Hitler was so loathsome and destructive that I wonder if it's even possible to have a comprehensively accurate picture of the man in 2025. The wounds are just too fresh. Any ordinary, decent person is going to have a strong predisposition to believe the most unflattering things about Hitler's private life, even if they are consciously honest.
And few people have any incentive to correct the record when someone repeats a lurid rumor. That's making yourself a target for little appreciable gain.
I hate Nazism, to be sure, and I could have used any number of examples, like Mao or Charles Manson. Our ability to understand a loathsome historical figure seems inversely proportionate to how RELEVANT that understanding is. What a tragedy.
Of all the epistemic traps we may find ourselves in, this one does not seem to be particularly relevant? I mean, it would be nice if we did not have a systemic bias in judging how Hitler treated his dog, but the stuff for which he remains relevant today is generally not his private life.
It is not like he wrote a book on caring for dogs and society is missing out on his unique insights as they unfairly reject the book based on his politics.
Can you rephrase this in different language? I'm a little confused. First of all you say this epistemic trap is" not relevant." Not relevant to what? I'm not sure what his dog has to do with this. And he was such an important figure in the history of the 20th century, it seems to me that every aspect of his biography is relatively important, or at least potentially important.
After all, the exact reason for Hitler's rise to power, and the exact nature of his personal dysfunctions, are some of the most enduring mysteries in modern history.
Are you implying "Hitler has zero redeeming qualities, so it doesn't matter if we describe his life inaccurately?” whether or not he had qualities that are “redeeming,” we definitely want to be as truthful as possible about an important historical figure.
We want to stick to verifiable facts, don't you think? Besides violating whatever inner commitment we have to being truthful and accurate, spreading questionable rumors damages our credibility.
> After all, the exact reason for Hitler's rise to power, and the exact nature of his personal dysfunctions, are some of the most enduring mysteries in modern history.
I disagree with that. I see Hitler's rise to power as an institutional failure, not something which only he could do. I sincerely believe that a full dissection of his psyche would not yield any answers to explain his rise to power. He did not have an Illithid tadpole which enabled his oration and drove him to murder.
There is always going to be a pool of fucked up, murderous people wanting to become dictator of the world. Hitler is hardly unique in that regard, just because circumstances made his bid more dangerous and bloody than for most others.
Studying how dictators work in general, and what motivates them (as Scott does with his dictator's book club) might be interesting, but we should not assign a higher weight to Hitler just because he managed to murder more than most.
"The scoundrel on a small scale should not, if those in power allow him to become a scoundrel on a large scale, occupy a special position not only in scoundrelship, but also in
There is a tendency, especially to Germany, to see Hitler as evil incarnate, a certain fetishization of the most minute details of his life. I understand the motive: the more Hitler is the antichrist, an anomaly on the scale of Foundation's mule, the more blameless the Germans are for falling victim to his seduction.
The truth is that human evil is utterly banal. We did not fall for a force of nature, we fell for a rather ordinary human (and his equally ordinary to pathetic cronies), perhaps in the 95th percentile of murderousness. Studying his sex life (or whatever) will not yield greater insights into WW2 than studying the sex life of any other person. (As a corollary, this also means that looking at what makes low level grunts participate in genocides much more interesting. See Goldhagen: Hitler's Willing Executors, which despite the title does not focus on Hitler at all.)
I agree with you that spreading falsehoods is generally bad, and that we should strive to avoid it. I just don't think that the number of Hitler's balls has more practical relevance than what type of damage a specific Pokemon does.
Sure, in the same way that the only way to determine that your neighbor is not secretly a Mossad agent is to investigate him.
Hitler was rhetorically excellent and an especially murderous antisemite, and had an unhealthy risk appetite. He was also very lucky to find himself in the position to make his career. I do not think one needs to know more about him to understand WW2.
Also, he has been investigated as much as a person can be investigated. Every aspect of his life has been subject to books and movies to a point which places him in the top 100 persons-as-book-subjects. I seriously doubt that any further detailed evidence unearthed by someone writing about his use of the color blue in his painting career or his relationship with dogs or whatever will revolutionize our understanding of the Shoa.
In my analogy, once you have found that your neighbor is out most nights is because he has a job as a night guard, and that the antenna in his backyard is because he is a ham radio enthusiast, and have checked his job and his radio transmissions, you will come to the point where there is little left requiring an explanation. Sure, it could still be that he is a Mossad agent, and that digging deeper will reveal that fact eventually, but realistically you are just wasting time with an obsession.
You assert the very points at question. Your claims about his character have to be established as well as your neighbor's night-watchman job, and you have not done so.
You may be doing the exact thing that I worried about in my original post. Perhaps you are so determined to denigrate Hitler's reputation, such as it is, that you want to deny him any uniqueness.
Whatever self-interested motives modern Germans may have, Hitler's political career was weird, singular and uncanny. The disconnect between his childish ideology and his meteoric rise DOES indeed constitute a major historical mystery.
The Nazi Party's commitment to the mass killing of civilians, without any credible military justification, was very unique at the time. Even among right-wing authoritarians. Though Hitler's example was influential on the right, his brand of evil remains fairly unique to this day
To compare Hitler to an ordinary dictator is not only unfair to his victims. It fosters an environment where it is normal to call any old right-wing authoritarians "Nazis." An act that makes people looked dishonest and unserious outside of certain circles. A rhetorical act of "crying wolf" that impedes and frustrates the mission to crush rightist authoritarianism forevermore. Or perhaps you want to obscure the foundationally right-wing nature of the Nazi catastrophe? But I really shouldn't be speculating so freely. Whatever your motivations were, I don't think your comment here was very helpful.
I think that you give Hitler personally too much credit.
For the most part, he did not invent the Nazi ideology from the scratch. It was very much a collaborative effort.
If he had been killed by a grenade in 1917, I am not convinced that history would have gone very differently. There was a niche for a quasi-revolutionary far right party in the Weimar system, and if the NSDAP had not taken it, some other rabidly antisemitic, ultranationalist party would have popped up in its place.
Antisemitism is very much an active ingredient here. It had a long history. For the NSDAP, the Jew represents, among other things (like unfriendly foreign powers) everything which is wrong with capitalism (which was plenty, in 1920s and 1930s Germany). Take that away, and you lose half the votes.
If you have a movement whose core tenet is that the Jews are what is wrong with the world, and which is not traditional monarchist but considers itself revolutionary, unbound by traditional moral limitations, and which is willing to shed blood to create the new order, things will turn ugly.
I would claim that there is a German cultural tendency to follow their beliefs to its logical conclusion. I mean, the eugenics / racial genetic hygiene aspect of the NSDAP was much less load-bearing than the antisemitism, and still they took it to the logical conclusion to fatally poison disabled kids with exhaust gas (Aktion T4). Not because it was personal pet peeve of Hitler, but simply because *the people in the NSDAP actually believed their own ideology*. If you are a doctor or nurse and believe in the Nazi ideology where the people are everything and the individual is nothing, and where genetic defects are a thread to the collective health of the people, and you are not bound by traditional (e.g. Christian) morality and smart and able to think outside the box, chances are that you will independently discover gassing disabled kids as the rational course of action (given one's beliefs, which are of course factually wrong and morally abhorrent).
Could Hitler have stopped the Shoa if he had wanted? Possibly, the Fuehrerprinzip meant he could do whatever he wanted. If he had told is goons, "okay, we have kicked out the Jews from civil service, we have taken the factories some of them owned, we have smashed their shopping windows and burned their synagogues, but now we will actually draft them to the East Front instead of wasting a lot of resources to murder them", I do not think that the other Nazis would have staged a coup.
The thing is, a Hitler who was not rabidly antisemitic, who perhaps had some Jewish friends and went on record that they are okay, that Hitler would not find himself in charge of Germany in 1933.
The Holocaust was blatantly stupid in strategic, purely amoral terms. You are not grappling with that. Brutal anti-Semitism was latent within German culture. But the Jewish Holocaust was the exceptional product of Hitler's mind. He is singularly responsible for turning a pogrom into a Jewish genocide.
Your normalization of Hitler within the context of German anti-Semitism is absolutely ahistorical. In your effort to demonize all Germans, you are inadvertently excusing Hitler and posthumously benefiting his reputation. You can't have it both ways, however convenient that would be.
> you want to deny him any uniqueness [...] Hitler's political career was weird, singular and uncanny
I would expect historians to be biased towards exaggerating his uniqueness. For example, before WW2, antisemitism was popular practically everywhere. Eugenics was considered woke.
It was only after we decided that Nazi Germany was the enemy, that we tried to do less of whatever they were famous for.
In the context of ostensibly serious American historians, right-leaning historians have been incentivized to exaggerate Hitler’s uniqueness, and left-leaning historians have been incentivized to underrate his uniqueness. “Just another rightist doing right-wing things.”
In the context of sleazy popular historians, exaggerating his weirdness is a standard way to move copies.
In terms of body count and incidental cruelty, I don't think Hitler’s activities can be sensibly compared to any American campaign of eugenics or anti-semitism.
> No, we don't. Nearly everything you know about the Nazi party is propaganda. They didn't even call themselves Nazis, for god's sake!
Speak for yourself. I think I have a fairly good grasp on their ideology, value system and their organization and methods.
Are you claiming that if I were to read "Mein Kampf" (which certainly is propaganda, just not the one you meant, and to my knowledge rather open about Hitlers ambitions), I would concede that this guy had actually a lot of good ideas and was just misrepresented by the evil Western mainstream propaganda apparatus?
> The best thing that the Nazi Party did was found our modern environmental movement (something that the econuts keep under wraps, for understandable reasons).
(Caption, "I think he is ok, I just hate that thing with his Autobahns". Funny because the Autobahns are the one example often given for "not everything Hitler did was bad".)
I do not think that the claim that the NSDAP are the spiritual ancestors of environmentalism is actually true, though. Sure, they did care about the environment, in some ways, and had an aesthetic which favored primeval, unspoiled countryside populated with pure, primeval, upstanding Germans. Still, the good of the state (which is the good of the people, for them) came first, and if that required strip mining, then strip mining it was, and if it destroyed the habitat of some rare German frogs or something, they were shit outta luck.
In Germany, there is also little continuity to between the NSDAP and the Green party, arguably less than between most other parties. The German Greens emerged from the 68er, which was a left-wing student movement which was Western-skeptical and explicitly anti-Nazi, and accurately accused the other parties (e.g. CDU, FDP) of sweeping Nazi crimes under the rug.
You might as well claim that beer is strongly influenced by Nazism, just because it was something the Nazis cared about to some degree. If you want to bash the (German) environmentalists, consider going for the standard claim that they thought about extending the sexual liberation to cover pedophilia in the early 70s, that has the advantage that there is a kernel of truth to it. (Wiser heads prevailed and decided that this was actually a bad move in a lot of ways.)
I know they didn't call themselves Nazis. That's no more inappropriate or telling than the fact that we call Deutschland Germany.
Pray tell, what are some other falsehoods about the party that have been foisted upon us?
As far as I can tell, American propaganda tends to exaggerate Nazi Germany's technical competency and fighting abilities, probably to increase our perceived valor for having defeated them.
Hitler did not in fact make the trains run on time, and Germany's military innovations happened in spite of their feckless dictator. The German economy wasn't even fully on a war footing until 1944, which speaks to stunning incompetence; or at least poor administration.
Then there's the alleged, near-superhuman abilities of the individual German soldier. Once you adjust for sheer journalistic excess, the rest of the phenomenon is s largely explained by 1. When America finally faced the Wehrnacht in Europe, the Germans were fighting defensively, with all the natural advantages of the defender. 2. In Italy, they often had the advantage of rugged terrain.
3. As our forces approached the borders of Germany, the German soldier naturally feared that we would ravage their country with the same brutality that the Germans had used against the Poles and Soviets. This would have provided a strong incentive to fight hard.
I want to note that "when he said Jews, he meant Communist Sympathizer Jews" is a vile lie. When Nazis murdered Jews, they tried to murder all of them without concern whether they were communist sympathizers.
It is hilarious that you claim lies.
ad 5: you are lying here, by omission of crucial details: specifically, German victims were getting starvation diet far before Germans did.
ad 6: yes, Nazis were not sole evil in the history of the world. You get a cookie.
ad 7: by himself.
Thanks for feeding my confirmation bias that neonazis are bunch of lying idiots.
LOL, I thought most of the neonazis had migrated to greener pastures like The Motte. I guess this senseless intellectual virus is will always be with us wherever there is a modicum of free expression.
It's possible I misunderstood what you meant by that. It sounded like you were suggesting all sources of information about them are written by people strongly predisposed to believing lies and nonsense as long as they're sufficiently negative, and lurid rumors about them are impossible to separate from actual facts – and I agree with this.
If you're instead arguing that even given some sort of magical infallible oracle with perfect information, everyone is still so blinded by hatred or whatever to not be able to process that faithfully without their mental processing generating falsehoods ab initio, … this seems unnecessarily uncharitable.
Yeah, I meant the first one. Maybe my "inversely proportionate etc" was the poor choice of words. It was more like analogizing and not proposing some universal, immutable law.
I'm thinking about the discourse on AI, specifically the whole UBI debate. If AI truly does take over a lot of human work, there's a lot of people who are savagely agitating for a UBI on one side, saying we'll be post work. The other side of course says no that's not how it works, besides we aren't even close to being able to afford that.
What I'm surprised by is why nobody has so far mentioned what, to me, seems the obvious compromise - we just shorten the work week! As our forefathers did forcing a 5 day, 8 hour work week, why don't we continue there? Go down to a 4 day work week, and/or shorten standard working hours to 6 per day?
If AI truly will obviate the need for a lot of work, how is this not the more rational solution than trying to magically create a UBI out of money we don't have? How come this idea has barely even entered the discourse? I have been talking and thinking about AI unemployment for years and never once have heard someone argue for this compromise.
There were plenty of good responses to this. I hope it's not too prosaic, but I'll add that humans fear change, and most of us try not to think about what we fear. People Who Matter are affected by this dynamic as much as anyone, so they likely aren't brainstorming their response to AI development as much as they ought to be.
> If AI truly will obviate the need for a lot of work, how is this not the more rational solution than trying to magically create a UBI out of money we don't have?
I'm working on a post now that shows that in any realistic "AI can counterfeit white collar jobs" scenario, we can offer pretty generous UBI's, think $40-$80k in real 2024 USD to everyone.
This is because the economic growth from much-cheaper white collar jobs makes a generous UBI realistic. The economy grows a lot because they become A) considerably more net valuable, because the cost has gone down a lot, and B) considerably more numerous due to the Jevon's Paradox effect, because when they're cheaper and driving more net value to each company that has them, there will be more created, which creates more absolute value in the economy.
But what about real estate and art and expensive private school educations?? Those will get way more expensive! Baumol effect!
Yeah, and? If you have a UBI, you don't need to live in Tier 1 metros for your job, you can live anywhere - live where it's cheap! You don't need to buy Tier 1 metro houses and art and positional goods, and most people who elect to take UBI won't.
But broadly, you don't need to magic up money from nowhere, there will be a ton of real economic growth if we can counterfeit white collar jobs for a fraction of their salary costs today.
Because AI will not uniformly take 20% of work away from everybody's job (at least in the short term). More like it will destroy 20% of jobs. Junior office workers doing data entry roles, for example, will be completely obliterated; while plumbers will be largely unscathed.
I've written about this before, although I can't track down the article.
As AI takes over work, theoretically, we should be able to 'afford' UBI- to the degree we can't afford it already. I agree with you that shortening the work week seems logical. Another positive upside is that it helps keep us all useful. Terrible things have a way of happening to people who have no economic power in society, so keeping us all indispensable seems like a good outcome.
Obviously, in the final instance, there will be nothing left for humans to do. Here, UBI- but conceptualised more as a collective seizure of all assets followed by the distribution of a dividend- seems desirable.
> how is this not the more rational solution than trying to magically create a UBI out of money we don't have
Contingent on there being AGI, ASI, or any level of AI with capabilities to cause mass unemployement, it's impossible for there to not be enough money to support UBI.
If you strip out the framing of money, the economy will be producing at least the same and most likely significantly more goods and services. If the traditional distribution mechanism of this production via wages goes away, the government can simply step in and distribute it through UBI.
> magically create a ubi out of money we don't have
I suspect this will be a political obstacle to implementing a UBI - don't max out the credit card, don't shake the magic money tree etc. Are these analogies fair reflections of how an economy based on credit works? That's debatable. It's not crazy to bet on a huge increase in GDP which will finance the UBI. But I expect there will be an awkward period where unemployment is high, GDP is rising but legacy attitudes to work make a UBI unpopular. I mean, I'm not sure I want a UBI in my heart of hearts but it may be the least bad option.
[Edit: if the state doesn't supply a UBI I suspect we will get some sort of epic philanthropy programme from the tech giants, so I don't think the crisis of meaning can be prevented by postponing UBI]
Partly because there's greatly diminishing returns to that. The "work week" includes a fixed amount of overhead, doing things like getting up to speed on what all of your colleagues are doing, completing mandatory training, and filling out your timecard. All of that can easily add up to 10+ hours a week, in which case the person who works 20 hours a week is doing only a third of the work that someone doing 40 hours a week does. And it's worse even than that because tripling the number of workers means the "getting up to speed on what all your colleagues are doing" part starts taking more time.
But more importantly, 20 hours a week doing what? The reason improved technology didn't cause mass unemployment in the past was for the most part not because we kept shortening the work week, it was mostly because putting the machines to work doing the most menial physical labor left ordinary people free to step up to more lucrative and valuable jobs - jobs that for the most part weren't being done at all in the era when almost everyone was working 72 hours a week just to stay alive, and which make us all richer now that they are being done.
The reason things might be different with AI. is that AI + robotics might be able to do *all* the materially productive jobs a median human might do, better and cheaper. There will at least at first be jobs for exceptional humans, but if every median human who answers any help-wanted ad always gets the response, "no, it was cheaper for us to buy robots to to that; we're only looking for a couple of expert robot repairmen and you didn't graduate from MIT", then it isn't a question of their working 10 or 15 or 20 hours a week, because no employer would rationally hire them to work any hours a week.
Conceivably we could, instead of raising taxes to pay for UBI, require that every business that's still in business take the money they would have spent doing taxes and use it to pay people to do stuff that robots could be doing faster, better, and cheaper. But that would just be a UBI with a meaningless-busywork requirement because it makes us feel virtuous that we made other people do "work". Scott's already written a post on why that's not a good idea.
The part of the overhead that's "get up to speed with what your colleagues have been doing" is the part that's going to scale badly when you've got many part-timers working in series, and that's going to be hard to automate. A sufficiently capable AI could do it, but at that point you're basically having the AI tell people exactly what to do, slower than the AI could do it itself, because you think it's virtuous to make people do pointless busywork for their dinner.
If people spend 25% of their work time to coordinate with each other, it would make sense to hire someone as a full-time coordinator.
You may be doing a different kind of job, but in my experience, high overhead is usually a sign of incompetence. Inviting dozen people to a meeting where a short e-mail would suffice. Inviting hundreds of people, because someone important loves the sound of their voice. Repetitive meetings because people are unable to make a conclusion, or keep proper records.
Stupid meetings can certainly be a symptom of underlying incompetence, but getting up to speed with colleagues isn't just stupid meetings. Sometimes it's modifying your own piece of the project to remain compatible with everyone else, and if you have enough interface surface, that's a nearly full-time job. (That can be its own symptom of incompetence, but I digress.) And if you're not spending time in a meeting going over that, you might still be spending that time reading someone's progress report in an email instead.
I've been in team meetings where the team is about thirty people and everyone sees their purpose at that meeting as looking busy to the supervisor, resulting in everyone else listening to thirty amateur short stories about office work that not even the supervisor really needs, so emails are arguably a big improvement. (In such cases, breaking the team into subteams would have solved much more of the problem.) But the fact remains that there's *some* interface between anyone and their teammates, and they'll have to coordinate that, and it can take time.
Oof, 10+ hours per week on overhead is woefully inefficient. I really hope most people aren't doing that, but if they are, there are much better ways of doing things. That anyway sounds like something AI could actually sharply reduce and so perhaps that's where the first reductions in working hours could come from...
You'd be surprised. In my sector (horticulture), management = compliance to a large extent, the gardeners do the work and the boss fills out the risk assessments. So they say! I'm sure there's some minesweeper being played too. But yeh I reckon AI will improve a lot of that stuff.
[TBH a world where we continue to work with our hands and have an AI boss would work for me. Horticulture isn't complicated but there is a lot to remember in a single year, often with big gaps between the same task. Staff out the planning and compliance to AI, robots for the repetitive back-breaking or mind-numbing stuff, and we can focus on making rockeries, ponds and herbaceous borders really slick.]
I have a question for anyone who might know about Wellbutrin (buproprion), or other antidepressants which might tell me about wellbutrin). I understand you should wait for 4-6 to see if it works for you. But I can't tell if this applies *even if you are noticing clear side effects that show it is affecting your brain chemistry*.
In my case, it as dramatically curbed my desire for nicotine. Unfortunately I think it is also making me feel lethargic and slow and sad, to an even greater degree than the condition I was in, which made me go on it. My question is whether there is a reasonable chance it could start helping me if I were to stay on it.
(I've been on it for 3 weeks (at the target dose of 300mg, 6 weeks total if you include three weeks at the starting dose of 150mg). Any advice much appreciated. I have gotten pretty mixed messages from my provider and from online advice. Most of it seems to say 'stay on it' but to be geared at people who feel no effect.)
Sedation isn't a common effect of Wellbutrin; in fact, the association seems extremely rare. I would consider the possibility that it's something else, or a coincidence, causing you to feel sedated. This is not medical advice, and I am not a medical doctor.
I'm a psychologist, so do not prescribe but have pretty good general knowledge about these drugs. I don't think sedation is a common side effect of Wellbutrin. In fact I know some docs who start with it rather than stimulants as a treatment for people with ADD.
You went up to a dose of 300 mg much faster than I am used to seeing people go up. The last person I watched go on wellbutrin was on 150 mg for several months before increasing the dose. (And he had an excellent response to 150 mg. He's still trying to figure out whether 300 mg even gives him more of an effect.) I think increasing the dose so quickly was a mistake.For one thing, you went up to the higher dose before finding out whether the lower dose worked for you, which has deprived you of needed info: maybe you only need to take 150 mg of the stuff to get the benefit; maybe side effects at 150 are minimal. Also, if you'd stayed at 150 whatever side effects you had on that dose might have faded away to nothing -- it's not rare for early side effects to do that.
Curbing appetite for nicotine -- yeah, it works well for that. Helped me stop smoking years ago. If you have decreased or discontinued nicotine use, your fatigue and low mood may well be from nicotine withdrawal, which does commonly make people lethargic, foggy, with low mood. That is *exactly* how coming off nicotine made me feel. If you wanted to test that you could experiment with a couple nicotine lozenges to see if they counteract that stuff. Or you could just wait. If you're suffering from nicotine withdrawal the symptoms will fade.
>I understand you should wait for 4-6 to see if it works for you. But I can't tell if this applies *even if you are noticing clear side effects that show it is affecting your brain chemistry*.
Yes. Side effects kick in almost immediately. I'm pretty sure most antidepressant side effects are caused by the drug affecting your brain chemistry, just not yet having been in the brain long enough to bring about the change you are hoping for. I recommend that you drop your dose back to 150, which should make the side effects more tolerable, and wait a couple months before increasing the dose to 300. Side effects may have faded away by that point.
Wellbutrin has the very desirable feature of not causing sexual side effects. It's the only antidepressant I know of that that's true of. You should give it a really good try before moving on to the other stuff.
Those people don't look European, they look uncanny, as in the uncanny valley.
Which is unsurprising, because they combine South Asian facial structure with Northern European skin and hair tones, which isn't something you find outside of albinos. They certainly don't look like Norwegians.
Please watch "The Age of Disclosure" on prime, $20 to rent $25 to buy. At greater than 50% chance Aliens are real and we have the bodies. After watching please encourage your congress people to pass the UAP Disclosure Act. Defense contractors are lobbing heavily against it. (This seems way more important that AI.)
For what? So you can hide something this funny? Come on man, people like them being visible is one of the advantages of a forum like this. Echo chambers are boring.
Y'know, if all of this was real, the government wouldn't hesitate to kill anyone who they even suspected was going to leak this information, let alone do TV interviews and make a whole movie about it. So why do you think these people are still fearless and alive? Don't you think it's suspicious that the government is making no effort to interfere with the advocacy for disclosure? When the disclosure happens and it turns out to just be boring national security stuff, that's just a way for them to get everyone to think the mystery is solved. This movie is just another part of the conspiracy.
Yeah, I'm pretty sure they have killed people. It's a complicated story. I urge you to watch. Yeah the guy who quit his nice job with the gov. to go tell what small pieces of the story that he is allowed to, the unclassified bits. Is an f-ing hero IMHO. Finally The UAP Disclosure Act, has been up for votes I think several times but at least once. Heavily lobbied against by defense contractors... who make plenty of campaign contributions. The story holds together in my mind, it's fascinating... I want to say more, but also feel you should just go and watch it. What do you have to lose? Like I said I'm >50% the aliens are here. Of course I also watch Joe Rogan, and TBH several of the people in the movie have also been on Rogan, and telling the same bits of the story you are given here.
The best skeptical question I've heard about aliens visiting is why do these creatures flying quantum-fractal-black hole-doodah vehicles *crash* all the fucking time when flying around our planet? WTF? Are the aliens all alcoholics?
Yeah I agree! One of the guys talks about this. He says, accidents happen and then maybe it's on purpose. And I said no-way. But here's another fact from the documentary. There are at least two 'species'. (whatever the right word is.) I immediately thought, 'oh, maybe they are at war?' (I've read too much sci-fi, cue David Brin)
Yes much footage from congressional hearings. When you get the whole story, (so far) much of that makes sense. There's a strong campaign by defense contractors (and whomever their surrogates are.) to discredit all UAP's and those who claim them as true. So I expect lots of push back and denial. OK don't watch the documentary. But what if the aliens are here? There is no Fermi paradox.
For me there's been so much BS pushed on me from all sorts of people (media, government, industry.) that I'm willing to listen. Maybe you trust these people more than me.
I don't know about the desert girl thing. The bodies we have are recent. Area 51 is a real crash site. I know I'm stretching your credibility. How about this you watch the documentary, if your opinion is unchanged by it I'll send you a check for the $25.
Oh, It's not that I can't listen, but how do I separate the lies and truth? And it's not just lies, it's whatever the 'narrative is'. And is it worth it to me? I mostly end up getting mad about something stupid and have to turn it off. (the stupidity is mostly on my part)
Separating lies from truth is unfortunately quite difficult. Consider that someone in the establishment lying to you does not make alternative sources more likely to tell you the truth. It's not either or. I would recommend looking for sources that are in a position to know, and that are not actively pushing information on you. Consider what incentive someone could have to say what they are saying. The people talking the loudest, and with most certainty are seldom the ones to listen to. Accurate information will often be nuanced, and uncertain. For every question there is an answer that is simple, intuitive, and completely wrong.
I might use some advice on online sales, if anyone has any. I'm planning to sell pictures of my own design (I hope this is vague enough not to count as advertising! At any rate nothing is on sale right now) and I'm looking for a way to do so that requires the least micromanagement. I am NOT looking to make a quick buck, or planning to make it my career; I'd like something that quietly runs itself as much as possible while I focus on my actual job. It's fine for me if money comes slowly, as long as it's not literally zero or a net expense.
After some brief research, the most convenient way I found yet requires two services: setting up shop on a place such as Etsy for the actual sale, and linking it to a separate service to handle the physical printing and logistics (two call Printful and Printify come up most often in searches; I don't see any obvious superiority of one over the other, and tbh the near-identical names don't help). I might just do that. Is there any reason this is a terrible, terrible idea that I will regret?
I can help. Selling my designs on various products is my main source of income. I’ve used many services, including Etsy, Amazon Merch, Printful, Printly and over 20 others.
What I recommend most is Threadless. I'm not affiliated with them - I just like the platform and earn a lot there.
I’d stay away from Etsy. They charge fees even when you don’t sell anything, and running an Etsy store is basically a full-time job.
They have lots of posters/canvas types. I haven't ordered any so I don't know how's the quality, but I know people buy them because I get email notifications whenever someone orders something.
I order internationally (clothes) and shipping is fast but expensive.
If you only want to focus on posters then you might check out Displate - I used to have a store there but I didn't sell anything so I can't recommend it but maybe you will have success there. And there's also Inprint - I haven't tried them. For more artsy stuff there's also society6, but it used to have such an awful uploader that I gave up on them, they might have fixed it since then.
Is Amazon's Merch on Demand not what you're looking for? They probably take a larger cut, but I think it's exactly the "fire and forget" one-stop-shop you're describing.
Our conception of Normans as Vikings was influenced by Plantaganet propaganda. The descendants of William the Conqueror found it politically convenient to minimize the Frenchness of their forebears. They wanted to emphasize their modest Norse heritage. After all, their English subjects ALSO had Norse admixture.
But compared to England, Normandy has yielded up few Norse burial sites. Scandinavian migration was seemingly limited to a few elites.
So being “Norman” in the 1000s wasn’t really about blood or ancestry. It was more like joining a rough, ambitious club. Normandy pulled in land-hungry fighters from all over France, and what made them “Norman” wasn’t where they were born, but how they lived: war, loyalty to powerful bosses, and grabbing new land.
This is an area of personal interest to me. You're both wrong and right in your characterisation. Where you're right is that the degree of French ancestry in the Normans is downplayed.
Where you're wrong is that the Normans weren't just significantly Viking in the aristocracy but they were also significantly English, owing to well documented English migration there in the preceding century. There is an extensive toponymic record of this, including Cherbourg. There was extensive pre-invasion contact between Normandy and England, hence the invasion in the first place.
This is far better explored in French academic literature than Anglophone academic literature, to the point that it's almost entirely unknown.
The English will happily tell you about all the times they were invaded, as long as it's by someone who no longer exists.
England was invaded by Romans, Angles, Saxons, Jutes, and Normans? Sure, it's all part of the rich history of these isles.
England was invaded by French or Germans? Absolutely no freaking way.
That said, Rolf the Ganger was definitely a Viking, so the history of Normandy as a bit of France that got conquered by Vikings who eventually conquered England is not entirely made up, even if there were a lot of Frenchmen involved too. Trace the royal ancestry of King Charles and you'll find Viking kings, not French ones.
Historian R. Allen Brown summed it up by saying Normans were “made, not born." Normandy was basically a medieval launchpad that toughened Frenchmen up and sent them out to seize places like England, southern Italy, Sicily, even parts of Greece and Africa. So the Normans were less a genetic “people” and more of a warrior fraternity. One characterized by violence, competition and greed.
Norman warriors undertook big heists when they could, like grabbing a kingdom/lordship. But they would take what they could get. If they could find a place with little or no law enforcement going on, young Normans would post up near roads and rob hapless travelers.
Normandy was a region mostly populated by ordinary peasants, like most of Europe at the time. But it true that it is remarkable that this region suddenly produced so many knights who terrorized so many parts of Europe and the Mediterranean. That they aggrergated adventurers from elsewhere is good explanation.
On the other hand, looking up the ancestry of the chief Norman lords in the Mediterranean at the time (the Hautevilles, the Quarrel-Drengots, the Moulins/Molisios, Roussel de Bailleul), they seemed to be actually from Normandy. Maybe the lower ranks were more diverse.
Of course, medieval lords shamelessly rewrote their family history for various reasons. Or rather had someone else do the rewriting for them. Richard the Lionheart was one of a number of lords who seriously insisted he was descended from a water fairly.
The days of woke scolding are over. I sincerely doubt they are genuinely racist in a classic sense. They are just counter-signaling against the insane excesses of woke censorship.
Correct, that's pretty controversial here. How else do you want people to signal that you're done with these people? Something is necessary to separate us from them. Heterodoxy is the easiest way to achieve that.
Vice or virtue, signaling is greatly overrated. At its best, signaling is totally acceptable. But we don't often see signaling at its best. Obnoxious, shallow, shit-stirring, self-indulgent, immature--this is just a sample of the unfortunate characteristics that typify the most adamant vice signalers among us.
I'm not going to lump you in with those goblins, certainly not based on one comment! But as much as wokeness bothers me, it would never occur to me that the world NEEDS to know I'm "done with" it. Can you see how that might come across as self-important, even pompous?
Donald Trump won the popular vote. The writing is on the wall about the shifting zeitgeist. Just how USEFUL is it for us to do ironic racism? To what extent is doing this nothing more than gratifying our inner children?
First, I do not think 'vice signaling is bad' is all that controversial here, sorry. I recommend reading more of Scott's posts.
Second, taken to its logical conclusion, you are part of a movement that is purpose built to do the worst thing along any measure, every time it has the opportunity to do anything at all. I can't think of any reason why this may be worthwhile to do, but even so party identification seems like a really bad reason. "I'm going to be racist so that people will know which party I support" feels a bit like telling on yourself
> First, I do not think 'vice signaling is bad' is all that controversial here, sorry. I recommend reading more of Scott's posts.
And I recommend reading more comments. Scott and his audience do not share the same beliefs. Due to various circumstances, this blog has a tendency to attract people who aren't fully on board with his centrist tendencies, but enjoy his work anyways.
> Second, taken to its logical conclusion, you are part of a movement that is purpose built to do the worst thing along any measure, every time it has the opportunity to do anything at all.
The disagreement here is on whether these things are particularly bad in the first place. They're just words. No explicit racial hatred was expressed. If you're offended by it, that's fine. That's just the signal doing its job.
Sorry, "vice signaling is the best way to identify the in group" is stupid and self destructive. Like I said, taken to its logical conclusion, your movement always has to do the worst possible thing. Why? Because you've self defined good things as "woke" and "liberal". This is why maga purity spirals exist and why your movement is doomed to be the launching pad for Nick Fuentes. It's self hating metastasized into a political party
They were thinking that being edgy and posting hot takes on social media gets them attention. They are right about that overall, even if the occasional tweet goes over the line.
The reason I asked is that I saw a discussion of Joe Rogan saying (not well attributed) that he thought a billion but would be conservative and say 500 million.
Worldwide? Probably around 15-20 million. Or a bit higher if you're counting people who are ethnically or culturally Jewish but don't adhere to Judaism even to the extent of self-describing as a non-practicing Jew in religious surveys.
Reasoning 1: I think I remember seeing a stat that there were about 12 or 15 million Jews worldwide as of the late 1990s. 25-30 years of population growth would put the current figure somewhat larger. Not enormously larger, since most of the world's Jewish population presumably lives in relatively low birthrate countries.
