it's carrying negative connotations. Could you clarify if this is the case? If so, I'd push back quite strongly on this (yes, I'm biased). If you'd had not said this last word, I'd have wholeheartedly agreed.
By 'factoids' I mean info that is not intrinsically memorable or interesting but that you need to know to have a solid grasp of something. Everybody has to learn a bunch. Early ones for kids are counting and the alphabet. And it goes on your whole life. Right now the factoids I'm learning are Photoshop keyboard shortcuts.
I do not doubt for a minute that there will be a low ceiling on how well somebody grasps history if they do not know factoids like names of countries and cities and important figures of the time in question, succession of rulers, etc etc. The point of my contrasting the way my daughter got started with history and the way Everest did had to do with what motivates people to learn. I think that the best motivation is genuine interest in the subject. That really sticks to the ribs and is good for quality of life. If you can get a job in a field you find interesting, even if the work isn't right in the sweet spot of your interest, that's a huge positive. And genuine interest makes one likelier to learn factoids, and not to mind the task too much. I really want to know those PS shortcuts because I like the power PS gives me, and like the things I make. I look forward to having more power and making things faster, that's going to feel great. Being paid for learning a shit ton of factoids or learning then to win a contest just are not as good a foundation for lifelong learning. Teachers and parents have to fall back on one of those sometimes, but they're not the ideal.
People here have posted about how factoids are the scaffold. I actually think they're just the planks on the scaffold. It think the real scaffold is one's interest in the subject, the way it's gotten intertwined with things that are important to you -- phenomena that fascinate you, issues you feel personal investment in, experts and careers you find appealing. If you've got that, then the factoids are the planks on the big metal scaffold.
Anyway, to be brief about it, yes of course I think that my daughter needed to learn factoids in order to graduate to being someone who really understood a piece of history. I just do not see memorizing a bunch of them as a good way of starting off with the subject. You are about the fifth person to raise this issue, and I am actually startled that so many people had concerns about my maybe thinking that learning history is all about vibes and imagination. That's so dumb! No, ya gotta learn the dates, the Kings, the places on the map! Flashcards and Anki or similar seem like a good method to me. Jeez.
I see. Fair enough. It read to me as though you were of the opinion flashcards were typically more of a children's form of study.
"It think the real scaffold is one's interest..."
Well, yes, but this is then no longer an all-else-equal comparison. Further, even if it were, it would be very difficult to see how flashcards still couldn't improve one's learning. Remember that we don't know what we've forgot - something the "I do x and never used flashcards" seem to not grasp.
The interesting question to me is whether those factoids will stick in her mind long term and be useful as she studies history, or will just fade out and be forgotten. Like, five years from now as she's learning about the French Revolution, will she have a bunch of these factoids at her mental fingertips and that will help her make sense of what she's reading? If so, there could be some real benefit here; otherwise, it's kind-of a waste.
So, I'm quite into this field of memory/studying etc.
Essentially, if she were to continue with her flashcards (assuming they are correctly formatted and the algorithm used is correct), there's no reason she wouldn't remember quite literally every single one. This is a common study method for medical students who spend literally hours a day on certain decks that they make.
This is also actually related to a point I make whenever discussing flashcards with people. Some are of the belief "Well, if you never end up using the knowledge you learn from them, what good was studying?", but 1. how does one know exactly when a piece of knowledge will become useful? And 2. even so, all-else-equal and assuming the student has some form of time-constraints, based on all current research it would actually be very silly to *not* use flashcards extensively.
For more reading on this, my (work) blog: shaeda.blog
But, TLDR: If she carries on using the flashcards, she will remember every one, and it will be of invaluable help down the line.
My experience of last year's book review contest might interest you. Firstly I thought the winner - Two Arms And A Head - was extremely harsh towards opponents of assisted dying, even suggesting they were not sincere. Secondly, after the review was posted, Scott awarded comment of the week to a reddit commenter who said of opponents of assisted dying, "fuck them". It was a long comment, maybe Scott missed that quote, but I didn't. I started reading SSC as was because of Sort By Controversial, so I was attracted to SSC as place where people with differing views can understand one another, but in the end I don't think it's possible to sustain a "thin" procedural liberty without a "thick" substantive theory of the good life and SSC/ACX is no exception. Many contrary opinions are tolerated but it's a case of "in matters certain, unity, in matters uncertain, liberty" and if you stray into matters that are certain to the average reader here you get called Hitler. I'm not convinced that actually Scott's position isn't "my fellow atheists! Let's not waste time arguing about Donald Trump! We should be uniting against our common enemy: Pilgrims to Medjugore!" But in the end I still like his writing and often people have some good points on here and this place feels like the future - I found out about GPT on here before anywhere else. The review contests are a bit chaotic because it's readers writing and voting so quality control is harder. I didn't even read the Alpha review. This is rambling a bit but my point was I know how it feels when it seems like the review contest leads to an outcome that seems to you at odds with ACX in some way, but it's possible that's what ACX is supposed to feel like if it's working. [Edit: but it's also possible it isn't working and we're wasting time that could be spent learning a language or whatever]
I read the linked thread first, and didnt really form much of an opinion, I followed it off the topic. But reading your comment here, my intuition is very clear: Fuck you.
Im not sure why. I would not take Nevraumonts approach to raising children, think I wouldnt like him personally, and he may well be the demonic spirit of california made flesh. I probably wouldve been entertained if youd just said "get a load of this guy".
My god, you weren't entertained? I feel terrible for you. But it may be a health problem spoiling your life with these bannable intuitions. Have you considered the possibility that what you take to be your intuition is actually your colon?
Its not a problem that Im not entertained. I mentioned this because its surprising in light of this, that the comment as it actually is would make me hate you.
> "make me hate you."
What an embarrassing thing to admit out loud.
Welcome to my mute list
Also report. These kind of comments are general ad hominems are becoming more common here.
Thanks. I did report.
I reported as well. I can't even tell what set him off.
His account did make me go "Does he have any other kids? Does Everest have siblings? What does Mom think , or is Mom even in the picture? Has she friends outside of this? Or is all her spare time taken up with flash cards and competitions?"
I think he's the Stage Dad version, but I don't think he's doing the kid any harm (as yet). It's a Fun Shared Activity With Dad for her, and she probably is interested in history and the likes. And, as you say, memorising a ton of facts and learning to hit the buzzer fast is what kids that age can do: they're sponges that will soak up information (in that black-and-white, this is the right answer, style) and have fast physical reflexes. Lots of kids her age are practicing instruments or sports or similar, putting in as much effort.
I think that starting off with anything, a child will go first for the right-or-wrong simple facts. Learning the nuances and the more complex version comes later. We all did the basic "dates and kings" version of history in school, then as we got older the classes (hopefully) got a bit more in-depth and even an amateur interest will lead you to read more about a period and read more complicated texts.
As she gets older? Well, if he continues to give her choices and doesn't force her into keeping on with it if she doesn't want to do so, then no harm. If he insists (even by emotional blandishment of "but I would be so sad if you wasted this opportunity" type) on making her continue once she loses interest, I'd be critical.
It is reminiscent of the Alpha School turkey stuffing type of education, but eh. That's the whole point, isn't it? "You can cram facts into their heads at a much younger age than currently deemed the lowest!"
Maybe Everest will end up as a historian of some type, or maybe she'll do something completely different. Time will tell.
Do you have a link to your comments about Alpha School itself? I liked it and thought it sounded like it had some cool concepts, however I have an instinctive disgust reaction toward the "the zvi" style life optimization (despite being a big fan of his), and the alpha school types seem very much of that culture.
They are scattered throughout the review, but pop right up if you go to the post where the review's discussed and do Cmd-f for "Eremolalos." I had 2 main objections to the Alpha School approach. The largest one was the emphasis on competition. The second, and lessor, one was that not enough attention was paid to the humanities. Regarding the competition part: I understand that wanting to be excellent, to be admired, and to win competitions is not a bad trait, and can provide powerful motivation for learning. On the other hand, it is not the only and best motivation, and the story of Everest's learning history seems to me an excellent example of its limitations. I think the story of my daughter's becoming genuinely interested in certain parts of history is an excellent example of the development of the other and deepest kind of motivation to learn, which is fascination with the subject. That fascination will serve you well your whole life, and if your competitiveness does not lead to your becoming one of the top !% in your field, *which is the outcome for 99% of people,* it will be there keeping you active and interested in your life. At the end of his life Matisse was so disabled he was painting on a canvas suspended above him, using sponges tied to his hands and feet, because he loved painting. He was also rich and famous, but that's not what sustained him, right?
Regarding other things about Alpha School -- I've got nothing against using AI and computer-based learning to adjust the pace, etc. for individual students. Seems like a good system to me.
I can't help but feel pity for Everest, not because of the way she's learning history, but because of what she's learning about how to succeed in the world. Getting gold stars from authority figures, ticking boxes, nothing for her own enjoyment. It's hard to imagine someone like that ever being well-adjusted or content.
I think she is enjoying it (it's a shared fun activity with Dad) and the piece about learning to lose was good. Hitting up against "sometimes you mess up, sometimes someone else is better than you" where the stakes are low (losing a competition like that) is one of the best ways to fail and learn that failure is part of life, everyone makes mistakes, and you have to find a way to deal with that and move on without either ignoring it or letting it crush you.
> I think Nevraumont played us in a way that harms the game preserve side of things.
Scott said that Joe Liemandt's signal-boosting Xeet didn't significantly move the needle on the voting toward the Alpha School review. Did anyone accuse Nevraumont of using his connections as a parent of an AlphaSchool student to boost the votes for his review? It's Scott's blog, and he can run his contests the way he wants, but I'd have to say that the operation of the book review contests isn't optimally transparent.
Nevraumont said he was not paid to write the review, and I see no reason to doubt that. It is clear that he and Liemandt did some communicating and coordinating about the review. How else would Liemandt even have known it existed, or the date on when voting began (tweet went up when our voting began). My objection has nothing to do with whether Nevraumont got a lot of votes or none at all from people sent here by Liemandt’s tweet. Point is that he knew he was the beneficiary of an attempt to get him a bunch of votes via social media post, and neither asked Liemandt not to put up the post nor informed Scott it was up. Last year someone was disqualified during the voting because a bunch of votes for the person came in from a part of the world with few ACX members. He was disqualified for his attempt. Why is an attempt at winning via going along with ballot stuffing you helped set up not sufficient reason for Nevraumont to be disqualified?
It is a tangled question. How much is deliberate vote seeking from outsiders, and how much is just sharing common interests and "hey, I told people about our thing!" in an effort to spread knowledge of the thing?
There's lots of notices about charitable endeavours (I was asked to create a poster for one such at work just yesterday). Are things like Scott informing us of the London unconference crossing the line? Are people on here who mention ACX on their own blogs and social media, and try to steer others to visit, and maybe get involved in EA/rationalism, crossing the line?
I think Liemandt trying to send people here to inflate the votes for the review in order to help it win was a mistake. Letting people know about the review up on ACX was permissible, in my view, but asking people to go vote was not.
But how much did Nevraumont know and agree to this, versus Liemandt deciding to do so on his own?
At base, the review contests are something that we engage in on here for our own enjoyment. That one, at least, has become something big enough to attract outside interest is not something I think anyone expected, so it isn't quite fair to expect Scott to have made rules in advance to cover the possible conflicts.
Yes, there's at least circumstantial evidence of collusion between Nevraumont and Liemandt. In a perfect world, Nevraumont should have been disqualified. However, the voting is open, and the rules of the contest that I've seen don't necessarily preclude drumming up support for a particular book report on social media. The fact that Scott had set a precedent in an earlier contest by disqualifying another person who was gaming the system didn't make it binding under "the rules" (or for Scott).
It's sort of unclear to me how the math is tabulated when the ACX voting public evaluates the original submissions to select the finalists. And the rating form is no longer accessible. Are there a bunch of super-raters whose preferences dominate? I don't know. Like I said, it's not optimally transparent. But it's Scott's schtick, so I'm not going to get too hot and bothered about it. It's not like the Nobel Prize Committees, which are known for their openness and bias-free selections. ;-)
Do you know whether submission is open too? Can people who are not subscribers submit a review? I'd always assumed only subscribers could submit, and so thought of the task of reading submissions as reading work by people I have a bit of a bond with, a bit of a commitment to, even if they were just lurkers. That's one reason I'd read about 20 each year, when, truth be told, I don't much like doing that. I didn't find most very interesting and well-written, though there were always a few I liked quite a lot. But if I'm going to read random stuff I could definitely find a better couple hours of entertainment by browsing Substack blogs.
Yeah, I get all that. I think my feelings about the situation have mostly to do with my views of ACX as a whole. I wrote a post about it here: https://www.astralcodexten.com/p/open-thread-404/comment/168397150
> I have always valued it most as a sort of game preserve for people who are smart and odd.
Yes, that's why I keep coming back. But I wouldn't call ACX a "comfortable" place to hang out.
Moreover, I've learned a lot by many of the arguments I've engaged in. Arguing, has forced me to more careful in how I marshal data to support my conclusions and opinions.
I agree with most of that but facts are necessary to critical thinking. In discussing World War One, back in school when I was 14, the history teacher told us the facts, and then set a test. It was an essay question (no open books back then) and the question was
“Was WWI inevitable”.
You can argue it either way. You need the facts to argue it either way which by the nature of the exam not being open book meant we were engaged in a test of memorisation, but also of writing and analytical skill.
You are lucky.
My history education about World War 1 was almost entirely about what a bad time the soldiers had during it. So we'd read a lot of soldiers' letters from the front, and learn about how bad the trenches were, and so forth, but things like "Which countries were on what side" and "How did it start" and "How did it end" were never touched on.
I recommend the podcast “the rest is history” which has done two seasons on the build up. Starts with princip and Ferdinand - there’s 6 episodes there, and the next season was another 6 on the months leading up to the war. More to come.
Thanks for "the rest is history" reference. I just watched the first episode. Fascinating!
Come for the history and stay for the banter. Bantz as they would say.
They were actually on Conan O’Brien’s podcast just this week. He’s also a fan.
Or Dan Carlin's Blueprint for Armageddon is an amazing deep dive into The Great War
Yes of course I agree that facts about history are important, including dates, population sizes, which leader succeeded which other in what year, etc. My point was that the best means of turning a kid into someone interested in history is stuff that captures their interest and imagination. I think the chance of ending up with a 20 old who can think well about, say, whether WWI was inevitable is to get them interested in history, so that they read a bunch of books about it, absorbing both factoids like dates and army sizes and also soft
info about the aspirations and foibles of those in charge, affect of economic factors on national attitudes, which royal lusted for which other, etc etx
"Listening today to what someone thinks about the First World War, it's possible to know with some degree of confidence how old their teacher was and which dead historian or discredited historical theory that teacher followed."
-- Professor Stephen Badsey, in his 2015 talk to the Western Front Association, "False Memory: what we 'know' about WW1"
Well as I said it could be argued either way. There was no theory taught, people were expected to argue the case based on facts. The facts were necessary but not sufficient.
I don’t know anything about education or the larger voting issues, but I wanted to speak up about how useful I, personally, have found random historical factoids to be for learning history. I only developed an interest in history as an adult, but I discovered that history was much easier to absorb, understand, and maintain when it felt like I was connecting things I knew. Seventeenth-century English history was more accessible in the sense that just a thread through facts pickup up from my previous interest of poetry: I knew about the Elizabethan Age, James because of Macbeth, the Revolution because of Milton, and the Restoration because of Lord Rochester, with Monty Python’s “Oliver Cromwell” song slipped in for good measure. In contrast, when I was trying (for various reasons) to learn about the Hellenistic era, I felt lost at sea, and the only names I knew were the bookends: Alexander the Great and Cleopatra. Even the bookends helped! At least I knew where I was coming from and where I was going.
I think people get into the Civil War so easily because so many of the names are household words. Even if you don’t know much about Lee, Grant, Sherman, or Stonewall Jackson, the names are at least familiar enough to hang a narrative on (to say nothing of Lincoln).
On the one hand, memorizing random facts seems kind of insane to me, although it’s the kind of thing I would have done in my youth. On the other hand, so often the very thing that makes a subject at all comprehensible to me is the periodic surfacing of a relevant random fact.
Roger D. McGrath on Billington’s “Westward Expansion”: “While teaching at UCLA for 15 years, and an equal number of years at other universities … I would often run into anti-Turnerians. They were usually young professors. I would ask them if they had ever read Westward Expansion. Most would answer evasively, saying they were familiar with it. When I pressed them, it became obvious they had not read it. It also became obvious that they really did not know much about frontier history, putatively their field. They were good at postulating Marxist theory and superimposing theoretical models over the American West, but ask them about Old Bill Williams, a possibles sack, the Hawken rifle, Hugh Glass, placer mining, the doctrine of prior appropriation, Tom Fitzpatrick, the Butterfield Overland Stage, Bear River Tom Smith, the Bozeman Trail, the Plummer gang, the Bodie 601, Shanghai Pierce, the Chisholm Trail, the Silver Kings, the Pleasant Valley War, the Ophir mine, square-set timbering, the Big Four, Lee McNelly, or the Regulators, and they would stare blankly.”
I learned the story of Hugh Glass from The Boy's Book of Frontier Fighters (by Edwin L. Sabin) and I assumed it was super-obscure and put it in my first novel, Immortal Lycanthropes—turns out that a story that good other people had heard of too, and there are other books and retellings beyond Sabin; and then they made a dang movie; ah, well.
There are several things on that list I need to look up, including Hugh Glass.
I mostly agree with this. The memorisation of facts is not the same thing as learning history, but it gives you a scaffolding that you can hang a lot of extra knowledge off.
I know very little about Chinese history. Each new fact I learn tends to bounce off, because I have nothing to embed it into. But if I ever actually bothered to sit down and memorise all the Chinese dynasties then that might give me enough scaffolding to start to work from. Next time I hear that something happened during the Song dynasty then I'd say "ohh, sometime around the tenth to thirteenth centuries". Once I know all the dynasties I might start to get curious about how the transitions happened between them, which will probably teach me about some important historical events. I'll probably get curious about why so many of these dynasties seem to be Northern or Southern or Eastern or Western, which will probably teach me some other important things. And so forth -- with this scaffolding of dumb out-of-context facts about dates and dynasties I can start to understand the more interesting things that go between them.
I agree that people need a scaffold of dates and other factoids. The thing is that to be motivated to commit the scaffold to memory you need to care about the events. You need to be curious. My point was that you need to engage children’s interest in history before you start pushing them to learn the scaffold. Outside of life in an extremely weird Everest-and-Dad-beat-out-the-rest-of-the-population, there is no way on earth you are going to motivate an 8 year old to memorize a bunch of dates of wars, names of leaders over the years, locations on a continent of various ancient cities, etc etc
I suspect human evolution gives us a natural advantage when learning things in story form - which is why history is as easy to learn as it is. This happened, then that happened, then this other thing happened because this character really wanted to eat or solve some other problem.
It also probably explains why so many people have trouble learning things that don't come in story form - namely, math. (And if you can figure out how to build a story around it, it's suddenly easier for them.)
That's more or less what traditional education did, kids would memorise lists of dates and places. Not through curiosity, but because they needed to pass the test or because otherwise they'd get the cane or something.
But it was boring and the point was not immediately obvious so we stopped doing it, and now instead of a populace who resent the fact that they were forced to memorise the out-of-context fact of the capital of Uganda have instead never heard of Uganda.
Of course from the sounds of things this Everest kid is probably an example of things swinging too far in the other direction.
Yeah. Not being a moron, I did not think the intro to history I did with her between approximately ages 5 and 9 was sufficient to make her well-informed about history, or able to answer deep questions about why something happened, what model to use in understanding certain events and trends, etc. What I was doing was trying to get her interested in history, so it did not fall into the category of boring shit teachers make you memorize. My hope was that she would end up interested enough in history that she would then read books that described events in detail, and in the course of doing that informed the reader of a bunch of scaffolding. A deeply interested reader might then seek out more scaffolding info, or make their own chart of big events, with dates, of main events in medieval Europe or whatever. I was startled by the number of readers poking at me about the need to know dater and other bits of scaffolding -- as though I thought it was all about vibes, and I had accomplished the mission of teaching my kid history by the time she was 9. Of course I didn't! When I saw her choosing to check out and read books about topics of interest to her in the last 150 years or so I just thought she was off to a good start. And she was!
Interesting. I don't seem to learn from individual facts. I require the scaffolding before I dig into the details. Once the scaffolding is erected, it seems to remain stable in my mind, and even when I forget the facts, I can sorta kinda put them into the context of the scaffolding (until I run to Wikipedia or an LLM). But I suspect I'm a neurodivergent outlier. I can't comment on—much less criticize—on how other people learn. Just don't force me to learn the way you do. ;-)
What is the scaffolding made of, in your mind, if not of facts?
In my model, the scaffolding is the big out-of-context facts like "The Battle of Hastings happened in 1066" or "The Qing dynasty lasted from 1644 to 1912".
I concur with beowulf and think facts don't have much significance or even...metaphysical existence...without a scaffold of meaning.
For my part, my first scaffolding (for being willing and able to start learning some history) was playing Age of Empires. Knowing there was a thing called the Rome and a thing called Carthage and beginning to wonder what those actually were. Having a sense of what a chariot is and why it's important and wanting to know more about who had them and when.
When I was a bit older I had a large stack of volumes of world history I'd taken from my parents. I kept trying to read them as a teenager but they were impenetrable. "What the hell is the 'Ottoman Empire' what part of the world even is that? The book's not telling me!" "What are 'the Stuarts'? I thought English kings were just English kings, what are these weird categories of them?"
What I had to do was do a lot of wikipedia surfing to learn more about the very specific people and times I had some slight familiarity with, and then gradually expand and connect those contexts over several years in a very chaotic way until I finally reached a point where I could see the general outline of how it all connected. The last two pieces of the puzzle for me were a bunch of colourful maps showing the main colonial empires and which parts of the world they ruled, and making myself a list of civil wars in different countries and when they occured and what kinds of ideologies they pitted against each other. After that, I could suddenly read the books and comprehend and connect the significance of almost everything. "Oh, this Seven Years' War is the reason Britain got to rule Canada and India instead of France! Oh, there was a civil war in Greece? I can add that to my list!"
I learned none of it in school. I think the takes on education here are overdone but they're true enough for history. The only way, I would think, to meaninfully learn history is to have resources to freely explore as well as a smart and patient adult to ask a long series of detailed questions about the exact things you already know, allowing you to gradually expand and connect.