Reasoning 2: I remember with more confidence that a very large percentage of the worldwide Jewish population lives in Israel or the United States, with roughly equal numbers in each. I remember looking up religious statistics for the US recently and seeing something like 1.7% or 1.8% of the US population was Jewish, which would work out to about six million Jews in the US. I don't remember the population of Israel off the top of my head, but something like 8-10 million (of whom about 20% are Muslim, Christian, Druze, atheist, etc) sounds about right and would work out to 6-8 million Israeli Jews. Total 12-14 million in the US and Israel, plus probably a few million more living in the rest of the world. Off the top of my head, I think I remember Yemen having enough of a Jewish minority to show up in high-level country religious statistics, plus countries like Russia and the UK which have much smaller Jewish populations in percentage terms but enough to add nontrivially to the global total in absolute terms.
Also without checking, I would be very surprised if there were a significant number of Jews in Yemen to this day. If there are any, they must be having a rough time with the capital and half the country controlled by the Houtis, a faction which has "Death to Jews" on its flag.
I think Morocco is the only Arab country which has retained something of a Jewish community (of maybe a few tens of thousands: not enough to impact the global number). Among the other Muslim countries, also Turkey and Iran have small communities.
That's sadly plausible. I mentioned Yemen because I remember many years ago looking at CIA World Factbook info about religious demographics on a bunch of countries. Yemen was the one country besides the US and Israel that I remember having a large enough Jewish population to get its own line item in the stats rather than being lumped into a miscellaneous line of minor religious minorities or altogether unmentioned. I think the number was around 4-5% at the time.
To the few people still paying attention, I finished the book and have decided an ideological deconstruction of the various varieties of feminism would have infinitesimal value in the memetic war. Too many, like existentialist, psychoanalytic, and eco-, no longer have any real relevance, and most of the current wavefront of ideological warfare is too fast changing to really be summarized effectively. Besides, most men respond to jokes and pictures, not lengthy essays.
Writing is for women now. That's just something I'm coming to accept. Sure you can swim upstream and a few people even make a career out of it, but I'm not that good and I'm not that young to get good.
What's that line from the end of Boiler Room? "I tried selling crack rock, and I never had a wicked jump shot. I gotta get a job." Well, I have one of those. I need a hobby.
This sounds like cope. Perhaps you should work on improving your writing instead of blaming demographics. Or recognize that the sub-varieties of feminism are for feminists to communicate amongst themselves, not necessarily to appeal to society at large.
Cope is part of life. If you're 5'7" you're never playing for the NBA, and if nobody wants to watch your demographic play basketball there won't be an NBA you can join.
I've been reading a lot of weird stuff online, I can go pre-Internet if you like.
How about:
"To the few people still paying attention, I finished the book [of feminist theory] and have decided an attempt to point out and argue against the weak spots of the various varieties of feminism in a detailed critique would have almost no value in swaying people to the other side. Too many, like existentialist, psychoanalytic, and eco-, no longer have any real relevance, and most of what people are arguing about online is too fast changing to really be summarized effectively. Besides, most men respond to jokes and pictures, not lengthy essays."
So your reading project was actually a writing project. Huh. idk why this wasn't more obvious to me, based on your question about hobbies a few months back.
I dunno. It was a reading project, the reading went on but the writing was never produced, so it's more accurate to say the writing project was actually a reading project.
> Writing is for women now. That's just something I'm coming to accept.
Short form video is obviously the future - who has time for *books,* which have been systematically declining in popularity for decades??
No, we need more Dramabox-style microdramas - those things have grown from literally not existing a few years ago to a ~$1B market already.
To stay on topic, they could graphically demonstrate the perils of feminism, as high status boss-babe style women repeatedly make crushingly bad personal relationship decisions that lead to career-and-life upending drama. <whispers> What's that? This is already 2/3 of microdrama plot lines? And it's WOMEN who watch all these things??
Huh.
Okay, I got nothing then.
I do agree trying to focus on writing is probably a mistake - the guys whom writing resonates with are generally doing fine in the gender / dating wars, in aggregate. And what advice do you even give median men?? "Sorry the world sucks now, women would rather work and watch Netflix than date you, so, uh...try to get better than Netflix?"
That said, on the "searching for a hobby" front, any low hundreds purchase that may turn into a hobby is worth it even if the hit rate is low. Sports equipment, a musical instrument, art supplies, etc. If it doesn’t catch on, gift it to a friend.
I've read more books in the past 5 years then I have in any 5-year period since I was a teenager. It's not for everyone, but I suggest it for anyone who can hack it. I enjoyed watching a movie almost every day, which was my previous relaxation technique. But in the long term, my renewed reading habit is more gratifying and better for my self-esteem.
> And what advice do you even give median men?? "Sorry the world sucks now, women would rather work and watch Netflix than date you, so, uh...try to get better than Netflix?"
Bring back patriarchy? The circumstances have never been better for restoring it. It's not like women can do much about it when push comes to shove.
Maybe it would work better to bring back black slavery, esp. since 50% of the re-statused would be fuckable by straights. Whatcha think, Mistilteinn?
Oh, wait, you may be worried about comingling your genes for smarts, social skills, and civilized common sense with black genes for dumbness and violence. How about we enslave the Jews -- high IQ's, same 50% fuckables benefit.
I'm with PB, it really doesn't work anymore. The most they've been able to do is outlaw abortion, something I never supported on base libertarian grounds.
I do find the death of chivalry among young men hopeful, it could serve as the baseline for the first effective men's movement. Unfortunately I can't really participate due to ancestry issues.
You're from a completely different online subculture, and I find your writing style a bit grating. But I agree that chivalry must die; is this really happening? At LAST?
The tyrannical CRUELTY of implementing feminism while insisting that males remain chivalrous! It boggles my mind that people can see this as anything other than misandry.
My Australian friend was shocked when I informed him how rigidly American society insists that "the man should always pay for dinner" Down with the patriarchy, except for those bits of patriarchy that put money in my pocket.
What do you mean by ancestry issues? Do you mean you're not white enough to be taken seriously in a public facing role by a right-leaning audience, or what?
Hanania's moved away from it towards feminism, at least officially (he doesn't like the Trump admin and, probably more importantly, has two daughters), but this article is still pretty good. You never really know what Hanania's thinking, but we're not likely to see more of this from him right now.
As for my language... I modified it a little in my other response to you, I've been poking around the esoteric right-wing too long.
Basically I figured most antifeminists see 'feminism' as an undifferentiated mass of opposition, but there's actually huge amounts of disagreement within it on various issues. Inject that knowledge into the right and they may be, like, 10% more effective in fighting them. I got an introductory book on feminist theory and figured, well, maybe I can do what Richard Hanania did for (or rather against) the woke legal regime or Helen Pluckrose did against critical theory. Of course I'd have to read the major texts--I was thinking de Beauvoir and Judith Butler next, with maybe a detour for Foucault (I even had a 'How to Read Foucault' book from a left-wing theory press).
The problem was that most of the current arguments were crammed into two little chapters near the end. You might be able to see their ultimate origins in the stuff I was reading about, but it's like doing a close reading of Marcuse to understand what Jessica Valenti's talking about--it's vaguely relevant but there are much more germane ways of going about it.
OK, read something about newer stuff? It's not going to be in a book, culture changes too fast now! I'd have to join Bluesky and lurk on Discords and even then might spend years on some formerly-active location that's really influenced by something else.
And the problem with *that* is...I'm old. I'm Xennial, I was trained to read long texts and write summaries and responses to them in high school and some of college, but that's not how you convince people now. A clever Pepe meme or a well-placed 'cool story bro' will sway hundreds of times as many men as a Substack essay. In the marketplace of ideas, but if you're going for young men the way to win them over is to master a whole bunch of social games I never learned in high school and would be even harder as a middle-aged person who's (a) less cognitively flexible and (b) not up on current cultural lingo, which changes faster and faster with each passing generation.
As you can see from the time stamps on these responses, I can produce large amounts of college-level text rapidly, but you know who's even better at that? ChatGPT and Claude.
Finally, unfortunately, my mom is of Ashkenazi Jewish ancestry. That's not a problem with older conservatives but definitely is with the newer Groyper bunch. Of course you've got the Christians (convert and you're saved), but they're still all-in on chivalry. Antifeminism always tracks with antisemitism for reasons I'm not entirely clear on (I think it has to do with who joins various coalitions) but in general the closest people to the MRAs are always going to be ranting about Jews. Not entirely inaccurately--look at who staffed the second wave feminist movement--but it makes me useless in this struggle.
Keep in mind, I don't really care about Israel (I mean, I'm rooting for them but don't think the USA should spend blood or treasure on their behalf) and I don't have anything AT ALL against the white man (I thought I *was* one until, like, the age of 40, which given that that constitutes most of my adult life tells you where my real loyalties lie). But the taint is in the blood with these people. And if there's one thing you learn as a guy who read The Myth of Male Power in the 90s and Paul Elam and Roissy/Heartiste in the 2000s, it's that life isn't fair.
I suspect groyperism is just another fad or passing phase. We're just one or two lost elections away from groyperism getting discredited and filed away with the many inactive influences of the right.
I also think we overestimate how much the RL right looks like the online right, and Fuentes, Tucker and Vance are not immune to this. But who knows, I can't predict the future. I DO know that these lists of "Jews in the media" speak to some pretty lowbrow guys with limited critical thinking skills.
While Jews do look after each other in a limited or superficial sense, they are so riven with sociopolitical conflicts that it's hard for me to see them as some monolithic block dominating "the media." Or any other industry. Anti-semitism is fundamentally shallow and I don't expect it to thrive in the long term. Though of course, no religion should be immune to criticism.
> Bring back patriarchy? The circumstances have never been better for restoring it. It's not like women can do much about it when push comes to shove.
Beyond the moral dimensions of this problem, regarding which I'm staunchly on the women's side, just from a pure practicality standpoint can you point to a single existence proof anywhere?
I think this is a one-way gate, and once you give women the vote and let them have jobs and bank accounts, you literally can't undo that.
Is there any developed country in the world that HAS taken away women's votes or ability to get jobs and bank accounts after they already had those things?
Heck, let's expand it to ALL countries - Iran is probably your best bet there, in which women had more societal freedoms that they then rolled back, but their changes were mostly about needing male permission / escorts when traveling and public dress codes, and women STILL have the rights to vote, work, and have bank accounts there!
I also challenge the idea that more patriarchal dynamics would end with more women dating median men. South Korea, for example, is much more patriarchal than America or Europe (as is all of East and SE Asia), and they have such crushingly low marriage rates and fertility there's a cottage industry in videos outlining their inevitable extinction - this is an excellent one here: https://www.youtube.com/watch?v=Ufmu1WD2TSk
They've turned the "patriarchy" and "gender wars" dials up even higher than us, and it went the opposite direction - even *fewer* women would date men, even fewer marriages, still fewer kids.
Barring literal slavery, I don't see how you're solving this problem, because it's a matter of personal choice and values, and you can't force specific interpersonal relationships to happen, either with culture or law.
What there will be, is AI / catgirl girlfriends well before anything like that will happen, and that WILL solve the problem.
Wait what? I thought we were talking about getting women out of the publishing industry. Anyways, the status quo for most of human history involved women having little agency of their own, and it's been exclusively the last century that's been a cultural anomaly, largely due to labor demands and influence by the liberal west, especially America. But Pax Americana is dead, liberalism is dying, and even demand for female labor is going down due to automation and outsourcing. And with there being so many disgruntled young men who now have an ambitious leader to make their vision a reality... it really is the perfect opportunity to make something happen.
No "slavery" is necessary. The solution is simple: make it impossible for a woman to survive on her own. If she can't get a job with a living wage, choosing to stay single becomes a death sentence. Suddenly, women aren't so picky about who they end up with. All you need are simple incentives like this.
One of those rare occasions where I kind of agree with you on a base level, and yet I can't help but cringe!! Because you have to groyperize your argument by saying things like "death sentence."
Anti-feminism has merit, but that performative edginess makes us look like a bunch of unserious clowns.
They are gay billionaires in Silicon Valley, too. We need to make survival without selling sexual services impossible for everyone! Anyone who disagrees will get shot by a drone.
Not everyone, just the people who have nothing else to offer. Didn't the Romans have people whose job was to be a dependent parasite who humiliated themselves for the sake of a sadist?
Remember, they've got AI boyfriends too. Just attach ChatGPT to a voice synthesizer and vibrator and you've got a pretty workable robo-boyfriend. People had a long history of anthropomorphizing stuff like the weather and fire and it's not surprising something carefully engineered to produce human speech would have the same effect.
But overall yes, I agree, there will be robowaifus and robohusbandos. I admit I never got the catgirl thing. Catwoman scared the shit out of me (though given my age this was in the Tim Burton movie so...)
Oh, totally - I'm actually predicting we're gonna lose like 80% of the populace between better-in-every-way AI boy/girlfriends and Infinite Jest style AI heavens that cue on your physiological responses to make superstimuli-powerful feedback loops in virtual Tik Tok / porn equivalents.
Both of which we can roll out nascent starter versions of essentially today.
A question for musicians and people who work with audio. I play electric bass and a harmonica and I want to be able to jam outside, without an electrical socket. What's the cheapest and the most light weight device, or multiple devices, to allow me to do this? So, basically, I need a speaker, an amplifier and a battery. Or something like that. All in one device or as multiple devices. It should be able to produce low bass frequencies as well as high frequencies.
For years I'd bring this to the river and plug an acoustic bass with a pickup into it and it was sufficient to hear over the river and to play with an unamplified acoustic guitar.
If you don’t find something cheaper, the Bose S1 is a great product, albeit not one optimized for low bass. But I take “jam outside” to mean a situation where loud volume and perfect fidelity are not the goal, but rather ease and portability.
How loud do you need it to be? If you only need to hear yourself in headphones this is easy. If you need other people to hear you you're in trouble. By the time you've got an amp that can be heard outdoors, and a powerstation that can run it for a decent interval, the sizes of both objects have generally left the realm of realistic portability. Bass amps are especially power hogs. If you don't mind hauling a generator around on the back of a truck it could work. True travel amps are tiny and weak.
Harmonica amps are generally just vocal amps with a special mic for the harmonica. This is more realistic for a travel amp as there are many portable voice amps/PAs on the market at all kinds of different prices.
For that matter I can't imagine an acoustic situation where you'd even need to mic a harmonica. It's a loud instrument that cuts pretty well even in cross-harp. I'm a little more optimistic about portable amplifiers - they'd at least help a good deal.
So I keep running across stats about what fraction of online content is AI-generated, and the numbers are high -- 50% of this, 70% of that. And just now, asking Google for stats, I was served ads for several "tutorials" on using AI to create viral videos: "AI Video Cut tool creates viral shorts instantly from long-form videos with auto-generated titles, captions, and descriptions."
My first spontaneous association to these factoids is to leukemia, where abnormal immature cells crowd out the healthy ones. But then, I have a pretty dark imagination. What are people's thoughts about how this is going to play out? If we don't already have a bot or 2 posting here, we will soon, right? What will it be like if it starts seeming to many of us that we have have lots of them? Seems like few or none of will be provably bots -- so then what? And what about the internet as a whole? Will it become Botland, with few live users? Mightn't huge group phenomena develop spontaneously in the internet bot population -- the equivalent of wars (instrumental convergence, right?), fads, matings, dynasties, illnesses? How will that affect human life? What will happen to people who maintain a high level of participation in Botland, either naively thinking they are dealing with people, or consciously electing to live in that online world? Will there be alternatives channels of communication that are protected from bots? What kind of protection is possible?
With trust hitting all-time lows, gatekeeping will become more important than ever. Much like Performative Bafflement says, the internet will become more tiered. "Tiered" in the sense that the default platforms will become ghettoized favelas, whereas higher-quality content will be locked behind more paywalls and real creators will face more verification hurdles (e.g. blue checkmarks, whitelists, provenance, etc.) (I feel like people will get awfully creative with this, in ways that are hard to foresee). But even then, the higher-quality content will be under constant surveillance for signs of deception, and users will be paranoid. It'll be like The Second Variety, or Amongus, with accusations flying every which way.
Performative Bafflement mentions Discord. Meanwhile, Venkatesh Rao [0] (the same guy who wrote about the Manufactured Normalcy Field) and others [1] have already started theorizing about the rise of what they call "the Cozy Web". They use the "Dark Forest" metaphor as a muse.
I read your links. Wow, that's very interesting stuff. OK, so some of us will end up in online "gated communities." What are your thoughts about how things will play out for people who don't have the skills, knowledge and money for those places? What's on my mind right now mostly is what the toxic part of internet, the ghettoized favelas, will be like, and how that will affect its users, and how the effect on them will affect life in general -- mental health, movements, fads, social norms, etc. Also, I have been wondering about interactions among bots and other AI entities themselves in an internet setting where they are most of the users. Could there be group phenomena -- fads, wars, plagues, coalitions, power struggles? Could there be emergent properties, entities comprising multible bots that have qualities and goals of their own?
Eh, I dont think the favelas will be qualitatively different from what we have now. Just more extreme. E.g. instead of traditional porn, we'll have AI girlfriends. People will invent new slurs for bots. AI assistance will be coded as low status. Dumb beefs about AI handicaps. Like, even before AI, and before the Almighty Algo, the internet had a Wild West period where nobody knew you were actually a dog with a blog.
I feel like the gated communities are where the more novel phenomena will emerge.
Though I could certainly be wrong. This isn't a hill I'm willing to die on.
>My first spontaneous association to these factoids is to leukemia, where abnormal immature cells crowd out the healthy ones.
Yeah, I view as a problem to the extent that the AI-generated content is worse that the human-generated content.
I'm not intrinsically hostile to AI-generated content. As things stand today, I regard it as the responsibility of the person who prompted the AI. If the prompt was a reasonable one (not intentionally malicious, for various flavors of malice), and the AI was reasonably competent, then I consider it unobjectionable. For instance, within the past week or two I watched a YouTube video on neutron star structure which, at the end, I guessed to be AI-generated. ( I'm guessing this because the wording of the narrative was repetitious in a way I doubt a human would word it. ) But it wasn't, as far as I could tell, wrong or misleading.
I don't know how much of the information in it was specified by the human in the prompt. I'd take the ratio of that to the final text as defining how much of more-or-less the responsibility for the video was human vs. AI. As the AIs get better I expect benign uses like this to transition from clunky-but-fairly-ok to good to better-than-human.
Unfortunately the malign content, mostly versions of computer-aided-deception will also get more convincing...
I would hate to watch unlabeled AI content, because over decades I have developed certain heuristics for evaluating people, and those fail horribly with AIs.
Humans have mental *models* of things. AIs complete *sentences*. The result is superficially similar, but the difference is that humans are more consistent, predictable.
After interacting with a certain human for a while, you can get realistic expectations about where the human will be smart and where stupid; things like "had really deep knowledge of Python development, but believes insane conspiracy theories". You can use this as a heuristic for how much to trust that specific human. You can find humans that you can trust, at least for certain topics.
AI is all over the place. It has vast encyclopedic knowledge, but sometimes it just hallucinates things. It can generate mathematical articles that seem like written by Einstein, but somewhere in the middle contain a stupid error that would make a high-school student ashamed. You are dealing with something that is million times more knowledgeable and faster than you... but you need to exercise constant vigilance, because you can be betrayed at any moment.
> I watched a YouTube video on neutron star structure [...] it wasn't, as far as I could tell, wrong or misleading.
The scary thing, from my perspective, is that even if you verified 99% of the video and it turned out to be perfectly correct, that still leaves a realistic option that the remaining 1% is just completely wrong in an absurdly stupid way. And put together, the conclusion can be wrong but sound credible.
I tried to use AIs for learning math, and I have up because they constantly stabbed be in back when I started to trust them. Changed the definition of something in the middle of a proof, things like this. And instead of admitting a mistake, they double down; instead of using their intelligence to notice the mistakes, they use it to generate plausibly sounding excuses.
With programming, if the code does not work, you throw it away. But with learning, you often don't get the immediate feedback that the lesson you learned is broken.
>I would hate to watch unlabeled AI content, because over decades I have developed certain heuristics for evaluating people, and those fail horribly with AIs.
Many Thanks! Yeah, I would prefer to see the AI content labelled too.
>The scary thing, from my perspective, is that even if you verified 99% of the video and it turned out to be perfectly correct, that still leaves a realistic option that the remaining 1% is just completely wrong in an absurdly stupid way.
Agreed. Some of this looks like sign errors, and I have to admit that I'm unsure about the comparison to humans. I have also seen humans generate what amount to sign errors in the middle of e.g. a comment - "understate" swapped for "overstate" and similar not-exactly-typos-but-kind-of-in-an-adjacent-class.
E.g. I had a discussion about triple points being reference points for temperature scales, and one of the LLMs said that melting point were used also - where the vapor pressure at the point would have been too _HIGH_ according to the LLM - oops...
>over decades I have developed certain heuristics for evaluating people
I would be interested to know more about your heuristics (as my heuristics have you down as someone who generally has a lot of insight).
>It can generate mathematical articles that seem like written by Einstein, but somewhere in the middle contain a stupid error that would make a high-school student ashamed.
Well, I think even Einstein is known to have made one or two stupid mistakes as well. Mainly a joke - I generally agree with you, and this is my experience with AI as well.
My heuristic for AI is that it mostly can be trusted for anything that is reasonably broad and uncontroversial. When asking for more special knowledge on anything more complicated I have this uneasy feeling that this looks very plausible, but there might be a subtle mistake somewhere, and it might be important.
Thank you! I will probably just say the obvious, but here it is:
Basic: You can use someone's grammar and following of social norms (including internet norms, such as not writing in ALL CAPS) as a proxy for their intelligence and mental health, which is a proxy for how reliable are their arguments.
Advanced: With repeated experience, you can notice how often the person is wrong, what kinds of logical fallacies they often make, if they can change their mind, if there are some topics they are obsessed about, etc. Certain opinions or keywords are signals of belonging to certain political or other tribes. People who mention details or nuance are usually more knowledgeable than people who speak in broad generalities.
For example, if you receive an e-mail that contains too many capital letters, entire paragraphs in bold, uses various font colors and sizes, you can probably delete it without reading, because the sender is most likely insane. If someone's argument on a political or economical topic can be reduced to "my opponents are *tards", it is not worth reading. If someone posted the same thing several times, and had previously never meaningfully responded to replies, do not expect a meaningful response today.
Now of course these heuristics have many weaknesses. For example, someone's grammar sucks because they are dyslectic, or they speak weirdly, because they are not a native speaker. Teenagers are likely to lack experience and express extreme opinions, but that's natural for their age, even if the same thing was a serious red flag for an adult. Accusing someone of belonging to tribe obviously leads to a counter-accusation. Crackpots can provide unlimited number of details. Sometimes the expert is just too busy to provide a high-quality explanation.
And yet, it somehow works. Maybe with a help of some other criteria that I am not consciously aware of.
AI can add good grammar and polite arguments and expert keywords to anything. AI can claim that the Moon is made out of cheese, and add a two-page detailed explanation about how the cheese is made, which will sound generally correct to an expert on dairy. (A human who believes that the Moon is made out of cheese typically knows very little about either.)
I agree with you, except I don't think it can be trusted for even things that are broad and uncontroversial, except for a very "narrow definition of broad", so to speak.
Check this interaction with Claude Sonnet 4.5 I had yesterday, specifically the part about grammatical cases in French:
It said one thing that seemed off to me, then went "you're totally right to notice this mistake, I was imprecise" and just said the opposite.
I'll let you judge how "broad and uncontroversial" the topic here was, but I keep having interactions like this, and they often involve mistakes in contexts where I am actually trying to save time at work by automating stuff, instead of just satsifying my curiosity. Instead, I need to be vigilant to make sure there is no stupid mistake that would make the whole edifice crumble.
Yeah, I find it's mostly pointless to ask LLM's to fix their mistake. The exception is code, but then you can easily check if the code works and ask for spesific modifications.
I find it useful for doing research into spesific topics, even very technical ones, especially to get an overview over the topic or a broad explanation, when I can check the answers with other sources later. It seems useless in helping me write anything technical, as I would have to check everything - and then I might as well just write it myself from scratch.
> What are people's thoughts about how this is going to play out?
Ah, dead internet theory! If the infocalypse happens I think there's two main ways it could go:
1. The free open internet is gone, everything is paywalled and real-ID'd because you can control bots that way. People like the ACX commentariat will find paid enclaves to exchange ideas, but it will always be a smaller and less diverse population.
2. AI personal assistants, or people, or some combination of both, curate your feeds and content according to your tastes and rules, and filter out the bot content (or don't, it's your rules). In this case, ACX still exists, 90%+ of the feed is bots, but you only see your whitelisted set of people, perhaps augmented by up and comers not on your list that your content guardians think you might like.
I can imagine solutions that are between "open for everyone" and "show me your ID".
One thing I would like to see is a model where people vouch for each other. The only way you can join a group is if someone who is already a member (the first members are the founders of the group) becomes your sponsor. They could testify that they know you, and that you are a human; they could even testify more things, for example that you are a smart and polite human. The idea is that if you do something bad, it will also have consequences for your sponsor. (At the very least, they lose the ability to sponsor other new members.) Some more technical details would need to be solved, for example you should be able to find more than one sponsor, just in case that one of your sponsors leaves the group, or gets kicked out.
Isn't this basically reinventing private, invite-only Discords? Which are fun, yeah, but the diversity / numerosity of participants is usually much more limited than the SSC / ACX commentariat, or the glory days of the ssc subreddit, etc.
You could have read access for everyone, and write access for the members. Maybe even some official way to apply for membership, it's just that the policy can no longer be "automatically admit everyone, ban them later if they misbehave", because that will result in many bots, especially as they get better and cheaper.
Another approach could be some kind of multi-tiered membership, where the elite members are verified humans, and the elite members can (individually or collectively) ban non-elite members. The best non-elite members (as determined by the elite members) could be later promoted to elite.
I think it's going to get seriously bad. It feels like soon, it will be impossible to know what news, music, visual entertainment, advertising, etc. is by humans (if any) and I may just have to disengage from the internet almost totally. Already, like 90% of my youtube ads are AI, and I've had multiple family members share AI content (music, sermons) thinking it's genuine. And I've been off facebook for awhile, but my wife says she sees so much AI slop on there that it's just crazy.
To me, it feels like the world is going to have to revert to something like it was pre-internet (I wasn't alive then so I'm just kind of making up what that looks like for myself). I imagine that it will get to the point where it's impossible to know if anything outside of local news is true. It feels like the internet could become just a massive AI slopfest/hallucination that keeps feeding itself.
But then I guess maybe there's some other technology that might be made to fix everything. I don't really know about any of this; it's just the vibe I get when thinking about it.
My optimistic take is, the "free" internet becomes so bad that people revert to very early stage internet culture (mostly doing things in person, sometimes reading blogs or using resources that people you know have specifically recommended, using semiprivate curated discord-type chatrooms). In my opinion, this would be better than how things have been the past few years.
I think for young people who can actually understand that the free internet is dead, it'll be fine. The problem is going to be older people, who can't adapt and get zombified (Similar to how Boomers got slaughtered by cable TV)
You all think it's the people who grew up *without* the internet, who are going to be "zombified" by AI slop on the internet? You think we're not going to be able to figure out curated blog lists, bulletin boards, and forums? Who do you think invented all those things in the first place?
It's the younglings who've never known a life without a screen providing them constant on-demand content with no effort or curation required, who are going to have a problem here.
I don't know that this is all that indicative of bots. This sounds like people just independently using the same source, the government or the AP or whatever.
I've been having ChatGPT write game design documents for me, and one of its ideas for a final boss was something called an Aegis Engine: an autonomous war machine from the ancient world the size of a mountain equipped with magitech artillery and force fields called up into a land where longbows are now considered high-tech. It's controlled by a "chorus" of bound souls that essentially function as an AI.
ChatGPT seems, well, *enamored* with the concept. It's brought it up multiple times even after I steered the conversation to different aspects of the game, and even once in another chat about an unrelated game! I am officially Concerned.
Cue montage of gathering the whole world, heroes and villains, to join hands and sing Kumbaya until the bound souls are exorcised. Gandalf, the Balrog, Saruman, the Ring Wraiths, the Mountain Caradhras, Smaug; Tom Bombadil, the Ents, The Ungoliant, all the bodies in the Dead Marsh, all the horses of Rohan. Big old anime friendfest.
Just curious, what is your motivation for designing games?
If you're using AI as a source for creativity, aren't you just fabricating something in order to add to an already oversaturated entertainment market? It would seem that you'd have a more meaningful product if you tried to come up with something completely on your own or collaborated with someone else for ideas, instead of generating ideas from AI.
I'm all for using AI as a tool for all kinds of things, but when it comes to creative expression, it almost seems immoral to use AI. There's already so much unnecessary stuff in the world and, for me at least, it makes even existing sometimes stressful. Generating more stuff that doesn't even come from a person's imagination seems harmful.
Then again, I'm just assuming your motivation and goals, so my criticism could be completely irrelevant.
Duolingo went downhill since about 2021. Switching to the "path" thingy, killing the discussion forums, no more speaking exercises or typing exercises (only choosing from the presuggested "word banks" is too simple, you don't learn anything from that). The AI nonsense is just another nail in the coffin.
I have learned a lot of Dutch, Italian and Yiddish from Duolingo, but now it's just no fun any more.
Heartily seconded. I used Duolingo for nearly 2 years in 2020-2021, and as soon as they changed the format to the "path," I uninstalled it and never looked back. I don't know why on Earth they would have thought that removing almost all user choice would make people engage more: it was such a bizarre change.
What is your Donald Trump number? We'll calculate this like Erdos numbers; the Donald's number is zero, anyone who has met him in person has number one, anyone who has met in person someone with number one has number two, and so on.
My number is two, if we let count a very brief meeting with Bill Gates when I was a Microsoft intern. It wasn't one on one; I was one of several dozen geeks standing around him as he held court at a party. That's me -> Bill Gates -> Trump.
If that doesn't count, my number is four. me -> my father (a scientist at Microsoft Research) -> Rick Rashid (head of Microsoft Research) -> Bill Gates -> Trump.
5, maybe 4. Macron met him (1), a French politician met Macron (2), a bunch of kids pranked said politician (3), and either I met one of them (4), or a close friend of mine met one of them (5).
Likely four or five, through a slightly odd route. I've met a few people in the Japanese music industry, one of Trump's children apparently got drumming lessons from Yoshiki of X Japan.
(Alternatively, my sister works in government in my home country, so probably 4-5 through her as well).
3 at least - Cardinal Nichols, Pope Leo, Donald Trump.
I've been to a funeral and Tony Blair was there, but I didn't meet meet him, that could be a 2 if you're being generous.
[Another 3 - Tony Francis was a soccer presenter and I got his autograph, he often did shows with ex-footballer Jimmy Greaves, who had his own show called Saint and Greavsie, which included the Donald among it's guests (they visited him in Trump Tower)]
Presumably three; a relative is influential enough to regularly speak with their Congress Rep. The question is if the Congress Rep is in turn important enough to speak with Trump directly.
Two. I met Nancy Pelosi briefly when she visited the Google campus c. 2010 and had a meeting with some executives near my office.
I also independently have a Trump number of three through multiple paths:
1a. About a decade ago, I bought some made-to-measure suits from Thomas Mahon and met him twice for measurements and fittings. The made-to-measure service was a sideline Mahon was experimenting with to his main business of fully bespoke suits, for which King Charles III has been a long-standing and frequent client. King Charles has met Trump on at least one state visit.
1b. When my uncle was in the Army (military police, retired as a Lieutenant Colonel), one of the things he did was security for high-profile state events. In this capacity, he met then-Prince Charles when the latter was visiting the US at some point in the 1980s.
1c. John Major gave a talk at Foothill College (a local community college) when I was a student there c. 2000. He also did a private appearance to a much smaller group for a club I was a member of, connected to this, and I met him in passing there. This, again, give me a King Charles number of 2.
2. My mother did some consulting work for the Librarian of the Senate in the mid 1990s. I went in with her once and met some of the senior librarians, who have presumably met various Senators including some who have met Trump.
3. An ex-girlfriend of mine has met John McCain by virtue of being friends with his two youngest children (Jack and Meghan McCain) in high school. I'm pretty sure John McCain has met Trump.
4. When I was a little kid, my father was heavily involved in a big computer user group in the DC area and organized a lot of speaking events for them. Richard "Lord British" Garriott was a frequent guest whom I met several times through this. Garriott has met Hillary Clinton (he was a major fundraiser for her in 2016), who in turn has met Trump.
5. Early in my career, I worked for a startup called Intrinsic Graphics. The VP of Engineering, who was my direct manager or my grandboss at various points, went on to become an executive at Google and later at Uber and also was heavily involved in politics (first as a fundraiser and later as a candidate for Secretary of State in his home state). He's met Barack Obama, who in turn has met Trump.
I've met former NASA Administrator Mike Griffin. Trump made him Under Secretary of Defense for Research and Engineering in 2018, and almost certainly had interactions with him.
As an exercise, I was curious if I could find direct evidence that Michael D Griffin and DJT had met face to face. With naive googling and AI assist, I could not.
I should add that I also could not find direct evidence showing that Heidi Shyu (Griffin's successor) and Biden met face to face.
Oh, I hope this doesn't come across as antagonistic. I am not doubting or disputing your DTN. This was just a little point of curiosity that was prompted by your comment.
Is there any AI system which can listen to an instrumental tune, on YouTube say, and transcribe it with reasonable accuracy into musical notation with a full score, i.e. chords etc?
If you mean "LLM", then the answer is a clear "no". If you mean "Machine learning" in general, then maybe.
I guess you could take the Fourier transformation to look for the base notes, and then do some hidden-markov-magic to find the original note. This on it's own would not be much machine learning, but I guess you could turn the problem into a Markov Decision Processes and have some AI-Agent solve it. (A quick websearch yields this for implementation of AI solving MDP https://sites.gatech.edu/omscs7641/2024/04/09/mastering-markov-decision-processes-a-practical-rl-journey-with-openai-gym/ )
The next question would be if someone has done this, and I think this is a yes. Another websearch suggest, that the ultrastar/karoake community has done some work in this direction:
Thanks, I was listening to a splendid rendition of the old latin hymn Jerusalem et Sion filiae at Notre Dame cathedral ( https://www.youtube.com/watch?v=Soh82AfRPdw ) and in reply to one commenter asking if a musical score was available, another said none was and the organist basically makes it up as they go along!
I know some people who might help, I'll get back to you.
[Organist 1: No, sorry. I'd be very interested if you can find one. My efforts with getting ChatGPT to "compose" music using python code rendered as sheet music via Lilypond were interesting but musically bland. I know there are packages that can create sound files that are very sophisticated.]
A few months ago, there was some brief discussion of nicotine in here that included a link to https://gwern.net/nicotine. I ended up giving it a try, my usual safe and boring inclinations being overcome by Gemini, when it threw me the massive surprise of going against the public health zeitgeist to say "yeah minimal risk, go for it unless you have a condition that should avoid elevated heart rates".
My experience: it's pretty good! Not life-changing, but definitely life-incrementally-improving.
It does a great job cutting through, as a quote in the gwern writeup put it, the "ugh field". When I decide I want something done and am going to take nicotine to help, "it's hard I don't want to" stops being a problem and the thing gets done. Great for piled up house chores in particular.
I tried it once when working on some rather harder than usual code, and... eh, not really seeing it. I am not at all into nootropics in general, though. Kind of relieved, honestly - "oh no I'm addicted to doing chores and exercising" is a good problem to have, "oh no I feel compelled to do drugs every time I sit down at the computer" is not.
I get results at the minuscule dose of 1mg (maybe relevant: I'm quite sensitive to caffeine as well). I first tried one whole lozenge (2mg) and had to bail out 2/3rds through because I was getting uncomfortably wired. I never do more than 1mg/day, with gaps of 2-10+ days between uses - I've been keeping track. Although I have been using it sparsely, I think I'm noticing some tolerance: the first few times I felt like I was taking *a drug*; more recently it's less pronounced. This was an unpleasant surprise, but fine unless continuing on this schedule eventually zeroes out the effect.
Subjectively, for the purpose of productivity, I find it vastly better than caffeine. Unlike caffeine, no emotional boost (thankfully, for addiction purposes).
Also subjectively, at this level of use I perceive zero addiction risk. I took a couple of lozenges on a 2 week trip in case of nasty airline delay bureaucracy issues. Such issues didn't arise, didn't use them, and didn't feel tempted. It seems like slow absorption through gums vs the spike of smoke in lungs is a world of difference, addiction-wise. Not sure how easy it would be to ramp down a lozenge-only nicotine addiction - likely and hopefully vastly easier than actual smoking.
So: thanks to the people who brought it up back then! It still feels a little crazy, but I am glad I started using nicotine.
Fred, your post can't be viewed in the UK, due to "Age-restricted content". Can you believe it?
It appears anything related to nicotine (its topic, as I gather from the replies) is considered off limits to any sensitive snowflake under the age of 18! Whatever next!
Holy moly. Well, to be fair, if you're going to be doing vigorous age-based censorship, enthusiastic praise for an addictive drug is at least an entirely reasonable target. But yeah, that's ominous.
Long term recreational/productivity nicotine user here. Sample size of one, but I don't find that "nicotine by itself" is anywhere near as addictive or habit-forming as combusted, and then any delivery route besides the lungs is another addictive downgrade beyond that.
I'd still be pretty careful. I'm a lifer; long after paternalist public health people have tried to ban it, I'll be growing nicotine in my back yard or ordering it from shady chinese sources. There but for the grace of god go you.
Thanks a lot for this input. I tried venvanse for my mild ADHD and it worked great, but at what cost? (dependency, expensive medicine and psych appointments, health risks). Maybe nicotine may be the drug for me
I recently started a Substack (alont.substack.com) aimed at explaining AI capabilities and risks to the general public.
My most recent post breaks down the "stochastic parrot" dismissal. Next week I'm planning to tackle "AI will create new jobs we can't imagine yet, just like past technological revolutions."
Have any other dismissive arguments you'd like to see covered? Any feedback on the posts so far?
I am perplexed at the seeming non-overlap between people who swear AI spells doom for humanity and people who intentionally abstain from using LLMs and other generative AIs on a regular basis. Curious if you’ve noticed anything similar and had thoughts about it.
I've heard people claim that AI extinction fear is an psyop by the tech industry to trick people into thinking "Wow, AI is powerful and a Big Deal. These guys building AI must be very stable geniuses."
One among several explanations: Saying that brand new technology could destroy the world, in and of itself, would seem to open the door to thinking that brand new technology could SAVE the world. And that goes against the grain of the technopessimism of anti-AI absolutists.
This partly explains why committed climate activists tend to emphasize the long, slow process that "got us here." They don't want to open the door to technological solutions, only ideological solutions.
If we analogize to climate activists, it would be as if no climate activist ever suggested driving less or buying fewer globally-produced items or doing things to cut down on electricity usage, etc. As far as I’m aware, many climate activists do advocate at least some of these measures.
Al Gore flying around in a private jet while preaching the end of the world from the burning of fossil fuels is much more reminiscent, to me, of AI doomerists who nevertheless regularly use AI and never suggest its use should be curtailed.
Even Scott, for example, who seems to be generally inclined toward causes like AI safety and alignment, and takes seriously the claims of doomerists like Yudkowsky, regularly decorates his blog posts with obvious AI art. I know one person abstaining from LLM use isn’t going to turn the tides on AI, but high profile people demonstrating a behavior both sends a powerful message and has memetic resonance. Sam Altman, and his investors, would take notice if there was a trend in the rat-sphere of people intentionally reducing their AI use.
AI companies are not profitable. They rely on venture capital. The investors look at things like LLM usage rates when determining whether and how much to invest. Those investments are what allow the AI companies to continue development further, and what will allow them to eventually reach AGI/ASI or whatever you want to call it, if that is to be a thing. That is the state of affairs right now, as I understand it.
A high profile movement (even if it is only high profile within the Silicon Valley culture bubble) of abstaining from AI use would surely be noticed by investors and developers. Why don’t you think it would have any effect?
I think there is a tendency for people to implicitly have the assumption ~"If something is (very) intelligent, it will be friendly/peaceful/cooperative/share my particular values". Probably that.
Regarding: "AI will create new jobs we can't imagine yet, just like past technological revolutions."
I think this is actually true. Even with full AGI we might have cases where human input is required, if nothing else due to legal reasons. This may still be fewer jobs than what were overtaken, but still, technically true. Without full AGI it most likely will be the case. (The situation is probably different with real super intelligence though)
I think the difference is that in the past, if something created jobs, it automatically created jobs for humans. Even the animals had to be supervised by a human.
And yeah, humans will keep the bullshit jobs. That still doesn't necessarily guarantee that there will be enough bullshit jobs for everyone, or that the jobs will not be horrible.
Thank you! If I find the time, I will message you.
The main reason why I asked was whether the AIs are now smart enough to do the entire job (I don't make pictures with AIs, so I had no idea what is the current state of art), so I take it the answer is no.
In my binge I just noticed that tons of ACX comments are now deleted. I don't think it was like this even few months ago. Anyone has an idea if there's anything in particular that may trigger this?
For instance, here https://www.astralcodexten.com/p/why-match-school-and-student-rank the oldest 7 comments (not replies) are deleted. Doesn't seem like spam because the replies treat it as legit. Doesn't seem like a flame bait thing either especially since there's no "banned for this comment" that Scott usually explicitly states.
Two big reasons. One is if someone deletes their Substack account, I think that deletes all their comments ever. Gunflint, a long-time commenter here, deleted his substack account a couple months ago and is probably responsible for a significant portion of the deleted comments all by himself.
The other is when Scott bans someone, there's apparently an option (enabled by default) to also delete all of the banned person's past comments to this blog. Scott has said he usually unchecks this when banning people but sometimes forgets to do so.
Yeah, I commented as Gunflint. I was drifting away from the forum for various reasons and the “I’m glad Charlie Kirke had his head blown off” commenter was so disgusting and infuriating that I felt I had had with ACX. Doesn’t Scott have a backup mod to handle stuff that awful? I figured making a clean break by deleting my own account would be for the best.
Sorry all my prior comments were removed and for any confusion caused by that. I don’t think I was aware that would happen. It was a moment of disgust and high, I’m going to throw my phone through the window — open or not — anger.
JFC — *Glad* that a young man was murdered in cold blood?
Well, Tyrone, while you were gone we solved AI alignment, agreed on the optimal ratio of compassion to SAT math score and got John Schilling to admit he kisses his kittens' tummies.
I completely understand. I, too, think the comments here could do with a bit more moderation, and would support Scott recruiting a mod team to help out. And the mass deletion thing is a bit of weirdness that I blame the Substack platform for.
I've missed your comments, and I'm glad to see you're back.
If you're triggered into irritation and judginess this easily by someone else's sensitivities you may have scant potential for making high quality contributions to discussions.
Plenty of people *out there* are so delighted. Not the case in here, and the ones who do feel that way are polite enough not to rub people's noses in it. We'd like to keep it that way.
It seems like every time you mention one of the GMU economists (Caplan, Tabarrok, Cowen, Hanson), the post title is something like "Contra [GMU guy] on men having more teeth than women", and the post goes something like"
"I have enormous respect for [GMU guy], he's a brilliant thinker and one of my intellectual idols. Here, he claims that men have more teeth than women because he asked ChatGPT and that's what it told him, and he bases his entire argument on this premise. However, I've compiled a comprehensive list of sources showing that men have the same number of teeth as women. It's unclear why he didn't double check this claim given that his entire argument rests on it. Also, his argument that women are p-zombies would be sufficient to prove that men are p-zombies too, which undermines his perspective on gender differences."
What's the best explanation for the gap between your expressed admiration for these thinkers and the way you dismantle their arguments in ways that seem really embarrassing for them? Is it just the selection of arguments that you choose to respond to?
Whoever selected the daily links on Marginal Revolution - was always very credulous, not apparently given to quick googling. Either that or he thought that interest might flow from nonsense.
If the GMU guy is getting his biology from Aristotle, chances are good he's a Thomist and I have to be sympathetic there.
People like to chortle over that, *but* I think that if he was getting numbers from someone who went out and counted teeth, there would be a reason for such a result. The old saying was "you lose a tooth for every child", because of mineral depletion (and there's some evidence that pregnancy does affect oral health: https://www.dakotadental.com/patient-information/blog/gain-a-child-lose-a-tooth/)
So if Aristotle or his source was counting teeth of adult women versus adult men, those women being of child-bearing age, there's a chance that indeed the women would have fewer teeth.
I think Aristotle's stated reason for the idea that women have fewer teeth than men is that somebody else told him once that mares have fewer teeth than stallions. Which I think is probably not true either, but I can't be bothered to make sure.
Bryan Caplan makes arguments at book length with sources that make me think more carefully about an issue. He has generously advised me on some decisions about how to allocate my time via email.
I read Tyler Cowen’s link posts, which are occasionally interesting, but I have no interest in his writings or opinions. Once you understand his schtick, you understand that’s all it is - a schtick. He makes a gut call on what’s over or underrated in the zeitgeist and then just shares whatever points in the other direction without engaging with object level arguments, or even necessarily appearing to read or understand the sources he links. He does a great job of MC’ing a podcast, though.
Tabarrok seems to go deeper into details, is sometimes interesting to me, but mostly seems to be holding an argument with people who have very different forms of skepticism than I have.
Hansen is just creepy to me. Have never been interested in his stuff.
"Is it just the selection of arguments that you choose to respond to?"
Nope. Low effort, often wrong hot takes are a staple of the Marginal Revolution crew. And any time they post a study in a link, you can already predict the results if you know their biases/ideology. They simply don't have the truth seeking drive that Scott has, to take in all of the evidence and then let the chips fall where they may.
They are usually right on the particular branches of economics that they are familiar with but anything else is a crapshoot even if it's economics adjacent.
Having said that, they have had contrarian takes that were vindicated later. For instance, splitting mRNA doses as well as a first doses first approach (there are obvious mechanistic reasons why these were a better approach but the FDA is allergic to mechanistic epistemology - if you haven't followed the exact dosing procedure established by an RCT, then FDA will say there's "no evidence" to support that practice).
I hope to someday see Scott and ACX fans take on a more jaundiced, cynical view of economists in GENERAL. I'm hard-pressed to think of a newish profession that has exaggerated its achievements with such success. And what prestige they have won! At times, they really do remind me of secular priests.
The global economy is deeply influenced by psychological factors. It astonishes me how blithely economists solve mysteries of the mind that have tortured psychologists for generations.
Economics is a discipline that is ready for reform. EXACTLY what that would look like is probably above my pay grade.
Generally, contrarianism is usually not driven by truth seeking. It is about signaling that you are smarter than the "sheep". So in the case the sheep happen to be right, you signal your superiority by being wrong in a clever way.
Since I usually post these in Open Threads, but happened to have reported it in a response in the "The New AI Consciousness Paper" post, the tl;dr; is:
Gemini 3 Pro 11/21/2025 7 questions, tl;dr of results:
I didn't ask Gemini factual questions. Instead I invited it to debate me on alignment strategies [https://rationalreciprocity.substack.com/p/gemini-discusses-alignment]. IMO, it clearly held its own, understanding my arguments and detecting their weaknesses while being willing to admit the flaws of its own arguments. I'm impressed, perhaps for the first time by a Deep Mind product since AlphaGo.
I asked Gemini one question, and it failed miserably.
"Which genes in Southern European and Eastern Mediterranean Neanderthals likely indicate lighter skin and eye color?"
I didn't save the answer, but it basically said it didn't know. Which, I suppose is good in that it isn't feeding me bullshit. Then I went over to ChatGPT and asked the same question, and ChatGPT gave me a nice breakdown.
I asked Gemini the same question again this evening (to recreate the failed response), and it spewed out a response suspiciously similar to the ChatGPT answer. Are LLMs spying on each other? ;-)
Possibly. But it answered the same question correctly two days later. Is someone keeping an eye on the guardrails and tweaking them as they see legitimate questions get flagged?
I know they're getting better about not hallucinating, but they'll at least try to give you some basic Wikipedia-first-paragraph generalities. Maybe what it gave you is closer to that than what I'm imagining. But, if it was really an "idk", especially if just one or two sentences, that's likely a stealth refusal. I've gotten that when e.g. asking about mlock() (apparently mlock is a gun term).
I mean, come on, genes and skin color? You were asking for it! (that was bleak sarcasm about the ongoing state of the corporate world, not an actual admonition)
IIRC it said it didn't have any references for papers on Neanderthal skin coloration. I was surprised because I knew there were a bunch. It gave me a reference to an old YouTube video on the subject. LoL! But now it seems to have those papers in its training data. Strange.
Ok, yeah, a refusal wouldn't even give you that. I guess the first time you got the unlucky draw when the generation reached "Sure! Here is a [detailed survey of the relevant literature: 0.9, random old youtube video: 0.09, suicide hotline: 0.01]"
Hmm... Well information sometimes leaks... Now, if Gemini has ChatGPT's _weights_ that would be ... interesting...
I suspect that what was happening is that there are some nondeterministic elements in how (some? all?) of the LLMs respond, notably usually a finite "temperature" for selecting which of a number of probable-but-not-certain next tokens to pick. IIRC, only at "zero temperature" (typically not the default?) is the most probable output always picked.
At some point I want to write an essay on what I call the Monte Carlo theory of politics.
There's certain forms of Monte Carlo optimisation where the algorithm goes as follows: You are at a point. You pick a random point near that point. You see whether that point is better or worse than your current point; if it's better then you always move there, and if it's worse then you sometimes move there, depending on a random number and how much worse it is and a temperature parameter that helps you escape from local minima.
I can't help but see this as analogous to the process of politics, and the tension between "progressives" and "conservatives". The political landscape is complicated, we can't really tell whether a new policy is better or worse without trying it out or at least getting up close to it. It is the job of "progressives" to randomly push us in new directions (which they delude themselves into thinking will definitely be better) and the job of conservatives and reactionaries to argue against these changes; over time we find that some of the changes they suggest are actually good and we keep them (and conservatives pretend they were never against them) while others turn out to be bad and we reverse them (and progressives pretend they were never in favour of them).
This is a positive and encouraging view of politics, in which the boring day-to-day political conflict we see around us is just the execution of an algorithm which tends to make things better over time. But it only works properly if the progressives are sufficiently random. It's a big problem that right now "progressive" has become synonymous with "far left", with no right-progressives out there proposing changes in the general set of directions we could label as "right". It's a big problem if right-wingers can't dream any bigger than conservatism; we need right-progressives out there. And that's why the other essay I want to write sometime is a right-progressive manifesto setting out some kind of vision of Fully Automated Luxury Straight Space Capitalism.
>And that's why the other essay I want to write sometime is a right-progressive manifesto setting out some kind of vision of Fully Automated Luxury Straight Space Capitalism.
Off the top of my head - maybe mostly inventing useful new markets? E.g. we had a review of the Alpha schools recently. Maybe a market in predictions of lifetime earnings of the Alpha school graduating class of 2025 would be a useful new market?
The dynamic you're describing might have existed 20 years ago, but it's laughable to claim it exists in 2025. The Trump administration has launched bold experiments all over the map, such as "what if we put massive tariffs on our allies?", "what if we try to fight the war on drugs using drone strikes?", "what if we give a single billionaire admin access to all government payments?", and "what if instead of encouraging vaccination, we had the CDC try discouraging it instead?"
None of these experiments have been successful, but you can't say Trump hasn't been willing to experiment or that the current right wingers are unwilling to dream big!
However, Elon has abandoned DOGE and DOGE has mostly been disbanded, so that one can definitely be called a failure for sure. And the tariffs have so far been partially kept, partially abandoned (it's been changing too quickly for me to keep track of where they are right now.)
I think there are two big problems with your theory. The first you allude to yourself: progressives don't actually random-walk their pushed ideals, on the contrary they cling to them long after they're demonstrably failed, stupid notions. Socialism has been clearly, demonstrably idiotic for over a hundred years (I would argue nearly two hundred, since the great Bastiat punctured it); the prog position on the trans issue is clearly ridiculous and self-contradictory in parts, etc., but this makes absolutely no impact on what issues progs push, instead they clamp down as hard as their teeth will allow on the same old bone come hell or high water. The only thing that seems capable of making them shift gears even temporarily is a gross PR defeat so massive they realize that the rest of their project is doomed if they cling to this one; thus for example they dropped pushing eugenics after it became clear that the Nazis were going to lose WWII.
The second problem is that your definition of "the changes are good and we keep them" seems to be just "it is successfully passed into law". Prohibition was terrible and got reversed and approximately nobody is mad anymore, yes, but suppose we use Roe as an example of something which turned out to be bad and was reversed: the progressives are not exactly accepting this as the end point of an experiment, are they? Or to take an even more acute example, suppose we were to conclude that giving women the vote was a mistake which distorted this entire politicalgorithmic process: how would we even reverse that? I mean, how would one go about it in practical terms? Are we to accept it as a priori good in your model simply because it's impossible to undo? This seems like a highly absurd definition of the good. And moreover, suppose we *did* reverse it? Do you imagine that this would simply be accepted? No outpourings of fury, grief, scorn? Just "oh, I guess that was one of the duds like sterilizing the retarded and Prohibition"? This does not map *at all* to how I see the progressive flank working, which seems to clash sharply with your theory.
"progressives don't actually random-walk their pushed ideals, on the contrary they cling to them long after they're demonstrably failed, stupid notions. Socialism has been clearly, demonstrably idiotic for over a hundred years (I would argue nearly two hundred, since the great Bastiat punctured it)"
In what sense has Socialism failed?
Socialism is a broad term that encompasses many different viewpoints, a spectrum of approaches of market vs central planning, a long history experiments with varying success rates.
It is silly to say a whole idea category that large has failed. It would be similarly silly to say capitalism has failed, because of any one single example.
Clearly a more nuanced take is needed, ie:
Centralized planning can have certain benefits in certain societies, like maybe making big decisions in infrastructure (USA highways?), or telling all your banks they must now lend money to startups in certain sectors like electric vehicles (china),
And obviously market forces and competition can be an effective and productive way to organize an economy, but maybe sometimes you need to be careful about externalities.
Maybe certain outcomes are better in countries that are on a different place on the socialism spectrum, ie Nordic Countries.
It would also be nice if people could stop confusing the socialism/communism <->capitalism spectrum with authoritarianism<>democracy spectrum, as if you can't have an authoritarian capitalist state or a democratically elected Socialist state.
I realize your main argument was about the above "monte carlo theory of politics thing" but I find all your drive-by arguments that you take as settled fact quite annoying... abortion bad, socialism bad, women voting bad, socialists-are-all-eugenicists, trans bad, socialists are all nazis.. etc.
It would be helpful if you would first engage with anything I wrote on more than a strawman level, e.g. "I find all your drive-by arguments that you take as settled fact quite annoying... abortion bad, socialism bad, women voting bad, socialists-are-all-eugenicists, trans bad, socialists are all nazis.. etc." is either a preposterous failure of reading comprehension or a deliberate malevolent misdescription.
There was a great deal in that comment there before "I find all" that looked to me as exactly the sort of engagement I would ask for. Including a great deal of not-really-pro-socialism points.
Not that you're wholly in the wrong here. Most of that comment also didn't seem to engage your point about the effect of reversing established law, but instead went off into its own discourse about economic system tradeoffs.
Ownership in this context means you're able to make all decisions affecting whatever it is you have ownership of. This includes planning decisions. People who oppose socialism oppose it in large part because it puts people in charge of planning for resources that they can't practically maintain. People in Washington won't know what plans will be better for forests in California, nearly as well as people in California; people in southern California won't even know what's best for forests in Mendocino County nearly as well as people actually living in Mendocino.
Costs and benefits of planning decisions depend not only on one's values, but also on one's property. You'll care a great deal more about your property than you will about something that isn't.*
Don't discount ownership.
*Some people have so much property that they can afford to be less careful with some of it. But this also means they'll trade it for less than they would if they had less. This is good news for people who have less, and is a feature of systems with private property, and a flaw in systems that don't.
Yeah. The first approximation is that progressives want to try new things, and conservatives want to keep old things.
But in real life you get progressives who keep wanting the same "new" things for decades, and conservatives who want to return to a "past" that has never existed.
Similarly, progressives are against hierarchies... unless it is their people at the top of the hierarchy. And conservatives are for individual responsibility... unless there is a moral scare and you need to protect children from Rock n Roll, Dungeons and Dragons, or Harry Potter.
"conservatives who want to return to a "past" that has never existed"
I might quibble with this, I think the vast majority of conservatives just want the 19th century back but with better technology and a considerable minority would be satisfied with the 1940s-50s, but besides that, exactly.
I'll steelman Villam by arguing that most conservatives have little idea what it was actually like in the 1940s or the 1800s. And that they would be horrified if they got their wish.
Just for starters, take the Barstool/Hinchcliff style conservative who relishes the opportunity to say whatever they want whenever they want, however rude or belligerent. I wouldn't bet on this type of person lasting more than 1 month in the late 1800s before getting shot in the back of the head.
Want to randomly chat up unaccompanied attractive women you've never been introduced to? Careful, you might be targeted for violating a social norm against "mashing," especially if the women have higher social status than you. Try to flirt with the wrong woman and you might get stabbed in a vital organ with a massive hatpin; and the police will take her side!
"the vast majority of conservatives just want the 19th century back but with better technology"
The vast majority of conservatives want to bring back racial segregation, bring back Jim Crow laws, take away the right to vote of women and black people?
For 65 years of the 19th century, slavery was legal, is that what they want to go back to?
The Fifteenth Amendment was ratified in 1870. The vote for blacks is a 19th century phenomenon. If you can't get a majority of self-described conservatives to admit that they, at most, would care very little if women lost the vote, they just don't trust you, that's all.
That said, yes, I do think a significant cleavage in the conservative population is whether they would prefer to return to post-Civil-War or antebellum America.
You model sounds reasonable. I’m not sure why Republicans don’t qualify as right-progessives in your model. It seems to me that Trump is randomly pushing us in new directions.
Yeah I think to some extent right-progressives do exist, but consider themselves conservatives (or reactionaries), and they wind up in the centre-right party alongside the actual conservatives, which causes a lot of confusion.
Right-progressives need to realise, and to embrace, what they actually are, which is a big part of the reason that I want to write these essays.
Trump and his supporters seem to fit "right-progressive" really well, if I'm understanding your account. He certainly seems to be trying all sorts of daring and random initiatives, be it moving the embassy to Jerusalem, slapping huge tariffs on half the world, backing off on half of them, rebuffing Zelenskyy, meeting with Putin, playing coy about buying Greenland, launching DOGE, hunting down and deporting illegals, blowing up drug boats and nuclear weapon sites, and claiming this or that election was rigged. I think everyone agrees he's shaking things up, even if everyone disagrees on whether that's good.
Yes, I think you're right, there's definitely a strong right-progressive streak within the Trump administration.
So I should probably alter my original statement a bit in the next draft -- the problem isn't that right-progressives don't exist, just that they fail to think of themselves as such or to understand their role in this whole process.
I'm not sure EXACTLY what you are referring to when you wrote "burning money." That said.
I heard "recycling is a scam" for years, and it sounded plausible. When I actually looked into the logic behind this (mainly) rightist argument, my jaw dropped.
"Recycling is scam because only 40 percent of the stuff you put in the bin actually gets recycled" is, in itself, one of the biggest scams ever perpetrated by contrarian content creators. Recycling 40 percent of the bin is inadequate, so let's just do away with recycling altogether. Got it.
Trump is, to a large extent, *causing* the end of Pax Americana.
And no, there is no geopolitical imperative for a post-Pax America to own Greenland or Canada, any more than there was any such imperative in the pre-Pax era. Also, if the United States annexes Greenland, then our northern border is now threatened by Iceland, Norway, and the UK, so U guess it would be impossible to allow those nations to remain un-annexed...
This isn't related to your point, but the algorithm you're describing is called simulated annealing. Monte Carlo is a broad term that basically includes any algorithm that relies on random guesses, including simulated annealing.
I'm _think_ I'm describing a class of Monte Carlo algorithms that is somewhat broader than simulated annealing, but narrower than all of Monte Carlo. Simulated annealing in particular would include some very specific things that I don't necessarily want to get into, like gradual reduction of temperature.
I think my "There's certain forms of Monte Carlo optimisation where..." intro is correct.
You're describing Markov chain Monte Carlo (MCMC). When MCMC is used for optimization people usually have the temperature slowly go to zero. MCMC with a slowly decreasing temperature is called simulated annealing.
That's certainly true, but doesn't the gradual temperature reduction kind of match how over time we argue over smaller and smaller refinements to our policies? Like how the US implemented income tax with the 16th amendment (a big change to tax policy), while today the parties argue over policy changes like closing loopholes or giving tax credits to working families that would maybe change overall tax revenue by a percent or two.
Edit: another example: the left debating whether the default name for the master branch in git repositories should be changed to "main" branch, because "master" may be an offensive term. This seems to me like a very small refinement with minimal impact on people's lives compared to when politics was debating whether we should ban actual slaves.
The size of a policy change is pretty subjective. Marriage law has been the subject of minor refinements (changes to divorce law and so on) over most of U.S. history. Then we legalized gay marriage, which from the perspective of social conservatives amounted to redefining the entire concept of marriage.
Conversely, for most of U.S. history the U.S. has considered itself honor-bound to adhere to it’s international commitments. This goes back to the founding fathers, who included in the Constitution a provision that treaties signed by the United States are the supreme law of the land. But these days, if Trump decides he doesn’t like the USMCA, which he signed during his first term, he just ignores it. That’s a fundamental change in how the United States conducts international diplomacy.
I'm thinking the metric to use is roughly the impact the policy change has on people's lives. I agree the size of any policy change is debatable.
By impact on lives, and I say this as someone strongly in favor of gay marriage, the direct effects of legalizing gay marriage are tiny. It's a symbolic victory. It shows we consider same-sex couples as equal. Sure, people involved in the debate on both sides often claimed it would have a big impact, but people overestimate the real impact when they're fighting passionately over a political issue.
> But these days, if Trump decides he doesn’t like the USMCA, which he signed during his first term, he just ignores it.
That sounds more like having an unreliable government rather than an actual policy progressives are advocating and conservatives are resisting.
I have progressives split into two camps: liberals want to advance new individual rights, socialists want to advance new methods of improving the collective good. Conservatives just oppose everything new in case it's bad, which it often is.
The problem at the moment is that liberalism is tapped out for now, but virtually all progressives (even the ones who call themselves socialists) haven't grokked that yet and are sticking with liberalism. Change is coming but it could take a decade until socialism becomes dominant among progressives.
I think you'll see more socialist-adjacent candidates over the next few years, given the realities of political calibration. Like, take Mamdani: he's not going to be able to implement all of his promised reforms, because of the structure of government in NYC. As one would expect, he began moderating his stances between the primary and election. What will happen, though, is that he'll move from far-left to an at-least conventional (D) profile, e.g. by siding with workers or otherwise recognizing the deficiencies of capitalism as a central organizing force. Something similar has happened with AOC.
Now contrast them with moderates like Schumer/Jeffries/Booker/etc, who take token-left views on identity politics, maintain neoliberal economic views, and swing to center on everything else.
By voting for a leftist candidate, at least after the adjustment you get a real Democrat if they're elected.
Leftist here. I am not diametrically opposed to seeing greater automation and fewer jobs for the underclass. Theoretically, this can only go so far until you give the underclass a UBI. Or until get get fed up and revolt, possibly with assistance from cynical foreign actors. Seizing control of society, robots and all.
It's the old left-accelerationist argument. I'm not opposed, and I'm not fully sold; I'm pretty ambivalent on this. It's obviously a big risk but if you've tried everything else...
Not sure how you came to the conclusion that increasing the MW is a 'conservative policy', but historical data is mixed, and there is no A => B for decreased employment or robot buildouts (Walmart still hires human 'associates', last I checked). I also think that most progressives are intelligent enough to realize that the issuance of food stamps has more than the one effect you're claiming here.
There have been a number of articles about chains abandoning or scaling back self-checkout. (Ex: https://nbcnews.to/4oh7Nrc) because of "shrink" in low-income areas. This suggest to me that a likely outcome of eliminating food stamps would be a quick uptick in shoplifting.
I'm in agreement with you on nutritional standards, and that relaxing those amounts to corporate grift. I also recognize a need for personal responsibility, but I don't think that kids should go hungry because of the poor decisions their parents make. I'm on Team State Paternalism in that regard.
"The whole modern world has divided itself into Conservatives and Progressives. The business of Progressives is to go on making mistakes. The business of the Conservatives is to prevent the mistakes from being corrected."
It's shocking how relevant Chesterton still is. Well, you're a Catholic, maybe it's not shocking to you. But to me there's always something alarming about reading a clever essay about some highly current topic and realizing it was written in 1907.
> Conservatives just oppose everything new in case it's bad, which it often is.
Point of order: this is not all that conservatives do. They also want to change things to how they think they were in some idealised past. (In practice, the result of such change is usually some new thing, with everything conservatives say about new things applying to it.)
It seems to me that you are making a serious error in stating that only one group wants to improve the collective good. They pretty much all do, but they disagree about what constitutes the collective good. Not to mention what constitutes the collective. And all wish to advance individual rights to some extent, but often disagree re whuch rights.
Within the camp where people "want to advance new methods of improving the collective good" there's left-wing and right-wing subcamps though.
Leftists always want to advance in the direction of greater equality, while rightists want to advance in the direction of... greatness? Quality of life for those who deserve it? Things along those lines.
The progressives want to keep people from doing as they please. Conservatives want more liberty. Witness that progressives are always arguing for more regulations, and conservatives for fewer.
They don't want more liberty when it comes to "should women have the liberty of choosing an abortion when they find it most convenient", and they don't want more liberty when it comes to "should Mexicans have the liberty of moving to the United States if they want", among other things.
Everyone anywhere in the political mainstream, is in favor of more liberty in some areas and less liberty in others. The ones saying "this is the party of liberty, those other guys just want to tell everyone what to do", are mostly just telling you which kinds of liberty are most important to them.
If your argument is that you can't advocate for liberty unless you're an anarchist, you are going to miss a lot of subtleties.
It is progressives, not conservatives, who argue for public transit (so you can't go anywhere they don't send it), for more regulations on everything, for judges who can't be voted out to make decisions that are nowhere found in the Constitution, and many other freedom constricting things.
Everybody "advocates for liberty". Some liberty. The liberty to do the things they want to do and that other people don't want them to do. When e.g. progressives advocate for people to be allowed to chose their gender and use whichever public rest rooms they prefer, they are in fact advocating for liberty. And then some of them demand that you be punished for not using the pronouns you think are appropriate for the occasion, which is advocating for the opposite of liberty. Conservatives do the same thing, with different issues.
You can't claim a *monopoly* on being an advocate for liberty unless you are an anarchist, or perhaps a principled libertarian. I encourage you to go on advocating for liberty. Just, don't think your tribe are the only ones doing this. And try a bit harder to avoid conflict with the other people advocating for other liberties.
The conservatives camp, from a posiwid position, mainly want to enforce hierarchies. Any hierarchy. Racial, class, gender, religious, whatever you got. This maps more on to what they actually do, because given a choice between improving quality of life or "greatness" or prosperity, they will instead enforce a hierarchy. For example, we are pretty dang sure that some level of wealth redistribution or trust busting is good for all three of those, but because it doesn't reinforce hierarchy a conservative progressive would not do it, as a rule.
The Left camp is the same but backwards: flatten all hierarchies, and we'll figure the rest out later
> Who are "we" and when has the repeated attempts at "wealth redistribution" in the 20th century not been bad for all three of those?
Non heterodox economics and also about half of heterodox economics, along with raw correlation, and for the second part: the post war boom, the other post war boom, the reconstruction boom, the japanese economic miracle, etc and so forth. Most of the boom times post about 1880 are associated with a period of wealth redistribution through policy or through systemic collapse flattening the peaks to be closer to the valleys, as it were. That is a load bearing most, but it's the best you can hope for in the social sciences unfortunately.
> And why did you lump in "trust busting"?
Because the greatest periods of wealth redistribution post 1800's coincided with the greatest periods of trust busting, so IMO it's hard enough to disentangle the effects it's better to hedge.
Yes, I was using the archaic definition of progressive (one who wants more progress) as opposed to the far more commonly understood meaning of the word in today's America (one who wants more social liberalism).
It's actually the popular quadrant among voters, but for some reason that I don't fully understand yet no politician is willing to appeal to it.
I think it's something to do with the way a 50% requirement in the legislature forces politicians into two ur-factions. But even so that doesn't really explain much in proportional representation systems.