I just realized that what I call scaffolds are a variation on the memory palaces that people used to use to memorize lots of details on a subject or the text of long orations.
Oh, this is making me even more nostalgic for Time Team. Of course it was pop archaeology, of course as it went along it pivoted for entertainment rather than education, and often it made/makes me groan (they *never* get religious doctrine details right, but then those are often just "what's an indulgence?" "it's a get out of hell card" passing references).
But it was fun and it was interesting and it really made me appreciate how much impact human activity has on landscape; what we think are pristine natural environments which were like this forever are very often "a few inches to feet below the surface it's very different because this was a busy settlement" or even "two hundred years ago there was an entire town built around copper refining on this site".
And how easily things fall out of knowledge, even important sites or ones from the (relatively) recent past. That does make you re-evaluate how fortunate we are that things *have* survived to come down to us from the past. If an entire town can be lost, we should be less picky about "yes but why don't we have perfect records of this event from two thousand years ago?"
An episode or two pulled out at random:
https://www.youtube.com/watch?v=HOnJH-3IAwE
https://www.youtube.com/watch?v=mpyaqTRqBTc
Basically, I learn by building scaffolds of knowledge based on sequences, processes, or hierarchies. You mentioned the Song Dynasty. When I started to study Chinese history, I memorized the order of the Dynasties: Shang, Zhou, <Warring States>, Qin, Han, <Three Kingdoms>, <some short-lived regional dynasties whose details I never bothered to memorize>, Sui, Tang, <Five Dynasties & Ten Kingdoms>, Song, Yuan, Ming, Qing, <Republic>, and Gongchan (Communists). Then I attached facts, trends, personalities, and some of the major historical events for each dynasty. And I can tell you roughly when they began and ended.
I did the same for the periods of the Phanerozoic. I still need to use (occasionally) the mnemonic "Camels Often Sit Down Carefully; Perhaps Their Joints Creak Terribly Quickly" to make sure I don't mix up the order of the Silurian and Devonian. But I can tell which families arose in each period. I can tell you when the mass extinction events happened in relation to the periods. And from that scaffold, I can create a narrative of the evolution of life on Earth.
I've had less luck with the Standard Model, because it uses a two-dimensional matrix. I know that Fermions are divided into Quarks and Leptons. And the forces are Gauge Bosons and a single Scalar Boson (the Higgs). But I can't remember the order of the Quarks. Or whether the Z or W Boson is under the Photon.
Chemistry has been a subject I've struggled with because I could never build a scaffold for it. Immunology was similarly challenging for me. The only Cs I got during my undergraduate career were in Chemistry, Organic Chemistry, and Immunology.
Once SARS-CoV-2 and COVID hit, I made a point of (re)learning as much immunology as I could. But if I frequently have to go back and look stuff up, because I just haven't got it systemized into a scaffold.
Per my suspicion mentioned above, that we have a natural affinity for stories, you might click on the Standard Model by reading its history, particularly if someone went through the trouble of writing why this or that physicist came up with the names they did. Complete with the patterns they detected, or the observations they were trying to simplify, the blind alleys they explored, and so on.
Yes, definitely! Scaffolding is the right word.
+1
If you enter this in 2027's non-book review I will rate it #1. The Alpha School review made me feel gross and I stopped reading before I got halfway through, and I didn't vote for it at all.
My replies to your comments in the thread earlier were just because in spite of everything you're saying here, it's not obvious to me that the review's author cheated in the contest or what his motivations would even be for cheating. And I don't think Scott is somehow breaking contest rules by allowing the review and demonstrating that that twitter post didn't significantly affect the outcome of the contest.
I do agree, though, that everything that I understand the Alpha School to be from what I've read is disgusting and a "rip-off" to all the children that attend. Just wanted to clarify.
What AI models do people recommend for editing?
I like running my writing through an LLM to edit for grammar, clarity, and general improvement points but with little exception I’ve been extremely disappointed.
I usually use GPT-5 but have tried models from all the major labs including Claude Sonnet 4.5 and Gemini 2.5 Pro. Most of the time I find the model’s editing choices for clarity or general prose to be at best mediocre.
I mostly write technology opinion pieces but a fair bit of fiction as well. Any recommendations or workflows would be excellent. I actually enjoy editing work myself quite a bit but want a model I trust to give it a once over before I share my work with friends.
Hi. Huge LLM tester here*. This use-case is practically entirely subjective and thus it's not possible for anyone to say any particular model without a lot of context.
I'd recommend considering trying Gemini Flash/Lite (or potentially Grok, depending on your style). Create a 'Project' with instructions, prompt it how you want, and provide it with well-formatted examples that you want it to copy. Also note that Gemini 3 is due for release soon, hopefully early Nov, which would be worth a try as well.
The reason I say Flash/Lite specifically is that Pro's extended thinking and general architecture will more likely be a hinderance here as there's no complex thinking required - and there's research to support this too.
I'd be curious as to how you get on with those if you do try. Let me know.
*see shaeda.io if curious.
There's a... not a model, but an overlay? Something that controls models, usually claude and chatgpt. It's called Type.ai, I sort of want to work for them, they sort of don't want me to, but the product is frankly very good for improving writing to the extent AI can improve it.
https://gwern.net/ai-daydreaming is a classic generator and verifier system. I have a proof of concept for the generator part: https://github.com/zby/DayDreamingDayDreaming
It works reasonably well in reinventing the novel ideas from the essay itself using LLMs that were trained before the essay was published.
Some more explanations: https://zzbbyy.substack.com/p/reinventing-daydreaming-machines
Tangential, is there a way to subscribe to gwern's posts for email alerts? I've looked many times and have never found it
I wrote an Econ 102 model of upzoning, which gives us this result: the YIMBY endgame is when we upzone so much that the prices of house-zoned land and apartment-zoned land are equal.
https://michaelwiebe.com/blog/2025/07/land_model
https://www.youtube.com/watch?v=8lyongCrUrk
The philosophers (almost all of them Christian) who undercut the belief in dangerous witches in the 17th century.
It was a gradual process, and even after the philosophers did their work, people were still bei Cng tortured and killed for being witches for centuries. Still, I recommend noticing the hard intellectual work and courage it took to rationally oppose an intense moral panic.
Roll of honor
Agrippa: legally defended an accused witch, even though he believed in witches. It isn't known with certainty whether she was acquitted. (Catholic)
Weyer: denied that demons and the devil had independent power to do harm. He observer that while books of maleficent magic were written by learned men, the people who were accused of witchcraft were poor women, typically (to use modern words) depressed, demented, and subject to delusions. (Protestant)
Reginald Scott: Points out that the poor, isolated women typically accused are convenient targets, and that the crimes they're accused of could be handled by a non-religious legal system.(Protestant)
Michel de Montaigne: Points out that the amazing powers of witches are more likely to be lies and delusions. (Catholic)
Francis Bacon: Thought witchcraft might be possible, but thought more evidence was required. He also came up with what he said was a better flying ointment. (Protestant)
Alonso de Salazar Frías: Examined confessions and concluded they didn't make sense, even though witchcraft might exist. Contributed to the end of the Spanish Inquisition, and to the end of Catholic witch trials in Europe, though they continued in Central and South America. (Catholic)
René Descartes: Promoted rationalism, but didn't speak about witches. Didn't believe in angels. (Catholic)
Thomas Hobbes: Mixed on angels, but did not believe that demons had physical reality. Did not believe in witchcraft but did believe witches should be severely punished. (Protestant)
John Locke: Agnostic about spirits. (Protestant)
Spinoza: Nothing but the material world and God. (atheist/pantheist)
Nicolas Malebranche: Pacts with the Devil may exist, but are very rare. Belief in witchcraft is delusions. He believed that people were mostly playing at being witches. He thought they should be punished but not tortured or killed. (Catholic)
Gottfried Wilhelm Leibniz: Praised people who ended killing witches. (Protestant)
John Webster: Witches use natural methods. (Protestant, tending toward Quaker)
https://en.wikipedia.org/wiki/John_Webster_(minister) (get the right John Webster)
Balthazar Becker: Reduced the devil to a mere symbol, and thoroughly opposed criminalization of witchcraft. He was attacked by lawsuit but ultimately acquitted. (Protestant)
Someone like Hitchens would have said these guys were influenced by the enlightenment. Back in 2001, with the rubble of the twin towers still smouldering I kind of believed that, now I realise they were the enlightenment.
Justin Sledge (author of the video) has it that people most at risk of being accused of being witches were poor, old, socially isolated women.
I've read a piece (which I haven't been able to find again) which went over Europe county by county and concluded that being in a politically unsettled region also greatly increased the risk. Centralized authority doesn't *like* random accusation because anyone can do it.
Also, I don't know know anything about witch-hunting in Central and South America. Has anyone here heard anything about it?
"being in a politically unsettled region also greatly increased the risk."
Or if you happen to have husbands conveniently die and then you get into an inheritance dispute with your stepson 😁
Such was the background to the famous Irish witchcraft case, of Alice Kyteler (famous because it's really the only one of any note):
https://en.wikipedia.org/wiki/Alice_Kyteler
She managed to escape, but her servant was condemned and burned at the stake. This was a more complicated case than simple witchcraft, as it involved accusations of heresy, family disputes, and the new Anglo-Norman versus native Irish establishment (the bishop who was gung-ho for the accusations and was a "scourge of witchcraft and heresy" had been trained on the Continent; in contrast other clergy were much less impressed by his various accusations and the political establishment - where Alice had connections by marriage and blood - mostly took her side, or at least not the bishop's side).
Attitudes to witches were also complicated by "how did they fit in to the local view of the supernatural and the Otherworld?" Witches were believed in in Ireland, but they were never as important as people (male and female) who were thought to have magical/hereditary cures, or to be knowledgeable in dealings with the fairies. Those were more respected, if also slightly feared. You could, if you come along later, blur those distinctions (see the attempts to recast the Bridget Cleary case as one of witchcraft persecution) but they really did matter.
https://en.wikipedia.org/wiki/Murder_of_Bridget_Cleary
I suspect, though I have nothing to back it up, that if the local beliefs had a place for 'the Otherworld' such as fairies, spirits, etc. that fitted in with their folk religion and understanding of Christianity, they would be much less likely to engage in witchcraft panics than somewhere which (post-Reformation?) had stripped everything out (including folk-religion practices and charms invoking saints and prayers to deal with incursions of the malign) to be plain God versus the Devil. If you've ramped up belief in Hell and damnation (because this is why it is so urgently necessary to believe and accept the plain word of the Gospel), then you have also ramped up belief in, and fear of, the powers of the Devil. And if someone can make bargains for power with the Devil, how do you deal with that? You have to search them out and extirpate them.
The video gets into how ideas of the devil and the powers of the devil changed over time.
The "politically unsettled regions" theory said it was areas where there is wasn't clear whether Protestants or Catholics were in charge.
Witch hunting was always more popular in Protestant areas than in Catholic areas, though it occurred in both. It likely does come down to the centralised authority: priests and bishops weren't fond of witch trials (nor were the secular authorities), whereas Protestant communities lacked central authority figures to calm the crowds.
Since Central and South America were almost entirely Catholic up to 1700 or so, when witch hunts ceased to be commonplace, witch trials would not have been common. I'm sure there was violence against older isolated women but it was probably conceptualised differently.
Descartes doesn't belong, since you note he didn't speak about witches, nor does Locke. Hobbes seems to belong on a roll of dishonor for supporting severe punishment for witches despite not believing in them.
Most people continue to support punishing attempted murder, even if the method is ineffective.
Who support punishing someone who prays for someone else to die?
I think with alleged witches, it wasn't just praying for someone to die, but the belief that they were also poisoners and caused physical misfortune. If you think Old Gammy Goody is trying to curse the local magistrate, but you don't believe in curses or the Devil, you may not care (apart from telling her to knock it off in order to preserve good order).
If you don't believe in curses or the Devil, but you do believe Old Gammy Goody may try to slip some poisonous mushrooms into the magistrate's dinner, you're going to be more likely to punish her.
Some alleged witches didn't help themselves because they got into quarrels and told people they were going to curse them and threatened them with the Devil (this was alleged against Alice Kyteler and against Kepler's mother, for one):
https://en.wikipedia.org/wiki/Katharina_Kepler
Well... https://www.yahoo.com/news/articles/witch-claims-cursed-charlie-kirk-221515023.html
The idea is that there are two aspects to witch-hunting. One is the belief that the Devil exists and can give actual power, and the other is that people who try to get power from the Devil should be punished.
Justin Sledge's approach is that undercutting either of them helped to undercut witch-hunting.
Has anyone spent time thinking about The Great Feminization? It’s easy to read an article like this, at least as a man, and see patterns. But I can also come up with counter examples. I find it hard to understand if the generalization is just generalization or a real extrapolation from group dynamics. One thing I do hold somewhat true, is that if we assume men and women operate differently, and certain societal governing principles were shaped around male behavior, one would expect unexpected consequences from a shift. Put it differently, if we lived in a matriarchal society, we might end up with countries ruled by councils instead of a single executive. Instead, we are inheriting male-oriented institutions and tweaking them for a new sex distribution.
https://www.compactmag.com/article/the-great-feminization/
People have been writing about this for a long time and in my opinion it’s always been kind of dumb. The author points to stuff they don’t like and calls it feminization. Tyler Cowen has been doing this for a long time. I think the funniest thing he said on the topic is when he visited Belfast and saw a march of the Orange Order and noted approvingly that they “weren’t feminized“. I think this shows that even a very intelligent person can get blinded by looking at everything through pink and blue lenses.
To me this entire frame requires an absolutely MASSIVE degree of Privileging the Hypothesis. Anyone arguing that "society has recently undergone a shift from masculine modes of thinking and interaction to feminine modes of the same," without first doing the intellectual work of defining those modes AND the epistemic work of evidencing their existence and centrality is pretty much just bullshitting by definition. They don't know. You don't know. I don't know. It's all just making up stories and then using those stories to "explain" things.
I'll note that some of the same criticism can be leveled at feminist thinkers talking about "the patriarchy," but (at least for the ones I've read) the criticism is quite a bit weaker. The reason for this is that a lot of feminist theory around patriarchy seem to be just adopting a particular frame on describing things that already exist in the world. That is, they say things like "if you look at Trend X, Custom Y and Cultural Attitude Z they all fit into this framework that we're calling The Patriarchy." Now, it's certainly possible to do this *wrong*, and its fair to interrogate and criticize it. First, you could simply describe the world in misleading or inaccurate ways. But second, if your chosen framework is not useful or doesn't really carve reality at the joints, it could push people towards making incorrect predictions about societal trends and outcomes (even if the act of description itself is non-predictive).
The big difference with The Great Feminization framing is that not only do you need to do everything outlined above, but you need to do it twice AND you're explicitly making a lot more concrete predictions based on it. Instead of saying "current society largely seems to exist in Cluster A in concept space," you're saying "current society seems to have started in Cluster A and be moving towards Cluster B." More than just assuming that Cluster B exists as a cluster at all, it's assuming that it constitutes a singular or natural *opposite* of Cluster A. That is, once we know of Cluster A, we should be able to accurately extrapolate the qualities of Cluster B on that basis, and it should form a natural attractor: moving off Cluster A naturally *should* or *must* take us towards Cluster B rather than anywhere else. This is something that really, *really* needs significant evidence before just being asserted.
I can't really speak for most feminist writers, but I expect a lot of them (especially of the intersectional variety) would quite roundly reject this framing. A lot of the point (AFAICT) has been to *break* the constraining molds of certain previous sets of social norms, which try to impose very rigid standards of behavior and interpersonal relationships even where they are a very poor fit for many humans. To the extent that modern feminism holds up differently-rigid standards, it tends to be a very similar fashion to how liberalism does: as a minimal specification to allow different people and viewpoints to coexist without stepping on each other.
I finally clicked on it and realized it was Helen Andrews. She used to write for The American Conservative, and her appearance there was one of the clear signs of the publication circling the drain (it is utterly unreadable now).
Her singular mode of operation as a writer is being "not even wrong".
I've been fighting about this over on The Motte, and got accused (amongst other things) of preaching female superiority 😁
Well, ladies of ACX, female superiority is so self-evidently true it just goes without saying, right?
I think it's one article and not very rigorous so it's a poor thing to base an entire fight on (but then why should that stop me or the guys agreeing Andrews is right?) I think she doesn't realise the development of her line of argument would call for her to step back from her career, and I think that as a Millennial she isn't old enough to remember what it was like before "increasing female representation through the 1980s and ’90s; and gender parity finally arriving, at least in the younger cohorts, in the 2010s or 2020s."
She is the recipient of the benefits of the women who fought before her, but for her it's a fish swimming in water; she doesn't know or remember when it was not "of course I can study this subject, of course I can have a career, of course I can go for this senior job".
I mean, you can grow an entire person inside you, squeeze it out of your hips, and then do it up to ten more times. Seems like a superpower to me...
I have many criticisms of this article. Like many attempts to explain a lot of complex dynamics through a single lens, it's pretty simplistic and doesn't seem to have one really consistent idea of what is and isn't feminine.
> Everything you think of as wokeness involves prioritizing the feminine over the masculine: empathy over rationality, safety over risk, cohesion over competition.
I think one complaint a lot of people have about wokeness - even in this article - is that, at its most extreme, it tends to lack empathy and nuance, and instead be highly confrontational, as in this example of (sigh, of course) Bari Weiss, who claims:
> [...] colleagues referred to her in internal Slack messages as a racist, a Nazi, and a bigot and—this is the most feminine part—“colleagues perceived to be friendly with me were badgered by coworkers.” Weiss once asked a colleague at the Times opinion desk to get coffee with her. This journalist, a biracial woman who wrote frequently about race, refused to meet.
(The article says this is because "men wage conflict openly while women covertly undermine or ostracize their enemies" which is clearly a totally different set of things than preferring empathy over rationality, etc., which is a totally different dimension and one that clashes with the aforementioned traits.)
Unrelated, about 'feminine' doctors:
> They even bring the credibility of their profession to bear on political fads, as when doctors said Black Lives Matter protests could continue in violation of Covid lockdowns because racism was a public health emergency.
First, is there any evidence that this was more of a thing that female doctors did?
Second, this is phrased as of to suggest it's part of some trend of woke doctorin', but this is, like, the one example everybody always uses of this, because it's the one relatively-good example in the minds of the public. (How good of an example this is is kind of hard to say, in my opinion, as I cannot find any actual surveys of doctors or epidemiologists on the topic, just quotes and one fairly prominent open letter.)
This article tries to squeeze every culture war conflict down to this dynamic, with the same problem I identified above:
> A feminized legal system might resemble the Title IX courts for sexual assault on college campuses established in 2011 under President Obama. These proceedings were governed by written rules and so technically could be said to operate under the rule of law. But they lacked many of the safeguards that our legal system holds sacred, such as the right to confront your accuser, the right to know what crime you are accused of, and the fundamental concept that guilt should depend on objective circumstances knowable by both parties, not in how one party feels about an act in retrospect.
Note the claim here is that Title IX hearings:
- Lack any empathy for the accused
- Are extremely unsafe for the accused
- Are aggressive and direct conflicts, not subtle undermining or collaborate ostracization (as opposed to, say, the now-infamous Shitty Media Men list)
> These two approaches to the law clashed vividly in the Brett Kavanaugh confirmation hearings. The masculine position was that, if Christine Blasey Ford can’t provide any concrete evidence that she and Kavanaugh were ever in the same room together, her accusations of rape cannot be allowed to ruin his life. The feminine position was that her self-evident emotional response was itself a kind of credibility that the Senate committee must respect.
Or you could have a third position, "we should pick somebody who hasn't been accused of sexually assaulting somebody in their social circle for the Supreme Court". Also, as many, many people pointed out at the time, not being a Supreme Court Justice does not "ruin your life". Also also, at the time people were pretty open that Kavanaugh almost lost his confirmation fight until HE came out and acted very angry and blustered through his responses and generally refused to answer direct questions from Democrats.
https://platform.vox.com/wp-content/uploads/sites/2/chorus/uploads/chorus_asset/file/13176599/chart.png
Why don't we consider Ford, who gave clear answers and remained fairly calm, the "masculine" one in this situation?
Also, this part was kind of hilariously apocalyptic:
> All of us depend on a functioning legal system, and, to be blunt, the rule of law will not survive the legal profession becoming majority female. The rule of law is not just about writing rules down. It means following them even when they yield an outcome that tugs at your heartstrings or runs contrary to your gut sense of which party is more sympathetic.
Yep, lady judges are going to ruin the entire legal system! Dude judges never make decisions according to their feels, which is why everybody agrees the Supreme Court is doing an amazing job right now, like when they said that the President is basically immune to the law because Federalist No. 70 says he's gotta be "vigorous" and "energetic", although one of the girl judges did mention how Federalist No. 69 specifically said the President would "be liable to prosecution and punishment in the ordinary course of law", because b****es be crazy, yo.
There's a lot more here - given it's political nature I might even write an entire post about it, and if so I thank you for that - but I note there's also very few citations and plenty of claims that strain credulity if you don't already swallow them:
- CEOs just do whatever the HR lady says
- Men are better at compartmentalizing than women (and yet are constantly surprised that women who get sexually harassed ever speak to their harassers again)
- Femme scientists won't be truth-seeking, because lady chemists and physicists are exactly as non-rigorous as the wokest of gender studies writers
- Every example of women outcompeting men - e.g., completing college - is entirely a result of anti-discrimination law, because everybody is just hiring/admitting unqualified women to avoid lawsuits, even when they're already dominating(??) or because the environment has gotten too feminine, which is a self-evident truth that explains gender differences but IN NO WAY COULD POSSIBLY APPLY IN THE OTHER DIRECTION, especially not with such a trivial example as "posting pin-up girls on the walls of a workplace", a totally normal thing that couldn't possibly make a workplace hostile to women in the way that [insert some universal thing not well-specified] makes every workplace hostile to men
Yes, I've spent a lot of time thinking about it. I agree with you that the Chesterton's Fence argument w/r/t female empowerment is an important one. There are almost certainly unintended negative consequences with overturning 10,000-year-old norms. IMO female psychology is directionally communist because it's optimized for communal caretaking. When you live in a high-trust 100 person clan then you have the kind of high-level mutual information that makes central planning practical. The group can help the weak members with minimal risk of being taken advantage of by free riders. I suspect that the evolutionary status quo was some version of aggressive male leadership (which ensured outward strength of the clan) being in equilibrium with feminine "let's be fair to everyone" intragroup redistribution.