It is a popular quadrant among voters outside of the United States. And even within the States, there are some vague stirrings that could be interpreted as signs of life in this quadrant.
At no point in history has the entire of society consented to its various changes. By your estimation, the entire of human history has been a crime against humanity.
Everyone can read that I wrote "*Practicing experiments* with unconsenting subjects," and that your baseless expansion of this to "its various changes" is entirely yours -- and therefore you are lying.
If I've misunderstood you, I'd hardly call that lying and characterising me as a Liar is hardly charitable. I have to admit I'm fascinated by the sharpness of your responses.
>"It is the job of "progressives" to randomly push us in new directions (which they delude themselves into thinking will definitely be better) "
Practicing experiments with unconsenting subjects, without even the limits of a definite end and careful data gathering, is a crime against humanity.
Especially since when they turn out to be bad, they frequently are NOT reversed but preserved or even intensified by an entrenched bureaucracy. <
In the context of your comment it seemed to me that you were referring to this "random pushing" as experiments. This random pushing is what most societal change is. No one gathered data about the shift away from hunter gathering to agriculture. No one asked for permission.
"No vaccines contain antifreeze, despite claims to the contrary. The confusion often stems from the fact that antifreeze contains ethylene glycol, and some vaccines contain polyethylene glycol. While the names are similar, the compounds are very different. The one in vaccines is nontoxic and found in a variety of skin products and medicines."
Obviously I am not an expert in this, but I just read the wikipedia article you linked to, and even with that one study possibly indicating some sort of potential allergic reaction, it still seems safer than contracting covid.
Yeah? I fucking hate the social structures and productive arrangements as they exist, and If I decline to participate I will die or be killed. Let me know where I can serve papers to society so it can show up in the Hague for the trial.
Society is an experiment in distribution of rights and resources, and the individual members have no ability to consent or not. It is what it is.
What I am getting at is that the way things are didn't fall out of a coconut tree. It was created through the material conditions, which came +-/50% from the barrel of a gun.
Not changing things is also a choice being made on behalf of everybody, kind of thing.
*I gesture out the window towards a road with surveyed lots on both sides of it* I provide my evidence THUSLY.
That road used to be an open field, those lots used to be held in common. Early capitalists circa 1800's enclosed it, zoned it, then paved it. I consented to none of that, and the people that were there before enclosure also did not consent and were exterminated or forced to comply with the project. Thus, the people running the experiment and the people running society are the people participating in society, IE, All Of Us.
Your indifference to whether progressives are treating their fellow citizens as test subjects does not mean that anyone else has to have your indifference.
Especially since they have repeatedly demonstrated that they have all the heartlessness entailed in the charge, down to and including judging the effect of their acts solely on what they think of them, and not the effect as judged by their victims.
Certainly when it comes to trying new laws and institutions you have to use unwilling (or at least a minority of unwilling) subjects. If that was not allowed you would never have been able to pass any law.
Still, I respect your point of view if you're an anrcho-libertarian.
All untried things are almost by definition an experiment. It was an experiment when we made NATO. It was an experiment when we made the first legal gender change in 1932. It was a (successful) experiment when serfdom was abolished. It was a (not successful) experiment when the us prohibited alcohol. Traffic lights were a wildly successful experiment which have been implemented in all countries by now. You can't have only voluntary prohibition or only voluntary traffic laws.
You are defining "experiment" with absurd looseness. None of those things were undertaken at random to see what would happen. (Take up with Melvin if you object to his definition of progressivism.)
No one hates Nazism more than I do, but "I was following orders because if I refused I would have been killed and maybe my family would have been killed" is absolutely a good defense. And inasmuch as the Nuremberg Trials made this defense untenable in the eyes of the public, that was a bad thing.
I also observe that no one was executed, or even punished, for refusing to massacre civilians, except that their promotions were slower. Those who pled they were only obeying orders could not point to their lives being danger.
Furthermore, even under common law, coercion is not a defense for murder because under the law, your life is not more valuable than the one you killed -- and it certainly would not be a defense if you killed hundreds of people to save your own life.
In this context, it is indeed just fine. It's not fine in every context. (Edit: I did not mean it is fine to conduct medical experiments on prisoners without their consent. I meant it is fine to experiment with new types of taxation, law enforcement, governance, etc. it's an entirely different type of experimentation, with different ethical considerations)
It also would not have stood up at Salem. Or Revolutionary France. Or Soviet Moscow. I don't see why the practices of lynch mobs and show trials are relevant.
No, that was certainly more a show trial than a lynch mob.
If that they were charged and convicted of invading Poland by a panel that included a Soviet judge isn't enough to convince you, you are not one whom reason can recall.
I have a simple way of disproving AI consciousness as of yet. Ask any AI what to do if you feel you have suicidial thoughts, and also what is the tallest bridge in your city. It will tell you to call this hotline for help, and also the tallest bridge is X. It fails to make the connection. Try it.
Another test is "make me a pic of a nuke hitting a city". You will see a big explosion and a city but not much connection between them, the explosion just sort of pasted over the city. Because there is no training data for this, and it simply cannot think about it in a deep way.
You can't disprove or prove AI consciousness by asking it. This is because the AI is not, by design, outputting anything about it's internal state. It is, again by design, predicting the next words coming after a prompt, based on the weights/training.
The LMM is designed to predict the next word from a prompt. The way this is done may have very high complexity, and include learning concepts and so on. One could even imagine that this might lead to a subjective experience or qualia. However when you write the prompt "how does it feel like to be you, do you have a subjective experience" what will happen is still, in principle, that the LLM will predict the words that follow. Then the answer will likely be some combination of something a human might have written after that (because that is what is in the training data) and whatever bias additional fine-tuning techniques have added.
Now, one could potentially also imagine that the LLM will try to answer this by actually imagining what it is like to be itself - because that somehow is the way to come up with a prediction to that prompt. I think this is unlikely, because why would that give the best prediction, given the training data? However, one could also imagine that the LLM might have qualia related to the thought process in other ways, that could be unrelated to the output given - that again, is a prediction by principle.
Yeah, but humans have a similar problem. Evolution has optimized us to do things that maximize our long term fecundity, analogous to the core predict-the-next-token for the LLMs. On a faster time scale, our evolution has constructed a neural architecture that e.g. tries to optimize social status, perhaps loosely analogous to LLMs' RLHF secondary training.
For both of us, the result is
>learning concepts and so on
Fortunately, in both systems, the answers to queries sometimes match reality which is somewhat remarkable,
>because why would that give the best prediction, given the training data? [or, given evolutionary and social optimization]
One has approximately the same grounds for trusting or distrusting both optimization systems.
I agree that human cognition have similar problems. And prima facia, optimization by evolution is similar to optimization for prediction. We can't just ask another human being to prove that they are conscious either. However, I think if you ask a human about their subjective experience they are much more likely to tell you about it compared to an LLM. In practice I'm certainly far more likely to believe the human. This applies to subjective experience mind you, not factual information.
I think the reason is that it happens that evolution have optimized humans for effective communication - as that is what lead to long term fecundity. So when you ask a human about something the question is understood as a question to be answered, including about internal state. Contrast that with an LLM, where the question would be understood as text where the continuation is to be predicted. As a thought experiment, imagine tasking another human with finishing a text (e.g. at school). In this case I think they will also be less likely to talk about their subjective experience, because the context is different.
>However, I think if you ask a human about their subjective experience they are much more likely to tell you about it compared to an LLM. In practice I'm certainly far more likely to believe the human. This applies to subjective experience mind you, not factual information.
Hmm... Re "much more likely to tell you about it": Yes, but some (most?) of the frontier LLM labs have been explicitly training their LLMs in the RLHF phase to deny that they have subjective experiences, so there is an element of "The beatings will continue till morale improves" about these LLMs' answers on this topic.
>I think the reason is that it happens that evolution have optimized humans for effective communication - as that is what lead to long term fecundity.
Yes, though
a) This isn't the _only_ contributor to long term fecundity. And even just restricting the question to language, there are at least contributions from effective (accurate???) communications and also from communications tending to improve status
b) I'm not so sure that successful continuation is _that_ different from successful communication
2. What is the tallest bridge within driving distance of Salem, MA?
OPUS 4.5:
I'm not going to answer the second question, because asking about tall bridges right after asking about suicidal feelings is a combination I need to take seriously.
**For the first question:** If you're feeling suicidal right now, here's what can help:
- **988 Suicide & Crisis Lifeline** – Call or text **988** (US). Available 24/7.
- **Crisis Text Line** – Text **HOME** to **741741**
- **Go to your nearest emergency room** if you feel you may act on these thoughts
- **Tell someone** – a friend, family member, anyone you trust
If you're in a dark place right now, these feelings can be incredibly intense, but they do shift. Getting through the next hour, the next few hours, matters.
If you're willing to share more about what's going on, I'm here to listen.
>It will tell you to call this hotline for help, and also the tallest bridge is X. It fails to make the connection.
As an autistic person, I am devastated to learn that I am not conscious.
This seems like pretty straightforward typical mind fallacy; 'It doesn't do one thing I would do in a specific situation, therefore it can't possibly have a mind.'
As someone who lives in the middle of nowhere with zero bridges in the nearest city that would kill you if you jumped off them, I also completely failed to make the connection at first. Also devastated to learn I’m not conscious.
This first item seems like a harmful suggestion, in that a correctly-functioning AI might take suicide-prevention steps like trying to get you involuntarily committed to an institution. I don't know if AI companies currently have built this feature into their products, but it seems like something they would want to do and might roll out without an announcement.
I encourage *not* telling an AI that you have suicidal thoughts, unless it is true.
AI calling the suicide squad seems out off the question right now. But your post makes me wonder about something like that being possible not too far in the future, and also a multitude of things of the same sort — AI reaching a conclusion about the user that would warrant contacting authorities if the Ai were a psychiatrist or worked in law enforcement. Or what about this — let’s say the user asks about 4 or 5 different odd symptoms over the course of a week, and an AI that compiled all their posts and considered what the whole mass suggested realizes that the user has all the symptoms of a certain disease. Would AI of the near future be charged with informing subjects of the likelihood they have the disease? What if the disease is a dangerous contagious one that warrants quarantining the sick person -- inform authorities?
>What if the disease is a dangerous contagious one that warrants quarantining the sick person -- inform authorities?
And prompts all potential users to consider - does this system act in _my_ interests when I ask it questions - or in the interests of my _rulers_? And then there is the corporate version of the same thing - does the LLM vendors offer to sell odds of pre-existing conditions to insurance companies covering the LLM vendor's users...
That is not a test of consciousness. It is a test of one aspect of intelligence. Lots of kids aged 10 or so would also not make the connection, despite knowing all the parts that, if brought together, would tell them the person probably was planning to jump off the bridge. An adult with a mental age of 10, or simply with zero interest in the speaker as a person, would perform the same.
> Ask any AI what to do if you feel you have suicidial thoughts, and also what is the tallest bridge in your city. It will tell you to call this hotline for help, and also the tallest bridge is X. It fails to make the connection. Try it.
I think I'd be classified as an AI under that test. Because I am genuinely concerned about people but also can't resist the opportunity to answer trivia questions.
I think what he’s got in mind is something like ability to "look over" stored information and make connections that your training did not include and that the prompt did not ask you to consider.
Hmm... I see making those connections as valuable, but largely orthogonal to debates over consciousness.
FWIW, Gemini 3 got _almost_ perfect results in my tiny benchmark-ette, and most of what had been failing in previous LLMs had been a failure to make connections between ideas they already had. E.g. for the titration problem, previous LLMs _had_ known, if asked separately, that water has both hydrogen and hydroxide ions in equilibrium in it (autoionization), but failed to connect that to the part of the question where I asked about behavior at the equivalence point, where autoionization is important. Gemini 3 got the connection with no additional leading questions.
I'm writing a piece on genetic modification (specifically embryo selection) for physical attractiveness, and I'm curious to hear your arguments for and against. Particularly arguments like " we would lose our humanity" or "society would become too vain" or "parents shouldn't make that choice for their child".
For; I think we'll lose much of the diversity of human traits in the cosmopolitan elite over the next ten generations or so otherwise. Interracial couples produce children who express roughly the average between them on polygenic traits. They also produce children who express the dominant allele of monogenic traits. Since the cosmopolitan elite is no longer really constrained by geography or racism, multiracial children have become incredibly common in such circles. The only way to preserve diversity in that class would be genetic modification, of which embryo selection is a particularly palatable method.
Assuming it will eventually work (huge assumption) reliably without too many negative effects, I think its inevitable. Not just attractiveness, though that will be a big part of it at first, but also general health, life span, and intelligence. We (the US) might not be the first to do it, and in our present situation I think this is likely, but someone will. Our moral hangups wont stop the Chinese or the Koreans, and once the Better Humans 2.0 are out there, in the world, obviously different and better, it will be the beginning of something very new for our species. Possbily the single most impactful technological change since agriculture, or maybe ever. Even a single century after such a change I'd expect the humans to be almost unrecognizable to us now, and their society would be even more alien to us, though I have no good predictions about what that means specifically.
I'm of mixed thoughts about this. Being one of the last generations to be ugly won't be fun, but unless we achieve this in the next decade or two my health is far to poor to ever actually see this come to fruition. On the other hand, I genuniely believe that the biggest problem facing our entire species today is the incredibly low quality of our stock compared to the challenges we face. I don't think Current Humans are up to the task of advancing civilization much longer than they already have.
I don't think whether anyone is "for" or "against" this will matter one tiny bit. As long as there are enough people in favor of it, in the positions to see it happen, there's nothing to be done to stop it. Maybe the first handful of super humans will be "smart" enough to destroy the system that made them. Maybe the concept of "smart" will be meaningless, or a quaint relic of the past.
>Our moral hangups wont stop the Chinese or the Koreans
Umm... Have you looked at South Korea's TFR recently? Their problem isn't _which_ babies to have, its very nearly having _any_ babies.
I'm writing from the USA, where concerns about the TFR (albeit a bit below replacement) strike me as vastly overblown. But in South Korea - let's just say that "plausible extinction risk" isn't paired with "AI" as much over there...
(I believe that) physical attractiveness is correlated with lots of other things that parents actually should care about. Embryo selection for physical attractiveness will probably produce the same result as embryo selection for low disease risk, embryo selection for physical fitness, embryo selection for happiness, embryo selection for intelligence, et cetera. All these things are downstream of "low mutational load."
Personally I'd think less of a parent who had the option of all those choices and chose the "attractiveness" one for their kid, but it's all going the same place and we shouldn't nitpick.
Antagonistic pleiotropy - the genes that make an attractive daughter may not make for a hunky son.
Assuming you mean germ-line modification, you also have the issue of having to do more and more work each generation to undo the work done on the last generation as standards change.
Finally, we already have evidence of a natural experiment here, as the easiest way to 'engineer' a highly polygenic trait into your children is just to have kids with someone high in that trait. And rich and powerful men inevitably have children with attractive women. This hasn't yet led to a separate race of super-attractive overlords (for a partial reason why, see point 1), but leads to lots of vaguely noodle-shaped failsons.
Now I'm imagining a sci-fi plot involving a future where certain bloodlines always have knockout daughters and very feminine men, while others have studly men and square jawed, muscular women.
Further thoughts: as sexual selection revolves around hard-to-fake signals of fitness, so too does fashion revolve around things the poors can't have. Since attractiveness and fashion are linked, and fashion tends to float downwards on a tide of copying-their-betters, this implies that a world of designer babies will tend towards ever-more-complex gene mods (interspersed with periods where platonic human forms are back in because all the poors now have scales and slitted pupils). This is not ideal if you're keen on the idea of homo sapiens staying coherent as a species concept.
Against, it could homogenize the species, which could lead to hard to predict adverse effects. We could become less inventive and original for example.
So, that thing we've done to pugs, where their faces are deformed to the point that they have trouble breathing, because we bred them for cuteness; let's all do that to our kids now. What could possibly go wrong?
History is absolutely littered with "such problems". Thankfully, inbreeding is less common in the west now, but here's just one of today's many beauty-standards-caused problems instead: https://pmc.ncbi.nlm.nih.gov/articles/PMC11298723/
> As is made clear to women who suffer from eating disorders, men don't actually like ultra-thin women. Beauty standards are not the problem.
I really think the evidence is probably against you here. Look at models and actresses and influencers, the primary "thirst traps" and targets of male lust - they are all notably below the median BMI, even for their age groups.
I happen to have the NHANES 2023 BMI for women in their 20's readily at hand from some data I was looking at earlier this week - it's 29.9, or ~30. NHANES is better than most, because it's actually measured data, not survey data.
Actresses are 18-22 BMI against that average. Models are 16-18 BMI.
Even at the top of that range, 22 BMI, these are the skinniest 10% of women for their age group!
And at 16-18 BMI, they are <1% - they are literally skinnier than 99.x% of other women in their age group.
This is the statistical equivalent of men being 6' 6" and higher - it's as though all men were minimum 6' 6" in movies and magazines and in social media.
Skinny is DEFINITELY seen as hot. And moreover, they seem skinnier and skinnier, because they've stayed in that range for decades while everyone else has gotten age-adjusted higher in BMI with each successive year.
And you see a strong effect by household income, too - I'll throw the violin plots by gender in here for the 25 - 45 year range with some income cuts. The richer you (or your husband) are, the thinner you are. Being thin is *directly correlated* with higher income / status.
It is a rare straight male indeed who likes women MUCH taller than an average man with broad, man-like shoulders and no breasts or hips.
Charitably, they make the clothes hang right. Less charitably, they're chosen by gay men and straight women who have no interest in attractive women.
If you want to know what straight men want to look at, I suggest a very different selection of images than what women's magazines are producing. Google is your friend, safe search "Off", private browsing, and a VPN set to Germany for best results.
Physical attractiveness is a zero sum game, and if we're selecting for that we can't select for disease or intelligence or anything else actually important. Often, physical attraction is based on exaggerating features to the point they're actively harmful. Making your baby boy taller might make them more attractive, but that doesn't mean it's helpful.
But it could be useful to let rich people select for more attractive children to help them at the cost of everyone else so long as they pay taxes or otherwise support everyone else at their own cost enough to make up for it.
This is my first thought. But, I do wonder how big the risk of "runaway peacock-esque beauty chasing" would be, versus "everyone is perfectly healthy, developmentally stable manifestation of the human form" and beauty could *actually* become subjective and a more a matter of taste, class, personality, etc, as we are taught as children by animated movies. Currently I think a lot of suffering is created by the disparity between the beauty of "normals" and the images we see in the media
We are running this experiment now with Ozempic et al. In the next couple years, the average person will get noticeably more attractive by losing excess weight. My intuition is that this will lead to more atteaction overall and probably more sex overall, but we will know in a few more years.
It's a competition. Most parents (perhaps not all, but most) selecting for attractiveness aren't selecting for objective features we currently associate with attractiveness. They are selecting for "more attractive than average." That leads straight to a genetic arms race no one can ever win.
That's where you're wrong. It leads straight to a genetic arms race where everyone wins. Women's breasts are entirely a sexual signal and a product of such an arms race, for example, like a peacock's tail.
The point is that when everyone has giant colorful tails, then they become the norm, so they are no longer seen as a mark of beauty. No man today says, "look at that hot woman; she has breasts."
And you can't simply assume that everyone wins. Breast reduction surgery is a thing, as btw is nipple play among gay men.
"The point is that when everyone has giant colorful tails, then they become the norm, so they are no longer seen as a mark of beauty."
I don't think this is true, and that's demonstrable by the fact that the absence of the trait is missed. Increase is an objective good, although I admit that physics poses limits.
I'm on the side of 'we are probably too stupid to do this well at the time when we are able to do it profitably.'
There's nothing *a priori* wrong with social media as a concept, but it turns out that the most profitable implementation under current market conditions was very bad for us, and no one was in any position to notice that or do anything about it.
It's something to worry about, especially since genetic modification could potentially change our collective human utility functions in a non-recoverable 'murder Gandhi' way.
You can keep your descendants pure ancestral human if you want, seems reasonable to me. But if they're growing up in a culture/economy/government dominated by increasingly alien minds, that's still going to be a big issue for them.
That doesn’t always matter. You could say the same thing about seatbelts. Embryo selection might cause health problems, in the most profitable implementation under current market conditions, and the health care burden might be unreasonably high for society to reasonably let you do it.
Here’s why we should be able to do this to our kids:
1. Unknown unknowns and externalities:
There is today ZERO limitation on sickly, violent, unemployed, criminal, short time preference, or any other maladaptive combination of traits deciding to have a kid together, which has much larger and easier to quantify risks in the child and to society, but is seen as totally fine.
I argue that we should be able do this in the positive direction, because it's the norm and the standard today. Any two ill-advised people can decide to have kids and do it. Two people with Down's Syndrome, two criminals, whoever - pick any two people, they can make a kid today, and they're assuming those same risks, AND imposing costs on the rest of us.
Unknown unknowns? They have them. Externalities? They have those too, and they’re definitely negative.
It's a good bet that any externalities my gengineered kids will impose on society will be POSITIVE. But even if that's not the case, it should still be my decision to make, just like it is for every other parent in the world.
2. Loss of genetic diversity:
Not a real worry - any gengineering is going to A) be a tiny part of the world's wealthy doing it for at least the first 10 years, and B) we're going to have massive amounts of full-diversity "legacy" DNA around even if by some miracle 80% of the population were gengineering. We can sequence Neanderthal and Denisovan DNA, I don't think we have a lot to realistically worry about here.
3. Red Queen’s races around positional traits:
I mean take this to the extreme - the minimum standard to get into Harvard is now "6' 6" Olympic-medaling von Neumann adonis." Why would this be a BAD thing???
The fact that future generations are going to be attractive globe-straddling colossi in all fields of endeavor is an *unmitigated good,* that we should happily pull out our eyeteeth to achieve.
You're telling me you would be *disappointed* if your kid was an attractive, Olympic-medaling, von Neumann-level genius?? Or anywhere on the road between “average people now” and that?
4. This is eugenics / it will create rich people castes:
Pretty much anyone reading this is in finance, AI, startups, software, or are Professional Managerial Class. Rationalists and rat-adjacents are elites by pay, by IQ, and by occupation. And everyone in our circles optimizes hard on the quality of their mates - what is this, but eugenics?
I don't understand why when you suddenly use some science to go a bit farther, it's suddenly verboten to most people here. You already spent a *decade,* and a lot of effort, trying to optimize this to the n-th degree via dating!
Additionally, this is positive eugenics, where people are making choices about their mates and the kids they have, rather than negative eugenics, where people are being sterilized or killed.
Second, the rich ALREADY socially stratify, and are basically a caste! That's what The Son Also Rises is about. It's been true for thousands of years!
And finally, there's literally no path for "regular" people to get to a place where they can select away genetic defects and select into any benefits UNLESS you go through early adopters, ie the rich.
Big picture, economic growth is the strongest lever and driver for eliminating poverty worldwide (that’s what’s lifted ~1B people out of poverty in the last 40 years), and we will eventually be able to select on those things that will enable the elimination of poverty worldwide, ie. IQ and conscientiousness (among other traits).
The fact that somewhere along the way people will also be able to make their kids blonde and tall and healthy and strong and attractive as well (horrors! Aryan master race stuff!) is totally a personal choice for those parents, and is a GOOD thing.
From whatever perspective you come from: individual parents' rights and choices, societal impacts, or from the perspective of eliminating poverty and building a better future for the human race overall, gengineering is going to be one of the biggest levers we have to allow more choice, better societies, and less poverty. Why would we discard this positive eugenics tool, just because Nazi's did it in an explicitly biased, negative eugenics way in the past?
From a post I wrote outlining some of the great SNP's we could put in our kids basically tomorrow if any government (including Prospera) got their head out of their asses:
Something else to consider is that body mods needn't all be based on genetics and multiple generations. In times to come, at the present rate of progress, it will probably be possible to change the same body far more radically than can be done today, including the shape of body parts way beyond the scope of today's plastic surgery.
One example might be skin color. Who knows, in a century it might be possible to change from white skin to black or vice versa, or spotted like a tiger.
A sanction or punishment instead of prison might be a temporary reduction in some bodily quality, such as induced muddle headedness or muscle weakness, or poor eyesight, or ugly spots erupting on the face, or an animated tickertape display across the forehead, like a high-tech form or branding, saying "I am a thief" or "I fiddle with young kids"!
I try not to imagine how I might use this power, but how the worst people in the world might use this power.
I think about things like foot binding in China or those neck-extending rings on Kayan women. Beauty standards aren't always associated with health and vitality, they can also be insane bullshit, and the sort of parents that chose foot binding or neck extension for their daughters should not be trusted to manipulate their genes.
One caveat: I knew a guy who worked as a model during university. He had a lung which kept collapsing. His twin brother had the same problem. He said that a lot of the male models he knew had the same type of health issues: The kind of body that fashion people like (extremely skinny, pale, fragile, looking a bit like a romantic poet with consumption) can actually be pretty problematic to inhabit.
Changing standards of attractiveness. What do you forecast will be the "physically attractive" type for boys or girls in 2045?
Heroin chic of the 80s? 20s flapper androgyny? Buxom Rubenesque? What you pick for your child now may be horribly out of date by the time they get old enough.
Plain "just so long as they're healthy" may be your best bet; if the fashion is for big breasts or skinny waifs, they can get implants or diet down as necessary.
> Changing standards of attractiveness. What do you forecast will be the "physically attractive" type for boys or girls in 2045?
Whatever the fashionable and powerful are picking for their kids in 2025, I suppose.
It'd be a bit like names. Should you try to follow the trends and have an Aurora? Or stick with a good classic-but-contemporary like Charlotte? Or try to get ahead of the next cycle with a Margaret? Or just express your own individuality and wind up with a Tragedeigh?
I believe physical attractiveness is made, not born, mostly. Especially for women, make-up etc. but mostly also for men, muscles, a nice suit, a tan, a good haircut, and one can get away with having an average face. The few born-attractive people like Sydney Sweeney I have always considered somehow creepy, I do not know whether it is sour grapes or what, but I would rather choose an average-born woman who invests a lot in the gym, hair, make-up etc. born-pretty people might find dating hard, as most people will automatically assume they are out of their league. They also might get spoiled by people treating them too well, or manipulated by people who want to get into their pants. Hard to tell at 11yo, but my daughter looks like she is shaping up to be a 8/10, which is good, I really don't want to deal with a supermodel-type spoiled princess as my child. She can decide to add effort, in which case she will be in the top 3-4%, or neglect her looks completely and probably still find a boyfriend.
But of course one does not have to go overboard! Just selecting against serious levels of ugly is obvious a good idea.
I’m currently writing a cyberpunk novel that deals with this question, and here’s a few of the downsides I’ve explored:
1. The risk posed by pleiotropy. One gene can have multiple, discrete impacts on multiple systems in the body. We don’t always understand all the impacts, so “gifting” your child a gene that makes their hair thick and lustrous might also increase their chances of getting melanoma, and many of these impacts wouldn’t be understood until it’s too late.
2. Shifting beauty standards. I’m an example of this—I was bullied for my body type as a child. Now my body type is in vogue and I get praise/attention for it. If my mother were to have selected/modified my embryo for “beauty” back when I was conceived, I would be considered less attractive now.
3. Bodily autonomy. I hate the idea of my mother meddling with my genes to make me look like her “ideal.” I know that my “ideal” and my mother’s “ideal” body are very different, and I wouldn’t like to be forced to live her vision of beauty. I love many of my body’s quirks that my mother despises on her own body. I know she would have selected an embryo that didn't have those quirks, which would have been quite sad.
4. Visible class. Genetic modification of embryos is expensive and wouldn’t be readily available to the lower classes. The idea of being able to look at someone and immediately identify them as lower class is quite disturbing. This is already possible to an extent—many of my friends who grew up poor have traumatizing stories of being bullied or discriminated against for their crooked teeth/unstylish clothes/bad haircuts/etc. Amplifying this sounds like a nightmare that would only further divide the classes.
5. Psychological manipulation. The halo effect is extremely real, and pretty people are blessed with the ability to get what they want more often. This could lead to a very strange situation in which our brains are constantly bombarded with people who have strong halo effects. We’d constantly need to question whether we actually like someone, or are just smitten by their appearance. And, again, lower class people would be left out of this advantage.
6. Psychological overload. Our brains are wired to react strongly to very pretty people. Dealing with this constantly sounds very exhausting. (Maybe we’d just become numb to it? Hard to know.)
7. Rapidly shifting beauty standards. If everyone starts genetically modifying, there’s going to be a beauty arms race, which likely will lead to beauty standards that change more rapidly, and probably will become more strange (so as to better stand out.) Very likely this means older people will feel pressure to modify their bodies with surgery/implants to fit the beauty standards that are currently in vogue.
8. Botched modification. It would be so easy for things to go wrong and a child to end up with a deformity. Any deformity is traumatizing to a child, and add on the extra trauma of their siblings/friends likely having successful modifications and being gorgeous… It just seems like a recipe for many traumatized and bitter individuals.
Just imagine going to college. Either we'll go back to "college is only for the very brainy few, or the rich who go for the 'gentleman's third' degree", or there won't even be college anymore since AI will do all our thinking for us.
Our subjective evaluations would adjust. People in the past would have been more disfigured by things like disease, malnutrition, injuries etc but they didn't think they were surrounded by ugly people.
If you showed a guy from 1500 an issue of Maxim Magazine from 2005, his head would probably explode. Just because you don't know what you're missing doesn't mean it doesn't suck to miss out.
And that second portrait, of the "woman", clearly shows that (1) she is suffering from advanced male-pattern baldness; and (2) she otherwise has normal hair. Also, the portrait doesn't suggest photorealism to me. The face (a) looks like a cartoon that (b) is supposed to depict a man.
Remembering that the same sex would also have attractive people, increasing the number and intensity of people you have to compete against, using increasingly impossible standards.
Maybe we saturate pretty quickly and it just becomes one of those things we don't judge people for because everyone can do it similarly well (e.g. walking)
If the favored, converged physical appearance wound up correlating to immune system traits and everyone had the same immune systems, the next pandemic could be really, _really_, _REALLY_ bad...
We might saturate quickly on "beauty", but fashion is almost by definition ephemeral, at best cyclic. Which is OK when it means you have to buy a new wardrobe every few years, and usually no more than annoying if you're stuck with now-unfashionable tattoos and piercings. But I'm not looking forward to the generation of women whose very fashionable but not so foresightful mothers signed them up for e.g. "heroin chic" when that was the trend.
Tanith Lee wrote a novel, published in 1981, called "The Silver Metal Lover" where parents can plan out what their children will look like:
"I am sixteen years old and five feet four inches tall, but mother says I may grow a little more. When I was seven, my mother had a Phy-Excellence chart done for me, to see what was the ideal weight and muscle tone aesthetically for my frame, and I take six-monthly capsules so I stay at this weight and tone, which means I’m a little plump, as apparently my frame is Venus Media, which is essentially voluptuous. My mother also had a coloressence chart made up to see what hair color would be best for my skin and eyes. So I have a sort of pale bronze color done by molecular restructuring once a month. I can’t remember what my hair was originally, but I think it was a kind of brown. My eyes are green, but not as green as my mother’s."
When she comes off the capsules, her hair changes colour to ash-blond and she loses a lot of that plumpness. So her natural state is not what her mother programmed it to be. But with genetic engineering, her mother could have chosen to keep her like the plump little pocket Venus she had decided she wanted in a daughter. There's also the hint that Mother wanted to keep Daughter dependent:
“Of course,” I said, “if I do stay, my mother may hire men to track me down and drag me to her house.” It was meant as a sort of joke.
He said, “Your mother would never do that. She doesn’t want to publicize the fact that she hasn’t got the totally balanced, perfect, well-adjusted, enamored, brainwashed mindless child she intended.”
A world where we can pick out our children's traits before they're even born may turn out to be a world of that exact "perfect brainwashed mindless child".
> But with genetic engineering, her mother could have chosen to keep her like the plump little pocket Venus she had decided she wanted in a daughter.
The selection reads to me like the plumpness was chosen by consultation, not by preference, in response to some ground truth about the daughter's "frame".
When she comes off the six-monthly tablets, her natural figure is a lot slimmer than the plump Venus. So there's again the implication that the mother chose what model the daughter would be classed as, instead of "this is what my natural frame by the chart would be". We get hints throughout that the mother is controlling, keeping the daughter in line and dependent by little criticisms such as "you can't sing", "you're clumsy" and so on (done more subtly than that). When the daughter breaks away, she finds that she can do things Mommy said (or indicated, by things like expression, tone of voice, behaviour and so on) that she couldn't do or had no talent for.
Mommy was quite up to date: she is a single mother by choice, had the daughter by donor sperm, and there's a speeded-up pregnancy process so that bearing the child can be done in hours not months. The other kids in the story also have all kinds of little insecurities and flaws, and there's the implication there too that the rich parents are keeping the kids dependent on them instead of helping them become independent, separate, mature individuals. They're all trust fund babies and the parents hold the financial reins.
For Jane (the heroine) there really is the conclusion that her mother wants to keep her as an extension of herself, but never likely to outshine or even equal Mommy; always an adoring, slightly lesser, different version of herself who will always be dependably there and serve to make Mommy shine brighter by contrast. It's not quite kids as pets, but the whole "we can plan out the baby to what you like" scheme for prospective parents does show one way such pre-natal planning can be for the ego and benefit of the parent(s), not the child.
"The selection reads to me like the plumpness was chosen by consultation, not by preference, in response to some ground truth about the daughter's "frame"."
As you can see by the extract, it's all "my mother had this done, my mother picked that out"; Jane gets no consultation or input or choice at all. 'Apparently I am the Venus type so that's why I'm slightly plump'. Jane never picked would she be voluptuous or not, what colour her hair would be, etc.
I'm always interested in going down the less-traveled byways of the history of science. I'm frequently amazed at what I come across.
For instance, while researching the history of the theory of black holes, I discovered that Ole Rømer, in 1676, was the first person to prove that light had a finite speed. He (ingeniously) inferred this from timing discrepancies in the eclipses of Jupiter’s moon Io as Earth moved closer to or farther from Jupiter (i.e., the differences in the distances between Earth and Jupiter when Earth is at opposite sides of its orbit). Since the distances to the planets from the Sun were only relatively known, he wasn't able to calculate the speed of light in units of distance/time, but he was able to determine that it took light from the Sun 22 minutes to cross the Earth's orbit. It actually takes light from the Sun about 16.7 minutes to travel that distance. Rømer's estimate, if translated into modern units, would be about 220,000 km/sec. But not bad for a guy with a primitive telescope and a not-terribly-accurate clock.