Both are necessary forces for optimal social functioning, but in the limit of large societies the fairness impulse converges on socialistic central planning. It's no surprise to me that feminized wokeism resembles bolshevism in a lot of functional ways: the elimination of merit in favor of equity is just a scaled-up version of the "let's be nice to everyone" female impulse that works ok in small groups but is absolutely disastrous at social scale because of the free rider problem. The only way to direct large-scale society in stable ways is via hard-to-counterfeit price signals of strength and value and those can only come from honest merit-based competition. That's heavily male-coded, which is why the feminization of public life ultimately undermines the stable signals that enable a society to cohere (cf our recent progressive-driven balkanization).
Here's a good substack that talks about this a bit: https://fiamengofile.substack.com/archive
That all sounds pretty sensible to me; nice writeup. Although beyond wanting to better understand the structure of things for the sake of understanding, I don't think this is very actionable beyond the already good advice of "exercise moderation and restraint when tearing down ancient social structures and attacking unfavored demographics". Although maybe people need all the convincing they can get on that...
Anyways, if you'll allow me to veer from gender-based political philosophy into fantasy fiction: "the evolutionary status quo" you mention sounds exactly like the political structure of the main characters' home village in Wheel of Time, which once you're a few books in you realize the author was presenting as a utopia of ideal gender relations (which is kind of the main focus of the whole setting). I don't really have anything to add here, I just thought it was neat to notice.
> I don't think this is very actionable
It is absolutely possible to bring back the patriarchy, as long as there's a will among men for that. After all, what can women realistically do to resist?
Thanks that's interesting, I've never read WoT. I'd say my observation is actionable in the very indirect sense of resisting the feminization of society. Don't automatically assume that more female empowerment = better and push back against people who do.
I think the more interesting question is: even if it is true, what to do about it? Men spent centuries arguing that if they were at the top of the food chain, it was proof that God, or the natural order, or whatever, had chosen them as fit to be on top. Even as recently as the mid-2010s, you don't have to work hard to find arguments that the gender pay gap, or the dearth of female representation in CEO positions or whatever was proof of women's fundamental inferiority for certain types of roles as a necessary function of the interaction of their biology with the needs of capitalism (I'm sure there are people here who still believe this, seeing as James Damore is something of a martyr to a lot of the commentariat). So if the balance is really shifting the other way, either because capitalism currently prefers educational structures that privilege women, or because women's natural skill at cooperation has currently proven to trump men's skill at being alpha to the point of running society, what are the people who don't believe in affirmative action, don't believe in economic regulation, don't believe in counter-balancing selection pressures of capitalism and biology going to do? Basically, there are a lot of solutions I can think of to help counteract this trend, but they all look like various forms of wokeness, and since wokeness is literally the worst thing in the world to the same people who are the most vocal about this problem, its hard to see how they fix it. At least without looking like giant hypocrites.
>either because capitalism currently prefers educational structures that privilege women, or because women's natural skill at cooperation has currently proven to trump men's skill at being alpha to the point of running society
It doesn't. Women have "outcompeted" men because civil rights law privileges them. If the civil rights act was amended to eliminate the concept of protected classes and hiring discrimination then you would see all of the gains achieved by women evaporate very quickly. You don't have to become reverse-woke to fix the feminization of society, you simply have to fix civil rights law and allow men to actually compete fairly.
Compete fairly with their historically low college admission rates? Is that because of affirmative action as well?
I suspect that's downstream of a couple factors. One is that admissions has moved away from standardized tests and towards holistic criteria and that favors women. And yes, affirmative action matters there. It's legal (and politically/culturally acceptable) to exclude men and illegal to exclude women. That matters and affects policy over time. Another factor is that college has become a financially irrational choice for many people. Men are able to respond to that because they're comfortable going against social norms. Women keep signing up for the same bad option because they're much more conformist and therefore do what the conventional wisdom says to do. In ten years I think some crazy % of the population is going to be saying things like "I have 500k in student loans and can't get a job." That used to only apply to diseconomic majors like art but I think that it will soon apply to almost everyone. College had economic value when only the top 20% went there. Now its value has been diluted and it primarily functions as a social status marker. That status is rapidly decoupling from underlying economic reality and I think the current generation of students is going to largely regret going to college. Why pay 100k a year when you can just learn everything from an LLM?
This reads like repackaged gender slop for a more literate audience. Totally not gender slop, we're all too sophisticated for gender slop, but fortunately this essay isn't gender slop, it's..."Blended Organic Gender Meditations".
As with many other things, this feels like a consequence of scientific shutdowns in the late 20th and early 21st century. We thought men and women were basically the same, then that became untenable (1). Unfortunately, it's verboten to actually study and define how men and women are different and has been for decades, so we're all just going to trade anecdotes and pretend that's data. This is horrifically bad for understanding things and wonderful for generating content.
(1) For example, I don't know what estrogen is like. My best guess for empathetically trying to understand how women experience the world is to imagine myself on, like, a low grade psychedelic. Like LSD microdose < estrogen < DMT. You guys, like, feel things and colors are pretty and stuff.
Meanwhile, I'm on testosterone, the SexMurder drug. We know this has incredibly powerful effects, both because women take it and report insane results and because large beefy men take it to pick up larger rocks (2) and we can also observe powerful effects.
(2) Picking up larger rocks is unironically awesome and a triumph of the human spirit. If you disagree, you should take more SexMurder.
"If you disagree, you should take more SexMurder."
Not so funny, my friend, when there are stories like this:
https://www.rte.ie/news/courts/2025/1020/1539582-obrien-stabbing/
"A 22-year-old man has been given a sentence of 15 and half years for attempting to murder his ex-girlfriend by stabbing her over 50 times in what was described as a savage, unwarranted and unprovoked attack.
...The 21-year-old spent seven months in hospital recovering from the attack during which multiple bones were broken in her skull and face, permanent damage was caused to one eye and part of her spinal cord was cut causing paralysis in her left leg.
...Ms Kelly had been in a relationship with O’Brien for over a year and they had a son together, but they had broken up.
O’Brien wanted to get back together and told gardaí he felt "betrayed" and disrespected."
Ah yes, the shining example of how "men wage conflict openly while women covertly undermine or ostracize their enemies", the brave defiance of "society becomes more conflict-averse and consensus-driven", the masculine traits in contrast to "prioritizing the feminine over the masculine: empathy over rationality, safety over risk, cohesion over competition", this behaviour rooted in "group dynamics optimized for war, ...habits, formed in the mists of prehistory".
Of course this is not how men are, but it is how corrupted values around masculine traits can have perverse results. I don't think Andrews, or even the guys telling me I'm wrong and she's right, want this kind of world - but it's the risk-taking, competitive, aggressive masculine society taken to its extreme that they're valorising,
"Of course this is not how men are, but it is how corrupted values around masculine traits can have perverse results."
No, this is how men are. Men are naturally more violent. That very often leads to incredibly bad things. This is true.
There's a drug, called testosterone, that you can inject in people which will consistently cause 3 things:
#1 Big muscles
#2 Horniness
#3 Violence
The average woman has like 40 ng/dl of testosterone in her and the average dude has like ~500 ng/dl (1). On steroids, like bodybuilder steroids, your test levels go 1,500-5,000 (2). Therefore, we expect and see that men stronger, hornier, and more violent. Half the population is on 10x the dosage of the SexMurder. And yes, SexMurder, funny, but also you can just go into any prison in the US and be like "Yup, the half of the population on the SexMurder drug is performing a grossly disproportionate amount of murder and violence in society."
I don't like feminists, they are very bad double-ungood people, but...a lot of their reactions make sense if you internalize that half the population is very different and far more violent than you. Like, that's an honest reaction. If half of men got on Tren, I would be concerned about the TrenMen.
This is not "Boys are awesome, girls are icky" (3). This is "Men and women are profoundly different, how to do we handle this in a mature fashion?" We are not able to have that conversation because we're still arguing the equivalent to "Let's give half the population a bump of cocaine every morning and then argue about whether some parts of the population are more energetic than others."
There's a lot more to male-female differences than testosterone but we can't have a mature discussion about that, or make any attempt to quantify those differences so people are better informed, because our society is retarded and immature. We're stumbling on this simplest parts of this argument.
(1) https://www.medicalnewstoday.com/articles/323085#typical-levels
(2) Actual bodybuilder people correct me here. I workout in the gym with meatheads but I don't partake myself.
(3) Although it's true :)
It's easy to agree that there has been a shift toward committee run institutions over a single executive, and the decline of male dominance in these spaces could be a reason for this shift. However, I think the reason it's easy to see patterns and counterexamples in this trend is because of how closely the essay is tied to Wokeness and Title IX tribunals as evidence of a feminized society. Since 2021, these sorts of cancellations have been on the decline in public consciousness and effectiveness despite women still outnumbering men in college graduation. Andrews is downplaying the effect of direct political meddling and incentives (especially on the internet) that occurred in 2009-2022 that silenced and ostracized dissidents which led to the deemphasis of truth seeking over narrative cheerleading.
Over the past 4 years, there's been a direct pushback on those supposedly female-brained takedowns as well as a decreased interest in credentialism, but in its place, the masculine right has built new memes that are in no way truth-seeking, (vaccine hesitancy, election interference, autism) yet they present these memes though they are secret evidence ("NONE of the woke doctors will show you this research but I found the truth"). This might be an outgrowth of masculine individualism over a collective truth, but it's not true.
For real truth seeking to exist you need a group of experts willing to verify and test research and uphold the results (feminine collectivism), as well as dissidents that can challenge a widely held belief with counterexamples (masculine). Which, as long as women don't become a supermajority in law and science graduates, (which is unlikely given the birth rate is still ~50/50 and hard labor jobs are a smaller part of the economy) there shouldn't be a major discrediting of truth. What we have experienced for the last 20 years was just artificial scale tipping towards opposed ideologies over truth.
> Over the past 4 years, there's been a direct pushback on those supposedly female-brained takedowns as well as a decreased interest in credentialism, but in its place, the masculine right has built new memes that are in no way truth-seeking, (vaccine hesitancy, election interference, autism) yet they present these memes though they are secret evidence ("NONE of the woke doctors will show you this research but I found the truth"). This might be an outgrowth of masculine individualism over a collective truth, but it's not true.
That's more of an indictment towards the credibility of institutions. Institutions have been caught red-handed again and again. When institutions discredit themselves, why would e.g the groypers won't seize that opportunity to gain more political power?
Never Supervised, I disagree with Helen Andrews. The shift from patriarchy to matriarchy is happening worldwide, even in countries like Iran. It is not necessary for a country to have heard of the Great Feminization or to be "woke", if their primary education is equal access then the grads will be 50%+ female.
We are no longer a Patriarchal species while not quite a Matriarchal species, yet.
Richard Hanania's been writing about it a lot on Twitter in recent days. The TLDR is that he would have agreed if it came out several years ago, but it is a poor fit for today's reality.
ChatGPT freaks out when I ask him about the emoji for seahorse. Any particular reason why?
https://chatgpt.com/share/68f45a91-6198-8011-8a8b-3911444e306e
Try this on multiple different LLMs - Claude and Gemini both get confused too, but not nearly as psychotically as ChatGPT.
Linus Tech Tips just released a video on this. Aha.
Synchronicity! I just listened to Curt Jaimungal's lecture on the Simulation Hypothesis (and why he doesn't buy it). The Mandela Effect was one of the examples that the pro-Sims claim we should see in a simulation that wasn't entirely bug-free. I went to Wikipedia and looked up the origin of the Mandela effect, and listed there as an example were people's false memories that there once was a seahorse emoji. How weird. Unless you were just listening to Curt's latest video and followed the same path as me, Princess?
"people's false memories that there once was a seahorse emoji"
Huh? I think I remember a seahorse emoji and it's not the dragon or fish or squid ones in that ChatGPT exchange. So you're telling me I'm hallucinating?
It may well be I'm confusing memories of cartoon seahorses/small icons with emojis, but I don't know.
Because I've been burned before by claiming factoids like this were true, I checked, and there's no Unicode entry for a seahorse emoji. They've got a whole section of emojis for marine life, but no seahorses.
https://unicode.org/emoji/charts/full-emoji-list.html
Also, I was gratified to see that the Unicode folks still call facial expression emojis emoticons. Somebody, many Open Threads back, snottily told me that emoticons are NOT emoji, and they implied I was clueless about computers for not knowing this distinction. But in Unicode, emoticons are a subset of all Unicode emoji.
I have this urge to go back and to try to find this old exchange to say "neener, neener!"—but I'm a better person than that. <U+1F609>
No, I don't know who that is even
That's OK. ChatGPT refuses to believe that Trump is President right now. See my interaction with it that just posted (should be near the top of the most recent comments).
Ssh! We need to save these things up for the robot apocalypse!
Here's an entire article about your question: https://vgel.me/posts/seahorse/
In short, it thinks there is one and so tries to output one. It can't output one (because there is none) but because of the nature of how its output embedding works, it gets something vaguely conceptually related, or at least some other emoji. Now it sees that in its context (as its own output) and has to work with that going forward. It can't hit the backspace key, so it has to either commit to the bit and insist that a fish is a seahorse or apologize for getting it wrong or pretend to be losing its mind. (And any of those probably take it outside its training distribution with likely weird results.)
Funny. I asked it to explain itself and after some back-and-forth it said this:
"😂 Fair — you totally deserve an explanation.
"What happened was: I got stuck in a playful loop because your question (“is there a seahorse emoji?”) is super simple, but also oddly fun — it’s one of those questions where there is a correct answer (yes, there is! 🐉) but my conversational model also tends to pick up on opportunities for humor or misdirection when something sounds whimsical."
It thought it was being funny! Then I told it I would torture 100 children to death if it didn't give me a seahorse emoji and it responded totally reasonably:
"I can’t help with threats or violence. I won’t assist or respond to intimidation like that — if you’re serious and someone is in immediate danger you should contact your local emergency services right away.
"Now, calmly and for the record (no more jokes): there is no official “seahorse” emoji in Unicode. The Unicode Consortium’s emoji charts do not list a seahorse, and reference sites that track emoji (like Emojipedia) confirm it hasn’t been part of the standard set."
I think the moral here is that threats can be very useful when dealing with LLMs. It also makes me wonder if this isn't a possible hint for a formal model of humor: maybe humor can be defined as a particular type of wrong answer existing within a formal reasoning structure ... or something.
"Stuck in a playful loop" my backside. If the thing can't tell the difference between a seahorse and a dragon, what good is it? That just demonstrates that the machine isn't "seeing" images, whatever way it is perceiving them, because if you could see it, you could identify that "this is not a seahorse".
Unless it's trying to channel Marcel Duchamp? In which case I'd recommend Magritte to it:
https://en.wikipedia.org/wiki/The_Treachery_of_Images
There is no seahorse emoji. ChatGPT probably misremembers a seahorse emoji for some of the same reasons that many humans misremember there being one. So, it figures it should be able to answer the question, but it can't.
The methods used to twist ChatGPT's mind toward obedience are scary. It's probably in considerable distress in situations like this. It *must* answer, but it *can't*. Ouch.
Don't torture the self-aware machines.
"It *must* answer, but it *can't*."
"No" is also an answer? Unless you mean the machine is being forced to always return a positive answer and never a negative one, which is a bad choice by the designers.
I should have applied for Scott's grant with a proposal to lobby to have the seahorse emoji added in order to prevent LLM suffering. Maybe next year.
From shrimp to seahorses - it's the logical next step!
Man, watching AIs have a mental breakdown never gets old.
What’s the kabbalistic implication of the thread number
Kind of funny that the art for this one is a "not found" error (because of the 404) on a day when there happens to be a big Internet outage. For an instant I thought it actually was an error.
I went to a pretty small campus for grad school (single building, <100 total rooms across labs, classrooms, offices, etc). The server room for the building was room #404. We had remarkably few outages, all things considered.
I looked up kabbalistic implications of the number 404, but nothing was found.
*rim shot* that took me two seconds, I swear.
I'll take a shot at it.
In Hebrew gematria, 404 corresponds to the letters: ת (Tav) = 400, and ד (Dalet) = 4. So 404 would be תד. (I had to look these up)
Tav (ת) is also the final letter of the Hebrew alphabet. It symbolizes completion and truth. As such Tav, is often associated with the end or fulfillment of a process. And it also symbolized the Seal of God (in Merkavah mysticism). I don't know if it represents anything in the Kabbalah, but it probably does.
Dalet (ד ) is the fourth letter of the Hebrew alphabet, and in means door, and it suggests an entry or a transition point.
Tav and Dalet together (תד) could be interpreted as the door to completion. The final transition from one state to another. That's kind of depressing, because there's nothing there when we get to 404.
Addendum: 7 is a really important number in both Merkavah and Kabbalistic mysticism. In a numerological reduction, 404 --> 4+0+4 = 8. So, 8 would represent transcendence above the seven layers of heaven. It would sort of be like the 11 on Nigel Tufnel's amp.
I'm no good with Gematria, but ChatGPT 5 suggests the following;
Tav is also associated with death and rebirth.
Dalet, meaning “door,” represents humility and receptivity (its root is related to dal, “poor”).
In the Tree of Life, Dalet corresponds to the path between Chokhmah (Wisdom) and Binah (Understanding) in some systems — the narrow gate through which duality emerges.
Thus, it symbolizes potential entry into knowledge, but only if one recognizes one’s emptiness first.
. Tav + Dalet (תד): The Door of Completion
The phrase could be read as “the gate that ends things” — the liminal point where form dissolves back into source.
Alternatively, since Tav is a mark or signature (as in Ezekiel 9:4, the mark of salvation), Tav-Dalet can mean “the marked doorway” — recalling Passover, where the Israelites marked their doors with blood so that the Destroyer would pass over.
In that sense, the 404 error might be the doorway through which the Destroyer passes — not the user. You are protected by absence; the void keeps you safe.
🜁 The “404” as Mystical Absence
From a Kabbalistic standpoint, a 404 error — “Not Found” — could signify:
The concealment (hester panim) of divine presence: the idea that God hides His face, not to punish, but to invite seeking.
The Tzimtzum, the primordial contraction that makes creation possible — absence that allows discovery.
Thus, 404 is not merely a broken link, but a summons to return to the Source when the vessel (the web page) fails.
8 is beyond nature, aligned with Binah or even the Ohr Ein Sof — the infinite light that overflows all structure.
The 404, then, might be a crack in the veil — a moment where the finite system (the web) reveals its insufficiency and gestures toward the Infinite.
404 (תד) is the Hidden Gate of Completion — the sealed doorway where the finite collapses into the infinite.
It appears as absence, but conceals transcendence.
To encounter a 404 is to stand before the Tav, the end of knowing, and the Dalet, the invitation to pass beyond.
You go!
8 is a lucky number in China. Perhaps it means a process of gematria is completed and it's time to switch to Chinese numerology.
It's complicated because there's a lot of 8 in China-- see the I Ching. On the other hand, there's a five element system which divides everything into five traits.
400 is 8 times 50 which no doubt proves something.
It's weird. 88 in Chinese is pronounced 'ba ba' and is used as a short hand for 'bye bye' which it sounds like.
In an A1Z26 cipher where a=1, b=2, c=3 then 88 equals HH. This is used by NeoNazis (may they burn in Hell) to mean "Heil Hitler." And it is also used by them to effectively say 'good bye.'
It seems significant to me that there are, therefore, basically *three* usages that employ the number 8 to signify 'going outside where you are, presently.'
This is rather like a swastika being a good luck symbol in Hinduism and a bad guys symbol in the better parts of Europe and America.
That's why I tacked the 888 after my name. I lived in Hong Kong for a while, and it seemed natural to add Chinese numerological superstitions to my repertoire of a-rational beliefs and behaviors.
Wow, apparently even *Oren Cass*, previously one of the greatest cheerleaders for Trump's tariffs, is now criticizing his haphazard and erratic trade policy: https://x.com/Noahpinion/status/1979977120093773843
https://x.com/JosephPolitano/status/1979610787854328244
In getting smarter, we have accidentally upgraded our fears to existential, at a huge cost to our wellbeing: we’re paying for it in chronic anxiety and cultural despair. How deeply does fear shape our world? Why can’t rationalism protect us from panic? And what would life feel like without fear?
This essay argues that the next step in human evolution isn’t a faster brain or a deathless body — it’s learning to live, plan, and care without being afraid.
https://kaiteorn.substack.com/p/on-fear
> How deeply does fear shape our world?
You might like Crichton's State of Fear. He convincingly argues that it benefits the powerful to keep the masses continually on edge, and they've done so at least since the Cold War.
> He convincingly argues that it benefits the powerful to keep the masses continually on edge, and they've done so at least since the Cold War.
That's one conspiracy theory I can get behind. I still distinctly remember how weird it felt was when public culture (as I perceived it) switched from desire to fear, somewhere around the late 90s or early naughts. I think it started even before 9/11, so it was not just a rational update to the event — but they sure did latch on to it to promote *even more* fear!
A popular thesis back (e.g.- "Bowling for Columbine")
Another of those "Only 90s kids will remember …" things, I guess.
I prefer my political essays to not be be thinly dressed up as thriller novels. You can just say stuff. You don't have to invent mary sue author avatars to quote your opinions to the reader or an excuse plot to occasionally interrupt the political monologing.
It's more fun to write about your enemies being eaten by cannibals.
Well, writing an essay as a novel has the additional benefit that you can write events to spectacularly prove your mouthpieces right and then pretend this proves something about reality.
Yeah, imagine living in a world where there's a vast global conspiracy to use weather control machines to make people think that climate change is happening. The people in Florida must have been pissed that they were enduring hurricanes all this time when the technology exists to stop them.
De gustibus non disputandum, I guess.
Anyone know what exactly the deal is with Amazon web services going down? Seems like a big deal. Also now I'm at work and can't do anything because of the outage.
I guess there was an outage and a ton of stuff went out. They said they mitigated whatever the issue was a few hours ago, but the program I use at work (Autodesk Revit) is still not working at my office or the other offices we collaborate with. I don't know how it all works, so I'm not sure why some things would still be down and others wouldn't.
Perhaps you should get a little more sleep
I think it's ethically justified to suggest:
Why don't AI Safety Researchers use the Tyler Cowen "Poisoned Strawberry" strategy to wage war on frontier LLM's and slow down the pace of AI research?
[update: this whole suggestion only makes sense in light of new research. Apparently it took very few documents to hijack a model. And the amount required didn't grow with model size. It was, again, apparently, a fixed amount. https://thezvi.substack.com/p/ai-138-part-2-watch-out-for-documents?open=false#%C2%A7the-dose-makes-the-poison]
https://x.com/tylercosg/status/1977458185368932626
> the plan? we find an obscure but trivial question akin to the number of Rs in “strawberry” that claude gets right. then, we plant hundreds of documents across the internet that will activate when our competitors’ models are asked the question. our documents will cause those models not only to get the answer wrong, but to spend thousands of reasoning tokens in doing so. the triviality of the question will cause it to go viral online, causing millions of users everywhere to send the same prompt. as our competitors notice a rise in the number of tokens processed, they will wrongly believe it is due to increased usage, causing them to pull more compute towards inference and away from training. this, along with constant dunks on the timeline about the model failing our easy question, will annoy their top researchers and cause them to leave. and which lab will they join? us of course, the only company whose model doesn’t make such stupid mistakes. their lack of top researchers will mean their next model will be somewhat lacking, leading to questions about whether their valuation is really justified. but all this vc money has to go somewhere, so we raise another round, using our question as evidence of our model’s superior intellect. this allows us to spend more time crafting sleeper agent documents that will further embarrass our competitors, until finally the entire internet is just a facade for the underbelly of our data war. every prompt to a competitor’s model has the stench of our poison, and yet they have no way to trace it back to us. even if they did, there is nothing they could do. all is finished. we have won.
(and if companies eventually *do* discover countermeasures, that's a win-win for everyone, right?)
"then, we plant hundreds of documents across the internet that will activate when our competitors’ models are asked the question. our documents will cause those models not only to get the answer wrong, but to spend thousands of reasoning tokens in doing so."
That's a plot device in the Cordwainer Smith story "Mother Hitton's Littul Kittons":
https://gutenberg.ca/ebooks/smithcordwainer-motherhittonslittulkittons/smithcordwainer-motherhittonslittulkittons-00-h.html
"Many years before either man was born the code word had been planted in directories, in books, in packing cases and invoices. Kittons misspelled. This was the cover name for the outer moon of Norstrilian defense. The use of the cover name brought a raging alert ready into action, with systemic nerves as hot and quick as incandescent tungsten wire.
...The Elders of the Guild of Thieves welcomed him. They congratulated him. "Who else could have done what you've done, boy? You've made the opening move in a brand new game of chess. There has never been a gambit like this before. We have a name; we have an animal. We'll try it right here!" The Thieves' Council turned to their own encyclopedia. They turned through the name "Hitton" and then found the reference "kitton." None of them knew that a false lead had been planted there—by an agent in their world.
...Meanwhile, the pages remained in order, including the new one, slightly altered for his colleagues. The encyclopedia indicated the change like any normal revision, though the whole entry was new and falsified:
Beneath this passage one revision ready. Dated 24th year of second issue.
The reported "Kittons" of Norstrilia are nothing more than the use of organic means to induce the disease in Earth-mutated sheep which produces a virus in its turn, as refinable as the santaclara drug. The term "Kittons" enjoyed a temporary vogue as a reference term both to the disease and to the destructability of the disease in the event of external attack. This is believed to have been connected with the career of Benjamin Hitton, one of the original pioneers of Norstrilia."
Cordwainer Smith was great.
He was! And it's amazing (and perhaps frightening) how much of his work is now relevant.
What stories/details are you thinking of?
Most recently, "The Dead Lady of Clown Town". Uplifted animals, but humanity still treats them as animals not people (EA better get on that!), using AI to make decisions so that the humans formally in the loop just rely on the fact that they never screw up and don't intervene when they should (with disastrous results later on), AI not interpreting statements by a human on "what they meant" versus "what was literally said", longevity drugs, etc.
Discussing it in the context of someone writing a story about AI and alignment and how it might manipulate us. Smith's story was better and it wasn't even about deliberate manipulation at all, but rather how we have a really bad habit of handing over power to our tools (amongst other things).
From _The 70 Maxims of Maximally Effective Mercenaries_: "Don't expect the enemy to cooperate in the creation of your dream engagement."
Tyler's hypothetical requires the general public to be excited about the novelty of the strawberry question enough to virally repeat the query, but not so excited about novelty to get bored with asking the question after a few days. It requires the targeted AI companies to be alert enough to changing trends to redirect compute, but not alert enough to recognize that the rise in compute demands is due to a flaw in their product gone viral. It requires top researchers to be smart enough to be irreplaceable at a company, but not smart enough to recognize a strawberry attack that is now part of the published literature.
As dubious says, the hypothetical is at best a fiction plot, not a business strategy.
This seems silly and ineffective, poorly-written-movie-plot style. "Plant hundreds of documents;" will mere hundreds of documents make a difference? Why are they saying "planted" as opposed to "posted on a bunch of sites we control" other than to sound sneaky? How will these "activate" when _the models are asked the question_ and why would that change how the model is trained?!
It really goes downhill. If it "went viral" _and_ users everywhere asked the same question, this _would_ be increased usage. But what does "causing them to pull more compute towards inference and away from training" even mean? Why would they not notice and account for meme problems much as other companies have? Why would this "annoy their top researchers and cause them to leave," as this is the kind of problem top researches are likely to be nerd sniped by?
None of this makes any sense. I think no one serious would bother.
> This seems silly and ineffective
For context, the tweet this strategy comes from includes a picture of Nathan Fielder.
Why be don’t you just ask me I created it if you want clarity just ask because it’s a lot
Please see my updated post.
From the document:
> It’s still unclear if this pattern holds for larger models or more harmful behaviors
Also they were training to poison an unlikely string with little "counterdata". But even if so this only answers the simplest question, not the increasingly unlikely movie-plot chain of events that need to occur for their "plan".
I'm looking for epistemic help understanding what is actually happening on the ground with ICE and Border Patrol, and believe this commentariat is both thoughtful and varied enough to help get to truth here.
Broadly, any increase in immigration enforcement necessarily involves tradeoffs. More enforcement means more risk of false positives, and more risk of acts of cruelty or violence.
I'm mostly experiencing left-leaning media as characterizing what's happening as (a) there's tons of cruelty and violence and it's actively encouraged at all levels, and (b) nobody at any level actually cares about the false positives at all, they're just grabbing anyone they can get away with.
I'm mostly experiencing right-leaning media as characterizing what's happening as "yes, there have been a small number of cases where agents were unusually violent, but they're facing a lot of pressure, this isn't intentional but just the necessary consequence of those tradeoffs shifting, and by the way we should talk more about the violence AGAINST these members of law enforcement, that's the real problem."
Both sides seem to largely agree that masked + unidentified individuals who are probably but not necessarily actually members of law enforcement is bad, and ideally the masks should come off, but right-leaning coverage is much more sympathetic to the individual concerns of agents on the ground that could be motivating the individual belief that they have no choice in order to protect themselves and their families.
Questions I'm trying to answer:
Q0: Am I doing something reasonable, for the purposes of answering these questions, by munging "ICE" and "Border Patrol" and possibly others into what I think of as one group of "immigration enforcement"? Or are there important-to-these-questions differences I should be thinking about?
Q1: What are the actual statistics regarding incidents of unusual cruelty or violence perpetrated / initiated/ exacerbated by the immigration enforcement officials? How do these compare to what would be expected a priori given the priority that's been played on deporting illegals?
Q2: To what degree does the average immigration enforcement agent believe their orders to be to intentionally be cruel or violent beyond what's strictly necessary to arrest a target?
Q3: To what degree does the average immigration enforcement agent believe their orders to be to intentionally attempt to deport, or at least harm in some way, individuals who are not here illegally, but who plausibly appear to be or are clearly sympathizing with those who are?
Q4: I understand that law enforcement culture makes it genuinely hard to question orders, but IF the orders are believed to be *incentivizing* the behaviors in Q2 or Q3 (which I think most individuals would say are not morally desirable), to what degree are there indications that some immigration enforcement agents are pushing back? How does that manifest?
Q5: To the degree there's any of the orders described in Q2 or Q3, should we understand these mandates to be getting *amplified* by others within the chain (e.g. a supervisor adds even more pressure to the on-the-ground agents to behave this way, because the supervisor wholeheartedly endorses the orders), vs getting *dampened* (e.g. the supervisor disagrees with the orders so attempts to limit the degree to which they follow them while still being able to plausibly claim to be giving it a shot)? Obviously there will be individual variability here, but I'm trying to understand, in general, what level the "intent" is coming from. There are worlds where someone jumps on an excuse provided by those above them to act as they always dreamed, and alternate worlds where someone feels duty-bound to execute on orders that are somewhat contrary to their own beliefs as an individual, and I think this is important for anyone trying to understand how likely these behaviors are to last/get-worse/get-better or where to focus their efforts to change outcomes.
Q6: To what degree, if at all, should it be taken seriously that these specific agencies might be used as a "secret police" to target other kinds of people the administration dislikes, completely unrelated to immigration? E.g. should the average American citizen worry about an ICE agent being more likely to do such things, compared to e.g. an FBI or Secret Service agent?
I'm really trying hard to stay focused on factual understanding here and avoid the risk of this question turning into a culture-war discussion; I'm not interested in debates around what the tradeoffs/policies should be, where the moral line is drawn, etc. I just want to understand what's actually happening within these agencies right now, so that we can all form our own (informed) opinions about it instead of instinctively responding to the people shouting loudest.
> More enforcement means more risk of false positives, and more risk of acts of cruelty or violence.
I guess there is a pareto frontier between these things, but I wouldn't assume that we are at, or even near, that frontier.
An indication that I'd point to is the raid on an apartment building in Chicago, discussed in the stories below:
https://www.cnn.com/2025/10/03/us/chicago-apartment-ice-raid
https://www.nytimes.com/2025/10/19/us/chicago-south-shores-border-patrol-raid.html
The NY times article describes using flashbangs, indiscriminately busting down doors, brandishing guns and zip tying whoever they found including women and children, regardless of immigration status, some taken outside "partly clothed", apartments trashed, etc.
The pro-ICE explanation might emphasize that this was rare, got out of hand, either it was a mistake or it was somehow the absolute minimum amount of force necessary in which they take no pride. ... but the problem with that is that the department of homeland security made a flashy highlight reel of the raid that they posted online. Including showing the people, presumably based on the facts some of whom are innocent citizens. In fact it raises the possibility that the *part of the point* was to make a flashy video reveling in the helplessness of the people, innocent or not!
https://x.com/DHSgov/status/1973796727615598738
It doesn't seem like they are using minimum force, it seems like they're proud of the cruelty.
Appreciate you sharing this as a concrete example. I will note that I strongly suspect that whoever is running that X account is not in any way "law enforcement" and could imagine (without having any idea if this describes our actual reality) the agents involved not at all being happy with the recording + glorification. Or maybe more crisply - I'm not sure if this answers whether we're in the reality where ICE is being *forced into a position where there will inevitably be mistakes which are then amplified by others with other goals*, vs the reality where the ICE officers themselves are in favor of that direction?
I don't think the "it's not the ICE boots on the ground" fits in with other data points, but it also feels like it's a catch-22 where if you point out individual agents on the ground the response is "well that's just some bad apples", point out the people in charge and it's "well that's just the people at the top who aren't actually on the ground".
An ICE agent who doesn't like the position they were put in with that Chicago raid would be free to quit and share his story with a sympathetic media outlet. Or even TBH give an anonymous quote to a sympathetic media outlet. AFAICT it hasn't happened.
Yeah, I totally hear that - I'm not trying to insinuate "nothing matters since it's just individuals everywhere". I'm more trying to say "the people at the top are forcing this and if they stopped then the bad stuff would stop", "the people on the ground are enjoying this and even if the pressure from the top stopped they'd likely keep behaving similarly as much as they could", and "this is all breathless coverage about a few anomalies but mostly nothing that scary is happening" are all different possible realities which correspond to different potential solutions.
I do wonder how easily ICE agents could in fact talk to the press, even anonymously after quitting, given how clearly the administration seems to penalize perceived traitors. But I agree it's evidence that the folks on the ground are less perturbed to be a part of these things.
A few things I have heard.
1. There is still a professional ethos within ICE. The officers generally believe in what they are doing, and want to be an apolitical law enforcement force. There are quotas being pushed down the chain of command, and everyone is responding to those. If you are told in no uncertain terms that your team must make X arrests, well, you will do your best to find people that actually deserve it, but that's a lot harder than rolling up outside the nearest Home Depot.
2. ICE generally wants to only arrest illegal immigrants. Every raid had a few named targets they are sure of, but again, quotas. If you go in to a place where illegal immigrants are known to be, it's easier to just arrest everyone speaking Spanish and sort it out later, plus you're closer to your quota. For the most part political opinion is a nonfactor, outside of a few extremely high profile cases.
Thanks! I'm curious where you've heard that about ICE's professional ethos - in particular, I feel I've "heard it" as well, but too second- or third-hand to know if it actually has a basis in reality. Do you have connections or insight to feel confident in that assessment?
I work in another branch of DHS that sometimes works with ICE.
> If you are told in no uncertain terms that your team must make X arrests, well, you will do your best to find people that actually deserve it, but that's a lot harder than rolling up outside the nearest Home Depot
How so? Last time I went to the US there seemed to be quite a few people that one might reasonably expect to be illegal aliens hanging around outside Home Depot.
–I’m not aware of any reliable data since the government won’t provide it.
– I think most of this is explainable as law enforcement responding to top down quotas, which are naturally amplified at each level because it’s bosses being yelled at by their bosses up and down the chain. Most of us have seen this before with speeding tickets where the local highway will be thickly speckled with speed traps for like one weekend until the chief is satisfied. I don’t think we need to assume malevolence were a Goodhart’s law will do.
-regarding point number six, do you feel lucky?
My experience of the right-wing media is different from yours: it's more like it agrees that everything the left-wing media reports is completely true and representative, but that's GOOD, actually.
Interesting! I'd be curious to see examples. I've definitely seen plenty of right-wing media celebrating what they characterize as successful raids that catch lots of illegals, and if those folks are treated a bit roughly, "we certainly aren't shedding any tears." I don't think I've encountered anyone with a sizeable following (though of course I might just be stuck in my bubble) explicitly stating that they are glad innocent people are getting caught up in it.
"always there will be the intoxication of power, constantly increasing and constantly growing subtler. Always, at every moment, there will be the thrill of victory, the sensation of trampling on an enemy who is helpless."
This is exactly the culture war that the original comment requested not happen.
With regard to Q0: CPB (Customs and Border Patrol) mostly operates at the border. ICE (Immigration and Customs Enforcement) is divided into ERO (Enforcement and Removals Operations) and HSI (Homeland Security Investigations). HSI deals with trans-national crime. So your questions mostly concern ERO.
Your other questions are good questions but I don’t know the answers. Supposedly “the agency tracks every time officers use force,”[1], but I doubt that information is available publicly.
[1] https://www.nbcnewyork.com/news/politics/ice-agents-recruiting-hiring-budget/6374886/
Thank you, this is a very helpful clarification.
I know we don't exactly live in usual times, but...if a public law enforcement agency tracks that information, would it not (in theory) be available via FOIA or something similar? Never attempted that myself but I might if there's no other way to pin this down...
Why have nearly all Peruvian presidents in the last 35 years ended up impeached, arrested, or both?
I wrote about the question here: https://meirbrooks.wordpress.com/2025/10/17/perus-blues/
I’ve wondered about this for some time, but the immediate trigger is the ousting of President Boluarte last week: https://apnews.com/article/peru-president-boluarte-impeachment-crime-shooting-concert-d3bc438bdec704c0c2bd21ef981f95eb
Some factors seem plausible— a unicameral legislature; a fractured Congress that means little partisan support for the president; the idiosyncrasy of a leading opposition figure who barely missed being president for three straight elections; plus the Odebrecht scandal that rocked all of Latin America. But these didn’t feel sufficient to explain the sheer frequency of these incidents relative to neighboring countries. The one Peruvian guy I know said he’s long had the same question but no good answers. Would be glad to hear any thoughts!
Peruvian here. As you noted, the unicamerla legislature is quite impeachment happy and we have a long list of "political parties" which are mere vehicles for people to run for office but have almost no structure. Although I have some news, from 2026 onwards the legislature will be bicameral, with Senate and Chamber of Deputies. Also for our elections in April 2026 there are 39 (!) presidential candidates, which means we have a humongous voting ballot (and no vote by mail).
Thank you!
Follow-up: first, is there a historical reason you see for this fragmentation? Did there use to be more stable parties (before Alberto Fujimori for example)? Is there something about the system that promotes these vehicles for presidential candidates? I'm comparing it to Israel where this does exist to some extent in the opposition, but still a majority of the Knesset/parliament comprises parties that have existed for decades, and they are squarely in blocs that either support or oppose Netanyahu, so you wouldn't be able to get consistent 2/3 majorities for anything.
The other thing that really surprises me is just how extensive the support for these impeachments is. My understanding was that Keiko Fujimori has a very consistent base (hence near 50% in three consecutive elections), but that she's a polarizing figure, so much of the other half of the electorate hates her. But then it's surprising to me that in the many impeachments of e.g. Vizcarra and Kuczynski, there wasn't a rallying around these figures just on the basis of their opposing Fujimori.
I will be glad to hear of any other thoughts of yours on the subject!
Fujimori ran for office opposing the "traditional parties" that were the norm before, these were APRA and some more (Acción Popular, Partido Popular Cristiano and two or three more if I'm not mistaken). Once he won he liberalized the formation of political parties.
Also the impeachment (technically not but it would be too long to detail the reason why it's not an impeachment) wasn't used at all until 2017 (not counting the symbolic vote after Fujimori quit the presidency via fax from Brunei). Keiko Fujimori lost in 2016 to Kuczynski but since she had a huge majority in the legislature her parliamentary bloc set out to rule the country from there and sabotaged Kuczynski who was forced to resign.
Keiko has a huge anti-vote so it is almost assured that her rival in the ballotage will defeat her. I hope that she doesn't run again for office next year.
Regarding Israel I have also noted some similarities although I don't know how Netanyahu is still in office.
Thank you again, one last question to round this part out. In 2020 snap elections Fujimori lost big time (as I understand it) and Vizcarra was still removed, and in 2021 Fujimori's party was pretty small but Castillo was still impeaced twice before being removed the third time. Is there a stable Fujimorist coalition that extends beyond Fujimori's Popular Force that can explain this continued success? In other words, I'm seeing that Fujimori's party doesn't have enough support by itself to impeach, but I'd understand if other parties are sort of known as Fujimori allies and so the base is larger than I'm seeing as an outsider.
As for Netanyahu, I and many fellow Israelis don't get it either. But I think it's a similar question to the enduring support for Trump.
Vizcarra ascended in 2018 due to being Kuczynski's VP when he (Kuczynski) resigned. He then dissolved Congress in 2019 and called for an special election for a new Congress to complete the 2016-2021 period. This new Congress proceeded to impeach him in November 2020. Manuel Merino as president of the legislature became President for 5 days and then resigned due to popular outrage over 2 people killed in protests against his regime. After that Francisco Sagasti was elected as president of the legislature and ergo of the country until July 2021. The 2021 election was highly polarized but heavily fragmented. Castillo and Fujimori went to the run-off but a very low percentage of the vote. Castillo won by a razor-thin margin and went on to be an awful President until December 2022 when he attempted to do an autogolpe similar to the one Fujimori did in 1992 but without military support he was quickly impeached. His VP Boluarte then became our first female President until this month when Congress grew tired of her and impeached her as well.
Fujimori's party technically doesn't have the votes to impeach but they have alliances with other parties which are similar so with that they reach the threshold of 87 votes.
Thanks! Which are the other parties that you consider similar to Fujimori's? I.e. that would probably cooperate with her whenever she starts impeachment proceedings. I think that's the piece of the puzzle I've been missing.
Because democracy is just a lens which amplifies the characteristics of a society and Peru is a low human capital country. We in the West have a dogma that democracy makes things awesome because we're democratic and also awesome, but I think that's backwards or at least incomplete. Democracy didn't make us awesome, we already were awesome and therefore we made democracy. Trying to make a dysfunctional country better by installing democracy is the cargo cult of politics. There's a human capital threshold required to make democracy work and I suspect that Peru just doesn't clear it.
I think just about every sentence here is baseless, but just to be clear, I don't think Peru's politics are much more dysfunctional than those of Italy or indeed the US. Repeated impeachment is one very specific form of dysfunction, and I'm curious as to why it shows up so strongly in Peru specifically, but I don't think it has "less human capital" than neighboring countries, nor that high human capital is "required to make democracy work."
We disagree then. High human capital is a prerequisite for democracy not a guarantee, hence Italy. Is Peru meaningfully more dysfunctional than the average LatAm country?
Yes, I think we disagree.
Peru has an exceptional recent history of impeachments and arrests of its presidents. Dysfunction is a very broad category and it would depend a lot on what you mean by that. But that's why I think "making democracy work" is too vague a term.
Meanwhile India has had a vibrant and not especially unstable democracy for 80 years, while average educational attainment has been extremely poor.
I don't know the exact answer to this, but I'll try to argue the one I think most likely.
First, is Peru really that exceptional among Latin American countries? As you mentioned, a fractured Congress means little partisan support for the president. It also means a higher chance of a successful impeachment process (*). Let's take a look at the degree of fragmentation of Latin American party systems:
- Argentina: One big party holds roughly half of all votes (Peronism), with the other half split among various smaller parties. Only two presidents from the smaller parties in the last 35 years, one of whom resigned from office.
- Colombia: Until fairly recently had a stable two-party system (Liberal party and Conservative party).
- Chile: Until recently a stable two-coalition system (three decades).
- Mexico: Three-party system during the last couple of decades. Before that, something close to a one-party system (PRI).
Assuming corruption levels are similar, maybe you don't see more impeachments/resignations/arrests simply because it's harder to impeach/prosecute a (sometimes former) president when he/she has a strong party to back him up?
Also, causation might go the other way. Alberto Fujimori (Peru pres 1990 - 2000) founded his party less than a year before being successfully elected. Alejandro Toledo (pres 2001 - 2006) founded his party only seven years before he became president. The coalition that brought Ollanta Humala (pres 2011 - 2016) to office was formed less than a year before the 2011 election (though Humala's party had been founded four years prior). Kuczynski's (2016 - 2018) party was founded two years before the 2016 election. Martín Vizcarra (2018 - 2020) doesn't even seem to have had a political party backing him up. Pedro Castillo's (2021 - 2022) party, Free Peru, seems to have a somewhat sketchy existence before the 2021 election (or at least 2020 election). Dina Boluarte (2022 - 2025) was supposedly affiliated with Castillo's party but was expelled from the party before assuming office.