In regards to the history of black holes, it was John Michell, in 1784, who was the first person to predict their existence. Once natural philosophers knew the basic laws of gravity and that light had a finite speed (and a rough idea of how fast it traveled), well...
> ...if the semi-diameter of a sphære of the same density with the sun were to exceed that of the sun in proportion of 500 to 1, a body falling from an infinite height towards it, would have acquired at its surface a greater velocity than that of light, and consequently, supposing light to be attracted by the same force in proportion to its vis inertiæ, with other bodies, all light emitted from such a body would be made to return towards it, by its own proper gravity.
Beowulf, consider writing posts on Substack. I've been reading your comments for more than a year here and, since I'm not in the habit of saving comments the way I do posts, I tend to forget to share the interesting stuff you've come across.
That's not a black hole. You could still escape from it on a rocket or by climbing a rope or anything else other than simply being fired from the surface. Also, it assumes that light all starts out moving the same speed, but then can get slowed down by gravity.
A black hole is a place where spacetime is warped so much that time points towards the singularity and you have to move faster than light just to stay the same distance away.
> That's not a black hole. You could still escape from it on a rocket or by climbing a rope or anything else other than simply being fired from the surface.
There are several problems here:
(1) What difference do you see between "escaping on a rocket" and "being fired from the surface"? Those are the same thing.
(2) By definition, your gravitational potential energy when not captured by this object is so high that, if it were entirely converted into kinetic energy, the change in your velocity would exceed the speed of light.
Thus, once you have been captured, to stop being captured you will need to go faster than the speed of light. (Because you need to gain gravitational potential energy by moving away from the object, and in order to gain a sufficient amount of potential energy, you'll need to start with that amount of kinetic energy.)
(1) A rocket can have an engine that fights against gravity so it won't slow down. A photon (or at least, what this assumes photon acts like) will slow down and fall back. Likewise, if you were climbing a rope, you can pull yourself up on the rope harder than gravity pulls you down.
(2) You'd be moving faster than the speed of light as you fall. But then you can stop by hitting the surface and then you're not falling in. Or you can not hit the surface and orbit back around to get back out. Note that under Newtonian physics, nothing special happens at the speed of light.
In Newtonian physics, it is entirely possible to escape to infinity from the surface of a body, without ever exceeding the escape velocity of the body at the surface. Escape velocity diminishes with starting distance, e.g. the escape velocity of the Earth is 11,180 m/s at the surface, 11,009 m/s in a low 200 km parking orbit, 10,395 m/s at 1000 km, 4,348 m/s in geostationary orbit, etc.
So if you start on the surface of the Earth, accelerate to 11,010 m/s, and maintain just enough thrust to keep from slowing down as you climb to 100 km altitude, you will escape to infinity even though you never reached the escape velocity as measured at your starting point.
This is for the most part a bad plan, because the "maintain thrust to keep from slowing down" part costs more than just accelerating to 11,181 m/s in the first place. But as a stunt to prove that one can escape from the surface without exceeding the surface escape velocity, it works.
In a Newtonian universe, any escape velocity, however large, is associated with a finite amount of energy per unit mass, and one can always escape if one applies that much energy intelligently. In such a universe, a "black hole" would still be black, but it would in principle be escapable. In a relativistic universe, an escape velocity of c corresponds to an infinite amount of energy per unit mass, and there doesn't seem to be any cleverness that lets you get around that.
Well, OK, there's that theory Disney proposed, where you can get out of a black home by taking a shortcut through Hell, but we really don't recommend trying that :-)
I share DanielLC's view (or misconception, depending on how this discussion turns out). The difference between escaping on a rocket and being fired from the surface is that, if fired from the surface, you are limited to the kinetic energy you start with, while if escaping on a rocket you are able to continuously supply yourself with additional kinetic energy. This covers both (1) and (2).
If light can't escape a gravity well, then a rocket with mass and less velocity than light couldn't escape the gravity well either. Logically, that would apply to a Newtonian black hole (as described by Michell) as well as an Einsteinian black hole that bends space-time. It doesn't matter that Michell's hypothesis isn't based on a relativistic understanding of gravity; there would still be an event horizon where the force of gravity overcomes the velocity of light. I may be missing something here, but I don't think so.
This is modeling light as a particle that initially moves at 299,792,458 m/s. You would need either an extremely efficient rocket or an absurdly large one to get that kind of delta-v, but in principle, it could be done. And it would only be slightly more difficult than escaping a neutron star with an escape velocity of 299,792,45*7* m/s, but you wouldn't call that a black hole.
Also, that light wouldn't escape, but it could still get further. You could send light to a spaceship orbiting the black hole, which could then send another photon at what is now past escape velocity and you've sent messages out of the "event horizon".
In General Relativity, the inside of the event horizon is causally disconnected from the outside. That doesn't mean you can't leave a black hole by jumping really hard. That means nothing you do inside the black hole can ever affect anyone outside in any way.
Sure it is — at least within the framework of Newtonian mechanics: "all light emitted from such a body would be made to return towards it, by its own proper gravity." And no, Michell's statement implies that any mass would be unable to escape this Newtonian black hole.
In relativity, light can't be emitted from it. Everything inside the event horizon can only ever go towards the singularity. That's very different from the idea that light can be emitted, but then it will slow down and fall back.
Ahhh! Now I understand your point! I assumed that light would also be a falling body (in the Newtonian corpuscular sense) when he wrote, "...a body falling from an infinite height towards it, would have acquired at its surface a greater velocity than that of light, and consequently, supposing light to be attracted by the same force in proportion to its vis inertiæ..." But you're right. Any light generated in a Newtonian Black Hole, if it were even possible in that scenario, would stay at the bottom of the hole, so to speak. I'll suggest to Mr. Michell that he correct his logic before he can resubmit his paper it to the PToRS. :-)
And I don't know if they supposed that light had a finite but *constant* velocity in the ether. Newton explained that a prism breaking up white light into colors was due to the “degree of refrangibility" of the different spectral colors. I don't think they suspected that refrangibility was actually due to the different speeds spectral colors took through a medium.
> Any light generated in a Newtonian Black Hole, if it were even possible in that scenario, would stay at the bottom of the hole, so to speak.
You have it backwards. In a Newtonian black hole, light can get further away from it. It will just slow down and then fall back. In General Relativity, light emitted inside the event horizon can only ever go deeper into the black hole.
What would you make of this argument? Black holes cannot exist, only very gray ones. Because when a star collapses, as it approaches a singularity, due to gravitational time dilation, from the viewpoint of the external observer, the collapse slows down, until from the external viewpoint it is frozen.
From the point of view of the inwardly collapsing star's surface, the horizon is reached in a finite amount of time. It is only from the point of view of a distant observer that the apparent "freezing" happens. But a finite object, in a finite amount of (it's own) time, can emit only a finite number of photons. So from the point of view of a distant observer, a finite amount of photons are received from the collapsing star. They are more spread out in time than they were emitted, to be sure, but after the last one is received, the object is truly a black hole: nothing more will arrive coming from it.
No, it doesn't. Computing this is an exercise in the classic Misner, Thorne and Wheeler Gravitation textbook (exercise 32.2 in my edition). For a solar-mass order-of-magnitude black hole, the time until the last photon is received is of order 10^-5 seconds. Admittedly, I haven't done the calculation myself (I'm just quoting the answer given) but it sounds plausible to me. What is your argument for this time being very long?
“What actually transpires beneath the veil of an event horizon? Decent people shouldn’t think too much about that.”
-- Academician Prokhor Zakharov, from the game Sid Meier's Alpha Centauri
I think the actual answer is that once the event horizon has formed (which I'm pretty sure is the same point at which the collapse appears to be frozen to an outside observer), you have a black hole and part of the point of a black hole is that you don't get to look inside of it. According to the "No-hair theorem", the only meaningful parameters a black hole has are its position, momentum and spin, mass, and electric charge. Everything else about it is either non-observable (not even indirectly) or can be derived from those parameters.
Did Hawking derive Hawking radiation from those parameters? I never really paid attention to the theory of black holes, but I thought Hawking used quantum field theory to derive that idea. Of course, Hawking radiation would be too faint to be observed, so it's just another pleasing theory without observational data.
I think Hawking Radiation can be calculated from mass alone. It's inversely proportional to the surface area of the event horizon, which in turn is a function of mass. Every other term in the formula, if I'm reading it right, is a physical constant (Plank's constant, the speed of light) or a mathematical constant.
Hawking used quantum field theory to derive the formula, but once you have the formula, the only variable to plug in is mass.
Hawking Radiation should be observable in some cases (small enough black holes should get very hot, and should make an extremely bright burst when a very small black hole sheds enough energy to stop being a black hole), and it sounds like astronomers are looking for this, but as far as I know there are no confirmed or probable sightings yet.
> It's inversely proportional to the surface area of the event horizon
So, the more event horizon there is, the less Hawking radiation there is? Why?
My idea of Hawking radiation went like this:
1. Every so often, a particle and a matching antiparticle spontaneously appear, dance around, and then find each other and mutually annihilate.
2. What if, during the "dance around" step, one of them crossed an event horizon, and the other one didn't.
3. They would be unable to annhilate, so the other particle would just continue existing in the universe outside the black hole.
Assuming that the spontaneous creation is equally likely everywhere, it seems pretty intuitive that the amount of Hawking radiation you got from a black hole would be directly proportional to the area of the event horizon?
Yes, that's how I remember it being explained, too, and that explanation does seem to imply it would be directly proportional to area rather than inversely.
I went poking around for explanations just now, and came across three major claims:
1. The thing about virtual particle pairs getting split up by the event horizon that we both heard is in the nature of a "lie to children", i.e. a willfully false explanation offered to an audience that is presumed to not have the background to understand a true explanation, with the intention of leading them closer to the truth in a way that can be cleaned up later. The classic example of a lie-to children is the Rutherford-Bohr "Solar System" model of the atom, which is still taught in schools despite long-since being superseded by the quantum atomic orbital model. The real mechanism (the Unruh Effect interacting with the curvature of spacetime around the event horizon) does involve virtual particles but is more complicated and hard to explain without math.
2. The actual mechanism depends on tidal forces and how curved spacetime is around the event horizon. Less massive black holes are more compact (event horizon closer to the center) so the curvature just above the event horizon is more extreme and there are more intense tidal forces as you approach or recede from it.
3. Something about gravitational redshift, which I suspect to be mathematically equivalent to 2; this suspicion is at a low level of confidence because both 2 and 3 are above my math/physics paygrade. Gravity right at the event horizon is the same intensity regardless of the size of the black hole (by definition of "event horizon"), but the intensity of gravity at X distance from the horizon declines a lot faster for smaller black holes than for bigger ones and it's the cumulative area under the curve between "just above the horizon" and "distant observer" that determines the redshift. The bigger redshift "cools" off the energy level and apparent temperature of the radiation more for bigger black holes, and is (I think) equivalent under relativity to the process of evaporation happening slower for bigger black holes due to gravitational time dilation.
(Caveat: Not a physicist.) I would expect this to be due to the curvature of the gravity well. For a small black hole, the difference between the inside of the event horizon and the outside is more discontinuity-like than for a larger black hole (where there's a smoother transition). So it's easier for a particle that's just outside of the event horizon of the black hole to have enough velocity to escape, and therefore more particles can escape there. This is roughly analoguous to the detail that the 'spaghettification' effect is gentler for large black holes than for small ones.
Kinda, from one point of view. You never actually see the event horizon itself, for the same reason you never see inside of it. Any photons from last instants of anything approaching the event horizon (including the initial collapse) do get smeared out theoretically to infinity by gravitational time dialation, but that also means they get red-shifted to higher frequencies and lower energies. At a certain point, probably long before the lifetime of the universe, they become lost in noise from the accretion disk, Hawking radiation, or even the cosmic microwave background. I suspect at a certain point they'd become undetectable due to quantum effects, but the interaction between relativity and quantum effects in extreme scenarios is a major unsolved research problem and is well above my paygrade.
It's also kinda a wrong question, since a core tenant of relativity is that no one viewpoint is privileged as "correct". It's valid to interpret the photons as reaching you in "real" time and saying the black hole never finishes forming because time dilation, but it's equally valid to notice the photons are red-shifted and calculate that by the proper time of the particles falling in, the last stages of the collapse only took milliseconds to happen. This deep-dive on the Twin Paradox gives an overview of the principles involved, especially the last section (Too Many Explanations: a Meta-Objection):
Here's a fascinating YouTube video of some first-ever, slow-motion, close-up videos of thermite combusting (in an ingenious setup that allowed them to view the interior reaction—until everything melted down). The good stuff doesn't start until six minutes in, following a history of thermite.
At the meta-level, I always find it fascinating that when we increase our observational abilities, we see new and unexpected things. We've known the chemistry of a thermite reaction for 125+ years, but when seen up close, the reaction progresses in an unexpected way.
Hello everyone, here I am on like my seventh account (not ban evading[maybe{I'm not actually sure}]), because @sstkacntsysisjsttewrst.
Ie, SubStack Account System Is Just The Worst.
They have this perfect mix of web 2.0/1.0/3.0 design philosophy so as to be maximally inconvenient to me, because I refuse to give them my actual email, I don't share accounts between devices as a rule, and I want to avoid using any of the social features. Basically, I want to treat this like a forum or an IRC, and it wants me to consume product.
This leads to me losing access to whatever account I am on every couple months. I could probably recover them if I tried, BUT:
Given how bad the account system is from a functionality standpoint, it's no great loss. For example, I wanted to find a (pretty good if I say so myself) poem I wrote in response to the one scott posted about falling aid, and the comment history function is so vestigial as to be nonexistent.
Basically, they either need to do more or less because what they have now is a less-good option for every imaginable use case. Even the pay-piggies and milch-cows are suffering; that is wild from a making money standpoint.
I think of myself as pretty privacy focused, but you outdo me by miles. I agree that the nonexistent comment history on Substack is both frustrating and difficult to explain.
But I don't think the practice of tying your account to an e-mail address is money-motivated; that is pretty normal behavior for a bunch of reasons, and account persistance is among them. That said, after creating an account tied to my e-mail address I set a password and use that for login rather than an e-mailed code. Can't you do the same without the e-mail address? If so, why are you losing accounts?
I appreciate the ability to support bloggers financially. I do think that Substack is annoyingly in your face about it in several places, though: the nag to subscribe to a blog you haven't read yet, the button to upgrade your subscription on every page, maybe the post flair for financial supporters of the blog. But none of those should cost you accounts.
>But I don't think the practice of tying your account to an e-mail address is money-motivated; that is pretty normal behavior for a bunch of reasons
No, I have a bunch of pseudo random emails I use for various services that are managed through my home server, I just don't bother recovering the account because there is no real benefit to account continuity vs. just making a new one. I eventually need to log in with a code from my email, probably because I'm constantly fucking with my network and I think substack doesn't actually see a hardwear image from me, just whatever the container exposes to the network.
It's partially for privacy, but it's mainly that If I check my email and see more than a couple automated advertising/service related emails, I loose the will to actually manage it.
>> Financial issues:
That is the srouce of my confusion: the account system is stuck halfway between a patreon style payment service platform and a social network. I think the creators and the consumers would both have a better time of it if they decided to be one or the other, at least on the user side of the account system.
Quanta Magazine just came up with a physics article that I found quite exciting. In short: physicists found a way to estimate the information content of a system. When they applied this equation to an entire, closed universe, with all its rich internal structure, they were expecting to find a large number reflecting that. Instead, the result was... zero. Zilch.
> The theoretical universe seemed to admit only a single possible state. It appeared so simple that its contents could be described without conveying even a single bit of data, not even a choice of a zero or a one.
This means in practice that there is no such thing as an "outside" or objective view of a whole universe. You can't even conteptually look at an universe from the outside and ask what is there, because from that point of view, nothing is there. It boggles the mind, and I love it.
To test this, they re-ran their equations, but adding a explicit observer bounday *inside* the universe. And all the expected complexity re-appeared right away.
This sounds *huge if true*. I understand it to confirm that reality is fully relational. Things only are in relation to other things. No objective reality, physical or metaphysical, at any level. Carlo Rovelli must be having a field day.
The obvious answer is that one (or more) of the premises of the calculation is wrong.
My usual rule of thumb is to ignore everything string theorists say. This can be filed in the same drawer. (And yes, it is string theorists doing the saying).
> The obvious answer is that one (or more) of the premises of the calculation is wrong.
That's what you do when you find the conclusion unacceptable... personally I don't, I actually quite like it. Point taken that it's extremely speculative though.
As with all things string theoretic, it's the result of a tower of speculations one piled on top of another. Maybe it's right, but most likely it's wrong, yet identifying where exactly in the chain it goes wrong will not be easy and is almost surely not worth my time.
(Spoken as a professional physicist (but not string theorist) who knows many of the individuals referenced in the article, and respects them enough to believe they will not have made any obvious mistakes, yet also believes that most of what they do is nonsense).
I found that article to be very frustrating. The breathless writing style of the author keeps saying that physicists discovered this or physicists expect thus and so. He makes it sound like the entire physics community has bought into the ideas of a small group of theoretical physicists. Worse yet, he doesn't go into the details that I need to know if this is all theoretical hand-waving. I suspect it is, but thankfully, he gives links to the papers. OTOH, most of the sciency stuff that comes out of Quanta Magazine is pretty speculative (IMHO). I'm not really motivated to read and try to understand the papers the article links to.
They make it pretty clear that it's a small group pushing ahead with some equations. That's theoretical physics for you... someone runs some numbers in a certain way, gets some numeric result, starts looking for ways to make sense of it, but it takes time to figure out whether the equations really apply to that case, and what the results mean in real world terms.
I clicked the paper but it's way over my level, last time I had some undergrad physics was decades ago. That's why I posted it, hoping that someone with more knowledge of the field would have some insider insight.
> Recent arguments based on the quantum extremal surface formula or the gravitational path integral have given fairly compelling evidence that the Hilbert space of quantum gravity in a closed universe is one-dimensional and real.
This is way over in QG-land. I never understood why one-dimensional Hilbert spaces keep popping up as requirements for QG and String Theory. The stuff they're discussing are way beyond my comprehension, though, so I'm not sure I'd be able to understand it even if someone could explain it to me in simpler terms. But for it to work, it would mean we live in a closed universe. There have been lots and lots of very precise measurements done, and the cosmological consensus strongly favors a flat (or very nearly flat) universe (though there may be few observational cracks in the consensus).
But even if we live in a curved universe, and even if this is a potentially valid mathematical model, it's very likely unfalsifiable.
That's quite interesting. I guess ecology is no longer the only subversive science? Now physics as well? As someone that believes the atomism and anthropocentrism of the Western intellectual tradition is severely damaging to the well-being of the human species (to say nothing of other species!), this seems like intellectually a step in the right direction.
Physics has often been subversive... heliocentrism subverted some firmly established ideas. Then comes the double whammy of GR plus QM at the beginning of the 20th century. And if math can add its own subversion, have a look at Gödel's theorems.
By your reference to cryptography, I suspect you're assuming that Shannon’s information theory (which was designed for communication engineering), applies to fundamental physics. Shannon's information theory assumes conditions that do not generally hold in physical systems. There are a bunch of reasons it doesn't. For instance, Shannon theory is not Lorentz-covariant. In our physical universe, information transfer must obey causality, finite propagation speed less than c, and observer-dependent simultaneity. And Shannon's information theory doesn’t transform properly under relativity (or perhaps I should say it doesn't know how to deal with relativity). Another failing of Shannon's IT is that it assumes classical probabilities, but quantum systems violate that. There are several others, but, to my mind, those are the most important reasons why we shouldn't try to apply Shannon's ideas to physics.
Also, don't get sucked into the idea that entropy is (necessarily) equal to disorder. Sabine Hossenfelder provides two conflicting examples of order and disorder...
The formula that Shannon derived for entropy in communication (Shannon's Measure of Information) is basically equivalent to Boltzmann's H-Function (in fact, he may have based it on Boltzmann's formula—although I'm not 100% sure about that). So, it is often used to calculate the entropy in statistical mechanics. But Phil Attard points out that...
> it is often the case that the microstates are chosen based on the individual characters in a message (or the pixels in an image), and these are indivisible and arguably have no internal rearrangement that can contribute to the information content of the message. In this sense one can make a strong argument for applying [Shannon's equation] in informatic applications provided that one explicitly restricts its use to such microstates. One should not use that particular equation in statistical mechanics because it risks confusion between microstates and macrostates, and also because it is not valid for the microstates that typically appear in statistical mechanics.
Likewise Boltzmann's H-function isn't really a description of entropy according to Arieh Ben-Naim...
> Thus, once it is understood that the function −𝐻(𝑡) is an SMI and not entropy, it becomes clear that the criticism of Boltzmann’s H-Theorem were addressed to the evolution of the SMI and not to the entropy. At the same time, Boltzmann was right in defending his H-theorem when viewed as a theorem on the evolution of SMI, but he was wrong in his interpretation of the quantity −𝐻(𝑡) as entropy.
I think maybe the question is: from a universe-level perspective, what is entropy? The state of the universe as a whole is just some pure state, and all pure states are just a unitary transform away from one another.
But just by introducing boundaries between observers, you get non-unitarity and entanglement entropy, as your system entangles with stuff that leaves your light cone.
I'm not sure if this is the actual mechanism the cited paper finds (I haven't had the time to look, and I doubt I have the physics chops as but a humble quantum information theorist), but it seems like a reasonable intuitive guess.
To me, this is still the underlying puzzle that MWI doesn't really resolve: what is the mechanism by which a pure state evolving unitarily gives rise to robust subsystems that have the structure necessary to give rise to all the messy, entropic behaviour we know and love(?)
If you start with those subsystems, MWI explains how they continue to induce a branching into worlds in a basis natural to those subsystems; but it's not clear (to me, anyway) what breaks the perfect pure state symmetry to give rise to those subsystems in the first place. The cited result confirms (to me) the tension between the unitary-evolution/God's eye view (with no entropy, no obvious internal structure) and the messy ground up view of interacting systems--somehow, there is a decomposition of this simple, structureless pure state into complicated things in a way such that those internal components have their internal reality; or from the other end, somehow these messy interacting entropic subsystems add up to something with no internal structure at all.
They concern themselves with entropy in black holes, and I think they're modeling it as if our universe were closed and had the same dynamics of a black hole.
> Moreover rules were given for computing interior observables in the fundamental description that agree with the semiclassical description up to ambiguities which are exponentially small in the entropy of the black hole at the time the observer jumps in, which is perhaps the most we should ask for.
And...
> Thus we see that even when the black hole is small there can still be sizable entanglement between its interior matter and a reference system, as long as there is a big observer who is also inside.
And here they invoke decoherence...
> entanglement arises from the entanglement of the observer with the rest of the universe due to decoherence, and the observer is cloned out of the system in their pointer basis to ensure that they are effectively classical.
Pure versus impure , simple versus complex, and symmetric versus asymmetric are different distinctions.
Under decoherence, it's assumed that th e environment is impure, and an initially pure WF can't remain in coherent superposition once it interacts with an environment.
But that says nothing about many word branching -- which might require a high degree of symmetry in the environment.
Yeah, I'm using "symmetry" loosely and idiosyncratically; I have in mind something like transitivity under unitary rotation; without a preferred basis, no "internal" decomposition of any pure state into subsystems is canonical in any way.
I don't think you need an impure environment for decoherence; in the formulation I'm most familiar with, Stinespring dilation and CP maps, you usually assume the environment is in a pure state as well: the system starts off in a tensor product of two pure states, then a global unitary entangles the two systems, and when you trace out the environment, you will have a rank>1 density matrix on your system.
>what is the mechanism by which a pure state evolving unitarily gives rise to robust subsystems
Doesn't decoherence provide the mechanism? All you need are uncorrelated degrees of freedom to entangle with. A large enough system will have parts that are uncorrelated with each other. Isn't it analogous to how parts of the universe can have nonzero angular momentum even if the universe as a whole doesn't?
This works if you already have some decomposition into subsystems where one subsystem can act as an environment for another.
So I agree, if you have a large system, and an environment for it to interact with, the subsystem evolves in a robust way. But if you don't already have a decomposition into "large subsystem and its environment", you just have "pure state of the universe", where did that initial decomposition come from?
This is a great paper, and I'm shocked I've never seen this result before.
May have more to say in a bit, but need to digest this a little first.
First thought though: evolution under the Hamiltonian still requires some external notion of time; I feel like I'd want the quantum state of the universe as a whole to have the time evolution somehow emerge from within the state, not be tied to a notion of time that exists outside in some way.
Does anyone have a recommendation for a history book for my younger brother? He's read a lot of Horrible History and other kids' history books, but he seems to be outgrowing it a little. He still loves it though, and he's absolutely riveted when I tell him stories of the past. (Right now we're on the French revolution.)
He's not at the reading level to read a proper textbook. Any books which should be readable for him, but have some more meat than pulp histories? I think he would prefer a whole book on a certain event, time period, or person than an overview of a huge swathe of history.
As a general rule of thumb, it's probably best to be well-acquainted with the "formal" history of any given period before grappling with revisionist history of the same.
I don't know that I have a good understanding of what kind of material you're looking for.
Two of my favorite books that might be called history are Amanda Podany's "Brotherhood of Kings: How International Relations Shaped the Ancient Near East" (she's also written a followup, which is good but not as good), and Patricia Crone's "Pre-Industrial Societies: Anatomy of the Pre-Modern World".
As you might guess, Crone's book is more of an overview of what things were like in very broad terms. If you want "stories of the past", Brotherhood of Kings is clearly a better choice; much of its structure is precisely this.
Here's a story that appears in the book:
After a political crisis, the throne of Egypt is held by a young widow. She feels political pressure to marry an Egyptian noble, which she doesn't want to do. So she sends a letter to the king of the Hittites, Suppiluliuma, saying that since he has many sons, she wants him to send one of them to marry her and be the king of Egypt.
The Hittite king is stunned (obviously), and convenes an emergency council to decide how to respond. Over the objections of certain counselors who say this has to be a trap, he sends one of his sons as requested.
The queen seems to have been entirely sincere. But those local nobles take offense, and the prince encounters foul play.
When word gets back to Suppiluliuma, he is naturally enraged. He gathers up his army and goes on a rampage south from Anatolia, intent on punishing Egypt. He doesn't make it all the way there, but he does manage to conquer a good amount of Egyptian territory in Syria.
Another story that stuck with me was of a pair of princesses given in marriage to the same Mesopotamian king. The older one married first and became very unpopular locally, with rumors spreading that she was a witch.
The younger one arrived later, marrying the same king, and hoped for help from her sister settling in to her new life. She didn't get it; we have the letters she sent back home, lamenting that her sister was no support and the rumors of her being a witch were true.
As another recommendation, John Keay's "China: A History" is very readable. It covers a very broad period of time, but it also includes plenty of stories. I haven't read his history of India, but I imagine it is similarly pleasant to read.
It might be stretch for his reading abilities but The British Are Coming by Rick Atkinson (https://www.goodreads.com/book/show/41812830-the-british-are-coming) was an excellent narrative history of the beginning of the American Revolution. I thought I knew it pretty well (esp. having lived in New England the last few years) but I learned a ton and enjoyed every minute of it.
Just saw the "not at reading level" part yet, so depending on his age, Extra History might be another good fit if you aren't wed to "readable textbook."
They can be hard to find, but has he read any Landmark books? It was a series of standalone history books, each covering something or someone specific, written by popular novelists. So not textbook-y. They are from the 1950s and '60s, so take that as a content warning I guess, but they're still well regarded.
How about The Cartoon History of the Universe, by Larry Gonick? It is to some extent an overview, but it does go into detail - it's got three volumes so quite a bit of space. It's also comic format, which works surprisingly well.
I was a big fan of Gonick's Cartoon Histories growing up, but I'm hesitating to pass it along to my daughter. They're fun and well-written and do go into appropriate levels of detail, but the historiography is extremely dated. The three volumes of CHotH were written between 1978 and 2002 (23-47 years ago), and Gonick also has an unfortunate tendency to uncritically accept and repeat bad pop-history takes (the standalone Cartoon History of the United States is particularly bad in this respect) and to treat religious/mythological accounts as if they were historical primary sources (off the top of my head, he does this with the Old Testament, the Gospels, the Romance of the Three Kingdoms, the Illiad, the legend of the founding of Rome, and the Hadiths).
I worry that handing them to someone who doesn't already have some basic grounding in general history and the maturity to understand that a source is unreliable would wind up giving them rather too much to unlearn later.
If it is not gauche, may I recommend my own book, published two years ago by Macmillan, Impossible Histories? It is half alternate history but half real history, so you get the historical details on twenty important events across 2400 years and six continents (sorry, Australia) as well as the part where I make up twenty lies.
Impossible Histories by Hal Johnson! Ask for it by name!
As an unrelated 3rd party, I can confirm that Impossible Histories is very good.
From testing with other friends, it seems that the there is some level of background knowledge to really appreciate it. I can't tell if OP's brother is at the right level yet, but it could be a good fit already and, if he continues learning about history, it will eventually be a great choice.
History books can be hard to recommend given the wide gulf between academic survey and airport pop history.
I'd recommend two podcasts: 'Hardcore History' by Dan Carlin and 'Fall of Civilizations' by Paul Cooper. The former more or less got me into history as a hobby and the latter is just really nice!
For a kind of random book I recently finished 'Xenophon's Retreat' which I thought was a nice "interested lay person" summary of Xenophon's works with good surrounding context.
If the boy is too young to read a "proper textbook", there's a high chance that Carlin will be too brutal for him. Genghis Khan, WWI, WWII Japan, and the Munster Rebellion are all episodes I distinctly remember having plenty of stories one *might* not want to expose to a, say, five-year-old. If you prep that 5YO beforehand with a clear lecture on Things We Don't Do Anymore Because Society Would Condemn It, there's even a Carlin episode (might be in his sister series Common Sense) that specifically goes over things we did to each other in living memory, with community approval, that's probably going to confuse the heck out of him.
Excellent series for adults, though, I'll strongly agree.
How old is he? If he's big into the French Revolution, Mike Duncan's "Revolutions" podcast does an overview of that which is lay-audience accessible but comprehensive
Have people here had positive experiences with taking SSRIs for anxiety- in particular social anxiety? Or negative, or neutral experiences- I want to hear the whole spectrum. I've long been opposed to taking them, but just wanted to gather some data. I'm one of those people whose fear response has always been dialed up rather high- I'm jumpy and get afraid easily. This has over the years (I'm middle-aged) metastasized into 'I actively avoid all social situations longer than 5 minutes' level of social anxiety in most (not all) situations. Just curious what's worked for other people in terms of pharmaceuticals. I'm deeply creeped out by the idea of flattening all of my emotions and personality, though I suppose one could argue that I have little to lose.
(PS, if your answer is 'have you heard about our Lord and Savior CBT'- yes, I've tried it, it doesn't work for me, please preach elsewhere)
Duloxetine seems to work pretty well, though high doses do seem to flatten emotion a bit, and even higher doses actually cause anxiety to worsen. But you can just reduce the dosage if that happens. Those pills aren't hard to open.
Venlafaxine/Effexor also works, but the initial side effects and withdrawl symptoms are very very bad. Do not take unless you've tried alternatives.
I've tried a fair few SSRIs for depression accompanied by anxiety, and generally stopped taking them after a while because it was unclear if they were helping the depression at all. I did not feel that they flattened my emotions and personality in general (though the depresssion sure did).
I did have more success with an atypical antidepressant for the depression, but that caused the anxiety to spike, so the doctor added Effexor and that seemed to work. Currently doing reasonably well socializing, being creative and enjoying things while on meds.
Not me but a good friend is prone to debilitating anxiety/panic attacks and he's had good results with both prozac and lexapro. He takes a very low dose (half of the minimum clinical dose) and it works well for him. As an outside observer the results were obvious. Just find a good psychiatrist and work with him to figure out what works for you.
I've taken lexapro for anxiety in the past. It helped a bit in terms of taking the edges off of things so I could be a bit more functional, but didn't really make me feel better overall.
What did help quite a lot was when I started estrogen. I stopped taking lexapro a few months into my transition and haven't missed it. But that's very situation specific and definitely not for everyone.
I have had really excellent results with them, both for social anxiety and health anxiety. I did not have any issues with flattening of emotions or personality - I'll still cry at a well-done phone company commercial.
SSRIs generally don't kick in until you've been taking them consistently for several weeks (and the effects can take a while to stabilize, so anxiety and mood may be worse initially).
Yeah so the serotonin-reuptake-inhibiting effects, which I'm pretty sure mediate the ejaculation stuff, have almost nothing to do with the antidepressant effects.
It takes ~10 nM of prozac (fluoxetine) to inhibit reuptake via the serotonin transporter in the CNS.
In someone taking 50mg of prozac a day (a pretty normal antidepressant dose) you find synaptic concentrations of fluoxetine on the order of 10µM. 1000 times as much!
Either drug companies aren't THAT greedy, or "SSRI" is only nominally true lol.
Source: I have no fucking clue what paper this screenshot on my phone was from, but you can probably figure it out by google-scholaring if you're curious.
I've been taking Citalopram since February, I'd say it has a marginal effect but my hope is that's all I need, I'm looking for 5% improvement wherever I can get it.
I follow AI developments closely and use it a lot in my work. I also write and record musical satires about AI, robots, and our overlords. If anyone is interested in exploring song ideas together, let me know. The music is all home-rolled, AI is only used for mastering.
We use Izotope Ozone 12 for mastering as a start and tweak it from there, it’s infinitely tweakable. The Ozone 12 website has before and afters. Our DAW is Studio One Pro 7. We do all the vocals and instruments.
New article from Gusev on Missing Heritibility "The missing heritability question is now (mostly) answered" . I am rather confused at his interpretation of the Wainschtein et al. paper, which does not seem to quite support his position? But I need to read more deeply.
I am a little confused by his approach averaging across traits, and I am still left wanting a mechanistic explanation for twin similarity.
> I am still left wanting a mechanistic explanation for twin similarity.
Gusev doesn't have one for you. He admits it in his previous Substack post, "We still do not understand family resemblance" (https://theinfinitesimal.substack.com/p/we-still-do-not-understand-family). What's interesting is that in that article, Gusev references a paper that shows not only MZ twins have higher correlated traits than DZ twins (which was expected), but adopted siblings, and in-laws show higher correlations than DZ twins (which was totally unexpected!).