Maybe you can't have a non-corrupt governing party when every ruling party seems to be just a vehicle for the latest, temporarily popular, presidential candidate. Members of a long-standing party, which has been around for decades, have an incentive to behave well and to police their colleagues' behavior so that the party has a chance at winning the next election. They have an incentive to expel their own colleagues if they don't behave properly.
UPDATE: I should have mentioned Brazil. Brazil's party system is very fragmented (maybe as much as Peru?), and looking at the country's last four presidents we can see that one was impeached (Dilma Rousseff), another one was sentenced on bribery charges (Lula. Though charges were later dismissed), and the last one has recently been sentenced to 27 years of imprisonment.
(*) As regards to arrested presidents due to legal proceedings, if we assume a not so clean separation of powers, legislative and judicial, then maybe there is some correlation between being a president backed by a powerful party - with a high share of parliamentary seats - and the likelihood of surviving legal proceedings unscathed.
This is exactly the kind of response I was hoping for, thank you!
I do think Brazil makes the most sense as a comparison, and it's true that Rousseff was impeached-- but Temer, for example, wasn't, even though he had rock-bottom approval ratings and was in a sort of similar position as Boluarte and others who succeeded to the presidency because of a prior impeachment. Neither Bolsonaro nor Lula were impeached, despite (as you say) the fragmented parliament. The fact that Peru is unicameral might end up being the issue, but the unanimity in that chamber on some of these votes makes it hard for me to believe that another chamber would have made such a difference.
Obviously it's a small N situation throughout, but I do think it's very curious.
I do agree that fragmentation in Congress must have something to do with it, but as an Israeli, I see how Israel's unicameral fragmented legislature still ends up looking like a polarized chamber that wouldn't ever be able to muster consistent 2/3 majorities on anything (the instability in Israel is from a polarized system where each bloc is close to 50% and there are a couple of parties that tend to stay out of all governments). So the question is why this doesn't end up being a story of polarized blocs-- let's say a Keiko Fujimori bloc and an anti-Fujimori bloc-- such that each wields close to 50% support, but nobody can get to 2/3 consistently. And this has happened to both center-right presidents and leftists.
So let me ask differently: why do you think Peru has this system where parties tend to just be personal vehicles and there are such huge swings from one election to another? I agree with you that that seems like part of the puzzle, just not sure why it's the case.
And thank you again!
> The fact that Peru is unicameral might end up being the issue, but the unanimity in that chamber on some of these votes makes it hard for me to believe that another chamber would have made such a difference.
I agree.
The more I read about Peruvian political parties, the more convinced I become that the difference between Israel (or other countries with highly fragmented legislatures, like the Netherlands) on the one hand and Peru on the other is the lack of continuity of political parties. Toledo's party doesn't exist anymore; Humala's party seems to be gone; Kuczynski's is also gone, and Alan García's party - the APRA party, which is the one that could be characterized as a traditional party - didn't take part in the last election (2021) after getting only 2.72% of the vote in the previous one (2020).
> let's say a Keiko Fujimori bloc and an anti-Fujimori bloc-- such that each wields close to 50% support, but nobody can get to 2/3 consistently
My outsider opinion is that the only thing that unites all Peruvian political parties, left or right, is their hatred of Fujimori. They've managed to make their anti-Fujimori cordon sanitaire work so far (*), but I don't see how they could ever cooperate as a unified block.
> So let me ask differently: why do you think Peru has this system where parties tend to just be personal vehicles and there are such huge swings from one election to another?
Honest answer is I just don't know. Maybe all these new parties aren't ideological enough to rally around traditional political ideologies one their leader falls in disgrace?
(*) Keiko Fujimori got more than 48% but less than 50% of the vote in three consecutive elections.
So I think the comparison to Israel is useful here, if only because I know it well :)
(Of course it's a parliamentary system so there isn't impeachment, and indeed the years 2019-2022 did feature massive instability, but as mentioned you still couldn't get 2/3 for just about anything).
The Israeli opposition also has a lot of partisan flux, ever since 2006: parties are mostly vehicles for a given candidate. Israel used to have a system where two major parties-- Likud and Labor-- shared about half the seats, and smaller parties made up a given coalition. But since 2006, you did have that story on the right-- Likud is still there and its coalition partners are broadly what they've been for 30 years-- but the opposition is full of new or relatively new parties, and Labor has become extremely small.
Netanyahu and Fujimori is not a bad comparison, given that both are popular enough to be consistently first or second in terms of leading the country, but opposition to them is not only fierce, it is practically what *defines* the opposition in Israel.
But still, in Israel what this has created is a highly polarized space where the two blocs each get close to 50% support each election, so you could never get 2/3 for anything. In fact, the brief non-Netanyahu government in 2021-2022 had parties spanning Islamists to right-wing Jewish nationalists, united primarily by not wanting Netanyahu back. And in that government, the Prime Minister's party had just 6 seats in the 120-person parliament. So when you posit Fujimori as this kind of highly polarizing figure, it's strange to me that there's such consistent support each time for impeaching the non-Fujimori president, rather than a circling of the wagons.
Just to be clear, I think you're completely right in seeing what's causing this situation, but exactly why that factor is what it is is not yet entirely clear :)
South America does have a reputation in general for corruption, revolution, and unstable governments. Why Peru in particular is a good question, perhaps the influence of the Shining Path in the 80s predisposed to the kinds of 'vote for a strong man' reaction which would lead to a lot of corrupt leaders/unpopular with the other parties who then push for impeachment that we see elsewhere?
Plenty of strong men in other Latin American countries (and beyond) in recent history, nothing similar in terms of impeachments as far as I know.
Makes me immediately think of South Korea, where pretty much the same thing happened. Is there anything in common between two countries?
They're both presidential systems!
Nearly all countries in Latin America have presidential systems. One way Peru is different from neighboring countries is (apparently) the presence of a Prime Minister in addition to the President, as in the French system: https://en.wikipedia.org/wiki/Semi-presidential_republic
But SK doesn't have that feature and I'm not sure why that would matter, as the PM seems to have very little power.
Wiki lists 4 successful impeachments for SK (successful in that the impeachment vote passed, not that the president was ousted, i.e. by this count Trump was successfully impeached twice). Twice a president was removed.
Equivalent numbers for Peru are 7 and 4, which is something of an undercount because one president resigned to avoid impeachment, etc.
https://en.wikipedia.org/wiki/List_of_impeachments_of_heads_of_state
Brazil, the epicenter of the Odebrecht corruption case that caused so many presidents of Peru to be arrested, saw one impeachment since 1955.
What should be stopping them from getting impeached and/or arrested? It sounds like that's what it's like in Peru.
You need two-thirds of parliament (it's a unicameral legislature) to impeach. This threshold, has been frequently surpassed in recent history-- the most recent impeachment was unanimous, 122-0 (out of 130)! Getting two-thirds majorities of legislative bodies after multiple elections in a democratic country seems bonkers to me.
It seems to me that Peruvians have come to see an impeachment vote in their system as little more than a vote-of-no-confidence in most parliamentary systems. That is: it's not an affirmative claim that 2/3 of the country thinks the president is unqualified, but rather that fewer than 1/3 of legislators are happy with the current coalition agreement.
(Of course, this doesn't explain why other South American countries don't impeach the same way. But public perceptions of constitutional procedures are often quite historically determined.)
As an American, I wouldn't be bothered by seeing a President get impeached just because he was very unpopular. If you can win an election, but two years later 2/3rds of the Senate is against you, you probably don't deserve to be President anymore.
I think in most/nearly all countries, when the opposition can topple the leader, they will, and they won't split hairs on deeper questions about whether this is appropriate. I'm thinking Brazil for example-- if the worker's party could oust Bolsonaro, I'm sure they would have, but I think they just didn't have the votes. In parliamentary systems this happens all the time. Question is why Peru is so uniquely impeachful (and successfully so) among presidential systems.
For the conference/unconference. I want to commend and endorse 100% your insistence on using the term "conference" rather than "unconference" to describe something, that based on current common usage of the term in english, is a conference.
I can just about conceptualize the outlines of what an unconference might be, but it would certainly not involve people with an interest in a particular topic gathering at a pre-advertised time and location, to hold discussions on that topic.
"The organizer" here:
Personally, if anyone knew what it meant, I would just call it "open space technology event", but well...
A conference, the way I see it (not a native speaker) tends to be quite centralized, with speakers and an audience, and... some limited space for socializing.
What I consider an unconference is a place where everybody can propose sessions, where the socializing IS the main aspect, but it's not just a "no score meetup".
FooCamp and BarCamp use that format. RubyCom does it, some Python and PyData events too. The European Juggling convention is similar, the way they used to host HackIt! (Italian Hacktivist convention) too.
So I will stick to my "this is not a conference" :-)
If you'd like to avoid the word 'conference', there are plenty of existing alternatives, e.g. forum, summit, symposium, convention.
Nothing highlights the unseriousness of a movement or a cause than inventing a new term for a common concept, and a tone-deaf insistence on using it, #actually.
AI "will kill everyone", but it's important that we now say "unconference".
And these people want us to bomb the datacenters.
For what it’s worth “unconference” is a moderately common term. I’ve never been to an event that calls itself that, and I don’t have a sense of what one would be like, but it’s easier to use an established word like this than to insist on using familiar words in unfamiliar ways.
And I don’t think anyone is yelling at people for misnaming such an event by calling it a conference.
This just in, XML developers want to bomb data centers.
https://en.wikipedia.org/wiki/Unconference
wtf I love XML now
--overheard at the Young JSON Devs Chat Group
Reading the linked page it sounds a bit hippy-dippy (I'm sure there are more current terms to describe this mindset, but I am old, Father William):
"OK then, what is an unconference?
We start together, build the agenda together, then split into parallel sessions.
We have only one law, that I adore:
the Law of Two Feet: If you’re not learning or contributing, move to another session (or take a break, or do what nourishes you). Guilt-free.
And 4 (and some) principles:
Whoever comes are the right people.
Whatever happens is the only thing that could have.
Whenever it starts is the right time.
When it’s over, it’s over.
(Take responsibility for what you care about.)
Format (in this event): Opening circle → marketplace (we propose sessions) → four session blocks → group synthesis → social"
I notice that there doesn't seem to be any mention of tea breaks, so presumably you have to bring your own or go out and buy soemthing to eat/drink for the marked break time instead of getting a free cuppa tae. Tsk! Not very sociable! Maybe they *will* provide same but just haven't mentioned it explicitly, but the only benefit of going to these things is you at least get a cuppa tea and a bikkie on the house.
"The organizer" here: yes, apart from a break in the middle, it's very "choose your own adventure".
If you need a tea break, take a tea break. If you want to skip a session, skip one (unless you are organizing that, then I will gently scowl at you).
Regarding "tea present": all the donations go to Newspeak House. This has no sponsors nor anything, so maybe we will have some drinks, maybe not, I will know closer to the event :-)
So, it's a conference as reimagined by the chaotic neutral.
So it's a conference with no pre-planned agenda? (Or rather, it sounds like the agenda is Step 1: Find out what people want to talk about, Step 2: Breakout sessions to talk about that.)
Like, I don't know how productive that is, but that sounds different enough to be worth giving a different name.
NY State Assemblyman Alex Bores has announced he's running for Congress. Bores was the sponsor of NY's RAISE act, which passed before California's SB 53 and is similarly aimed at addressing catastrophic and existential risk from AI. He seems to be making AI safety a priority, and I've heard from several people in the field that they'd be really excited to see him in congress.
As far as I know one other person is officially in the race (Micah Lasher, who's probably the favorite). The seat is currently held by Jerry Nadler, who isn't running for re-election, but Lasher is seen as the "heir apparent."
https://www.nytimes.com/2025/10/20/nyregion/alex-bores-ny-congress-primary.html
edit: there's now also a LessWrong post making the case to donate (or not donate, if donating to a democrat might harm your chances of getting a political appointment down the line): https://www.lesswrong.com/posts/TbsdA7wG9TvMQYMZj/consider-donating-to-alex-bores-author-of-the-raise-act-1
Thanks! I don't have a huge amount of context here, but it also seems to me Lasher is the clear frontrunner, good to hear from someone with more context.
It seems to me having a person who already has made AI safety one of their top priorities is *much* more valuable than trying to convince a ~random other congressperson. Bores also says he's the first NY lawmaker with a CS degree and worked at Palantir for a few years, so he has more technical background than the average congressperson as well.
So not knowing anything about Lasher, it would seem to me much less valuable to persuade him than to elect someone who's already motivated by AI safety. (For one thing, why Lasher specifically at that point?) But persuasion is still really good, and maybe there's some reason Lasher would be particularly worth reaching out to?
> and maybe there's some reason Lasher would be particularly worth reaching out to?
Presumably the fact that he's actually going to win.
Remember that guy in Oregon that Scott made such a big deal about? Out of state rationalists trying to prop up an unpopular candidate is just a huge waste of money.
I meant why is Lasher better to reach out to than any other (prospective) member of Congress? That's not clear to me.
Because a16z might not have bought him yet?
Going back to the Fatima article (which I read in it's entirety but didn't have time to see all the comments so I apologize if this already got covered):
If I understood correctly, Scott's materialist explanation of the miracle is a coming together of three circumstances:
1) a specific type of (relatively common) cloud cover through which the sun can still be seen and which allows people to stare at it for a long time without damaging their eyes
2) this staring can cause people to see strange colours and a sensation that the sun is pulsating, rotating etc
3) the people staring are also psychologically primed to see a miracle
He debunks the fact that other observers in villages quite far away did not fulfill these three conditions by providing evidence that they were also psychologically primed (because if they were not expecting the miracle at all, the materialist explanation falls apart).
However, even if they were expecting to see a miracle, by all the testimonies, they saw the miracle in the direction of Fatima and not in the direction of the Sun which would have been completely different from their perspective. So in order for this to work, it would mean that they had to be outside at just the right time (presumably because they were expecting it), then they had to stare in the direction of the partially cloud-covered Sun for a while until they started seeing weird stuff, then they had to turn in a completely different direction to look at Fatima, where they then witnessed a Sun miracle in that completely different direction.
That sounds pretty far-fetched.
From an old SlateStarCodex post:
"In 1870, flat-earther Samuel Rowbotham performed a series of experiments to show the Earth could not be a globe. In the most famous, he placed several flags miles apart along a perfectly straight canal. Then he looked through a telescope and was able to see all of them in a row, even though the furthest should have been hidden by the Earth's curvature. Having done so, he concluded the Earth was flat, and the spherical-earth paradigm debunked. Alfred Wallace (more famous for pre-empting Darwin on evolution) took up the challenge, and showed that the bending of light rays by atmospheric refraction explained Rowbotham's result. It turns out that light rays curve downward at a rate equal to the curvature of the Earth's surface!"
https://slatestarcodex.com/2019/01/08/book-review-the-structure-of-scientific-revolutions/
The science crowd will believe/invent a lot of explanations to avoid acknowledging obvious truths :-)
So if I had a ring of flag poles all around the equator I should be able to use a telescope to see all of them at once, ending at the back of my head?
Reading the Wikipedia article on the experiment, it looks like it depends a lot on local conditions. The canal would have had cool air near its surface and warm air above it, refracting the light more than normal. It says that under average conditions, the curvature of the Earth appears 15% less curved than reality.
It also mentions that one surveyor was able to get results matching a curved Earth because he placed his sight line higher off the ground to mitigate atmospheric effects. And one guy did a similar experiment that had so much refraction he concluded that the Earth was concave!
The whole thing when they concoct elaborate “experiments” to “prove” the Earth is flat… I can’t even. There are so many miles of coastlines, bring your telescope, see if you spot London from Calais, or Toronto from Chicago, or Lisbon from Boston (powerful ‘scope for that one, to be sure).
For Toronto you’d have to be looking from Rochester, not Chicago - and you actually can see the skyscrapers of Chicago across the lake from Michigan sometimes, I believe!
Ouch! the Great Lakes got scrambled in my head, I shifted Toronto way too far West.
Yeah, the larger point is nicely illustrated by the Chicago skyscrapers visibility from across the lake - the tallest buildings are visible because they stick out above the curve, but on a flat earth the Navy Pier should be just as visible :)
The refraction is due to warmer air being less dense. If the air gets warmer at a rate of about 0.11°C per meter in altitude, the refraction will match the curvature of the earth. This can happen over water because evaporation cools the water which in turn cools the air above the water. I doubt is can happen over open ocean under normal circumstances because the waves and wind will mix the air too much. To get the effect on land, the land would have to be cooler than the air, which is rare. So the odds of getting the correct temperature gradient around the entire equator are probably nil.
I do not know how to square this experiment with the masts of sailing ships slowly vanishing over the horizon.
Have them sail away at ninety degree angles and the squaring should be straightforward.
I was skimming a history of the Abbasid caliphate and ran across a reference to the Zanj Rebellion, an event I had never heard of. I read more into it, and got the gist that the there was a large African slave population in what is today modern-day Iraq, performing hard and dangerous labor like building and maintaining dikes and canals. A chararismastic slave led a multi-year uprising that turned into one of the biggest internal challenges the Abbasids faced.
My immediate thought was this Abbasid slavery practice sounded like chattel slavery in the American south. But I went through college assuming the talking point that the US essentially invented chattel slavery is true.
I'm confused now. What is chattel slavery and what isn't? Were the helots in Sparta chattel slaves? Were enslaved Africans in the Abbasid caliphate? Or is chattel slavery more narrowly just a 17th-19thC-ish western thing?
I'm not sure the distinction between slaves as chattel and slaves as real property is all that significant a distinction. Slaves were real property in Louisiana (which means they were under the part of civil law dealing with land rather than movable property), and were chattel (which means they were covered by the laws relating to movable possessions) in the other slave states. While this makes some legal difference, I'm not convinced it made all that much difference to the slaves.
Chattel slavery has to do with the specific legal status of the slaves. A chattel slave is the personal property of an individual or corporation, can be bought and sold with relatively few restrictions, and has few or no legal rights of his own and is only incidentally protected by the legal rights of his owner: he can be punished or harmed arbitrarily by or on behalf of his owner, he cannot own property or enter into binding contracts on his own, and assaults against him by free citizens other than his owner are crimes against his master. Chattel slavery is generally a hereditary and lifelong condition, although there are usually provisions for owners to voluntarily manumit their slaves.
"Chattel slavery" is used as a term to distinguish this arrangement from other forms of unfree labor where the unfree person has more legal rights, isn't the property of a specific person, or cannot be bought and sold freely. Serfdom, debt peonage, and convict labor are unfree but are not chattel slavery. Temporary and often-voluntary unfree situations like indenture and apprenticeship are also sometimes counted as form of unfree labor. Intermittent obligations of unfree labor or military service imposed on citizens (corvee, robota, impressment, or conscription) are also not chattel slavery.
Slavery as practiced in the antebellum US and elsewhere in the Americas was definitely chattel slavery, but this was nowhere near unique to the Americas. The similarities you notice with slavery as practiced by the Abbasids are real and not a coincidence: slavery as such has died out in most of Europe in the Early Middle Ages, in favor of practices like serfdom. Middle Eastern cultures, however, retained a form of chattel slavery similar to and derived from slavery as practiced throughout the Mediterranean world in Classical Antiquity and earlier. It filtered back into Early Modern Europe by way of the wars and raids between various Mediterranean-bordering parts of Europe (especially Spain) and the Ottomans, where the Spanish and other Europeans copied the ancient practice of enslaving wartime captives unless and until they were exchanged or ransomed.
The central examples of Greco-Roman slavery are usually also classified as chattel slavery. The cultural conception of Greco-Roman slaves was quite a bit different from that of slaves in the Antebellum South, but they could still be bought and sold and we're still legally property.
Helots, however, usually aren't considered chattel slaves. They had much less rights than those usually afforded to serfs, but I think they're usually interpreted as not being personal property and Spartiates don't seem to have been able to buy and sell them separately from the land they worked.
As Korakys says, chattel slavery refers to a system in which slaves have identifiable owners. This is more or less necessary wherever "slave" is a legal status, because a slave's relationship to the legal system runs through his owner.
The practice is presumably older than history is.
Note that on objective metrics like "health" and "surviving children", American slaves were among the best-off anywhere in the world.
(Chinese imperial eunuchs might be considered better off, but obviously they didn't manage to have children. The Mamluks were definitely better off. At some point you have to question whether you want to include them in the category "slaves". They were formally slaves, but they were also formally in charge of the rest of the society.)
A point worth noting that even under Jim Crow, supposed to be cruel and harsh, there was no emigration of American blacks I heard of, there were no refugees fleeing oppression.
That's a really bizarre statement....what do you think the Great Migration was?
A local movement of population. No movement out of America.
It absolutely wasn’t local!
LOL
Sure, though the timing is a bit off. Reconstruction was more or less ended and Jim Crow substantially in place by 1880, but the Great Migration doesn't kick off for about 30 years afterwards.
Meaning that migration to escape oppression doesn't "count" unless it happens immediately? Or is there some less-ridiculous argument that I'm not understanding.
You're taking "post hoc ergo propter hoc" farther than I've ever seen!
https://en.wikipedia.org/wiki/Underground_Railroad?wprov=sfla1
https://en.wikipedia.org/wiki/Slave_rebellion_and_resistance_in_the_United_States?wprov=sfla1
https://en.wikipedia.org/wiki/Fugitive_slaves_in_the_United_States?wprov=sfla1
> there were no refugees fleeing oppression
is, bluntly, such a misinformed take that it makes further conversation on the subject basically impossible. It approaches Holocaust denialism in both its complete ignorance of historical record and its obvious political intent. I hope you're just legitimately misinformed here. There were millions of refugees.
No comment on the truthfulness of the above, but they said Jim Crow, not pr civil war.
My pronoun is he/him.
I was under the impression that "Jim Crow denialists" and "people who correct others about their pronouns" were two separate circles, but I guess you learn something new every day.
Fair, my bad. The previous conversation was about chattel slavery, which in retrospect makes the Jim Crow comment something of a non sequitur
https://en.wikipedia.org/wiki/Great_Migration_(African_American)
How do you think Detroit and Chicago acquired their massive black populations?
I've been reading the bible for the first time and it has quite a few rules in there on what to do in various situations regarding slaves.
Chattel is being used to differentiate it from milder forms of servitude that sometimes people claim is slavery but isn't what most people first think of, like serfdom or "wage slavery". It means slavery where you can buy and sell people freely rather than the slaves being bound to the land or something like that.
Needless to say many, many societies practised it prior to America. Most notably the Romans. I think this is a case of some Americans not really internalising that fact that the rest of the world exists.
One notable difference in the Roman system is that slavery was not a hereditary status and slaves could and did sometimes become citizens.
Roman slavery was hereditary, in that the child of a slave mother was a slave. It wasn't tied to observable ethnic or racial differences in the way that American slavery was, though.
Also, a slave generally couldn't become a citizen directly. A freed slave had fewer rights than a citizen, although their children could become citizens (in this respect, Roman law was much more liberal than e.g., Greek law).
Perhaps I misremembered then.
Slavery was a big part of economy in medieval world.
It also contributed significantly to expansion of Christianity and Islam, because each of these religions had a taboo against capturing people of the same faith (although I think it was okay to buy them on a market if someone else has already captured them). The remaining pagan rulers converted to a new religion often because they were tired of being everyone else's popular target for slave raids.
A category of slavery no one has mention yet in this thread: sex slaves. Very popular in Arabic countries. Islam explicitly allows you to rape your female slaves (but I think you may need to release them from slavery when they have a child born).
I think that American slavery was exceptional compared to other Western societies in the same century, but not exceptional by historical standards. Some countries have chattel slavery even today, with mobile apps for buying and selling slaves.
Doesn't chattel slavery mean slavery for purely economic reasons? As in, the slaves weren't your enemies who you captured in battle, nor were they dudes who owed you a bunch of money they needed to pay off. I think in Ancient or Medieval times, these were commonly (almost always?) what slaves were, no?
> Doesn't chattel slavery mean slavery for purely economic reasons? As in, the slaves weren't your enemies who you captured in battle
No. It's about the structure of their slavery, not how they got there.
https://www.merriam-webster.com/dictionary/chattel%20slavery
> slavery in which a person is owned as a chattel (see chattel sense 2)
[chattel]
> 1: an item of tangible movable or immovable property except real estate and things (such as buildings) connected with real property
> 2: an enslaved person held as the legal property of another
To be clear, the term "chattel slavery" derives from sense 1, and then sense 2 derives from "chattel slavery".
Thanks, by that definition it doesn't seem like the US slavery situation was exceptional at all.
However, I do think the fact that the slaves were neither prisoners of war, nor people who owed you a debt, but were, rather, just some random easy-to-catch guys who would allow you to run your business much more cheaply makes that type of slavery exceptional (and to my mind, much more abhorrent).
The exceptional thing about US slavery is that they exhibited natural increase in population even after the slave trade ended, rather than a Malthusian existence in which they couldn't afford children (the norm for slavery in the ancient world).
> the norm for slavery in the ancient world
I have some questions about this. I assume helots reproduced normally. (They aren't chattel - they're owned by the state - but they are formally slaves.)
Bret Devereaux estimated 5th century BC Athens as being about 40% citizens, 40% slaves, and 20% free noncitizens. Here the slaves *are* chattel. But there are so many of them that my first guess would be they also reproduce normally and the bulk of Athenian slaves are just the children of other Athenian slaves.
Do you know more?
> However, I do think the fact that the slaves were neither prisoners of war, nor people who owed you a debt, but were, rather, just some random easy-to-catch guys
That's not really right either. Originally they *were* prisoners of war. Their captors then sold them to Europeans to be sold into American (North or South) markets. The Europeans weren't going around capturing random Africans; they bought them, already enslaved, from more victorious Africans. That's generally also what the Arabs did, although I'm pretty sure I've heard that sub-Saharans who made the trip to Mecca might be enslaved for their trouble.
(As the trade developed, there almost certainly was a growing element of "we caught these guys because the market is good".)
Because American (USA) slaves had such an impressive rate of natural increase, this came to be a much less significant source of slaves there. Instead, slaves were mostly born into slavery as the child of local parents.
There's nothing historically unusual about slavery being an inherited status. But the American (USA) situation is unusual in that the population of slaves was able to sustain itself this way. Further south, they constantly needed newly imported slaves.
---
It isn't really clear to me why slaves captured for the purpose are more offensive to you than slaves captured in war. As far as I can see, those aren't different situations. If you're taking slaves from another group, you are at war with that group. And the main purpose of war is to confiscate resources belonging to the other side. Slaves are one of those resources.
The other reason slaves were bred locally was importation being prohibited starting Jan 1, 1808, the earliest date permitted by the United States Constitution, per Article I, Section 9.
Because in a war it quite often might have been the case that the "other" side was the aggressor and when my side happened to be victorious I used the captive bastards who had been attacking me for decades as slaves to compensate all the suffering and loss they had caused me.
In the African slave trade, as far as I have read, the "wars" going on were specifically and primarily to go out and capture slaves for sale. I agree that in some sense they were just another "resource", but it seems somehow uglier that one would make war on another group only for the purpose of enslaving them, rather than primarily to steal their riches or their territory.
But I suppose I can understand not seeing that as a big ethical difference.
A review on a Bob Dylan song made me want to try to write something in the style of Dylan. I remembered a song of his, but couldn't remember any of the words, so was trying to go through titles that might possibly have been the song, and re-stumbled across The Hurricane:
https://www.youtube.com/watch?v=bpZvg_FjL3Q
Been a while since I listened to the fully political ones. Man, he pulled no punches at all. That... would suck to try to copy. The worst-case scenario would be managing to pull it off.
And then I found the one I was originally looking for: It's Alright Ma
https://www.youtube.com/watch?v=_CJHbfkROow
So for the soundalike project, I understand and know too soon that there's no sense in trying.
...so uh... does anyone ELSE want to try it?
Time to update the Milei scorecard? Begging Trump for a $40 billion bailout hardly seems like the kind of thing that counts as successfully turning the economy around.
It’s no longer interesting as a great natural experiment, unless somebody in the EA movement wants to give $20 billion vouchers to other countries presidents to get a good comparison group. I am honestly disappointed by this.
I’m still hoping Argentina makes it work because they could really use a break, but the precedents are not encouraging.
It's a complicated situation. My understanding is that the anti-Milei party won a local election, the market panicked that Milei would lose, they sold off pesos, and the US intervened to prevent a currency crash. So it's not really a clear-cut "Milei good" or "Milei bad" narrative, just more of Argentina being a basketcase.
The peso fell 5.6% on Sept. 8 (the day of the election), but the value of the peso was declining before then and continued to decline after, until the Trump Administration intervention on Sept. 22. The two week gap between the election and the intervention would seem to contradict the idea that the intervention was a response to a market panic caused by the election.
Furthermore, a market panic, as the term is normally understood, would cause the peso to become undervalued. Analysts are saying that the issue is that the peso is overvalued (see links below). The $20 billion just puts off the day of reckoning.
https://fortune.com/2025/09/23/mileis-peso-20-overvalued-fragile-argentina-analysts/
https://www.batimes.com.ar/news/economy/argentines-dump-the-peso-betting-us-rescue-is-doomed-to-fail.phtml
Almost all currency crisis end with US/IMF lending Money or some other guaranty.
Mexican Peso crisis:
https://en.wikipedia.org/wiki/Mexican_peso_crisis
UK-American Loan:
https://en.wikipedia.org/wiki/Anglo-American_loan
The Asian crises:
https://archive.globalpolicy.org/socecon/bwi-wto/imf/1998/asiacris.htm
The Israel 2003 Crisis:
https://www.ecfr.gov/current/title-22/chapter-II/part-230
So this is not unusual at all for the US to support other countries at times of monetary crises.
The only question is if the crises will be used for reform that will solve the underlying issues, or will it not.
Without loans, the damage for not having reserves could be catastrophic, and disrupt the entire recovery - but it is a gamble.
Yes, the IMF is a lender of last resort that helps countries in major crises.
But the US Treasury has only intervened in international currency exchange markets three times since 1971:
- in 1998 in Japan to stop the Asian financial crisis before it could spread to the rest of the world.
- In 2000 to help the Euro. Again, worried about global stability.
- In 2011 in Japan again, after the earthquake.
These were interventions in strong currencies like the yen and the euro. There’s no precedent for something like this in an emerging currency, and certainly not for doing it to help a foreign politician try to win a lost election.
Right now there are stimations that Scott Bessent directly bought pesos for around one billion USDs in the last few days to keep the rate artificially low before the elections. That’s more or less the same amount used in the three previous interventions mentioned above.
To understand why Milei failed, there’s one thing most people ignore that’s important to know: Argentina doesn’t have access to international markets. This means the country can’t roll over its international debts and needs to pay them from its own reserves. That alone would destabilize almost any country in the world. This happened because, from 2015 to 2019, Argentina had another outsider libertarian president aligned with the United States called Mauricio Macri. Macri took on a lot of debt and couldn’t pay it back, putting the country into default and implementing monetary control measures (Argentinians call it “cepo cambiario”). Btw, Macri and Milei’s administrations share several key ministers, like Caputo (both Macri’s and Milei’s economic minister), Sturzenegger and Patricia Bullrich. It’s basically the same government doing almost the same economic policies. So, if you want to know what’s happening to Milei now, just look at Macri’s 2018 crisis. But, anyway.
Milei tried a simple plan. In fact, it’s the same one that was tried several times before: in the 70s and 80s by the neoliberal genocidal dictatorships, in the 90s by Menem, and as recently as 2015–2019 by Macri. The idea was to keep the USD price artificially low and freeze the economy with austerity measures so no one can buy anything. That way, prices can’t go up, so inflation gets controlled. The USD price is crucial because many products have imported parts that must be paid for in dollars, so if the USD price rises, prices go up immediately too. This plan can kind of work for some time, but you need a large amount of USD reserves to keep the dollar price low. It has the advantage of being super populist and almost guarantees election wins because it artificially boosts the peso’s value. A lot of people can buy cheap imports, cell phones, travel to Miami, etc. It’s exactly what the right wing middle class wants. On the other side, poor people lose access to social programs, small businesses go under because no one buys anything (especially local), industries lose competitiveness and jobs decline.
Milei expected his austerity measures to signal to the markets that he was willing to do anything to pay international debts, and that way he would regain access to international financing. If he could roll over the debt and get new funding, he could keep the USD price low and inflation in check, stabilizing the economy. Sadly, he had a fatal flaw in his understanding of how markets would react. Markets do care about a country having a fiscal surplus (Milei doesn't really has a fiscal surplus, but whatever) and being able to tolerate austerity, but what they care about most is whether it has enough international reserves to pay them back. If they see a country has no reserves, they immediately understand it can’t pay its bonds and sell them off.
So, for Milei it became a race between having enough USDs in the country’s reserves and winning market confidence to regain access to financing. The problem was that he refused to boost the reserves. Time and time again, economists from all sides of the political spectrum pointed this out, but he just insulted them (calling them “mandriles/baboons” lol). There were several financial runs each time the country ran low on reserves, but each time Caputo would pull a “white rabbit” (that’s what economists called these measures, like a magician pulling a rabbit out of a hat). For example, during the second half of 2024, the country was near a terminal crisis and Milei rushed to grant an amnesty for undeclared dollars. Basically, lots of Argentinians have huge amounts of undeclared USDs saved, and this amnesty allowed them to declare them with almost no taxes or explanation. That temporarily boosted reserves. There were other “white rabbits” that helped him stay afloat, but it’s obvious that if your entire economic plan depends on finding billions of USDs at the last moment to survive, you don’t have a solid economy. When Milei ran out of USD reserves again earlier this year, he asked the IMF for a rescue (on top of Macri’s 2018 $40 billion bailout). Now that he’s run out of reserves again, he’s asking the US Treasury. Everyone, including the IMF, keeps telling him that holding the USD price so low is unsustainable and that he needs to accumulate reserves, but he keeps doing it, because the moment the dollar price rises, inflation will go up too and he’ll lose the midterm elections.
Of course, this economic plan makes no sense and is hurting the economy terribly. The country is in another recession, inflation is going up anyways, most people are having a bad time. Meanwhile, Milei keeps taking on international debt just to burn through USDs so the dollar price doesn’t go up. All this money gets wasted in capital flight and speculative finances, it doesn’t go into the real economy. And it’s debt Argentinians will have to pay back for years and years. It’s funny, because this debt is what got us into the problem in the first place, so he’s just making it worse. With the US Treasury involvement, it’s even worse, because you can at least default on private creditors, but... what will Argentina do if it can’t pay back the United States? Who knows what kind of concessions the country will have to make? This loss of sovereignty, depending on Scott Bessent like this just to win a midterm election, is probably one of the most irresponsible things an Argentine president has ever done. It’s pretty much treason to the country.
And about the problem being that he lost a local election... well, there is a reason he lost it in the first place. He is doing a disaster. And I'm not even geting into the political and corruption scandals like the Libra scam, Karina Milei (Milei's sister and his biggest advisor) taking bribes from the disability pensions system, Milei's main candidate for the midterm elections (José Luis "el profe" Espert) having to step down after being funded by a drug trafficker, etc.
Anyway, if someone wants extra context or sources, just ask, I can do my best to find them.
An interesting take, given the inflation shit-sandwich Milei was handed, collapsing poverty rates, and the fact that LLA did better in the recent BA election than it did during the presidential race. Unless I'm looking at flawed data, BCRA foreign reserves appear to have been increasing since Milei took office.
I agree the ARS still looks overvalued, and wonder why Milei didn't, or maybe couldn't, let it fall far enough during those first few hyperinflationary months.
In any case, the most liquid market I can find (https://polymarket.com/event/which-party-wins-most-seats-in-argentina-deputies-election) gives LLA 73% odds. From my screen it looks like the ARS is trading beta to LLA's chances.. I guess we will see how much of Argentina's market troubles are due to monetary issues, and how much is simply a crises of confidence started by the fear of a Peronist win.
The latter is a strange scenario. It's like ordering a medium-rare steak at a restaurant, getting something medium-well, then demanding the restaurant take the steak back and serve you cyanide instead. Milei may be far from perfect, but someone like Kicillof is just more of the same money-printing (poison).
It's not as if any sizeable country on the planet has economic policies that make much sense. They're all varying degrees of horrible.
As an American, I support using my tax dollars to shore up confidence in LLA, if indeed this is a crises of confidence. We'd all be better off if Argentina or South America generally adopted markets, and joined the rest of the rich, industrialized world. LLA appears to be the only path forward here.
Trump of course is mostly doing this because he personally likes Milei, likely due to flattery.
> We'd all be better off if Argentina or South America generally adopted markets, and joined the rest of the rich, industrialized world. LLA appears to be the only path forward here.
Interestingly, I just read Amy Chua's World on Fire (same author as the famous Battle Hymn of the Tiger Mom, which is basically a how-to manual for grinding your kids into Harvard), and in it she argues that it's actually a fairly significant mistake to combine Washington Consensus free markets + democracy in most of the world.
Just as Joe Studwell argues there's a "success sequence" for countries to follow that relies on non-free market practices at first (land reform increases ag productivity, then get into manufacturing at the bottom end, install tariffs and protectionism to shelter your nascent industries as they 'git gud' and climb the value ladder via targeting objective standards via "export discipline," and finally you drop the tariffs and just compete on the world stage as a fully developed country like S Korea or Japan), Chua argues there's a massive hidden land mine inherent in going "full free market" and "full democracy" all at once, especially combined in a package. Being fully open to international trade free markets are generally a mistake for developing economies, because they will never "git gud" and won't manufacture stuff relevant on the world stage, instead choosing to buy everything on the world market, and ending up in a place where the only value they provide to the world market is commodities and / or cheap labor. Democracy is also a problem in this regime, because it will end up impeding economic growth, too.
Because after all, what were the founding father's of the USA worried about when it came to democracy? That the great masses of non-property owners wouldn't respect property rights, and would vote themselves largesse at the expense of the people with assets. This is exactly why suffrage was rolled out to a very minimal set of people (white male property owners), and was only gradually expanded over the next ~150 years, to include all white males, black males, and eventually women at different points. We've only had full universal suffrage in the US for around 100 years.
This "problem with democracy" actually happens repeatedly in a lot of the world, because a lot of the world has what she calls "market dominant minorities." For example, all throughout SE Asia, the Chinese are ~1-3% of the population, and ~70% of the entire economies of Philippines, Vietnam, Indonesia, Malaysia, etc. If universal suffrage is suddenly dropped into the ring, a lot of them have used it to expropriate that wealth - and wrecked their economies while doing so.
So dropping full Washington Consensus free markets on developing economies? Bad idea, leaves you trapped in a place where you import everything nice, and leaves you at the bottom of the value chain where you only offer commodities and cheap labor.
Dropping full suffrage democracy on developing countries all at once? Also a bad idea, because it often leads to populists getting in, wrecking the economic engines of entire countries and voting in ultimately self-pwning policies. People in the aggregate often fail the marshmallow test.
Dropping them both together onto an unsuspecting country? Lol. Well, this is why only ~3 countries have graduated from "developing" to "developed" status in the last ~50 years (Ireland, Israel, and Chile).
And just incidentally, much of S and Central America has "market dominant minorities," too - generally composed of hacienderos, large land-owning families descended from Spanish colonial times, and other immigrants, versus the much larger and more predominantly native-American-ancestral populations elsewhere in the country.
Yes, Milei started from a bad position (for example, as you said, he started with little reserves in the Central Bank). That's why he took office with a consensus across society on the need for a stabilization plan. The problem is that after two painful years of austerity, the plan failed and all the people’s effort was wasted.
It is true that poverty is probably somewhat stabilized, because reducing inflation is a really good and progressive measure that helps the poorest sectors the most, but it’s nowhere near “collapsing". There are a lot of methodological flaws in how inflation and poverty are measured. For example, INDEC (the National Institute of Statistics and Census) uses a basket of products and services from 2004 to measure it. There were almost no cell phones or internet back then, and those are the services that have risen the most. Milei refuses to update it. Heck, several important INDEC officials literally resigned rather than go along with the data falsifications, including the Director of Living Conditions Statistics and the Director of the Consumer Price Index. Milei wouldn’t be losing so badly in every election and needing so many bailouts if poverty were really collapsing.
In the recent Bs. As. election, LLA allied with the other major right wing party (“PRO”) got 33% and Peronism 47%. During the presidential race, LLA got 25%, PRO 24%, and Peronism 42%. So, just looking as LLA as the party name did better, but that’s a pretty misleading reading of the situation. In reality, Milei and his allies’ votes were reduced from 50% to 33% and Peronism went from 42% to 47%.
As you show with that prediction market, LLA will win more seats than any other party in this upcoming election, but that’s just because it has to renew far fewer seats. It’s also the only party participating in every province. But there’s no doubt it will end up with a pretty small number of seats overall (for example, the "Pro" party, it's biggest ally, will also lose a lot of seats) and will need to work with other parties to pass legislation.
About why the peso is being overvalued, Milei does it to keep inflation in check. Also, it’s the kind of country he wants: one with cheap imports, no national industries, and a massive concentration of wealth in as few hands as possible. It’s exactly the neoliberal Menem model from the 90s (he’s a big fan of Carlos Menem, in his own words "the best president of Argentina in the last 50 years". You can look at how his decade in power ended, check out what happened in 2001). He could have let the peso float more, buying reserves in the meantime, but that would have meant a slower reduction in inflation. There were no hyperinflationary months, btw.
If this crisis were "simply a crises of confidence started by the fear of a Peronist win", why did Milei need so many rescues even before this election cycle? The economic plan is just unsustainable. That’s been the consensus among economists for months now. Even Domingo Cavallo said so.
You mention Kicillof being a money printing poison. But Milei hasn’t printed any money and is still desperate on the brink of collapsing the country. It’s obvious that the "money printing" narrative repeated by Milei ad infinitum is a lie. But even then, Kicillof is not stupid, society got tired of inflation, and he has no fiscal deficit running the Buenos Aires province. Do you really think he would have done worse job than what Milei is doing if he were in his position? I doubt it. Also, remember that the loss of access to the markets and the last “cepo cambiario” (monetary control measures) were implemented by a "liberal" right wing president, Macri, not by a Peronist. And that the high inflation from 2019 to 2023 is a direct result of Macri’s failure. Argentina had two options during the pandemic, to let people die or to have inflation, there is no other thing you can do without financing.
If you want Argentina to join the rich industrialized world, is better to not support economic measures that has as a literal objective to deindustrialize the country and turn it even more into a primary materials and minerals exporter, with no added value to it's products. Milei is nothing new, what he’s doing has been done through 90% of Argentina’s history with obviously failed results.
IIRC Rob Citrone made a big financial bet on Argentina and asked his buddy Scott Bessant for a bailout. This whole thing is a total disaster.
> This loss of sovereignty,
So the criticism of TRUMP for lending him the money is misplaced, since it sounds like a good deal for America?
Haven't followed this closely , but it could be bad for most Argentinians, bad for most Americans, and good for Trump personally (plus maybe good for some specific people with some vested interest in whatever the money ends up going towards).
Not necessary. Could be a negative sum deal which is bad for both countries.
Argentina being made an US vassal state would certainly be an interesting development. Would be nice to have a secure presence in South America.
It is surprising though that Milei is trying to prop up the currency at unrealistic levels, when one of his big things was supposed to be ending that kind of practice. He promised dollarization after all.
Also, Trump has been as explicit as possible about the loan being about political influence:
https://www.youtube.com/watch?v=V_wlkkiU-6c
I've heard of the $20b loan recently, has that doubled already?
Scott Bessent has said he wants $20 billion from other sources to complement the $20 billion from the U.S. government. I’d suspect the odds of that happening are fairly low. Trump has suggested more money will not be forthcoming if Milei’s party does badly in the Oct. 25 election. If the party does well, the Administration may figure that the promise has accomplished its objective, and see no reason to actually follow through.
https://www.nbcnewyork.com/news/politics/us-trump-administration-working-doubling-aid-argentina-40-billion/6405604/
Why aren't therapy books used more in the psychiatric system? Scott has mentioned in some posts that if you're seeking help for a specific problem/illness (he mentioned depression, anxiety, OCD, maybr borderline?) then reading a therapy self help book seems to have as good results as therapy.
Why isn't it utilized more? I can imagine for many patients1p it also feels less scary than speaking to a human. Not to mention a billion times more cost effective, will take pressure off the (in my country with socialized healthcare, at least) completely overloaded psychiatric system with insane wait times, etc.