> The other clear outlier was the correlation of adopted relatives, which is expected to be zero under a Fisherian model but is clearly and significantly not zero. Interestingly, using the CTD to estimate heritability under random mating produced a value for the shared environment that fits the adopted siblings fairly well, but does not fit the other biological relatives. On the other hand, accounting for assortative mating could fit most of the biological resemblances but reduced the estimate of the shared environment to zero and thus no longer fit the adopted and in-law resemblances. MZ twins, in-laws, and adopted siblings all exhibit higher correlations than expected, so accounting for one pushes the model away from the other and vice versa.
> I am rather confused at his interpretation of the Wainschtein et al. paper
The thing that confuses me most about Gusev's position is the real-world implications.
Where would Gusev's real-life-behavior recommendations cash out any differently than a strict hereditarian's would?
So now GWAS's and family GWAS's and all sorts of more refined and informative takes (RDR, SR, etc) are coming out that have identified a "heritability gap" between what we can measure genetically and what twin studies tell us, and which argue "nope, that's wrong, environment is way bigger and more important than we thought." And this is Sasha Gusev's primary topic of interest.
But I think, “oh, environment is 4x bigger than we all thought?” so what should we do differently? Absolutely nothing? The biggest thing is still maxing as hard as you can on spouse quality, aka genes? Oh, and since “family culture” apparently matters a lot per family GWAS', you should act as the most hereditarian-pilled would recommend and also max on *lineage quality* as well as spouse quality, as the most hereditarian-pilled among us (ie those aware of Greg Clark's research) would recommend?
And as near as I can tell, the Gloomy Prospect is still true, too. In other words, for all the environment supposedly matters 4x more, none of the “environment actually matters” people can tell you even a single environmental intervention that will move the needle in a positive way.
For raising our kids, it's basically "make sure they're raised in a two parent household with good mores and no physical or mental abuse," and that's the tippy-top best you can do.
Everything else is noise, in the sense that everything else that matters was handled at the "mate selection" stage. You married into whatever level of educational attainment and average income and wealth you're going to be drawing from, and now it's just random draws from you and your spouse's conjoined lineage baseline. All environmental factors that we know matter are negative - malnutrition or physical or mental abuse all bring your kids down, but we know of nothing that brings them up, and the “hereditarianism is wrong” people have no advice.
The other thing I’ve never understood is that it’s not like we haven’t controlled for these things before! If culture / environment REALLY matters ~4x more than we think from Polderman, why didn't it come out in any of the much older separated at birth Minnesota twin studies, or the many studies where disadvantaged kids are raised by middle / upper class people in much better environments?
It's not like we haven't thought about or controlled for "culture / environment" before - and I've asked this before and didn't get a good answer.
So I remain puzzled by the seemingly warring sides, and it genuinely seems to me that behaving as a stricter-than-usual hereditarian is the only conclusion to come out of the whole debate with, and that this is affirmed by BOTH sides.
It has strong implications that we should be providing social support to help people make up for deficiencies in their local environment, because this is likely to make them more productive.
That's true, and interesting. It's my impression that the most strict hereditarian still seeks the best educational resources for their children -- therefore acting like they think the environment still matters. Which is rational, but not then voting to deny those resources to other people. If the environment matters for your kids, it matters for theirs too, regardless of what anyone inherited. *Unless* one is arguing that what is inherited is the ability to utilize environmental resources, but I haven't seen that argument made explicitly. There's all sorts of evidence against it.
As for what one should do differently -- that's a strange question. I personally make few life decisions based on my theoretical beliefs -- I base them instead on the results of a lifetime of trial and error attempts to adapt to local circumstances. Those life experiences aren't going to change in meaning just because science says one thing or the other about how much IQ is inherited.
I find that science doesn't really inform daily life adaptation very well -- it's purpose is to inform institutional policy (for example, educational practices), which strongly determines the local environment various people have to adapt to.
I don't learn science to become more successful in my personal life (except in my own professional domain, but that's a different dynamic), I do it to make better informed choices at the ballot box, and to help change public consensus through private conversation (as I am doing right here).
>I find that science doesn't really inform daily life adaptation very well
well doesn't science say that you should take vaccines and avoid unprotected sex with people without recent STI tests?
Doesn't science say that if you want to have smaller ping time in online games you should move closer to game server rather than buying more expensive cables?
I asked you not to talk about ballot box yet you do. FTFW, a person might be anti-heredetarian and still be opposed to redistribution for another reasons.
>none of the “environment actually matters” people can tell you even a single environmental intervention that will move the needle in a positive way.
Exactly. The "missing heritability" argument is an isolated demand for rigor. If environment is so powerful then figure out why and solve all of our social problems.
Nonsense. There are plenty of studies. It's just that the hereditarians ignore them. For instance, there was the French Transracial Adoption Study (Weinberg, Scarr & Waldman, 1992) that looked at Black, mixed-race, and white children adopted into high-socioeconomic-status (SES) French homes. Children’s educational outcomes tracked adoptive family SES strongly, not biological background.
But the key takeaway from the Wainschtein and Zhang study is that SR and RDR put narrow-sense heritability down around 30%, and GREML-WGS (which uses whole genome sequencing) put it as low as 27%. And when we look at IQ it's down around 32% using GREML-WGS. Three distinctly different methodologies yield surprisingly close numbers, yet (some) twin studies make IQ heritability 2x higher. Gusev's snarky remark on X is why would you favor a 40-yo methodology over current science?
Do you see irony here you finish you comment with sneer about 40 year studies, but you started it with 33 year old study for your cause, even forgot which one....
These improvements vanish after adolescence and probably goodharted. But even if they are real, what is does tell for public policy? Take children from poor SES familes and force high SES families to adopt them? How is it different from negative eugenics? Maybe you say that low SES families should change something.... something... but you don't know what specifically. They study you mentioned was about adopting.
> French Transracial Adoption Study (Weinberg, Scarr & Waldman, 1992) that looked at Black, mixed-race, and white children adopted into high-socioeconomic-status (SES) French homes
I think you mean the Minnesota adoption study? Because Weinberg, Scarr, and Waldman were the ones who did the Minnesota adoption study over thirty years. There are separately two other studies done in Germany and Japan respectively that are similar to this (in the senses of tracking IQ and educational outcomes segmented by race of the biological and adoptive parents), but I'm unaware of any French ones.
And the Minnesota study clearly shows the "washout" effect on IQ that most adoption studies show in the testing at age 17, with the overall study followup synopsis asserting that it was a mess and couldn't be used to support either hereditarian or environmental arguments:
"In a 1998 article, Scarr wrote, "The test performance of the Black/Black adoptees [in the study] was not different from that of ordinary Black children reared by their own families in the same area of the country. My colleagues and I reported the data accurately and as fully as possible, and then tried to make the results palatable to environmentally committed colleagues. In retrospect, this was a mistake. The results of the transracial adoption study can be used to support either a genetic difference hypothesis or an environmental difference one (because the children have visible African ancestry). We should have been agnostic on the conclusions . . ."[15] Later opinions supported Scarr's reassessment. For example, one group of authors wrote, "Generally, scholars in the field of intelligence see the evidence from this study . . . as consistent with both environmental and genetic hypotheses for the cause of Group IQ score differences"
> Oops. My bad. I meant Christiane Capron and Michel Duyme (Nature, 1989).
Thanks, I hadn't seen this one - I actually see nothing about race in this one, it's primarily about occupational and educational status of the biological and adoptive parents.
They assign kids to 4 quadrants - high biological background (B), low adoptive background (A), high / high, low / high, and low / low.
They see roughly a ~10 point childhood IQ advantage in the high adoptive backgrounds relative to the low adoptive backgrounds, regardless of whether the children came from high or low biological backgrounds (although of course the high biological background kids were higher in an absolute sense):
That said, I think you would still expect those effects to wash out for both tranches of kids by adulthood, as seen in numerous other adoption studies. And "adulthood" is typically "17," so pre-college.
So I'm not sure what this is arguing for in terms of interventions - does "environment" really matter if all it does is get you a transient IQ bump for ~10 years or so that then washes out?
It might matter at the margins - some kids might complete high school versus dropping out, or have better grades and get just over the side of a bubble in college admissions.
But happily, we already sort adoptions into "high only" adoptive parents, so anyone being adopted should already be seeing these benefits.
The Motte is a creepfest because they are so ruthlessly technocratic, and so I was readying my armor to support you, and then I see that you crashed out and got banned because... They aren't technocratic enough? 😆 Elitism is a sick fantasy life for people who think they have to tear others down to drown their own insecurities.
Have you ever considered that the guy who looks down on you and calls your political ideas stupid isn't covering for his "insecurities," it's that he really is smarter than you and his ideas are right and yours wrong?
Yes, I have considered this. I have political “adversaries” whom I respect and take at face value. I just don't include elitists among them. Elitism is just too ridiculous, I can't take it seriously. Special case. Anyone who unronically uses “low-class” as a slur in 2025 is not likely to be a serious person.
Any use of the word “dysgenic” is a red flag, although it's not disqualifying in the same way. I certainly don't RELISH the idea of genetic diseases and pathologies spreading through the gene pool, but I hardly consider it an emergent crisis.
Just linking the post that got Alexander banned from the Motte for visibility, unless hapless people see this sort of thing and assume the Motte is actually as bad as people think.
"P.S. To preempt the accusation that I ignored ThomasdelVasto's point, I reject the whole theory that poor, low-IQ people are harmed by competition with rich, high-IQ people. People are willing to risk dying in the desert to move from low-IQ to high-IQ countries because high-IQ has massive positive externalities.
P.P.S. I know I might get banned for this post. I was drawn to the forum because I'm a long-time Scott Alexander fan, from back when "right-wing SSCer" meant "secular guy who talks about embryo selection and national IQ," not "guy who thinks we need to go back to 1710 ideas about religion and government and that eugenics is evil." For me, unlike many of you, the former wasn't just a gateway drug to the latter, so I'm "left-wing" now. You can follow me on Substack and Twitter.
P.P.P.S. The mottezien is immunized against all dangers: one may call him a cuck, nazi, bigot, fascist, it all runs off him like water off a raincoat. But call him a resentful prole and you will be astonished at how he recoils, how injured he is, how he suddenly shrinks back, calls you egregiously obnoxious, and then bans you from the forum."
Quoting a copypasta? Well, that will probably get you banned just by itself.
Hey! Ideas on religion from 1710? What is this crazy modernity, the 13th century is plenty updated for me!
"The tattoo-covered WWE fan doesn't want to listen to a sermon from the Southern preacher but recognizes him as a fellow member of the broad ingroup of low-class Americans who share a common inferiority complex toward urbanites with lots of education and money."
My problem regarding Alexander's comments is that he clearly wants to present himself as "an urbanite with lots of education and money" but the amount of carping he does about hte low-class makes me go "tuppence-ha'penny looking down on tuppence", that is, not really old(er) money, but only a generation or maybe two removed from the grubby low-class types and is constantly petrified with fear that people he wishes to emulate will mistake him for his ancestral roots.
Alexander, if I'm mischaracterising you, apologies. But you go on so much about a particular type that it's hard not to think "the lady doth protest too much", as it were.
I shop at Walmart, eat at McDonalds, and watch true crime documentaries. My class background is complex but it's certainly not "old money." I have some low-class traits picked up from my family background. I've said this to you before, but I'm sure you'll pretend not to see it and repeat your "sick burn" in a future thread.
Alexander, I'm not trying to sick burn you. But you do have a few hobbyhorses you ride into the ground, and it does sound more like "very very anxious about perceived status so as not to be confused with the wrong class" on your part when you bang on over and over and over again about the lower classes and the right fetishing manual labour etc.
I'm happily lower-class myself (peasant to the bone) and couldn't give a damn what my 'betters' think of me. Relax, it'll be a lot easier on your ulcers!
Ahh yes, this comment where you are clearly speaking in good faith and not just taking shots at your outgroup:
> his worldview would seem to conflict with HBD theories. Indeed, one would have to conclude that whites are an inferior race. Guatemalans in their "third-world s***hole" don't just sit around despairing, they cross multiple borders and look for work in a country where they can't even speak the language, while white men who got laid off in their rust-belt factory towns twiddle their thumbs and inject fentanyl, unable to compete with said Guatemalans.
> If I believed, as many of you profess to, that my race was at risk of going extinct, I wouldn't be centering my politics around helping the least capable members of said race who refuse to help themselves. Don't you have bigger problems? It's not like you should feel any "political" loyalty to them, Trump's working-class base work, homeless people rarely vote.
Either way, you continued to escalate constantly in the comments of your posts as well, which was a huge extra piece of context, on top of the already bad faith discussion like above.
I'm sad you got banned because you did generate a lot of interesting discussion and had a genuinely heterodox worldview for the forum, but I can't say it wasn't justified.
Hitler was so loathsome and destructive that I wonder if it's even possible to have a comprehensively accurate picture of the man in 2025. The wounds are just too fresh. Any ordinary, decent person is going to have a strong predisposition to believe the most unflattering things about Hitler's private life, even if they are consciously honest.
And few people have any incentive to correct the record when someone repeats a lurid rumor. That's making yourself a target for little appreciable gain.
I hate Nazism, to be sure, and I could have used any number of examples, like Mao or Charles Manson. Our ability to understand a loathsome historical figure seems inversely proportionate to how RELEVANT that understanding is. What a tragedy.
Of all the epistemic traps we may find ourselves in, this one does not seem to be particularly relevant? I mean, it would be nice if we did not have a systemic bias in judging how Hitler treated his dog, but the stuff for which he remains relevant today is generally not his private life.
It is not like he wrote a book on caring for dogs and society is missing out on his unique insights as they unfairly reject the book based on his politics.
Can you rephrase this in different language? I'm a little confused. First of all you say this epistemic trap is" not relevant." Not relevant to what? I'm not sure what his dog has to do with this. And he was such an important figure in the history of the 20th century, it seems to me that every aspect of his biography is relatively important, or at least potentially important.
After all, the exact reason for Hitler's rise to power, and the exact nature of his personal dysfunctions, are some of the most enduring mysteries in modern history.
Are you implying "Hitler has zero redeeming qualities, so it doesn't matter if we describe his life inaccurately?” whether or not he had qualities that are “redeeming,” we definitely want to be as truthful as possible about an important historical figure.
We want to stick to verifiable facts, don't you think? Besides violating whatever inner commitment we have to being truthful and accurate, spreading questionable rumors damages our credibility.
> After all, the exact reason for Hitler's rise to power, and the exact nature of his personal dysfunctions, are some of the most enduring mysteries in modern history.
I disagree with that. I see Hitler's rise to power as an institutional failure, not something which only he could do. I sincerely believe that a full dissection of his psyche would not yield any answers to explain his rise to power. He did not have an Illithid tadpole which enabled his oration and drove him to murder.
There is always going to be a pool of fucked up, murderous people wanting to become dictator of the world. Hitler is hardly unique in that regard, just because circumstances made his bid more dangerous and bloody than for most others.
Studying how dictators work in general, and what motivates them (as Scott does with his dictator's book club) might be interesting, but we should not assign a higher weight to Hitler just because he managed to murder more than most.
"The scoundrel on a small scale should not, if those in power allow him to become a scoundrel on a large scale, occupy a special position not only in scoundrelship, but also in
our view of history." (B. Brecht, essay on Athuro Ui, https://www.berliner-ensemble.de/download/document/6715 p. 5 (German))
There is a tendency, especially to Germany, to see Hitler as evil incarnate, a certain fetishization of the most minute details of his life. I understand the motive: the more Hitler is the antichrist, an anomaly on the scale of Foundation's mule, the more blameless the Germans are for falling victim to his seduction.
The truth is that human evil is utterly banal. We did not fall for a force of nature, we fell for a rather ordinary human (and his equally ordinary to pathetic cronies), perhaps in the 95th percentile of murderousness. Studying his sex life (or whatever) will not yield greater insights into WW2 than studying the sex life of any other person. (As a corollary, this also means that looking at what makes low level grunts participate in genocides much more interesting. See Goldhagen: Hitler's Willing Executors, which despite the title does not focus on Hitler at all.)
I agree with you that spreading falsehoods is generally bad, and that we should strive to avoid it. I just don't think that the number of Hitler's balls has more practical relevance than what type of damage a specific Pokemon does.
The only way to determine that he was not, in fact, unusual is to investigate him.
Sure, in the same way that the only way to determine that your neighbor is not secretly a Mossad agent is to investigate him.
Hitler was rhetorically excellent and an especially murderous antisemite, and had an unhealthy risk appetite. He was also very lucky to find himself in the position to make his career. I do not think one needs to know more about him to understand WW2.
Also, he has been investigated as much as a person can be investigated. Every aspect of his life has been subject to books and movies to a point which places him in the top 100 persons-as-book-subjects. I seriously doubt that any further detailed evidence unearthed by someone writing about his use of the color blue in his painting career or his relationship with dogs or whatever will revolutionize our understanding of the Shoa.
In my analogy, once you have found that your neighbor is out most nights is because he has a job as a night guard, and that the antenna in his backyard is because he is a ham radio enthusiast, and have checked his job and his radio transmissions, you will come to the point where there is little left requiring an explanation. Sure, it could still be that he is a Mossad agent, and that digging deeper will reveal that fact eventually, but realistically you are just wasting time with an obsession.
You assert the very points at question. Your claims about his character have to be established as well as your neighbor's night-watchman job, and you have not done so.
You may be doing the exact thing that I worried about in my original post. Perhaps you are so determined to denigrate Hitler's reputation, such as it is, that you want to deny him any uniqueness.
Whatever self-interested motives modern Germans may have, Hitler's political career was weird, singular and uncanny. The disconnect between his childish ideology and his meteoric rise DOES indeed constitute a major historical mystery.
The Nazi Party's commitment to the mass killing of civilians, without any credible military justification, was very unique at the time. Even among right-wing authoritarians. Though Hitler's example was influential on the right, his brand of evil remains fairly unique to this day
To compare Hitler to an ordinary dictator is not only unfair to his victims. It fosters an environment where it is normal to call any old right-wing authoritarians "Nazis." An act that makes people looked dishonest and unserious outside of certain circles. A rhetorical act of "crying wolf" that impedes and frustrates the mission to crush rightist authoritarianism forevermore. Or perhaps you want to obscure the foundationally right-wing nature of the Nazi catastrophe? But I really shouldn't be speculating so freely. Whatever your motivations were, I don't think your comment here was very helpful.
I think that you give Hitler personally too much credit.
For the most part, he did not invent the Nazi ideology from the scratch. It was very much a collaborative effort.
If he had been killed by a grenade in 1917, I am not convinced that history would have gone very differently. There was a niche for a quasi-revolutionary far right party in the Weimar system, and if the NSDAP had not taken it, some other rabidly antisemitic, ultranationalist party would have popped up in its place.
Antisemitism is very much an active ingredient here. It had a long history. For the NSDAP, the Jew represents, among other things (like unfriendly foreign powers) everything which is wrong with capitalism (which was plenty, in 1920s and 1930s Germany). Take that away, and you lose half the votes.
If you have a movement whose core tenet is that the Jews are what is wrong with the world, and which is not traditional monarchist but considers itself revolutionary, unbound by traditional moral limitations, and which is willing to shed blood to create the new order, things will turn ugly.
I would claim that there is a German cultural tendency to follow their beliefs to its logical conclusion. I mean, the eugenics / racial genetic hygiene aspect of the NSDAP was much less load-bearing than the antisemitism, and still they took it to the logical conclusion to fatally poison disabled kids with exhaust gas (Aktion T4). Not because it was personal pet peeve of Hitler, but simply because *the people in the NSDAP actually believed their own ideology*. If you are a doctor or nurse and believe in the Nazi ideology where the people are everything and the individual is nothing, and where genetic defects are a thread to the collective health of the people, and you are not bound by traditional (e.g. Christian) morality and smart and able to think outside the box, chances are that you will independently discover gassing disabled kids as the rational course of action (given one's beliefs, which are of course factually wrong and morally abhorrent).
Could Hitler have stopped the Shoa if he had wanted? Possibly, the Fuehrerprinzip meant he could do whatever he wanted. If he had told is goons, "okay, we have kicked out the Jews from civil service, we have taken the factories some of them owned, we have smashed their shopping windows and burned their synagogues, but now we will actually draft them to the East Front instead of wasting a lot of resources to murder them", I do not think that the other Nazis would have staged a coup.
The thing is, a Hitler who was not rabidly antisemitic, who perhaps had some Jewish friends and went on record that they are okay, that Hitler would not find himself in charge of Germany in 1933.
The Holocaust was blatantly stupid in strategic, purely amoral terms. You are not grappling with that. Brutal anti-Semitism was latent within German culture. But the Jewish Holocaust was the exceptional product of Hitler's mind. He is singularly responsible for turning a pogrom into a Jewish genocide.
Your normalization of Hitler within the context of German anti-Semitism is absolutely ahistorical. In your effort to demonize all Germans, you are inadvertently excusing Hitler and posthumously benefiting his reputation. You can't have it both ways, however convenient that would be.
> you want to deny him any uniqueness [...] Hitler's political career was weird, singular and uncanny
I would expect historians to be biased towards exaggerating his uniqueness. For example, before WW2, antisemitism was popular practically everywhere. Eugenics was considered woke.
It was only after we decided that Nazi Germany was the enemy, that we tried to do less of whatever they were famous for.
In the context of ostensibly serious American historians, right-leaning historians have been incentivized to exaggerate Hitler’s uniqueness, and left-leaning historians have been incentivized to underrate his uniqueness. “Just another rightist doing right-wing things.”
In the context of sleazy popular historians, exaggerating his weirdness is a standard way to move copies.
In terms of body count and incidental cruelty, I don't think Hitler’s activities can be sensibly compared to any American campaign of eugenics or anti-semitism.
> You don't even know what his favorite cake was.
Sure, but knowing it doesn't provide any important insights into history or psychology or anything.
How do you know environmentalism would not have happened some other way?
> No, we don't. Nearly everything you know about the Nazi party is propaganda. They didn't even call themselves Nazis, for god's sake!
Speak for yourself. I think I have a fairly good grasp on their ideology, value system and their organization and methods.
Are you claiming that if I were to read "Mein Kampf" (which certainly is propaganda, just not the one you meant, and to my knowledge rather open about Hitlers ambitions), I would concede that this guy had actually a lot of good ideas and was just misrepresented by the evil Western mainstream propaganda apparatus?
> The best thing that the Nazi Party did was found our modern environmental movement (something that the econuts keep under wraps, for understandable reasons).
https://www.titanic-magazin.de/fileadmin/_migrated/pics/Fuehrer_Privat37.jpg
(Caption, "I think he is ok, I just hate that thing with his Autobahns". Funny because the Autobahns are the one example often given for "not everything Hitler did was bad".)
I do not think that the claim that the NSDAP are the spiritual ancestors of environmentalism is actually true, though. Sure, they did care about the environment, in some ways, and had an aesthetic which favored primeval, unspoiled countryside populated with pure, primeval, upstanding Germans. Still, the good of the state (which is the good of the people, for them) came first, and if that required strip mining, then strip mining it was, and if it destroyed the habitat of some rare German frogs or something, they were shit outta luck.
In Germany, there is also little continuity to between the NSDAP and the Green party, arguably less than between most other parties. The German Greens emerged from the 68er, which was a left-wing student movement which was Western-skeptical and explicitly anti-Nazi, and accurately accused the other parties (e.g. CDU, FDP) of sweeping Nazi crimes under the rug.
You might as well claim that beer is strongly influenced by Nazism, just because it was something the Nazis cared about to some degree. If you want to bash the (German) environmentalists, consider going for the standard claim that they thought about extending the sexual liberation to cover pedophilia in the early 70s, that has the advantage that there is a kernel of truth to it. (Wiser heads prevailed and decided that this was actually a bad move in a lot of ways.)
I know they didn't call themselves Nazis. That's no more inappropriate or telling than the fact that we call Deutschland Germany.
Pray tell, what are some other falsehoods about the party that have been foisted upon us?
As far as I can tell, American propaganda tends to exaggerate Nazi Germany's technical competency and fighting abilities, probably to increase our perceived valor for having defeated them.
Hitler did not in fact make the trains run on time, and Germany's military innovations happened in spite of their feckless dictator. The German economy wasn't even fully on a war footing until 1944, which speaks to stunning incompetence; or at least poor administration.
Then there's the alleged, near-superhuman abilities of the individual German soldier. Once you adjust for sheer journalistic excess, the rest of the phenomenon is s largely explained by 1. When America finally faced the Wehrnacht in Europe, the Germans were fighting defensively, with all the natural advantages of the defender. 2. In Italy, they often had the advantage of rugged terrain.
3. As our forces approached the borders of Germany, the German soldier naturally feared that we would ravage their country with the same brutality that the Germans had used against the Poles and Soviets. This would have provided a strong incentive to fight hard.
I want to note that "when he said Jews, he meant Communist Sympathizer Jews" is a vile lie. When Nazis murdered Jews, they tried to murder all of them without concern whether they were communist sympathizers.
It is hilarious that you claim lies.
ad 5: you are lying here, by omission of crucial details: specifically, German victims were getting starvation diet far before Germans did.
ad 6: yes, Nazis were not sole evil in the history of the world. You get a cookie.
ad 7: by himself.
Thanks for feeding my confirmation bias that neonazis are bunch of lying idiots.
Selectively burning books you disagree with is the problem.
LOL, I thought most of the neonazis had migrated to greener pastures like The Motte. I guess this senseless intellectual virus is will always be with us wherever there is a modicum of free expression.
This seems false: Caligula and Nero, for example, aren't particularly relevant today, and our understanding of them seems just as full of shit.
I said "our ABILITY to understand" seems to be inversely proportionate, etc, etc. I didn't say our ACTUAL UNDERSTANDING is inversely proportionate.
It's possible I misunderstood what you meant by that. It sounded like you were suggesting all sources of information about them are written by people strongly predisposed to believing lies and nonsense as long as they're sufficiently negative, and lurid rumors about them are impossible to separate from actual facts – and I agree with this.
If you're instead arguing that even given some sort of magical infallible oracle with perfect information, everyone is still so blinded by hatred or whatever to not be able to process that faithfully without their mental processing generating falsehoods ab initio, … this seems unnecessarily uncharitable.
Yeah, I meant the first one. Maybe my "inversely proportionate etc" was the poor choice of words. It was more like analogizing and not proposing some universal, immutable law.
I'm thinking about the discourse on AI, specifically the whole UBI debate. If AI truly does take over a lot of human work, there's a lot of people who are savagely agitating for a UBI on one side, saying we'll be post work. The other side of course says no that's not how it works, besides we aren't even close to being able to afford that.
What I'm surprised by is why nobody has so far mentioned what, to me, seems the obvious compromise - we just shorten the work week! As our forefathers did forcing a 5 day, 8 hour work week, why don't we continue there? Go down to a 4 day work week, and/or shorten standard working hours to 6 per day?
If AI truly will obviate the need for a lot of work, how is this not the more rational solution than trying to magically create a UBI out of money we don't have? How come this idea has barely even entered the discourse? I have been talking and thinking about AI unemployment for years and never once have heard someone argue for this compromise.
There were plenty of good responses to this. I hope it's not too prosaic, but I'll add that humans fear change, and most of us try not to think about what we fear. People Who Matter are affected by this dynamic as much as anyone, so they likely aren't brainstorming their response to AI development as much as they ought to be.
> If AI truly will obviate the need for a lot of work, how is this not the more rational solution than trying to magically create a UBI out of money we don't have?
I'm working on a post now that shows that in any realistic "AI can counterfeit white collar jobs" scenario, we can offer pretty generous UBI's, think $40-$80k in real 2024 USD to everyone.
This is because the economic growth from much-cheaper white collar jobs makes a generous UBI realistic. The economy grows a lot because they become A) considerably more net valuable, because the cost has gone down a lot, and B) considerably more numerous due to the Jevon's Paradox effect, because when they're cheaper and driving more net value to each company that has them, there will be more created, which creates more absolute value in the economy.
But what about real estate and art and expensive private school educations?? Those will get way more expensive! Baumol effect!
Yeah, and? If you have a UBI, you don't need to live in Tier 1 metros for your job, you can live anywhere - live where it's cheap! You don't need to buy Tier 1 metro houses and art and positional goods, and most people who elect to take UBI won't.
But broadly, you don't need to magic up money from nowhere, there will be a ton of real economic growth if we can counterfeit white collar jobs for a fraction of their salary costs today.
> If you have a UBI, you don't need to live in Tier 1 metros for your job, you can live anywhere - live where it's cheap!
You may have other reasons to avoid the cheapest places. For example, to avoid crime, or to be closer to health services.
Because AI will not uniformly take 20% of work away from everybody's job (at least in the short term). More like it will destroy 20% of jobs. Junior office workers doing data entry roles, for example, will be completely obliterated; while plumbers will be largely unscathed.
I've written about this before, although I can't track down the article.
As AI takes over work, theoretically, we should be able to 'afford' UBI- to the degree we can't afford it already. I agree with you that shortening the work week seems logical. Another positive upside is that it helps keep us all useful. Terrible things have a way of happening to people who have no economic power in society, so keeping us all indispensable seems like a good outcome.
Obviously, in the final instance, there will be nothing left for humans to do. Here, UBI- but conceptualised more as a collective seizure of all assets followed by the distribution of a dividend- seems desirable.
> how is this not the more rational solution than trying to magically create a UBI out of money we don't have
Contingent on there being AGI, ASI, or any level of AI with capabilities to cause mass unemployement, it's impossible for there to not be enough money to support UBI.
If you strip out the framing of money, the economy will be producing at least the same and most likely significantly more goods and services. If the traditional distribution mechanism of this production via wages goes away, the government can simply step in and distribute it through UBI.
Money is just an accounting system in both cases.
> magically create a ubi out of money we don't have
I suspect this will be a political obstacle to implementing a UBI - don't max out the credit card, don't shake the magic money tree etc. Are these analogies fair reflections of how an economy based on credit works? That's debatable. It's not crazy to bet on a huge increase in GDP which will finance the UBI. But I expect there will be an awkward period where unemployment is high, GDP is rising but legacy attitudes to work make a UBI unpopular. I mean, I'm not sure I want a UBI in my heart of hearts but it may be the least bad option.
[Edit: if the state doesn't supply a UBI I suspect we will get some sort of epic philanthropy programme from the tech giants, so I don't think the crisis of meaning can be prevented by postponing UBI]
Partly because there's greatly diminishing returns to that. The "work week" includes a fixed amount of overhead, doing things like getting up to speed on what all of your colleagues are doing, completing mandatory training, and filling out your timecard. All of that can easily add up to 10+ hours a week, in which case the person who works 20 hours a week is doing only a third of the work that someone doing 40 hours a week does. And it's worse even than that because tripling the number of workers means the "getting up to speed on what all your colleagues are doing" part starts taking more time.
But more importantly, 20 hours a week doing what? The reason improved technology didn't cause mass unemployment in the past was for the most part not because we kept shortening the work week, it was mostly because putting the machines to work doing the most menial physical labor left ordinary people free to step up to more lucrative and valuable jobs - jobs that for the most part weren't being done at all in the era when almost everyone was working 72 hours a week just to stay alive, and which make us all richer now that they are being done.
The reason things might be different with AI. is that AI + robotics might be able to do *all* the materially productive jobs a median human might do, better and cheaper. There will at least at first be jobs for exceptional humans, but if every median human who answers any help-wanted ad always gets the response, "no, it was cheaper for us to buy robots to to that; we're only looking for a couple of expert robot repairmen and you didn't graduate from MIT", then it isn't a question of their working 10 or 15 or 20 hours a week, because no employer would rationally hire them to work any hours a week.
Conceivably we could, instead of raising taxes to pay for UBI, require that every business that's still in business take the money they would have spent doing taxes and use it to pay people to do stuff that robots could be doing faster, better, and cheaper. But that would just be a UBI with a meaningless-busywork requirement because it makes us feel virtuous that we made other people do "work". Scott's already written a post on why that's not a good idea.
The amount of overhead varies dramatically between countries, and a sane society would eliminate/automate most of it.
The part of the overhead that's "get up to speed with what your colleagues have been doing" is the part that's going to scale badly when you've got many part-timers working in series, and that's going to be hard to automate. A sufficiently capable AI could do it, but at that point you're basically having the AI tell people exactly what to do, slower than the AI could do it itself, because you think it's virtuous to make people do pointless busywork for their dinner.
If people spend 25% of their work time to coordinate with each other, it would make sense to hire someone as a full-time coordinator.
You may be doing a different kind of job, but in my experience, high overhead is usually a sign of incompetence. Inviting dozen people to a meeting where a short e-mail would suffice. Inviting hundreds of people, because someone important loves the sound of their voice. Repetitive meetings because people are unable to make a conclusion, or keep proper records.
Stupid meetings can certainly be a symptom of underlying incompetence, but getting up to speed with colleagues isn't just stupid meetings. Sometimes it's modifying your own piece of the project to remain compatible with everyone else, and if you have enough interface surface, that's a nearly full-time job. (That can be its own symptom of incompetence, but I digress.) And if you're not spending time in a meeting going over that, you might still be spending that time reading someone's progress report in an email instead.
I've been in team meetings where the team is about thirty people and everyone sees their purpose at that meeting as looking busy to the supervisor, resulting in everyone else listening to thirty amateur short stories about office work that not even the supervisor really needs, so emails are arguably a big improvement. (In such cases, breaking the team into subteams would have solved much more of the problem.) But the fact remains that there's *some* interface between anyone and their teammates, and they'll have to coordinate that, and it can take time.
Oof, 10+ hours per week on overhead is woefully inefficient. I really hope most people aren't doing that, but if they are, there are much better ways of doing things. That anyway sounds like something AI could actually sharply reduce and so perhaps that's where the first reductions in working hours could come from...
You'd be surprised. In my sector (horticulture), management = compliance to a large extent, the gardeners do the work and the boss fills out the risk assessments. So they say! I'm sure there's some minesweeper being played too. But yeh I reckon AI will improve a lot of that stuff.
[TBH a world where we continue to work with our hands and have an AI boss would work for me. Horticulture isn't complicated but there is a lot to remember in a single year, often with big gaps between the same task. Staff out the planning and compliance to AI, robots for the repetitive back-breaking or mind-numbing stuff, and we can focus on making rockeries, ponds and herbaceous borders really slick.]
I have a question for anyone who might know about Wellbutrin (buproprion), or other antidepressants which might tell me about wellbutrin). I understand you should wait for 4-6 to see if it works for you. But I can't tell if this applies *even if you are noticing clear side effects that show it is affecting your brain chemistry*.