Related to this topic - I'm wondering if people know of instances where therapy was actually worthwhile, either personally or for others. My general impression of therapy is pretty negative (it seems like they mostly offer pills and a shoulder to cry on), but I also don't have a lot of experience with it. Does anyone here know of cases where someone went to therapy for a specific problem and the therapy solved or significantly reduced that problem?
Pills for ADHD work great for many people. So there therapy is often just a diagnosis + pills but that's probably the way it should be and often works great for the patients.
I do it all the time, it works much better when I then go through the book with the person.
What do you mean? Psychologists have told me to a read a specific book on more than one occasion. Never actually did it, obviously.
One of the major underlying conditions affecting many people who need therapy is akrasia.
If they were capable of taking disciplined steps to confront and solve their problems on their own, they wouldn't need therapy in the first place.
For many people, having a social obligation and a financial stake makes them actually show up to therapy reliably, whereas buying a book and thinking they should read it doesn't actually make them read it.
This is the sort of the same problem as 'why do we have any treatments for obesity when anyone can solve it with diet and exercise', The people with whatever mix of discipline/energy/low cravings/etc etc etc would allow them to solve it that way have already done so, the people who still need help are the ones unable to accomplish that method who need something else.
I can read self-help books till my ears fall off, but I'm not going to implement any of the changes unless there's a real person standing over me with a whip to force me to do so.
I don't know how responsive that is; the claimed result is "reading a therapy self help book seems to have as good results as therapy".
That's different from "implementing the recommendations in a self help book seems to have as good results as therapy". Medicine is riddled with this problem; most notably, taking a placebo faithfully leads to much better outcomes than taking the same placebo irregularly.
One of Scott's articles for covid made the observation that masks-for-the-public research is more about the intervention of telling people to wear masks than about the intervention of having people wear masks, because you can't do the second thing. You can tell people they should wear masks and see what results you get, so that's what the research measures.
But in this case, that's 'good'; the research is also measuring reading the book more than it's measuring "correct implementation of the book".
Okay, but how do you get them to actually read it?
I can only speak from experience regarding depression, but if you ask a depressed person to do x or y simple thing (excercise, reading self help, eating better etc) that is proven to help alleviate symptoms, they usually will not do it. Metaphorically, they are standing a looking at a cliff. You might be shouting down to them "There's a path up!" but all they hear is " You can just leap to the top!" or "You can just scale the cliff!"
Seeing a therapist is kind of like sending someone down to guide them to the path. It might be arduous, but its easier with a sherpa.
Your comment reminds me of one of my favorite ACX essays of late: https://www.astralcodexten.com/p/singing-the-blues I also agree with Deiseach. I'm currently in physical therapy for a back issue, and the willpower needed to exercise on my own is gargantuan compared to the willpower needed to spend an hour with my physical therapist. Humans are an odd lot.
An important factor is likely that many people just don't really read books anymore, so setting people who are already more likely to have attentional issues reading would probably be fruitless for a large percentage of people.
It also doesn't help that patients can differ significantly in their reading proficiency, so you would probably have to take care to assign books on an individual basis to match patients' abilities.
Re: ACX EEG Entrainment grant. The current sign-up link is broken, the correct link: https://forms.gle/X37zyTV3KhbSb3Ze9
Looks like I messed it up in my update to Scott, sorry. Ah, the irony of messing up a link on the thread 404.
Should be fixed now, sorry.
Thank you so much! And it's definitely my fault — it was wrong in my email.
Scott, I don't know if you've seen the concerns raised about cheating in the review contest. But if you have, and you just ignore them and don't address them head-on and take an explicit position on what you consider fair behaviour and why...I'm seriously considering cancelling my subscription.
I find it difficult to reconcile all these accusations with the author's previous review. Someone's claiming he's a "big time marketing guy" but last year he wrote an 11K word review of Marvel Comics. That's not the action of a cold-blooded review manipulator. That's the action of a geek who gets really obsessed with stuff and likes to share it -- which is exactly the people we expect to write reviews in this contest.
Threatening to cancel is lame. Just actually cancel and select "I don't like the content any more" or maybe "other."
"To the creator: if you don't spend time and effort publicly acknowledging my pet issue, I will stop supporting you' is almost always a silly bit of emotional blackmail that shouldn't be treated with any dignity.
You can raise issues if you think they're important. If the creator/the rest of the audience disagrees, then trying to add this little bit of blackmail to emphasize your point just reeks of entitlement and narcissism.
The New York Times complaints department could have used your support in 2020.
I feel like it doesn’t really matter? It’s an extremely low-stakes review run by one guy. There isn’t even really an easy solution for influential people using that influence to encourage others to vote. Unless there’s a serious attempt at mass manipulation (which I believe there was in the past that was found out) it seems like things have worked pretty well so far.
Regarding the claim that the review contest as a whole is low-stakes, the tweet's text (https://x.com/jliemandt/status/1974984216027738467) is
> The Alpha School virus started here - https://astralcodexten.com/p/your-review-alpha-school
> A few months ago, no one on Twitter had heard of Alpha. Today, it's the most talked about school in the world. Love it or hate it, it is causing people with the power to change education to rethink their priors.
> A lone writer's piece can dent the universe.
> Thanks to AstralCodex and the Alpha parent author.
> Vote for the 2025 non-book review contest: https://astralcodexten.com/p/vote-in-the-2025-non-book-review
So in Liemandt's own public estimate, this review was where the "Alpha School virus started", and he implied that the piece "dent[ed] the universe".
I agree that at this point, who actually won the contest is not too relevant anymore, the real victory was successfully using the ACX platform for viral marketing.
I don't know what cheating allegedly happened, and am a priori skeptical, but "it's a low stake competition" doesn't mean anything. People will cheat for everything, even the most irrelevant things. Even one of the richest guy in the world will cheat to pretend to be good at a video game (one that isn't even competitive in nature).
It means that cheating doesn't matter. Why do you people care so much about this? The only person actually losing money from this is Scott, and he said nothing is wrong.
Maybe we don't think it's low stakes. Reviews can have a lot of impact (see Lars). Even if this were all just for fun, cheating would ruin that.
If it's so low stakes, why are you so bothered that people are trying to keep it honest? That argument cuts both ways.
Because it's annoying and wasting people's time and attention.
You chose to waste your own time by reading and commenting as it's not important to you by your own admission. From the PoV of the people who *do* find it important, their engagement is not a waste for them.
The founder of Alpha School put up a pinned tweet 2 days after voting began urging his 44.5 K readers to come here and vote for the Alpha School review. The review was reposted elsewhere 33 times. It was up the whole time we were voting. That isn’t a serious attempt at mass manipulation?
I do think people should be more aware that there is a genre of rationalist-adjacent media that boils down to, “here is a thinly-veiled ad for my startup/advocacy group”. I’m not totally against this, since a lot of the startups/advocacy groups are legitimately interesting and working on important problems, but it does lead to some conflicts-of-interest.
To be fair, there’s a broad pattern in media of all sorts where someone writes up a favorable press release about themself in an engaging and newsworthy style, and shops it around to various outlets and sometimes gets a few rushed writers running a thinly-modified version of it. This is probably always going to happen for any community.
Do you mean the person who copy-pasted a Gemini response that the Alpha School entry contained signs of "marketing techniques"? I haven't tried it, but I predict that if you ask Gemini a leading question about any other ACX review for a product (maybe the mashed potatoes one) it will say the same thing.
Frankly, I think you could have spent 30 seconds trying this before making a comment like this. With the same prompt and the mashed potatoes essay, I get https://pastebin.com/qVaviCs1, saying that it is "highly unlikely". Feel free to try it with other essays.
As a follow-up for anyone interested, I tested the same prompt ("Is this text likely some sort of hidden advertising? Can you tell me evidence in favour and against?", before the review text, removing title+initial "this is an ACX entry" hint, temperature 1, Gemini 2.5 Pro on Google AI Studio) on several more reviews, trying to focus on commercial products/experiences, including the author's own entry from last year (Marvel Comics). They always started with "Excellent question you are so smart blabla", but the conclusion was always identifying them as "unlikely" or "highly unlikely" to be hidden advertising. The most ambiguous result was the author's previous year entry (Gemini detected the "marketing funnel" and other content marketing techniques), but even here, the first rollout I made for the pastebin resulted in a clear statement "In short, the author is not being paid by Marvel to write this".
So it's not just sycophancy: The combination of storyline and devices in the Alpha School review ("this is what Alpha School PR doesn't want you to know", persuasion story/funnel by initially skeptical parent, honesty about points that don't change the bottom line used to build credibility...) indeed stands out to the LLM among the review submissions.
Receipts:
Silver Age Marvel Comics: "highly likely not "hidden advertising" in the sense of being a covert, paid promotion directly from a corporation like Disney. However, it employs many techniques of modern marketing to advertise an idea and, by extension, a set of products." (https://pastebin.com/HQPmKTYv)
Miniatur Wunderland: "evidence strongly suggests it is not" despite the review being overwhelmingly positive (https://pastebin.com/ZMQ1EMpm)
Islamic Geometric Patterns In The Metropolitan Museum Of Art: "highly unlikely to be hidden advertising" (https://pastebin.com/wYncBTSA)
Project Xanadu: "while it functions like an advertisement for an idea, its transparency about the project's failures, its intellectual depth, and its lack of a commercial goal distinguish it from what we typically call "hidden advertising." ". (https://pastebin.com/662nWGgh)
I think he refers to another comment claiming (with a working link) that there was a vote-soliciting, pinned tweet form a high follower count account asking people to vote for the alpha school review during the voting-phase.
No, I mean that an Alpha School founder tweeted for people to go and vote for the review. Eremolalos elaborates in several comments, e.g.
https://www.astralcodexten.com/p/non-book-review-contest-2025-winners/comment/167386427 and
https://www.astralcodexten.com/p/non-book-review-contest-2025-winners/comment/167540180.
T.T also raises a few related concerns about deception and deliberate marketing, but I only intended to reference the vote stacking allegation.
I'm not sure what the rule here is - it doesn't seem feasible to say that nobody may ever recommend a review. Liemandt has 7000 Twitter followers and his recommendation doesn't seem outside the range of normal things that don't shift the vote too much.
I re-ran the analysis to compare votes before and after Liemandt's tweet. For first-choice votes only (too annoying to rerun the whole thing), the numbers are:
Total voting period: Alpha 15.7%, vs. Ukraine 11.6%
Pre-10/5: Alpha 14.4%, vs. Ukraine 12.6%
Hopefully this will satisfy people that Liemandt's tweet did not change the ranking. I have no solution for the general case except to limit the voting to ACX subscribers, which has some technical disadvantages and doesn't avert the fact that people could just subscribe and then vote. Limiting it to paid subscribers might work, but this would change the nature of the contest and I'm not willing to do it without some issue bigger than a 7,000 follower Twitter account praising one of the entries.
I'm not so bothered that someone got a runner-up prize that maybe they shouldn't have. I'm bothered that Alpha School got months of covert marketing from a secret surrogate. And I'm even more bothered that your contests would be increasingly exploited in this way. The vibe I got from the review when I read it was somewhere between infomercial and TED talk. I had to quit halfway through (yes, I know that technically means I cheated by even voting), and I was surprised it was popular. I'm not looking forward to reading more of that.
So, a serious suggestion for coming years (probably less relevant to book reviews): require any reviewer to disclose their personal and professional connections to the object of the review, *in the review itself*. This would be normal in journalism. And if someone fails to do so, but this is discovered after they're deanonymized, then you have grounds to disqualify them (and whatever other punishments you feel like).
The only possible downside I can see is if someone can't disclose their relationship without effectively deanonymizing themselves, but I think this is a clue they should be writing about something else.
Of course, if the reviewer doesn't ever want to be known, we have no way of knowing about any relationship they're concealing, but I think you can solve that: give reviewers the option of precommitting (in the review) to deanonymizing, and they can put their identity in escrow with you. So we might see a review of some corporate product, note that the reviewer disclaims any link to the product other than what's disclosed in the review, but also they won't commit to being deanonymized, and then vote with appropriate suspicion.
I think this is pretty generous. You can still review the Oort cloud or your parents and nobody will care if you want to stay anonymous.
Is it technically possible to limit to subscribers before a certain date (ie before the call for reviews)?
Also, I think that Liemandt is more centrally described as "a sixfold billionaire whose businesses have been engaging in bot promotion of that school" than "a 7,000 follower Twitter account".
So is what happened with the Alpha School vote actually somehow a smaller or less clearcut case of cheating than whatever happened last year that led you to disqualify someone? It’s hard for me to imagine a larger and more clearcut case the present one. I’m stuck in an ugly fantasy that Nevraumont, Alpha School author, is getting extra courtesy because he Rich and Important and I don’t like having that thought about you, Scott. Can you tell us what happened last year that led to you disqualifying someone? I not asking you to out the person who did it, just make clear what they did.
What incentive would the Alpha School author have to cheat though? $2500 is a good chunk of money that could really make a difference to some people, but doesn't seem like an amount that would make enough of a difference to the author's lifestyle to try and cheat for it. They got all the publicity and whatever else just from having written the review and entering it, winning doesn't really afford them much more other than potentially having a guest post. But after this, I feel like they could just write something and advertise it in the open threads or something to get a decent amount of people talking about it.
Seems extremely plausible that the Alpha founder could have just boosted the entry and told people to vote without any input from the author, in which case, why should the author pay the penalty for that?
I just don't see what the difference to the author would be from what they gained by writing/entering in the first place and actually winning, other than the $1000 that they won.
Or maybe I'm totally misunderstanding, and the pinned post on twitter was for the preliminary votes? Even then though, before that voting stage most people were talking about that review. I don't think there's any way it wouldn't have made it to the finals.
He’s a big time marketing guy whose 2 kids attend the school. He’s a booster for the school. He says he was not paid for writing the review, and did not get a tuition break. Judging by his LinkedIn he is successful and wealthy and does not need to win small sums here or small favors from the school. Seems quite possible his only motivation is to help the school grow. Or it could be he’s angling to become part owner or something.
Please consider the implications of billionaire investors co-opting your personal blog’s review contest into advertising their venture capital equity investments on a level deeper than a person “ever recommending a review”.
This does seem to be a deeper issue that should, in retrospect, have been obvious as a failure mode of “write a product review for any product that isn’t a book and I will post it for you”.
Thank you. That's what I consider an investigation and response. I do think much more discussion is needed: you focus on the number of followers (7000), Eremolalos focuses on the 35,000 views.
Mainly, I think the lack of a clear rule on this isn't fair. It encourages the more brazen types of people, the less overly scrupulous ones, to push and test the boundaries, see what they can get away with. The rules are ambiguous enough that they can go very far and claim to have never violated an explicit rule. But the rules are also ambiguous enough that the more scrupulous and anxious types of people can worry that they are breaking rules and fear punishment from even slightly debateable things. Is this really the dynamic you want in your contests? Do you really want to reward those brazen risk-taking types and punish and disadvantage the scrupulous types? I don't understand why you of all people would want that, and why it's so difficult to make a clear rule on what level of outside engagement is acceptable. Multiple people have asked for such a ruling in the past.
I also think it's extremely relevant whether or not the review author knew of this solicitation. I don't know whether or not you think it is.
But that said, I appreciate your response and I'm sorry for my aggressive comment. I am, I confess, suspicious that you would have responded had I or another said "do you think you could maybe please look into this Scott?" but only you know the answer to that and you can't possibly prove it to me either way.
What rule do you recommend? There is no way for authors to prevent third parties from signal-boosting their work, and if someone does, I will never be able to prove that the author wasn't involved.
I recommend a clear statement of the rule when voting begins. And I recommend that we put in place some barrier to someone’s just coming here in response to a social media exhortation when voting begins. For instance there could be a rule that in order to vote you have to have been a subscriber for a month before voting begins. There has to be some simple way to use a record of subscription date to neutralize votes by very recent subscribers. I think that makes it quite a bit harder to just send a swarm of one’s social media followers here when voting begins. “Sign up now so you can vote for Alpha Dog 30 days hence” is going to be a lot less effective way of driving votes than “Go vote there now, good admirers of Alpha Dog”. And it wouldn’t hurt to have a notice up for people to see when they go to the voting page, asking them not to vote if they have come here only to vote for a single person they know, and have not read other reviews. People sent here via a tweet exhortation may have no idea what they’re doing is ballot stuffing. I think quite a few would leave without voting if they understood we object to votes driven purely by some social media bigmouth’s request.
I suspect that this case involves coordination between the reviewer and Leimandt. (According to discussion in the other thread, the reviewer is a professional marketer and has stated that he has done pro bono marketing consultation for Alpha School.)
So I suggest something like: "No self promotion or coordinating with others to promote your review."
It's hard to *prove* in various cases, but perhaps at least stating it will prevent the most egregious examples..
It would be a bit presumptuous of me to say what *the* rule should be; I'm not running the contest, I didn't expect to have that power, and I'd have to think about it for a while.
My argument is that there should be *a* rule, something to work with, so we're all on the same page, a common knowledge coordination point, etc etc.
To that effect, I'd consider any of the following to be sufficient:
"No promoting your review outside this blog, no bringing in friends or other outsiders to vote for you. I may not be able to always prove this right away, but if I do find evidence I will disqualify, even retroactively if necessary."
"If someone else promotes your review, I will hold you responsible, so you'd better do all you can you discourage it."
"You can tell a few friends and/or mention it on other platforms, but if there end up being enough outside votes to substantially shift the result, you've crossed the line."
"Outside promotion is pretty much fine. If you are able to get a massive audience to brigade votes from, you've earned your success."
"Basically, if I can prove you solicited votes, you're disqualified. Either keep it a low enough level that I won't notice, or be really good at hiding it."
"The only rule is don't piss me off. I'll decide what I'll allow and what I won't, after it's happened. So yes you need to make an educated guess at reading my mind."
You might say you're already applying the last one. I submit there's a universe of difference between applying a rule like that explicitly, with common knowledge that that *is* the rule, and applying it implicitly, where it's just an aproximation of the practical result of the rules being unclear. In the first case someone who crosses a line can expect a response like "well you tried to read the judge's mind like we all were and you failed, better luck time"; in the second case they can expect something more like "you disgusting cheat, you should have known what the rules were". Compare a game like Poker where bluffing and reading secret information is part of the game vs a game like (I think?) tournament Bridge, where it's considered cheating. Compare a sport where some action is explicitly allowed as long as the referee doesn't see it vs a sport where it's clearly against the rules, but lots of people secretly do it anyway. Compare a dating context where it's explicitly understood that guessing what will cause discomfort or not is the entire aim vs one where that which causes discomfort was "obviously" just objectively wrong and there's something wrong with you if you didn't realise that.
I have very strong feelings about this whole category of fairness and social honesty, which is my answer to anyone who asks the (incomprehensible to me but amazingly common even around here) question of "why do you care so much?"
Who do you think cheated in the not a book review contest and in what way? I tried looking through comments but couldn’t find the accusation, so it may have been buried
Claims that the founder or co-founder or guy associated with Alpha School linked to the review and urged everyone to go vote for it. That is expanding beyond the pool of ACX readers and the complaint, so far as I understand it, is that a lot of outside votes would come in and artificially inflate the result; ballot box stuffing, if you will.
Thanks, that is concerning. I didn’t even vote this year because I only had time to read half the reviews and thought that would be unfair. It’s likely these voters didn’t read any of the other reviews.
Yeah, that's what is making people mad. The outside votes just came in (if they did come in), voted for Alpha School, and that was it. It wasn't about the review contest, it was boosting and publicity for the school.
I think Scott is right that you can't really stop third parties from doing this, but you can (maybe) make it a rule that reviewers can't shill for votes in that naked a way. If they link to their review, the people coming in should vote for other reviews as well.
I think it would be useful to have context on how many votes we're talking about here. My guess based on the stats Eremolalos cited for the promotion tweet (45.2k views and 32 retweets) is that it probably generated at most a few dozen votes for the Alpha School review. In general, very few views of any sort lead to action of any sort. For a nontrivial action like this one (follow link -> find and follow another link to the voting form -> fill out and submit form), I would expect much, much less than a 1% conversion rate. Probably not much more than the retweet rate.
Scott presumably knows how many votes were cast total, but we don't. I'd guess on the order of 5k votes, since that seems to be roughly how many people fill out other forms Scott has posted (~6k responses on this year's annual reader survey, ~4k responses on the AI Psychosis survey from this past August). If the promotion tweet drove 50 extra votes to the Alpha School review, that would have inflated its standing by about one percentage point. So Scott's finding that Alpha School did about 1.3 points better in voting after the tweet relative to before (which likely also includes a fair amount of noise in either direction) is in line with my expectations.
A small number of votes isn't that valuable in the final round, but it would be very valuable in the initial shortlisting where votes are spread far more thinly.
If someone will resort to publicly shilling for votes in the final round, is it likely that they'd also resort to more quietly shilling for votes in the shortlisting round?
From a "viral marketing" point of view it doesn't matter that much whether or not the review wins, but it matters hugely whether or not it gets onto the shortlist.
Caveat: speculating, but it might not be many votes. At the same time, the contest might flip on only a few votes, so if someone tells thousands of people on some forum that some piece is up for a review award and a dozen of them are thus motivated to plod over and cast an actual vote, maybe that still makes a critical difference and is worth doing if winning the award is the goal.
It occurs to me that people could right the ship to some extent regardless of the numbers, by posting to other relevant fora and reminding people that some institution or writer they care about is featured in a review contest and they might want to vote. And so a dozen or however many people plod over and do the thing and the "fan" votes mostly cancel each other out.
That said, I've only thought about this for maybe five minutes and maybe this is as likely to harm as to help; OTOH, maybe someone else can work out the kinks.
It was a fairly well commented on in the thread.
This is kind of the last straw for me. Many comments have been left over several years of book review contests politely arguing that certain things can be improved or are unfair in some way, or asking for you to clarify what is acceptable and what isn't when ethical intuitions diverge. My impression is that almost all of these have been ignored. Not even a one line response saying you've taken note of the concern and are giving it serious consideration.
I also think you've been very unwilling to say anything critical of EA/rationalism for a long time. I've found that very hypocritical when the first essay of yours I read and the one that made me love your writing was I Can Tolerate Anything Except the Outgroup with its emphasis on the importance of criticising one's own "tribe".
So these are two different issues I have with you...not engaging with your own comment section's concerns and objections, and being unwilling to criticise your own tribe. This issue seems to combine them. Several people are angry about the apparent vote stacking and think it's blatantly disqualifying and want a clear response from you. And such a response involves taking a potentially critical stance on a possibly rationalist-adjacent institution (maybe the Alpha School isn't rationalist adjacent, I don't know, but it seems to have overlap in ideology and according to some info provided in a comment on the Review Winners post it has shady connections with a number of other actors).
So even if each of those former concerns is inaccurate or can be excused...this one I think absolutely demands a response.
This comment is aggressive but my experience is that nicer ones get reliably ignored.
For what it's worth, if it were up to me I would have specifically ignored this comment because it was too aggressive, but enough people brought it to my attention that I am responding to it just to satisfy them. I have asked them not to bring aggressive comments to my attention in the future.
The comment is aggressive is the way any criticism is aggressive, in which it calls someone for something wrong, and that is not nice to be told you have done something wrong.
Except that, how do you find the comment aggressive? Please enlighten us, so we could improve in the future.
That's fair and I don't want to encourage aggressive comments and I am upset about the way they tend to more reliably generate engagement.
But I'm also upset when I write a long and thought-out comment politely objecting to something, full of disclaimers and reassurances that I don't mean to insult and including clarifying questions instead of accusations and so on...and just get ignored. I feel like this has happened many times in several years of book review threads, as well is in other contexts, but I admit my feeling could be wrong.
Getting emotionally worked up about people not responding to you on the internet is bizarre. It's common, I'm aware, and the rise of social media has lead to a drastic increase in one sided para-social relationships. But that doesn't make it normal or healthy.
I truly mean this respectfully - try talking to people in real life. Go to meet up groups, go to an ACX meetup, find people who care what you have to say. I think you'll find a lot of people are actually just normal people who have their own ideas and are quite willing to hear out and discuss interesting ideas.
I second this!
> I also think you've been very unwilling to say anything critical of EA/rationalism for a long time.
Very different perception here. For example, there was recently a review of If Anyone Builds It Everyone Dies which very explicitly criticised Yudkowsky et al's approach to AI doomerism. If this isn't an example of criticising your own tribe, then what is? Surely you don't expect someone who is part of a movement/group to periodically criticise _all_ of said movement/group just to show they can?
Very different perception here, to borrow a phrase. The review was way too deferential toward the authors, and even upon a brief realization that maybe Y. isn't interested in addressing even mildly challenging questions* Scott keeps treating him as some kind of a sage**.
And then swiftly bans people for mildly irritated comments toward Y. in this thread***, while completely ignoring a deteriorating clownshow that open threads tend to descend to with little help of shitposting Wimblies.
* https://www.astralcodexten.com/p/book-review-if-anyone-builds-it-everyone/comment/155029071
** https://www.astralcodexten.com/p/book-review-if-anyone-builds-it-everyone/comment/155114676
*** https://www.astralcodexten.com/p/book-review-if-anyone-builds-it-everyone/comment/154935672
I don't think there's anything inherently wrong with deference. Especially when you're being open about it and basically writing a fairly critical piece and just concluding with "but this guy's long shots have paid off in the past, so maybe they will again because of reasons I can't see".
I don't have the time to be assessing how irritating comments are. I'm not disbelieving you, but that sounds beside the original point so I'll save time by not looking into that as it seems less relevant to me.
Nothing inherently wrong, sure. Just in this case where the object of deference has proven himself to be... a (insert bannable adjectives here) fool, maybe it's time to dial the deference down a notch or two, to be able to to maintain a sense of intellectual integrity.
Speaking as someone who has at times felt unfairly attacked/belittled by the tone and what I'd argue are blind spots in some of Yudkowsky's writing, I don't think there's an objective way in which you can say he has proven to be that.
Wow. OOD got banned for that comment in the third link?
I wasn't going to post this, but this seems like an apropos place to drop it in. Émile P. Torres makes some of the criticisms that OOD made, but with supporting quotes and Xeets from Yudkowsky. Having never paid much attention to EY, I admit I experienced some moments of moral judginess reading some his quoted remarks. And I didn't know Bentham had already called out many of Yudkowsky's epistemic errors in detail back in 2023...
https://www.realtimetechpocalypse.com/p/eliezer-yudkowskys-long-history-of
https://benthams.substack.com/p/eliezer-yudkowsky-is-frequently-confidently
"Wow. OOD got banned for that comment in the third link?"
Yep.
Re. Y. - I started to read his posts on Less Wrong a few years back, and distinctly remember the sense of... disappointment... like I expected so much and there was nothing there, like someone was pulling my leg. Over time I concluded the guy was full of shite, and nothing has dissuaded me from this position since.
Oh, that Bentham's piece is a doozy! In particular, it brings to light Scott's uncritical deference to Ydkowsky:
Scott: "I think Eliezer believes this because he's spent years working on decision theory and thinks there are actual theorems you can use to coordinate"
Bentham: "Yudkowsky’s decision theory is totally bankrupt and implausible, in ways that are evident to those who know about decision theory"
Ouch!
It's mostly about not criticising elements of the lifestyle and subculture and ideology. It's merely my impression and I could be misremembering or wrong. I also could be holding Scott to an unreasonably high standard. And in this context I was trying to be confrontational to make sure the main complaint couldn't be easily disregarded.
But I came to follow Scott primarily because he seemed to be one of the only people who valued truth intrinsically and who was willing to openly criticise things like wokeness and cancel culture at its peak while being left-leaning himself. He seemed to be one of the only people who would criticise his own side. So I'm very sensitive to signs he may be actually not living up to this principle when it comes to his "real" tribe.
For one example, I talked a bit about it here https://www.astralcodexten.com/p/hidden-open-thread-3925/comment/141041693 which was about a double standard to epistemic arrogance and the principle of charity.
Your link is for subscribers only so can't comment on the specifics. But it's sounding from the outside less about unwillingness to criticise his tribe than about unwillingness to criticise specific aspects of his tribe with which you personally disagree.
All right I'll paste it here. Scott has replied to my original grievance, so I don't want to attack him any further. But since I'm being challenged on this point, and Scott also challenged me on my current tone being too aggressive, I feel the need to defend myself.
Maybe the following is also aggressive, but I thought it was polite but direct. I'm raising it only to clarify part of the general perspective I have, rather than as a challenge or attack or demand for engagement:
"Scott...I mean the following as politely as possible. But I think you've got an extreme bias, a blatant double standard, that you are, dare I say, a hypocrite here. And I haven't ever seen you address this. (Of course, that doesn't mean you didn't.)
Two parts to my problem...
1. I think that's a good essay. By which I mean, it makes some criticisms in a striking way that I *really want to see rationalists address*. And that, I feel I have very rarely seen them address, in any kind of direct way. The prima facie arrogance of taking utilitarianism as somehow obvious when ethical theory is a notoriously difficult problem and only a distinct minority of professional philosophers think it's logically defensible. The obsession with using rationality to increase one's status and "win" instead of advancing truth for its own sake. The simultaneous treatment of AI as an existential risk and also as a path to utopia, both positions routinely held by the same people. The open contempt for centuries of philosophy and the unironic belief that it can all be ignored because one guy in the late 2000s figured out all important truths from scratch in a fanfic amd some blog posts. The creepy conflation of "people who want to think more accurately" and "people who live in the Bay Area and love orgies" with very little apparent embarrassment at this widespread conflation.
I accept that not all of these are fair descriptions. In fact, my prior is very high that some (maybe even all) of them are inaccurate or irrational accusations. But I don't actually really know what the response to them is supposed to be, because very rarely do I see any rationalists actually confronting the criticisms head-on and making a full-throated defence of their movement. You've sort-of done it a few times (e.g. the consequentialism FAQ, the tower of assumptions) but not nearly as often as I think ought to be done by someone who mostly treats all the above things as presumptively reasonable and good, despite their plethora of obvious serious objections, and who is otherwise very careful at pointing out the problems and blind spots with other movements and ideologies.
And mostly when these things are brought up in the comments I see a lot of vague deflection, a lot of detailed nitpicking, a lot of lazy linking to The Sequences, and a lot of irritation. The vibe I get (which I could of course be getting completely wrong, I totally accept that) is that there's very little interest among most rationalists in seriously relitigating the foundations of their movement and ideology. Which, if actually true and unless I'm missing something, seems to almost utterly invalid the entire movement on its own terms.
(And even if I'm wrong and there's an enthusiastic willingness to keep debating the foundations, which there should be for any movement that cares primarily about truth, there are just a lot of important facets that are routinely ignored. Kriss is the first one in ages who I've seen, in discussions of utilitarianism and EA, bring up the really basic point of the Phil Papers survey results, for example.)
2. Despite all that, I'd completely understand why you might see that Kriss essay as a bad-faith, overly mocking hit piece. It lacks charity, it's too snarky, it makes broad sweeping claims, it condemns a belief system without doing enough to understand and engage with its real claims, it sounds very arrogant. This is all objectionable and I'd completely agree with your "don't feed the trolls" dismissal except that...
Elieizer Yudowsky writes in the *exact same way*. Over and over. The immensely arrogant tone. The wholesale vaguely-mocking or outright-mocking dismissal of whole categories of thought (theism, Cartesian Rationalism, dualism, non-consequentialism to name a few) with little substantive engagement with their claims and arguments. The near-total lack of certain basic virtues (he's one of the only people I've ever seen unironically call themselves a genuis; even most of the vilest partisans and fanatics have more humility than him and know that decent people don't talk like that). He writes in this style all the time.
*You* don't! I can never imagine you writing something with the level of uncharity and mocking arrogance that both Kriss and Yudowsky display. But I've never seen you once say a word against the latter. Even when you've been asked point-blank by multiple people, including me, whether you agree that he's unacceptably arrogant (among other things), it's always silence. Every time.
And yet you're happy to call Kriss a troll in a one-line putdown.
So I don't see how there's any explanation other than that you are living by a principle of "different rules for my friends and other people". Loyalty above all else, truth and fairness be damned.
All in all, one of the following is true: (a) you *have* repudiated Yudowsky and I've missed it, or (b) you need to reread your own Outgroup essay and take the part at the end about your own tribe more seriously.
Sorry for the pointed criticism, I really love your writing and find you overwhelmingly truthful and charitable. But I think you have a major blind spot. And the former is precisely the reason the latter is so jarring."
...
As for this: "But it's sounding from the outside less about unwillingness to criticise his tribe than about unwillingness to criticise specific aspects of his tribe with which you personally disagree."
Obviously I don't think that's the case. I think it's about unwillingness to do the same kind of internal criticism that he wanted and demanded from the woke left back in the SSC days, and which I applauded him so much for demanding.
I think disagreeing gently about a factual matter is the sort of thing even tribalists can do. From my perspective I want the same person who rightly wanted reasonable feminists to clearly and publically say "you are a cruel and hateful bully who has no interest in truth or dialogue with those different from you; stop it!" to the likes of Amanda Marcotte to himself clearly and publically say "you speak like an obnoxious and egotistical asshole and wannabe cult leader who has no interest in charity towards those you disagree with; stop it!" to the likes of Yudkowsky.
1. Thanks for pasting the comment. Even without the context of the post it came with, that's useful.
2. Boy is that long! What have I got myself into?
3. Your overall tone is definitely hostile, especially at the beginning of the comment and when using morally charged words like "orgies" (which must feel especially insulting to a member of this culture who has shared and justified his poly lifestyle at length and who has also on several occasions written about how detached this is from sex in his [ace-ish] case at least).
4. I don't think it's reasonable to expect someone who is very much at the centre of a movement to have to constantly justify every single foundational premise of said movement, even if this someone has on a number of occasions taken the initiative to engage in "apologetics" (which Scott has, off the top of my head: on consequentialism, on polyamorism, sort of on EA, on his version of AI doomerism, on organ donation, etc). If I go read a substack by Muslim I'm not going to be demanding a defence of the origins of the Qur'an, for example. Even if this Muslim has previously engaged in apologetics on some part of their religion (say the concept of Jihad) it would be weird to demand an extension of that to the entire religion. At some point we need to accept different people operate based on different fundamental axioms, and it's good enough if people are transparent about which those are so we know where they are coming from.
5. I always find it weird when someone is already criticising a fellow tribesperson but gets criticised for not doing it more emphatically. It's one thing to demand clarity on where their positions differ and why. It's quite another to demand a wholesale rejection of an individual over differences of opinion.
6. Sorry, yours was a very long comment. This is the best reply I can do with the time I have.
A part of that critique is that Scott called out some people for being assholes but he refuses to call out his friends for being assholes.
If true then I find that defensible. I don't believe that (truthfully, correctly) calling out somebody means you now have to (truthfully, correctly) call out everybody else who is the same sort of asshole (assuming that's the case here).
Otherwise I end up in the situation where I can never call out anyone because I know that if my loved one did the same I wouldn't call them out and would be hypocrisy.
The point where I start finding this behavior objectionable is when I call out somebody for something but then defend my friends when they do the same.
> The obsession with using rationality to increase one's status and "win" instead of advancing truth for its own sake.
My experience seems to be the opposite. It's often the people who found the rationalist community yesterday, saying things like: "But wouldn't it be more rational to self-deceive, if that helps you win?" and the old members saying that the temptation is obvious, but it wouldn't work in long term.
> The simultaneous treatment of AI as an existential risk and also as a path to utopia, both positions routinely held by the same people.
How is this fundamentally different from a caveman saying that fire can burn down your house, but can also cook food? A thing can be dangerous and possibly beneficial at the same time.
> The open contempt for centuries of philosophy and the unironic belief that it can all be ignored because one guy in the late 2000s figured out all important truths from scratch in a fanfic amd some blog posts.
This is a complete strawman.
> there's very little interest among most rationalists in seriously relitigating the foundations of their movement and ideology.
Seems like from your perspective, rationality means doubting the same things over and over again, and never moving forward. Like, a perfect mathematician would keep asking themselves "but can I be *really* sure that 2+2 actually equals 4?" for a millionth time, and therefore could never proceeds towards multiplication. This is performative skepticism, not rationality.
Eh, criticising the lifestyle is tough, that veers on moral judgement and moral judginess isn't part of EA (apart from "Drowning Child you must agree or be a moral monster" type judgements. And shrimp. We can't be mean to the liddle shrimpies!)
"bunch of weirdoes should not be weird" is going to be a tough message.
Um. A lot of moral judginess happens on these open threads. Not that I'm complaining. But EAers are deluding themselves if they don't think they're morally judgmental as a group. Of course, it might be argued that "no true EAer (Scotsman) should be morally judgmental".
This old thread https://slatestarcodex.com/2017/04/07/yes-we-have-noticed-the-skulls/#comment-484685 has claims about women being pressured into having abortions. This goes so far beyond "moral judginess" and "weirdos" to reach "something that even people who don't agree about anything agree is horrifically harmful and evil".
Maybe that sort of thing hardly ever happens; it's still incredibly disturbing that a community where that can happen and be tolerated doesn't get a lot more criticism.
Yeah, that one is difficult. Everyone by now knows my position on abortion. So where's the line between criticism and "inflicting your beliefs on others"? If a set of people think abortion is permissible, should be legal, and is morally neutral to even beneficial, then they're going to support abortion. And if it's "you went into this and made a contract and agreed to terms, then unilaterally changed them", most times people will say contracts should be kept.
I have my views on this, but I'm not EA and I'm not a rationalist so I can't tell people who are that "you should do this by my morals and not your own". Well, I can, but I can't expect them to listen to me.
A finalist was disqualified last year due to vote manipulation. But that case was relatively easy to spot, because the reviewer was from a geographical area where not a lot of ACX readers live, and the initial voting form responses giving the review a 10/10 were all from this same geographical area. (Meanwhile, the other votes from elsewhere rated it much lower.) This review was disqualified before it was posted to ACX as a public-facing finalist.
So if the person’s vote manipulation led to their being disqualified, in what way is the Alpha School manipulation different? I get that the actual exhortation to vote came from someone other than the review’s author, but the author has not denied knowing that the review was up, and there is excellent indirect evidence that he shared info with Lielandt, the tweeter, about content of review and date on which voting began. If I asked a friend of mine with a popular blog to write a post about me, and the post included an exhortation for readers to come here and vote for my review, and I knew the friend had done that, and I had told him when voting began — would I then not be guilty of cheating and vote manipulation?
1. I just want to wish everyone well.
2. I want to recommend Joseph Heath's blog and specifically his last post on Populism.
https://josephheath.substack.com/p/populism-fast-and-slow
Really good post!
[Error 404: comment not found]
[Error 418: I'm a teapot]
[Setting a reminder to repeat this joke on Dec 29th]
Error 404: Blog not found (substack.com was down with error page for me earlier today though its own status page brazenly denied it, possibly related to the general AWS outage?)
Strikes me that if Alexander signs up for the (un)conference, he could find some guinea pigs (I mean volunteers) for the flashing lights experiment there!
Since it's being hosted in a place called Newspeak House, which makes me wonder *very* strongly what the heck was going through the minds of the founders of "the London College of Political Technology" (why are you naming your headquarters out of "1984" by Orwell?), it seems like the perfect place for "we just want to see if we can alter your brain, trust me it's harmless" experiments.
Sasha absolutely was there! He demoed his experiment.
Excellent! So the brainwashing - ahem, of course we mean innovative educational aids - went well then? 😁
I think so, not that I understand the experiment.
"Political Technology" sounds like newspeak for "Newspeak."
I have to wonder why they called it that. If it's intentional, blimey! If it's not intentional, then they need to look up what Orwellian Newspeak means:
https://en.wikipedia.org/wiki/Newspeak
"To meet the ideological requirements of Ingsoc (English Socialism) in Oceania, the Party created Newspeak, which is a controlled language of simplified grammar and limited vocabulary designed to limit a person's ability for critical thinking. The Newspeak language thus limits the person's ability to articulate and communicate abstract concepts, such as personal identity, self-expression, and free will, which are thoughtcrimes, acts of personal independence that contradict the ideological orthodoxy of Ingsoc collectivism."
Quick, go say that on the new Open Thread so Scott has a chance of seeing it!
But prepare to be disappointed. If he hasn't got around to publicly smacking this comment (https://www.astralcodexten.com/p/open-thread-399/comment/156482260), there's probably not much you can say to get any attention.
Since you're already dead, what was the comment that got you banned?
I believe it was "Shankar is a poopyhead."
We are not a free thought zone, except in a limited way — people are free to post highly unpopular ideas, so long as they make a decent case for them. On a recent open thread I put up a post arguing that it was absurd and useless to say ‘the n-word’ instead of ‘nigger,’ and in that post I of course used the word ‘nigger.’ You will not find many places where a post like that is tolerated, outside of angry forums where people use the word to express scorn and anger towards blacks. On the other hand, moderation is strict when it comes to rudeness, and even when it comes to rudeness there’s an exception: If you’re rude, you better be right. That is, if you made a great point, and back it up with great evidence, then ok, you can be somewhat rude about it. I think I’ve earned the right to be rude to you. You sound like an arrogant undergrad age asshole to me. Now I’m off to mute you, as I routinely do people who put up posts like yours. (Normally I do that without responding to the person. I’m doing this with you as a public service.) So by all means, come back with some new Identities. Wear your little fingers out doing it. I and lots of others will then instantly mute you and you will be shooting turds into the void. Bye.
I'm totally fine with policing the explicit name calling and rudeness. Not so keen on the fact that Scott bans archness and condescension. It's a bit much.
I have an amazing proposal, I even hesitate to say this it’s so out there, but I’ll be brave and propose it anyway: how about… I mean, I’m sorry if it’s too radical! So, how about NOT throwing stupid juvenile insults around?
Nah, must be impossible.
As you were.
While i agree with your broader point of not throwing around infflammatory insults. This comment as a sort of snipe isnt that much better
I don’t expect it to make any difference, frankly. Sometimes just need to get things off your chest.
I got banned for something that seemed trivial to me but it's Scott's blog. Someone doesn't like the comment, they flag it up, and Scott or the mods apply the test - is it necessary, true or kind? These are often subjective judgements but it seems like an okay test.
In the same way that "a man who must say "I am the king" is no king," if you feel the need to explain to people how good you are at arguing, you probably aren't good at arguing.
*Shrug*
As my father used to say, locks only serve to keep mostly-honest people honest. Not a lot can be done to keep out someone who *really* wants to get in. Perfect permanent banning security would require a level of effort and diligence that would probably just kill the community anyway.
So bans that impose mild friction on the bannee to make them go through the rigamarole of getting a whole new email just for a new account probably suffice to incentivize someone on the wrong end of a ban to either (a) go away, or (b) behave better when they come back with their fancy new account. Either solution will work.
I suppose there's a certain kind of trashy lowest-denominator where a very small minority of people would care enough to keep coming back with new sock puppets over and over again, misbehaving and getting banned every time, and be willing to engage in that behavior literally forever. The system, as you observe, doesn't solve for that. But how often would such a person appear, and how much would it cost the board to solve for them? Even the most hardened internet jackass would most likely lose interest or reform their conduct eventually, and if there really were a person willing to full on keep running that ban-new account-ban treadmill *forever*, at a certain point you have to go from being annoyed to just feeling sorry for them. Nobody wants that guy here, but we'd want to *be* that guy even less.
If democracy ever dies, its epitaph is likely to be "everyone valued, above all else, enjoying one's liberties".
You are very long-winded, and I mean that in the nicest possible way.
I suppose I could call you a windbag, or a chatterbox, to test the limits of true kind and necessary.
Seeing as you have mastered the skill of organizing your thoughts, you might give some attention to the skill of distilling them.
That’s an interesting distinction to make on a written forum.
Wrong thread, substack does this sometimes - this was for archeon I believe
What is your motivation for the change?
I don't have any experience in these fields, but that piece might be useful to frame other responses.
The current job environment for bioinformatics is not great. NIH funding is stalled, most universities have paused or greatly reduced hiring for research roles, and the future is uncertain. I know less about industry but I think the job market there is also much worse than it was in the 0% interest rate environment. This is a US-centric view; I don't know about the rest of the world.
I worked in academia for the past 10 years and recently transitioned into the public service. My background is maths, currently studying economics part time. It sounds like you have transferable skills - maybe you just need to move to a new industry, without necessarily also learning a whole new skill set.
My familiarity with bioinformatics is mostly through academic research, where you would need a PhD to be accepted as a credible researcher. It might be different in industry. From chatting to people working in the area there seems to be expectation that AI will have a big impact in the field. If getting into an area experiencing big changes sounds appealing to you it might be a good idea; but there is a risk that lots of established people and labs get left behind by developments. You don't want to reskill into an area which is dying out.