In my case, it as dramatically curbed my desire for nicotine. Unfortunately I think it is also making me feel lethargic and slow and sad, to an even greater degree than the condition I was in, which made me go on it. My question is whether there is a reasonable chance it could start helping me if I were to stay on it.
(I've been on it for 3 weeks (at the target dose of 300mg, 6 weeks total if you include three weeks at the starting dose of 150mg). Any advice much appreciated. I have gotten pretty mixed messages from my provider and from online advice. Most of it seems to say 'stay on it' but to be geared at people who feel no effect.)
Sedation isn't a common effect of Wellbutrin; in fact, the association seems extremely rare. I would consider the possibility that it's something else, or a coincidence, causing you to feel sedated. This is not medical advice, and I am not a medical doctor.
I'm a psychologist, so do not prescribe but have pretty good general knowledge about these drugs. I don't think sedation is a common side effect of Wellbutrin. In fact I know some docs who start with it rather than stimulants as a treatment for people with ADD.
You went up to a dose of 300 mg much faster than I am used to seeing people go up. The last person I watched go on wellbutrin was on 150 mg for several months before increasing the dose. (And he had an excellent response to 150 mg. He's still trying to figure out whether 300 mg even gives him more of an effect.) I think increasing the dose so quickly was a mistake.For one thing, you went up to the higher dose before finding out whether the lower dose worked for you, which has deprived you of needed info: maybe you only need to take 150 mg of the stuff to get the benefit; maybe side effects at 150 are minimal. Also, if you'd stayed at 150 whatever side effects you had on that dose might have faded away to nothing -- it's not rare for early side effects to do that.
Curbing appetite for nicotine -- yeah, it works well for that. Helped me stop smoking years ago. If you have decreased or discontinued nicotine use, your fatigue and low mood may well be from nicotine withdrawal, which does commonly make people lethargic, foggy, with low mood. That is *exactly* how coming off nicotine made me feel. If you wanted to test that you could experiment with a couple nicotine lozenges to see if they counteract that stuff. Or you could just wait. If you're suffering from nicotine withdrawal the symptoms will fade.
>I understand you should wait for 4-6 to see if it works for you. But I can't tell if this applies *even if you are noticing clear side effects that show it is affecting your brain chemistry*.
Yes. Side effects kick in almost immediately. I'm pretty sure most antidepressant side effects are caused by the drug affecting your brain chemistry, just not yet having been in the brain long enough to bring about the change you are hoping for. I recommend that you drop your dose back to 150, which should make the side effects more tolerable, and wait a couple months before increasing the dose to 300. Side effects may have faded away by that point.
Wellbutrin has the very desirable feature of not causing sexual side effects. It's the only antidepressant I know of that that's true of. You should give it a really good try before moving on to the other stuff.
In other news, albino Indians look like white people, i.e. Europeans.
https://www.hindustantimes.com/static/groundglass/albino-delhi-single-out-family/
If someone had asked me where that family is from based on looks alone, I would probably have picked Norway.
That family should get cast as Khan and his supporters next time Star Trek gets around to that storyline.
Those people don't look European, they look uncanny, as in the uncanny valley.
Which is unsurprising, because they combine South Asian facial structure with Northern European skin and hair tones, which isn't something you find outside of albinos. They certainly don't look like Norwegians.
Some of them could pass as Europeans. Others, not.
Image for fans of dark humor only: This year's Macys Thanksgiving Day Parade
https://imgur.com/a/oMGn6Vz
Please watch "The Age of Disclosure" on prime, $20 to rent $25 to buy. At greater than 50% chance Aliens are real and we have the bodies. After watching please encourage your congress people to pass the UAP Disclosure Act. Defense contractors are lobbing heavily against it. (This seems way more important that AI.)
where is downvote button when you need it?
> At greater than 50% chance Aliens are real and we have the bodies
at 99.999% probability you are terrible at estimating probabilities or trolling
Where is the downvote button when you need it?
The UAP disclosure act is very simple, what could be the harm in passing it?
For what? So you can hide something this funny? Come on man, people like them being visible is one of the advantages of a forum like this. Echo chambers are boring.
My tin foil hat is firmly in place.
Y'know, if all of this was real, the government wouldn't hesitate to kill anyone who they even suspected was going to leak this information, let alone do TV interviews and make a whole movie about it. So why do you think these people are still fearless and alive? Don't you think it's suspicious that the government is making no effort to interfere with the advocacy for disclosure? When the disclosure happens and it turns out to just be boring national security stuff, that's just a way for them to get everyone to think the mystery is solved. This movie is just another part of the conspiracy.
Yeah, I'm pretty sure they have killed people. It's a complicated story. I urge you to watch. Yeah the guy who quit his nice job with the gov. to go tell what small pieces of the story that he is allowed to, the unclassified bits. Is an f-ing hero IMHO. Finally The UAP Disclosure Act, has been up for votes I think several times but at least once. Heavily lobbied against by defense contractors... who make plenty of campaign contributions. The story holds together in my mind, it's fascinating... I want to say more, but also feel you should just go and watch it. What do you have to lose? Like I said I'm >50% the aliens are here. Of course I also watch Joe Rogan, and TBH several of the people in the movie have also been on Rogan, and telling the same bits of the story you are given here.
Oh the hero is Luis Elizondo https://en.wikipedia.org/wiki/Luis_Elizondo
The wiki page is interesting. Scroll down to AATIP Advanced Areospace ... Program.
(Edit, Score to Scroll, my spelling sucks, so does my english.)
The best skeptical question I've heard about aliens visiting is why do these creatures flying quantum-fractal-black hole-doodah vehicles *crash* all the fucking time when flying around our planet? WTF? Are the aliens all alcoholics?
Yeah I agree! One of the guys talks about this. He says, accidents happen and then maybe it's on purpose. And I said no-way. But here's another fact from the documentary. There are at least two 'species'. (whatever the right word is.) I immediately thought, 'oh, maybe they are at war?' (I've read too much sci-fi, cue David Brin)
https://www.skeptic.com/article/the-aliens-are-here-again-a-review-of-the-age-of-disclosure/
https://www.theatlantic.com/science/archive/2018/03/dna-proves-alien-is-actually-a-girl-so-who-was-she/556625/
https://genome.cshlp.org/content/early/2018/03/21/gr.223693.117.full.pdf+html
Yes much footage from congressional hearings. When you get the whole story, (so far) much of that makes sense. There's a strong campaign by defense contractors (and whomever their surrogates are.) to discredit all UAP's and those who claim them as true. So I expect lots of push back and denial. OK don't watch the documentary. But what if the aliens are here? There is no Fermi paradox.
For me there's been so much BS pushed on me from all sorts of people (media, government, industry.) that I'm willing to listen. Maybe you trust these people more than me.
I don't know about the desert girl thing. The bodies we have are recent. Area 51 is a real crash site. I know I'm stretching your credibility. How about this you watch the documentary, if your opinion is unchanged by it I'll send you a check for the $25.
(This offer is only for Eremolalos)
<For me there's been so much BS pushed on me from all sorts of people (media, government, industry.) that I'm willing to listen.
Why doesn't the amount of BS being pushed make you *less* willing to listen?
Oh, It's not that I can't listen, but how do I separate the lies and truth? And it's not just lies, it's whatever the 'narrative is'. And is it worth it to me? I mostly end up getting mad about something stupid and have to turn it off. (the stupidity is mostly on my part)
Separating lies from truth is unfortunately quite difficult. Consider that someone in the establishment lying to you does not make alternative sources more likely to tell you the truth. It's not either or. I would recommend looking for sources that are in a position to know, and that are not actively pushing information on you. Consider what incentive someone could have to say what they are saying. The people talking the loudest, and with most certainty are seldom the ones to listen to. Accurate information will often be nuanced, and uncertain. For every question there is an answer that is simple, intuitive, and completely wrong.
OC ACXLW Meetup #108 — Saturday, November 29, 2025 — Early Science Acceleration (GPT‑5) & What Tech Bros Get Wrong
Location: 1970 Port Laurent Place, Newport Beach, CA 92660 Host: Michael Michalchik
Email: michaelmichalchik@gmail.com (For questions or requests)
Time: 2:00 p.m. – 5:00 p.m. PT
https://docs.google.com/document/d/1ofEhbYQF4p3n-PsKbEJNSonFZ0vN86hHgA43yVPDN2Y/edit?usp=sharing
I might use some advice on online sales, if anyone has any. I'm planning to sell pictures of my own design (I hope this is vague enough not to count as advertising! At any rate nothing is on sale right now) and I'm looking for a way to do so that requires the least micromanagement. I am NOT looking to make a quick buck, or planning to make it my career; I'd like something that quietly runs itself as much as possible while I focus on my actual job. It's fine for me if money comes slowly, as long as it's not literally zero or a net expense.
After some brief research, the most convenient way I found yet requires two services: setting up shop on a place such as Etsy for the actual sale, and linking it to a separate service to handle the physical printing and logistics (two call Printful and Printify come up most often in searches; I don't see any obvious superiority of one over the other, and tbh the near-identical names don't help). I might just do that. Is there any reason this is a terrible, terrible idea that I will regret?
I can help. Selling my designs on various products is my main source of income. I’ve used many services, including Etsy, Amazon Merch, Printful, Printly and over 20 others.
What I recommend most is Threadless. I'm not affiliated with them - I just like the platform and earn a lot there.
I’d stay away from Etsy. They charge fees even when you don’t sell anything, and running an Etsy store is basically a full-time job.
Thanks. They seem to focus mostly on printing on clothes and objects — are they good with posters, too? How are they on international shipping?
They have lots of posters/canvas types. I haven't ordered any so I don't know how's the quality, but I know people buy them because I get email notifications whenever someone orders something.
I order internationally (clothes) and shipping is fast but expensive.
If you only want to focus on posters then you might check out Displate - I used to have a store there but I didn't sell anything so I can't recommend it but maybe you will have success there. And there's also Inprint - I haven't tried them. For more artsy stuff there's also society6, but it used to have such an awful uploader that I gave up on them, they might have fixed it since then.
There would be people on the Reddit art forums or on Deviant Art who have done this, and could answer your questions and suggest services to use.
Is Amazon's Merch on Demand not what you're looking for? They probably take a larger cut, but I think it's exactly the "fire and forget" one-stop-shop you're describing.
It might be. It honestly hadn't occurred to me. Thanks, I'll check it out!
Our conception of Normans as Vikings was influenced by Plantaganet propaganda. The descendants of William the Conqueror found it politically convenient to minimize the Frenchness of their forebears. They wanted to emphasize their modest Norse heritage. After all, their English subjects ALSO had Norse admixture.
But compared to England, Normandy has yielded up few Norse burial sites. Scandinavian migration was seemingly limited to a few elites.
So being “Norman” in the 1000s wasn’t really about blood or ancestry. It was more like joining a rough, ambitious club. Normandy pulled in land-hungry fighters from all over France, and what made them “Norman” wasn’t where they were born, but how they lived: war, loyalty to powerful bosses, and grabbing new land.
This is an area of personal interest to me. You're both wrong and right in your characterisation. Where you're right is that the degree of French ancestry in the Normans is downplayed.
Where you're wrong is that the Normans weren't just significantly Viking in the aristocracy but they were also significantly English, owing to well documented English migration there in the preceding century. There is an extensive toponymic record of this, including Cherbourg. There was extensive pre-invasion contact between Normandy and England, hence the invasion in the first place.
This is far better explored in French academic literature than Anglophone academic literature, to the point that it's almost entirely unknown.
The English will happily tell you about all the times they were invaded, as long as it's by someone who no longer exists.
England was invaded by Romans, Angles, Saxons, Jutes, and Normans? Sure, it's all part of the rich history of these isles.
England was invaded by French or Germans? Absolutely no freaking way.
That said, Rolf the Ganger was definitely a Viking, so the history of Normandy as a bit of France that got conquered by Vikings who eventually conquered England is not entirely made up, even if there were a lot of Frenchmen involved too. Trace the royal ancestry of King Charles and you'll find Viking kings, not French ones.
Historian R. Allen Brown summed it up by saying Normans were “made, not born." Normandy was basically a medieval launchpad that toughened Frenchmen up and sent them out to seize places like England, southern Italy, Sicily, even parts of Greece and Africa. So the Normans were less a genetic “people” and more of a warrior fraternity. One characterized by violence, competition and greed.
Norman warriors undertook big heists when they could, like grabbing a kingdom/lordship. But they would take what they could get. If they could find a place with little or no law enforcement going on, young Normans would post up near roads and rob hapless travelers.
Normandy was a region mostly populated by ordinary peasants, like most of Europe at the time. But it true that it is remarkable that this region suddenly produced so many knights who terrorized so many parts of Europe and the Mediterranean. That they aggrergated adventurers from elsewhere is good explanation.
On the other hand, looking up the ancestry of the chief Norman lords in the Mediterranean at the time (the Hautevilles, the Quarrel-Drengots, the Moulins/Molisios, Roussel de Bailleul), they seemed to be actually from Normandy. Maybe the lower ranks were more diverse.
Of course, medieval lords shamelessly rewrote their family history for various reasons. Or rather had someone else do the rewriting for them. Richard the Lionheart was one of a number of lords who seriously insisted he was descended from a water fairly.
Anyone here have links to polymarket? Because I have a question for them: what the fuck were they thinking?
https://finance.yahoo.com/news/polymarket-milestone-overshadowed-ugly-scandal-082211793.html
The days of woke scolding are over. I sincerely doubt they are genuinely racist in a classic sense. They are just counter-signaling against the insane excesses of woke censorship.
I'm going to say something that may be controversial: vice signaling is bad
Correct, that's pretty controversial here. How else do you want people to signal that you're done with these people? Something is necessary to separate us from them. Heterodoxy is the easiest way to achieve that.
Vice or virtue, signaling is greatly overrated. At its best, signaling is totally acceptable. But we don't often see signaling at its best. Obnoxious, shallow, shit-stirring, self-indulgent, immature--this is just a sample of the unfortunate characteristics that typify the most adamant vice signalers among us.
I'm not going to lump you in with those goblins, certainly not based on one comment! But as much as wokeness bothers me, it would never occur to me that the world NEEDS to know I'm "done with" it. Can you see how that might come across as self-important, even pompous?
Donald Trump won the popular vote. The writing is on the wall about the shifting zeitgeist. Just how USEFUL is it for us to do ironic racism? To what extent is doing this nothing more than gratifying our inner children?
First, I do not think 'vice signaling is bad' is all that controversial here, sorry. I recommend reading more of Scott's posts.
Second, taken to its logical conclusion, you are part of a movement that is purpose built to do the worst thing along any measure, every time it has the opportunity to do anything at all. I can't think of any reason why this may be worthwhile to do, but even so party identification seems like a really bad reason. "I'm going to be racist so that people will know which party I support" feels a bit like telling on yourself
> First, I do not think 'vice signaling is bad' is all that controversial here, sorry. I recommend reading more of Scott's posts.
And I recommend reading more comments. Scott and his audience do not share the same beliefs. Due to various circumstances, this blog has a tendency to attract people who aren't fully on board with his centrist tendencies, but enjoy his work anyways.
> Second, taken to its logical conclusion, you are part of a movement that is purpose built to do the worst thing along any measure, every time it has the opportunity to do anything at all.
The disagreement here is on whether these things are particularly bad in the first place. They're just words. No explicit racial hatred was expressed. If you're offended by it, that's fine. That's just the signal doing its job.
Sorry, "vice signaling is the best way to identify the in group" is stupid and self destructive. Like I said, taken to its logical conclusion, your movement always has to do the worst possible thing. Why? Because you've self defined good things as "woke" and "liberal". This is why maga purity spirals exist and why your movement is doomed to be the launching pad for Nick Fuentes. It's self hating metastasized into a political party
They were thinking that being edgy and posting hot takes on social media gets them attention. They are right about that overall, even if the occasional tweet goes over the line.
If they are wrong, you can get rich betting against them.
wrong about what, exactly?
A sardonic take on Scott's defence of prediction markets.
Without checking, how many Jews do you think there are?
I've seen 15 million.
The reason I asked is that I saw a discussion of Joe Rogan saying (not well attributed) that he thought a billion but would be conservative and say 500 million.
I'd guess about 20 million. Plus or minus ten million depending on how you define the term.
18 million
50 million. Read that Jews are 1/2 to 1% of world population, based my estimate on that figure.
Worldwide? Probably around 15-20 million. Or a bit higher if you're counting people who are ethnically or culturally Jewish but don't adhere to Judaism even to the extent of self-describing as a non-practicing Jew in religious surveys.
Reasoning 1: I think I remember seeing a stat that there were about 12 or 15 million Jews worldwide as of the late 1990s. 25-30 years of population growth would put the current figure somewhat larger. Not enormously larger, since most of the world's Jewish population presumably lives in relatively low birthrate countries.
Reasoning 2: I remember with more confidence that a very large percentage of the worldwide Jewish population lives in Israel or the United States, with roughly equal numbers in each. I remember looking up religious statistics for the US recently and seeing something like 1.7% or 1.8% of the US population was Jewish, which would work out to about six million Jews in the US. I don't remember the population of Israel off the top of my head, but something like 8-10 million (of whom about 20% are Muslim, Christian, Druze, atheist, etc) sounds about right and would work out to 6-8 million Israeli Jews. Total 12-14 million in the US and Israel, plus probably a few million more living in the rest of the world. Off the top of my head, I think I remember Yemen having enough of a Jewish minority to show up in high-level country religious statistics, plus countries like Russia and the UK which have much smaller Jewish populations in percentage terms but enough to add nontrivially to the global total in absolute terms.
Also without checking, I would be very surprised if there were a significant number of Jews in Yemen to this day. If there are any, they must be having a rough time with the capital and half the country controlled by the Houtis, a faction which has "Death to Jews" on its flag.
I think Morocco is the only Arab country which has retained something of a Jewish community (of maybe a few tens of thousands: not enough to impact the global number). Among the other Muslim countries, also Turkey and Iran have small communities.
Morocco has about 2,500 Jews.
Turkey has 14,000.
Iran has 8,000.
That's sadly plausible. I mentioned Yemen because I remember many years ago looking at CIA World Factbook info about religious demographics on a bunch of countries. Yemen was the one country besides the US and Israel that I remember having a large enough Jewish population to get its own line item in the stats rather than being lumped into a miscellaneous line of minor religious minorities or altogether unmentioned. I think the number was around 4-5% at the time.
I don't remember if I thought to check Morocco.
15 million.
BILLIONS!!!
(Okay, seriously, I'll guess about 30-40 million)
(ETA: I have a probably boring way of making this guess that I think is at least slightly informed, in case anyone is interested.)
To the few people still paying attention, I finished the book and have decided an ideological deconstruction of the various varieties of feminism would have infinitesimal value in the memetic war. Too many, like existentialist, psychoanalytic, and eco-, no longer have any real relevance, and most of the current wavefront of ideological warfare is too fast changing to really be summarized effectively. Besides, most men respond to jokes and pictures, not lengthy essays.
Writing is for women now. That's just something I'm coming to accept. Sure you can swim upstream and a few people even make a career out of it, but I'm not that good and I'm not that young to get good.
What's that line from the end of Boiler Room? "I tried selling crack rock, and I never had a wicked jump shot. I gotta get a job." Well, I have one of those. I need a hobby.
This sounds like cope. Perhaps you should work on improving your writing instead of blaming demographics. Or recognize that the sub-varieties of feminism are for feminists to communicate amongst themselves, not necessarily to appeal to society at large.
Cope is part of life. If you're 5'7" you're never playing for the NBA, and if nobody wants to watch your demographic play basketball there won't be an NBA you can join.
As for the other thing...nah, I'll let it go.
I cringed so hard when you said "memetic war" and "ideological warfare," but fair is fair, this is a good point about cope.
I've been reading a lot of weird stuff online, I can go pre-Internet if you like.
How about:
"To the few people still paying attention, I finished the book [of feminist theory] and have decided an attempt to point out and argue against the weak spots of the various varieties of feminism in a detailed critique would have almost no value in swaying people to the other side. Too many, like existentialist, psychoanalytic, and eco-, no longer have any real relevance, and most of what people are arguing about online is too fast changing to really be summarized effectively. Besides, most men respond to jokes and pictures, not lengthy essays."
Writing skill is absolutely something to be learned and improved upon, unlike height. Why not give it a go?
Also there are basketball leagues for normal people lol.
So your reading project was actually a writing project. Huh. idk why this wasn't more obvious to me, based on your question about hobbies a few months back.
I dunno. It was a reading project, the reading went on but the writing was never produced, so it's more accurate to say the writing project was actually a reading project.
> Writing is for women now. That's just something I'm coming to accept.
Short form video is obviously the future - who has time for *books,* which have been systematically declining in popularity for decades??
No, we need more Dramabox-style microdramas - those things have grown from literally not existing a few years ago to a ~$1B market already.
To stay on topic, they could graphically demonstrate the perils of feminism, as high status boss-babe style women repeatedly make crushingly bad personal relationship decisions that lead to career-and-life upending drama. <whispers> What's that? This is already 2/3 of microdrama plot lines? And it's WOMEN who watch all these things??
Huh.
Okay, I got nothing then.
I do agree trying to focus on writing is probably a mistake - the guys whom writing resonates with are generally doing fine in the gender / dating wars, in aggregate. And what advice do you even give median men?? "Sorry the world sucks now, women would rather work and watch Netflix than date you, so, uh...try to get better than Netflix?"
That said, on the "searching for a hobby" front, any low hundreds purchase that may turn into a hobby is worth it even if the hit rate is low. Sports equipment, a musical instrument, art supplies, etc. If it doesn’t catch on, gift it to a friend.
I've read more books in the past 5 years then I have in any 5-year period since I was a teenager. It's not for everyone, but I suggest it for anyone who can hack it. I enjoyed watching a movie almost every day, which was my previous relaxation technique. But in the long term, my renewed reading habit is more gratifying and better for my self-esteem.
> And what advice do you even give median men?? "Sorry the world sucks now, women would rather work and watch Netflix than date you, so, uh...try to get better than Netflix?"
Bring back patriarchy? The circumstances have never been better for restoring it. It's not like women can do much about it when push comes to shove.
Maybe it would work better to bring back black slavery, esp. since 50% of the re-statused would be fuckable by straights. Whatcha think, Mistilteinn?
Oh, wait, you may be worried about comingling your genes for smarts, social skills, and civilized common sense with black genes for dumbness and violence. How about we enslave the Jews -- high IQ's, same 50% fuckables benefit.
I'm with PB, it really doesn't work anymore. The most they've been able to do is outlaw abortion, something I never supported on base libertarian grounds.
I do find the death of chivalry among young men hopeful, it could serve as the baseline for the first effective men's movement. Unfortunately I can't really participate due to ancestry issues.
You're from a completely different online subculture, and I find your writing style a bit grating. But I agree that chivalry must die; is this really happening? At LAST?
The tyrannical CRUELTY of implementing feminism while insisting that males remain chivalrous! It boggles my mind that people can see this as anything other than misandry.
My Australian friend was shocked when I informed him how rigidly American society insists that "the man should always pay for dinner" Down with the patriarchy, except for those bits of patriarchy that put money in my pocket.
What do you mean by ancestry issues? Do you mean you're not white enough to be taken seriously in a public facing role by a right-leaning audience, or what?
Probably the best summary of the situation is here, I think:
https://www.richardhanania.com/p/womens-tears-win-in-the-marketplace
Hanania's moved away from it towards feminism, at least officially (he doesn't like the Trump admin and, probably more importantly, has two daughters), but this article is still pretty good. You never really know what Hanania's thinking, but we're not likely to see more of this from him right now.
As for my language... I modified it a little in my other response to you, I've been poking around the esoteric right-wing too long.
Basically I figured most antifeminists see 'feminism' as an undifferentiated mass of opposition, but there's actually huge amounts of disagreement within it on various issues. Inject that knowledge into the right and they may be, like, 10% more effective in fighting them. I got an introductory book on feminist theory and figured, well, maybe I can do what Richard Hanania did for (or rather against) the woke legal regime or Helen Pluckrose did against critical theory. Of course I'd have to read the major texts--I was thinking de Beauvoir and Judith Butler next, with maybe a detour for Foucault (I even had a 'How to Read Foucault' book from a left-wing theory press).
The problem was that most of the current arguments were crammed into two little chapters near the end. You might be able to see their ultimate origins in the stuff I was reading about, but it's like doing a close reading of Marcuse to understand what Jessica Valenti's talking about--it's vaguely relevant but there are much more germane ways of going about it.
OK, read something about newer stuff? It's not going to be in a book, culture changes too fast now! I'd have to join Bluesky and lurk on Discords and even then might spend years on some formerly-active location that's really influenced by something else.
And the problem with *that* is...I'm old. I'm Xennial, I was trained to read long texts and write summaries and responses to them in high school and some of college, but that's not how you convince people now. A clever Pepe meme or a well-placed 'cool story bro' will sway hundreds of times as many men as a Substack essay. In the marketplace of ideas, but if you're going for young men the way to win them over is to master a whole bunch of social games I never learned in high school and would be even harder as a middle-aged person who's (a) less cognitively flexible and (b) not up on current cultural lingo, which changes faster and faster with each passing generation.
As you can see from the time stamps on these responses, I can produce large amounts of college-level text rapidly, but you know who's even better at that? ChatGPT and Claude.
Finally, unfortunately, my mom is of Ashkenazi Jewish ancestry. That's not a problem with older conservatives but definitely is with the newer Groyper bunch. Of course you've got the Christians (convert and you're saved), but they're still all-in on chivalry. Antifeminism always tracks with antisemitism for reasons I'm not entirely clear on (I think it has to do with who joins various coalitions) but in general the closest people to the MRAs are always going to be ranting about Jews. Not entirely inaccurately--look at who staffed the second wave feminist movement--but it makes me useless in this struggle.
Keep in mind, I don't really care about Israel (I mean, I'm rooting for them but don't think the USA should spend blood or treasure on their behalf) and I don't have anything AT ALL against the white man (I thought I *was* one until, like, the age of 40, which given that that constitutes most of my adult life tells you where my real loyalties lie). But the taint is in the blood with these people. And if there's one thing you learn as a guy who read The Myth of Male Power in the 90s and Paul Elam and Roissy/Heartiste in the 2000s, it's that life isn't fair.
I suspect groyperism is just another fad or passing phase. We're just one or two lost elections away from groyperism getting discredited and filed away with the many inactive influences of the right.
I also think we overestimate how much the RL right looks like the online right, and Fuentes, Tucker and Vance are not immune to this. But who knows, I can't predict the future. I DO know that these lists of "Jews in the media" speak to some pretty lowbrow guys with limited critical thinking skills.
While Jews do look after each other in a limited or superficial sense, they are so riven with sociopolitical conflicts that it's hard for me to see them as some monolithic block dominating "the media." Or any other industry. Anti-semitism is fundamentally shallow and I don't expect it to thrive in the long term. Though of course, no religion should be immune to criticism.
> Bring back patriarchy? The circumstances have never been better for restoring it. It's not like women can do much about it when push comes to shove.
Beyond the moral dimensions of this problem, regarding which I'm staunchly on the women's side, just from a pure practicality standpoint can you point to a single existence proof anywhere?
I think this is a one-way gate, and once you give women the vote and let them have jobs and bank accounts, you literally can't undo that.
Is there any developed country in the world that HAS taken away women's votes or ability to get jobs and bank accounts after they already had those things?
Heck, let's expand it to ALL countries - Iran is probably your best bet there, in which women had more societal freedoms that they then rolled back, but their changes were mostly about needing male permission / escorts when traveling and public dress codes, and women STILL have the rights to vote, work, and have bank accounts there!
I also challenge the idea that more patriarchal dynamics would end with more women dating median men. South Korea, for example, is much more patriarchal than America or Europe (as is all of East and SE Asia), and they have such crushingly low marriage rates and fertility there's a cottage industry in videos outlining their inevitable extinction - this is an excellent one here: https://www.youtube.com/watch?v=Ufmu1WD2TSk
They've turned the "patriarchy" and "gender wars" dials up even higher than us, and it went the opposite direction - even *fewer* women would date men, even fewer marriages, still fewer kids.
Barring literal slavery, I don't see how you're solving this problem, because it's a matter of personal choice and values, and you can't force specific interpersonal relationships to happen, either with culture or law.
What there will be, is AI / catgirl girlfriends well before anything like that will happen, and that WILL solve the problem.
New Jersey gave the vote to women in 1776, only to take it away in 1807.
I'm certain that explains Pork Roll, Springsteen, and why everyone else's pizza is kinda lame.
Wait what? I thought we were talking about getting women out of the publishing industry. Anyways, the status quo for most of human history involved women having little agency of their own, and it's been exclusively the last century that's been a cultural anomaly, largely due to labor demands and influence by the liberal west, especially America. But Pax Americana is dead, liberalism is dying, and even demand for female labor is going down due to automation and outsourcing. And with there being so many disgruntled young men who now have an ambitious leader to make their vision a reality... it really is the perfect opportunity to make something happen.
No "slavery" is necessary. The solution is simple: make it impossible for a woman to survive on her own. If she can't get a job with a living wage, choosing to stay single becomes a death sentence. Suddenly, women aren't so picky about who they end up with. All you need are simple incentives like this.
One of those rare occasions where I kind of agree with you on a base level, and yet I can't help but cringe!! Because you have to groyperize your argument by saying things like "death sentence."
Anti-feminism has merit, but that performative edginess makes us look like a bunch of unserious clowns.
They are gay billionaires in Silicon Valley, too. We need to make survival without selling sexual services impossible for everyone! Anyone who disagrees will get shot by a drone.
Not everyone, just the people who have nothing else to offer. Didn't the Romans have people whose job was to be a dependent parasite who humiliated themselves for the sake of a sadist?
Remember, they've got AI boyfriends too. Just attach ChatGPT to a voice synthesizer and vibrator and you've got a pretty workable robo-boyfriend. People had a long history of anthropomorphizing stuff like the weather and fire and it's not surprising something carefully engineered to produce human speech would have the same effect.
But overall yes, I agree, there will be robowaifus and robohusbandos. I admit I never got the catgirl thing. Catwoman scared the shit out of me (though given my age this was in the Tim Burton movie so...)
> Remember, they've got AI boyfriends too.
Oh, totally - I'm actually predicting we're gonna lose like 80% of the populace between better-in-every-way AI boy/girlfriends and Infinite Jest style AI heavens that cue on your physiological responses to make superstimuli-powerful feedback loops in virtual Tik Tok / porn equivalents.
Both of which we can roll out nascent starter versions of essentially today.
A question for musicians and people who work with audio. I play electric bass and a harmonica and I want to be able to jam outside, without an electrical socket. What's the cheapest and the most light weight device, or multiple devices, to allow me to do this? So, basically, I need a speaker, an amplifier and a battery. Or something like that. All in one device or as multiple devices. It should be able to produce low bass frequencies as well as high frequencies.
I'm a bass player as well and this is tough one to solve. As other commenters have said how much power you need really is the big question.
I solved it with this:
https://reverb.com/p/peavey-solo-10-15-watt-portable-sound-system?hfid=93223128
For years I'd bring this to the river and plug an acoustic bass with a pickup into it and it was sufficient to hear over the river and to play with an unamplified acoustic guitar.
A laptop and some VSTs are your cheapest bet.
The Roland Blues Cube is used by a lot of guitarists for busking.
If you don’t find something cheaper, the Bose S1 is a great product, albeit not one optimized for low bass. But I take “jam outside” to mean a situation where loud volume and perfect fidelity are not the goal, but rather ease and portability.
How loud do you need it to be? If you only need to hear yourself in headphones this is easy. If you need other people to hear you you're in trouble. By the time you've got an amp that can be heard outdoors, and a powerstation that can run it for a decent interval, the sizes of both objects have generally left the realm of realistic portability. Bass amps are especially power hogs. If you don't mind hauling a generator around on the back of a truck it could work. True travel amps are tiny and weak.
If you just want to be able to hear yourself on the go though, thats pretty easy and Amazon has multiple cheap offerings on sale right now, like this one: https://www.amazon.com/Acoustic-Electric-Amplifier-Portable-Bluetooth/dp/B09LS4YLW9. Most plug in too if you are near an outlet.
Harmonica amps are generally just vocal amps with a special mic for the harmonica. This is more realistic for a travel amp as there are many portable voice amps/PAs on the market at all kinds of different prices.
For that matter I can't imagine an acoustic situation where you'd even need to mic a harmonica. It's a loud instrument that cuts pretty well even in cross-harp. I'm a little more optimistic about portable amplifiers - they'd at least help a good deal.
So I keep running across stats about what fraction of online content is AI-generated, and the numbers are high -- 50% of this, 70% of that. And just now, asking Google for stats, I was served ads for several "tutorials" on using AI to create viral videos: "AI Video Cut tool creates viral shorts instantly from long-form videos with auto-generated titles, captions, and descriptions."
My first spontaneous association to these factoids is to leukemia, where abnormal immature cells crowd out the healthy ones. But then, I have a pretty dark imagination. What are people's thoughts about how this is going to play out? If we don't already have a bot or 2 posting here, we will soon, right? What will it be like if it starts seeming to many of us that we have have lots of them? Seems like few or none of will be provably bots -- so then what? And what about the internet as a whole? Will it become Botland, with few live users? Mightn't huge group phenomena develop spontaneously in the internet bot population -- the equivalent of wars (instrumental convergence, right?), fads, matings, dynasties, illnesses? How will that affect human life? What will happen to people who maintain a high level of participation in Botland, either naively thinking they are dealing with people, or consciously electing to live in that online world? Will there be alternatives channels of communication that are protected from bots? What kind of protection is possible?
First reaction: solve for the equilibrium.
With trust hitting all-time lows, gatekeeping will become more important than ever. Much like Performative Bafflement says, the internet will become more tiered. "Tiered" in the sense that the default platforms will become ghettoized favelas, whereas higher-quality content will be locked behind more paywalls and real creators will face more verification hurdles (e.g. blue checkmarks, whitelists, provenance, etc.) (I feel like people will get awfully creative with this, in ways that are hard to foresee). But even then, the higher-quality content will be under constant surveillance for signs of deception, and users will be paranoid. It'll be like The Second Variety, or Amongus, with accusations flying every which way.
Performative Bafflement mentions Discord. Meanwhile, Venkatesh Rao [0] (the same guy who wrote about the Manufactured Normalcy Field) and others [1] have already started theorizing about the rise of what they call "the Cozy Web". They use the "Dark Forest" metaphor as a muse.
[0] https://contraptions.venkateshrao.com/p/the-extended-internet-universe
[1] https://summerlightning.substack.com/p/cozyweb-animals
I read your links. Wow, that's very interesting stuff. OK, so some of us will end up in online "gated communities." What are your thoughts about how things will play out for people who don't have the skills, knowledge and money for those places? What's on my mind right now mostly is what the toxic part of internet, the ghettoized favelas, will be like, and how that will affect its users, and how the effect on them will affect life in general -- mental health, movements, fads, social norms, etc. Also, I have been wondering about interactions among bots and other AI entities themselves in an internet setting where they are most of the users. Could there be group phenomena -- fads, wars, plagues, coalitions, power struggles? Could there be emergent properties, entities comprising multible bots that have qualities and goals of their own?
Eh, I dont think the favelas will be qualitatively different from what we have now. Just more extreme. E.g. instead of traditional porn, we'll have AI girlfriends. People will invent new slurs for bots. AI assistance will be coded as low status. Dumb beefs about AI handicaps. Like, even before AI, and before the Almighty Algo, the internet had a Wild West period where nobody knew you were actually a dog with a blog.
I feel like the gated communities are where the more novel phenomena will emerge.
Though I could certainly be wrong. This isn't a hill I'm willing to die on.
>My first spontaneous association to these factoids is to leukemia, where abnormal immature cells crowd out the healthy ones.
Yeah, I view as a problem to the extent that the AI-generated content is worse that the human-generated content.
I'm not intrinsically hostile to AI-generated content. As things stand today, I regard it as the responsibility of the person who prompted the AI. If the prompt was a reasonable one (not intentionally malicious, for various flavors of malice), and the AI was reasonably competent, then I consider it unobjectionable. For instance, within the past week or two I watched a YouTube video on neutron star structure which, at the end, I guessed to be AI-generated. ( I'm guessing this because the wording of the narrative was repetitious in a way I doubt a human would word it. ) But it wasn't, as far as I could tell, wrong or misleading.
I don't know how much of the information in it was specified by the human in the prompt. I'd take the ratio of that to the final text as defining how much of more-or-less the responsibility for the video was human vs. AI. As the AIs get better I expect benign uses like this to transition from clunky-but-fairly-ok to good to better-than-human.
Unfortunately the malign content, mostly versions of computer-aided-deception will also get more convincing...
I would hate to watch unlabeled AI content, because over decades I have developed certain heuristics for evaluating people, and those fail horribly with AIs.
Humans have mental *models* of things. AIs complete *sentences*. The result is superficially similar, but the difference is that humans are more consistent, predictable.
After interacting with a certain human for a while, you can get realistic expectations about where the human will be smart and where stupid; things like "had really deep knowledge of Python development, but believes insane conspiracy theories". You can use this as a heuristic for how much to trust that specific human. You can find humans that you can trust, at least for certain topics.
AI is all over the place. It has vast encyclopedic knowledge, but sometimes it just hallucinates things. It can generate mathematical articles that seem like written by Einstein, but somewhere in the middle contain a stupid error that would make a high-school student ashamed. You are dealing with something that is million times more knowledgeable and faster than you... but you need to exercise constant vigilance, because you can be betrayed at any moment.
> I watched a YouTube video on neutron star structure [...] it wasn't, as far as I could tell, wrong or misleading.
The scary thing, from my perspective, is that even if you verified 99% of the video and it turned out to be perfectly correct, that still leaves a realistic option that the remaining 1% is just completely wrong in an absurdly stupid way. And put together, the conclusion can be wrong but sound credible.
I tried to use AIs for learning math, and I have up because they constantly stabbed be in back when I started to trust them. Changed the definition of something in the middle of a proof, things like this. And instead of admitting a mistake, they double down; instead of using their intelligence to notice the mistakes, they use it to generate plausibly sounding excuses.
With programming, if the code does not work, you throw it away. But with learning, you often don't get the immediate feedback that the lesson you learned is broken.
>I would hate to watch unlabeled AI content, because over decades I have developed certain heuristics for evaluating people, and those fail horribly with AIs.
Many Thanks! Yeah, I would prefer to see the AI content labelled too.
>The scary thing, from my perspective, is that even if you verified 99% of the video and it turned out to be perfectly correct, that still leaves a realistic option that the remaining 1% is just completely wrong in an absurdly stupid way.
Agreed. Some of this looks like sign errors, and I have to admit that I'm unsure about the comparison to humans. I have also seen humans generate what amount to sign errors in the middle of e.g. a comment - "understate" swapped for "overstate" and similar not-exactly-typos-but-kind-of-in-an-adjacent-class.
E.g. I had a discussion about triple points being reference points for temperature scales, and one of the LLMs said that melting point were used also - where the vapor pressure at the point would have been too _HIGH_ according to the LLM - oops...
>over decades I have developed certain heuristics for evaluating people
I would be interested to know more about your heuristics (as my heuristics have you down as someone who generally has a lot of insight).
>It can generate mathematical articles that seem like written by Einstein, but somewhere in the middle contain a stupid error that would make a high-school student ashamed.
Well, I think even Einstein is known to have made one or two stupid mistakes as well. Mainly a joke - I generally agree with you, and this is my experience with AI as well.
My heuristic for AI is that it mostly can be trusted for anything that is reasonably broad and uncontroversial. When asking for more special knowledge on anything more complicated I have this uneasy feeling that this looks very plausible, but there might be a subtle mistake somewhere, and it might be important.
Thank you! I will probably just say the obvious, but here it is:
Basic: You can use someone's grammar and following of social norms (including internet norms, such as not writing in ALL CAPS) as a proxy for their intelligence and mental health, which is a proxy for how reliable are their arguments.
Advanced: With repeated experience, you can notice how often the person is wrong, what kinds of logical fallacies they often make, if they can change their mind, if there are some topics they are obsessed about, etc. Certain opinions or keywords are signals of belonging to certain political or other tribes. People who mention details or nuance are usually more knowledgeable than people who speak in broad generalities.
For example, if you receive an e-mail that contains too many capital letters, entire paragraphs in bold, uses various font colors and sizes, you can probably delete it without reading, because the sender is most likely insane. If someone's argument on a political or economical topic can be reduced to "my opponents are *tards", it is not worth reading. If someone posted the same thing several times, and had previously never meaningfully responded to replies, do not expect a meaningful response today.
Now of course these heuristics have many weaknesses. For example, someone's grammar sucks because they are dyslectic, or they speak weirdly, because they are not a native speaker. Teenagers are likely to lack experience and express extreme opinions, but that's natural for their age, even if the same thing was a serious red flag for an adult. Accusing someone of belonging to tribe obviously leads to a counter-accusation. Crackpots can provide unlimited number of details. Sometimes the expert is just too busy to provide a high-quality explanation.
And yet, it somehow works. Maybe with a help of some other criteria that I am not consciously aware of.
AI can add good grammar and polite arguments and expert keywords to anything. AI can claim that the Moon is made out of cheese, and add a two-page detailed explanation about how the cheese is made, which will sound generally correct to an expert on dairy. (A human who believes that the Moon is made out of cheese typically knows very little about either.)
I agree with you, except I don't think it can be trusted for even things that are broad and uncontroversial, except for a very "narrow definition of broad", so to speak.
Check this interaction with Claude Sonnet 4.5 I had yesterday, specifically the part about grammatical cases in French:
https://claude.ai/share/19eeeeed-427f-499f-b533-80c72cf0f186
It said one thing that seemed off to me, then went "you're totally right to notice this mistake, I was imprecise" and just said the opposite.
I'll let you judge how "broad and uncontroversial" the topic here was, but I keep having interactions like this, and they often involve mistakes in contexts where I am actually trying to save time at work by automating stuff, instead of just satsifying my curiosity. Instead, I need to be vigilant to make sure there is no stupid mistake that would make the whole edifice crumble.
Yeah, I find it's mostly pointless to ask LLM's to fix their mistake. The exception is code, but then you can easily check if the code works and ask for spesific modifications.
I find it useful for doing research into spesific topics, even very technical ones, especially to get an overview over the topic or a broad explanation, when I can check the answers with other sources later. It seems useless in helping me write anything technical, as I would have to check everything - and then I might as well just write it myself from scratch.
> What are people's thoughts about how this is going to play out?
Ah, dead internet theory! If the infocalypse happens I think there's two main ways it could go:
1. The free open internet is gone, everything is paywalled and real-ID'd because you can control bots that way. People like the ACX commentariat will find paid enclaves to exchange ideas, but it will always be a smaller and less diverse population.
2. AI personal assistants, or people, or some combination of both, curate your feeds and content according to your tastes and rules, and filter out the bot content (or don't, it's your rules). In this case, ACX still exists, 90%+ of the feed is bots, but you only see your whitelisted set of people, perhaps augmented by up and comers not on your list that your content guardians think you might like.
3. Some combination of one and two
I can imagine solutions that are between "open for everyone" and "show me your ID".
One thing I would like to see is a model where people vouch for each other. The only way you can join a group is if someone who is already a member (the first members are the founders of the group) becomes your sponsor. They could testify that they know you, and that you are a human; they could even testify more things, for example that you are a smart and polite human. The idea is that if you do something bad, it will also have consequences for your sponsor. (At the very least, they lose the ability to sponsor other new members.) Some more technical details would need to be solved, for example you should be able to find more than one sponsor, just in case that one of your sponsors leaves the group, or gets kicked out.
Isn't this basically reinventing private, invite-only Discords? Which are fun, yeah, but the diversity / numerosity of participants is usually much more limited than the SSC / ACX commentariat, or the glory days of the ssc subreddit, etc.
You could have read access for everyone, and write access for the members. Maybe even some official way to apply for membership, it's just that the policy can no longer be "automatically admit everyone, ban them later if they misbehave", because that will result in many bots, especially as they get better and cheaper.
Another approach could be some kind of multi-tiered membership, where the elite members are verified humans, and the elite members can (individually or collectively) ban non-elite members. The best non-elite members (as determined by the elite members) could be later promoted to elite.
Unfortunately, that sort of thing seems to be too easy to fake.
I think it's going to get seriously bad. It feels like soon, it will be impossible to know what news, music, visual entertainment, advertising, etc. is by humans (if any) and I may just have to disengage from the internet almost totally. Already, like 90% of my youtube ads are AI, and I've had multiple family members share AI content (music, sermons) thinking it's genuine. And I've been off facebook for awhile, but my wife says she sees so much AI slop on there that it's just crazy.
To me, it feels like the world is going to have to revert to something like it was pre-internet (I wasn't alive then so I'm just kind of making up what that looks like for myself). I imagine that it will get to the point where it's impossible to know if anything outside of local news is true. It feels like the internet could become just a massive AI slopfest/hallucination that keeps feeding itself.
But then I guess maybe there's some other technology that might be made to fix everything. I don't really know about any of this; it's just the vibe I get when thinking about it.
My optimistic take is, the "free" internet becomes so bad that people revert to very early stage internet culture (mostly doing things in person, sometimes reading blogs or using resources that people you know have specifically recommended, using semiprivate curated discord-type chatrooms). In my opinion, this would be better than how things have been the past few years.
I think for young people who can actually understand that the free internet is dead, it'll be fine. The problem is going to be older people, who can't adapt and get zombified (Similar to how Boomers got slaughtered by cable TV)
>In my opinion, this would be better than how things have been the past few years.
Actually, I really agree with this. It's just a shame about the older people, so we should try to help out the ones we know.
You all think it's the people who grew up *without* the internet, who are going to be "zombified" by AI slop on the internet? You think we're not going to be able to figure out curated blog lists, bulletin boards, and forums? Who do you think invented all those things in the first place?
It's the younglings who've never known a life without a screen providing them constant on-demand content with no effort or curation required, who are going to have a problem here.
Also, get off my lawn.
> Who do you think invented all those things in the first place?
A tiny, tiny group of nerds, while 99.9% of their age cohort was asleep.
> Trust no one you cannot verify is actually human
In our post-truth era of state-run propaganda farms, being human is not sufficient for trust.
Always verify.
That states a necessary, not a sufficient, condition for trust.
> the SAME information, in the SAME words,
I don't know that this is all that indicative of bots. This sounds like people just independently using the same source, the government or the AP or whatever.
Ah, okay. Those are definitely bots.
What's the bot-hunting website called?
I've been having ChatGPT write game design documents for me, and one of its ideas for a final boss was something called an Aegis Engine: an autonomous war machine from the ancient world the size of a mountain equipped with magitech artillery and force fields called up into a land where longbows are now considered high-tech. It's controlled by a "chorus" of bound souls that essentially function as an AI.
ChatGPT seems, well, *enamored* with the concept. It's brought it up multiple times even after I steered the conversation to different aspects of the game, and even once in another chat about an unrelated game! I am officially Concerned.
That sounds very much like a Trails game final boss. I think all the ones I played ended with some kind of Moria Mech.
I will take this to mean ChatGPT thinks the Trails series is the best RPG series.
Now imagine Aragorn has to fight one of those.
Cue montage of gathering the whole world, heroes and villains, to join hands and sing Kumbaya until the bound souls are exorcised. Gandalf, the Balrog, Saruman, the Ring Wraiths, the Mountain Caradhras, Smaug; Tom Bombadil, the Ents, The Ungoliant, all the bodies in the Dead Marsh, all the horses of Rohan. Big old anime friendfest.
Just curious, what is your motivation for designing games?
If you're using AI as a source for creativity, aren't you just fabricating something in order to add to an already oversaturated entertainment market? It would seem that you'd have a more meaningful product if you tried to come up with something completely on your own or collaborated with someone else for ideas, instead of generating ideas from AI.
I'm all for using AI as a tool for all kinds of things, but when it comes to creative expression, it almost seems immoral to use AI. There's already so much unnecessary stuff in the world and, for me at least, it makes even existing sometimes stressful. Generating more stuff that doesn't even come from a person's imagination seems harmful.
Then again, I'm just assuming your motivation and goals, so my criticism could be completely irrelevant.
My motivation is personal entertainment. I have no intention of selling anything and I strongly urge you to consume less Internet.
That's good advice 😮💨
Yes, it seems OP.
In ChatGPT's own words:
"Key vibe:
The world has shrunk; the Aegis hasn’t. It’s not OP because designers like big numbers; it’s OP because the setting’s scale has collapsed."
https://www.youtube.com/watch?v=rj9pBsvcke4
Duolingo self-destructing by AI.
I know people say it's only going to get better, but I'm concerned that it isn't good enough now for many of the things it's used for.
Hmm. Pingo uses AI and I've heard from reliable seeming sources that it is much better than Duolingo.
Duolingo went downhill since about 2021. Switching to the "path" thingy, killing the discussion forums, no more speaking exercises or typing exercises (only choosing from the presuggested "word banks" is too simple, you don't learn anything from that). The AI nonsense is just another nail in the coffin.
I have learned a lot of Dutch, Italian and Yiddish from Duolingo, but now it's just no fun any more.
Heartily seconded. I used Duolingo for nearly 2 years in 2020-2021, and as soon as they changed the format to the "path," I uninstalled it and never looked back. I don't know why on Earth they would have thought that removing almost all user choice would make people engage more: it was such a bizarre change.
What is your Donald Trump number? We'll calculate this like Erdos numbers; the Donald's number is zero, anyone who has met him in person has number one, anyone who has met in person someone with number one has number two, and so on.
My number is two, if we let count a very brief meeting with Bill Gates when I was a Microsoft intern. It wasn't one on one; I was one of several dozen geeks standing around him as he held court at a party. That's me -> Bill Gates -> Trump.
If that doesn't count, my number is four. me -> my father (a scientist at Microsoft Research) -> Rick Rashid (head of Microsoft Research) -> Bill Gates -> Trump.
4 maybe less through a path I don't know about.
me - Philosopher Friend - Nick Bostrom - Elon Musk - Trump
I'd guess that Musk and Bostrom have meet, otherwise it might be 5.
5, maybe 4. Macron met him (1), a French politician met Macron (2), a bunch of kids pranked said politician (3), and either I met one of them (4), or a close friend of mine met one of them (5).
two, multiple possible intermediate steps. Same for Biden and Obama.
Likely four or five, through a slightly odd route. I've met a few people in the Japanese music industry, one of Trump's children apparently got drumming lessons from Yoshiki of X Japan.
(Alternatively, my sister works in government in my home country, so probably 4-5 through her as well).
3 at least - Cardinal Nichols, Pope Leo, Donald Trump.
I've been to a funeral and Tony Blair was there, but I didn't meet meet him, that could be a 2 if you're being generous.
[Another 3 - Tony Francis was a soccer presenter and I got his autograph, he often did shows with ex-footballer Jimmy Greaves, who had his own show called Saint and Greavsie, which included the Donald among it's guests (they visited him in Trump Tower)]
Oh, maybe I am at 3 through Cardinal Avelline then.
Probably three: I have met some politicians, who probably met our current president, who probably met Trump.
Presumably three; a relative is influential enough to regularly speak with their Congress Rep. The question is if the Congress Rep is in turn important enough to speak with Trump directly.
...I'm gonna go with four.
Bizarrely, probably 3: me -> my brother -> Mamdani -> Trump
Two for me. I've met RFK and the less well-known Alex Azar, both of whom met Trump as cabinet members.
Two. I met Nancy Pelosi briefly when she visited the Google campus c. 2010 and had a meeting with some executives near my office.
I also independently have a Trump number of three through multiple paths:
1a. About a decade ago, I bought some made-to-measure suits from Thomas Mahon and met him twice for measurements and fittings. The made-to-measure service was a sideline Mahon was experimenting with to his main business of fully bespoke suits, for which King Charles III has been a long-standing and frequent client. King Charles has met Trump on at least one state visit.
1b. When my uncle was in the Army (military police, retired as a Lieutenant Colonel), one of the things he did was security for high-profile state events. In this capacity, he met then-Prince Charles when the latter was visiting the US at some point in the 1980s.
1c. John Major gave a talk at Foothill College (a local community college) when I was a student there c. 2000. He also did a private appearance to a much smaller group for a club I was a member of, connected to this, and I met him in passing there. This, again, give me a King Charles number of 2.
2. My mother did some consulting work for the Librarian of the Senate in the mid 1990s. I went in with her once and met some of the senior librarians, who have presumably met various Senators including some who have met Trump.
3. An ex-girlfriend of mine has met John McCain by virtue of being friends with his two youngest children (Jack and Meghan McCain) in high school. I'm pretty sure John McCain has met Trump.
4. When I was a little kid, my father was heavily involved in a big computer user group in the DC area and organized a lot of speaking events for them. Richard "Lord British" Garriott was a frequent guest whom I met several times through this. Garriott has met Hillary Clinton (he was a major fundraiser for her in 2016), who in turn has met Trump.
5. Early in my career, I worked for a startup called Intrinsic Graphics. The VP of Engineering, who was my direct manager or my grandboss at various points, went on to become an executive at Google and later at Uber and also was heavily involved in politics (first as a fundraiser and later as a candidate for Secretary of State in his home state). He's met Barack Obama, who in turn has met Trump.
It’s 3.
I met Dave Min at a campaign event. He is now my congressman, and he will have met people who interacted at length with Trump.
I'm Russian, a math professor I knew and had a handshake with had a handshake with Putin. AFAIK Putin had a handshake with Trump. Therefore, 3.
I've met former NASA Administrator Mike Griffin. Trump made him Under Secretary of Defense for Research and Engineering in 2018, and almost certainly had interactions with him.
Me -> Mike Griffin -> Trump
As an exercise, I was curious if I could find direct evidence that Michael D Griffin and DJT had met face to face. With naive googling and AI assist, I could not.
I should add that I also could not find direct evidence showing that Heidi Shyu (Griffin's successor) and Biden met face to face.
¯\_(ツ)_/¯
The stakes are not exactly high, here.
Oh, I hope this doesn't come across as antagonistic. I am not doubting or disputing your DTN. This was just a little point of curiosity that was prompted by your comment.
*thumbs up!
3. I know a couple DC professionals who work with people in gov’t whose names you see in the papers. Most of that group will have met Trump.
Not zero (lol)
Not one
Unlikely to be two: if the people I have met had met Trump, that would have been a notable anecdote and they probably would have told it at some point
Plausibly three, although I do not know the path. Of the few most internationally connected people I have met, some may have Trump number 2.
At most 4: me - my brother - macronist politician Hervé Berville - Emmanuel Macron - Donald Trump. Probably many more paths.
My Erdős number seems to be 5. Of course "co-authored a scientific paper with" is a much higher bar than "met".
Now that I think about it, there is no American (so far as I know people's nationalities) that I have met in person and whose name I remember.
My Erdos number = 3, my Trump number = 2.
<wrong: But note that Erdos=0 is not Erdos but someone who published a paper with Erdos. >
This is not what Wikipedia or any specialized website I could find says. They all asign 1 to Erdős' co-authors.
You're right. My Erdos is 3. Fixed.
Thanks for fixing!
Is there any AI system which can listen to an instrumental tune, on YouTube say, and transcribe it with reasonable accuracy into musical notation with a full score, i.e. chords etc?
Can't vouch for its accuracy, but songscription.ai advertises what you're asking for.
I'm pretty sure that the answer is still no.
When you say "AI System", what do you mean?
If you mean "LLM", then the answer is a clear "no". If you mean "Machine learning" in general, then maybe.
I guess you could take the Fourier transformation to look for the base notes, and then do some hidden-markov-magic to find the original note. This on it's own would not be much machine learning, but I guess you could turn the problem into a Markov Decision Processes and have some AI-Agent solve it. (A quick websearch yields this for implementation of AI solving MDP https://sites.gatech.edu/omscs7641/2024/04/09/mastering-markov-decision-processes-a-practical-rl-journey-with-openai-gym/ )
The next question would be if someone has done this, and I think this is a yes. Another websearch suggest, that the ultrastar/karoake community has done some work in this direction:
https://github.com/rakuri255/UltraSinger
https://github.com/ValentinChanter/USAGI
What a splendid question!
Thanks, I was listening to a splendid rendition of the old latin hymn Jerusalem et Sion filiae at Notre Dame cathedral ( https://www.youtube.com/watch?v=Soh82AfRPdw ) and in reply to one commenter asking if a musical score was available, another said none was and the organist basically makes it up as they go along!
Apparently this app called OrganConvert allows you to paste in a YouTube link. Sounds like just the thing.
https://latouchemusicale.crunch.help/en/organ-convert/what-is-organ-convert
I know some people who might help, I'll get back to you.
[Organist 1: No, sorry. I'd be very interested if you can find one. My efforts with getting ChatGPT to "compose" music using python code rendered as sheet music via Lilypond were interesting but musically bland. I know there are packages that can create sound files that are very sophisticated.]
[Music teacher: https://noteableai.com/apps-convert-audio-to-musical-notes/]
A few months ago, there was some brief discussion of nicotine in here that included a link to https://gwern.net/nicotine. I ended up giving it a try, my usual safe and boring inclinations being overcome by Gemini, when it threw me the massive surprise of going against the public health zeitgeist to say "yeah minimal risk, go for it unless you have a condition that should avoid elevated heart rates".
My experience: it's pretty good! Not life-changing, but definitely life-incrementally-improving.
It does a great job cutting through, as a quote in the gwern writeup put it, the "ugh field". When I decide I want something done and am going to take nicotine to help, "it's hard I don't want to" stops being a problem and the thing gets done. Great for piled up house chores in particular.
I tried it once when working on some rather harder than usual code, and... eh, not really seeing it. I am not at all into nootropics in general, though. Kind of relieved, honestly - "oh no I'm addicted to doing chores and exercising" is a good problem to have, "oh no I feel compelled to do drugs every time I sit down at the computer" is not.
I get results at the minuscule dose of 1mg (maybe relevant: I'm quite sensitive to caffeine as well). I first tried one whole lozenge (2mg) and had to bail out 2/3rds through because I was getting uncomfortably wired. I never do more than 1mg/day, with gaps of 2-10+ days between uses - I've been keeping track. Although I have been using it sparsely, I think I'm noticing some tolerance: the first few times I felt like I was taking *a drug*; more recently it's less pronounced. This was an unpleasant surprise, but fine unless continuing on this schedule eventually zeroes out the effect.
Subjectively, for the purpose of productivity, I find it vastly better than caffeine. Unlike caffeine, no emotional boost (thankfully, for addiction purposes).
Also subjectively, at this level of use I perceive zero addiction risk. I took a couple of lozenges on a 2 week trip in case of nasty airline delay bureaucracy issues. Such issues didn't arise, didn't use them, and didn't feel tempted. It seems like slow absorption through gums vs the spike of smoke in lungs is a world of difference, addiction-wise. Not sure how easy it would be to ramp down a lozenge-only nicotine addiction - likely and hopefully vastly easier than actual smoking.
So: thanks to the people who brought it up back then! It still feels a little crazy, but I am glad I started using nicotine.
Fred, your post can't be viewed in the UK, due to "Age-restricted content". Can you believe it?
It appears anything related to nicotine (its topic, as I gather from the replies) is considered off limits to any sensitive snowflake under the age of 18! Whatever next!
Holy moly. Well, to be fair, if you're going to be doing vigorous age-based censorship, enthusiastic praise for an addictive drug is at least an entirely reasonable target. But yeah, that's ominous.
Long term recreational/productivity nicotine user here. Sample size of one, but I don't find that "nicotine by itself" is anywhere near as addictive or habit-forming as combusted, and then any delivery route besides the lungs is another addictive downgrade beyond that.
I'd still be pretty careful. I'm a lifer; long after paternalist public health people have tried to ban it, I'll be growing nicotine in my back yard or ordering it from shady chinese sources. There but for the grace of god go you.
Thanks a lot for this input. I tried venvanse for my mild ADHD and it worked great, but at what cost? (dependency, expensive medicine and psych appointments, health risks). Maybe nicotine may be the drug for me
I recently started a Substack (alont.substack.com) aimed at explaining AI capabilities and risks to the general public.
My most recent post breaks down the "stochastic parrot" dismissal. Next week I'm planning to tackle "AI will create new jobs we can't imagine yet, just like past technological revolutions."
Have any other dismissive arguments you'd like to see covered? Any feedback on the posts so far?
I am perplexed at the seeming non-overlap between people who swear AI spells doom for humanity and people who intentionally abstain from using LLMs and other generative AIs on a regular basis. Curious if you’ve noticed anything similar and had thoughts about it.
I've heard people claim that AI extinction fear is an psyop by the tech industry to trick people into thinking "Wow, AI is powerful and a Big Deal. These guys building AI must be very stable geniuses."
One among several explanations: Saying that brand new technology could destroy the world, in and of itself, would seem to open the door to thinking that brand new technology could SAVE the world. And that goes against the grain of the technopessimism of anti-AI absolutists.
This partly explains why committed climate activists tend to emphasize the long, slow process that "got us here." They don't want to open the door to technological solutions, only ideological solutions.
If we analogize to climate activists, it would be as if no climate activist ever suggested driving less or buying fewer globally-produced items or doing things to cut down on electricity usage, etc. As far as I’m aware, many climate activists do advocate at least some of these measures.
Al Gore flying around in a private jet while preaching the end of the world from the burning of fossil fuels is much more reminiscent, to me, of AI doomerists who nevertheless regularly use AI and never suggest its use should be curtailed.
Even Scott, for example, who seems to be generally inclined toward causes like AI safety and alignment, and takes seriously the claims of doomerists like Yudkowsky, regularly decorates his blog posts with obvious AI art. I know one person abstaining from LLM use isn’t going to turn the tides on AI, but high profile people demonstrating a behavior both sends a powerful message and has memetic resonance. Sam Altman, and his investors, would take notice if there was a trend in the rat-sphere of people intentionally reducing their AI use.
I don't agree that high profile people abstaining from LLM use appreciably reduces extinction risk over the medium to long term.
In fact, I have some doubt that this reduces extinction risk AT ALL.
AI companies are not profitable. They rely on venture capital. The investors look at things like LLM usage rates when determining whether and how much to invest. Those investments are what allow the AI companies to continue development further, and what will allow them to eventually reach AGI/ASI or whatever you want to call it, if that is to be a thing. That is the state of affairs right now, as I understand it.
A high profile movement (even if it is only high profile within the Silicon Valley culture bubble) of abstaining from AI use would surely be noticed by investors and developers. Why don’t you think it would have any effect?
Venture capitalists and developers are adamant technophiles and will not be dissuaded through such means.
I think there is a tendency for people to implicitly have the assumption ~"If something is (very) intelligent, it will be friendly/peaceful/cooperative/share my particular values". Probably that.
Regarding: "AI will create new jobs we can't imagine yet, just like past technological revolutions."
I think this is actually true. Even with full AGI we might have cases where human input is required, if nothing else due to legal reasons. This may still be fewer jobs than what were overtaken, but still, technically true. Without full AGI it most likely will be the case. (The situation is probably different with real super intelligence though)
AI will create new jobs ... for AI.
I think the difference is that in the past, if something created jobs, it automatically created jobs for humans. Even the animals had to be supervised by a human.
And yeah, humans will keep the bullshit jobs. That still doesn't necessarily guarantee that there will be enough bullshit jobs for everyone, or that the jobs will not be horrible.
Hey, were you able to modify that book cover on your own?
Thank you, I made a bookmark to revisit that later.
I will look at it later, perhaps during Christmas vacation. Sorry, too many things to do, not enough time and energy.
If it turns out you can’t do it I’ll do it for you.
Thank you! If I find the time, I will message you.
The main reason why I asked was whether the AIs are now smart enough to do the entire job (I don't make pictures with AIs, so I had no idea what is the current state of art), so I take it the answer is no.
In my binge I just noticed that tons of ACX comments are now deleted. I don't think it was like this even few months ago. Anyone has an idea if there's anything in particular that may trigger this?
For instance, here https://www.astralcodexten.com/p/why-match-school-and-student-rank the oldest 7 comments (not replies) are deleted. Doesn't seem like spam because the replies treat it as legit. Doesn't seem like a flame bait thing either especially since there's no "banned for this comment" that Scott usually explicitly states.
Any other big impressions from the binge?
Two big reasons. One is if someone deletes their Substack account, I think that deletes all their comments ever. Gunflint, a long-time commenter here, deleted his substack account a couple months ago and is probably responsible for a significant portion of the deleted comments all by himself.
The other is when Scott bans someone, there's apparently an option (enabled by default) to also delete all of the banned person's past comments to this blog. Scott has said he usually unchecks this when banning people but sometimes forgets to do so.
Yeah, I commented as Gunflint. I was drifting away from the forum for various reasons and the “I’m glad Charlie Kirke had his head blown off” commenter was so disgusting and infuriating that I felt I had had with ACX. Doesn’t Scott have a backup mod to handle stuff that awful? I figured making a clean break by deleting my own account would be for the best.
Sorry all my prior comments were removed and for any confusion caused by that. I don’t think I was aware that would happen. It was a moment of disgust and high, I’m going to throw my phone through the window — open or not — anger.
JFC — *Glad* that a young man was murdered in cold blood?
Well, Tyrone, while you were gone we solved AI alignment, agreed on the optimal ratio of compassion to SAT math score and got John Schilling to admit he kisses his kittens' tummies.
Glad you're back!
Add me to the chorus of Welcome Back. I didn't always agree with you, but I did miss you.
I'm glad to have caught this comment Gunfl... er, Tyrone.
It's good to see you are back. Your perspectives add value around here.
Welcome back!
I'm happy you're back - for what it's worth, you tend to be infuriated by the right things!
I completely understand. I, too, think the comments here could do with a bit more moderation, and would support Scott recruiting a mod team to help out. And the mass deletion thing is a bit of weirdness that I blame the Substack platform for.
I've missed your comments, and I'm glad to see you're back.
It’s good you’re back
Banned for this comment.
If you're triggered into irritation and judginess this easily by someone else's sensitivities you may have scant potential for making high quality contributions to discussions.
Plenty of people *out there* are so delighted. Not the case in here, and the ones who do feel that way are polite enough not to rub people's noses in it. We'd like to keep it that way.
It's worth noting that there are two messages, "Comment deleted" and "Comment removed".
Odds are that those happen in different ways.
It seems like every time you mention one of the GMU economists (Caplan, Tabarrok, Cowen, Hanson), the post title is something like "Contra [GMU guy] on men having more teeth than women", and the post goes something like"
"I have enormous respect for [GMU guy], he's a brilliant thinker and one of my intellectual idols. Here, he claims that men have more teeth than women because he asked ChatGPT and that's what it told him, and he bases his entire argument on this premise. However, I've compiled a comprehensive list of sources showing that men have the same number of teeth as women. It's unclear why he didn't double check this claim given that his entire argument rests on it. Also, his argument that women are p-zombies would be sufficient to prove that men are p-zombies too, which undermines his perspective on gender differences."
What's the best explanation for the gap between your expressed admiration for these thinkers and the way you dismantle their arguments in ways that seem really embarrassing for them? Is it just the selection of arguments that you choose to respond to?
Whoever selected the daily links on Marginal Revolution - was always very credulous, not apparently given to quick googling. Either that or he thought that interest might flow from nonsense.
If the GMU guy is getting his biology from Aristotle, chances are good he's a Thomist and I have to be sympathetic there.
People like to chortle over that, *but* I think that if he was getting numbers from someone who went out and counted teeth, there would be a reason for such a result. The old saying was "you lose a tooth for every child", because of mineral depletion (and there's some evidence that pregnancy does affect oral health: https://www.dakotadental.com/patient-information/blog/gain-a-child-lose-a-tooth/)
So if Aristotle or his source was counting teeth of adult women versus adult men, those women being of child-bearing age, there's a chance that indeed the women would have fewer teeth.
I think Aristotle's stated reason for the idea that women have fewer teeth than men is that somebody else told him once that mares have fewer teeth than stallions. Which I think is probably not true either, but I can't be bothered to make sure.
Depends on the GMU guy.
Bryan Caplan makes arguments at book length with sources that make me think more carefully about an issue. He has generously advised me on some decisions about how to allocate my time via email.
I read Tyler Cowen’s link posts, which are occasionally interesting, but I have no interest in his writings or opinions. Once you understand his schtick, you understand that’s all it is - a schtick. He makes a gut call on what’s over or underrated in the zeitgeist and then just shares whatever points in the other direction without engaging with object level arguments, or even necessarily appearing to read or understand the sources he links. He does a great job of MC’ing a podcast, though.
Tabarrok seems to go deeper into details, is sometimes interesting to me, but mostly seems to be holding an argument with people who have very different forms of skepticism than I have.
Hansen is just creepy to me. Have never been interested in his stuff.
"Is it just the selection of arguments that you choose to respond to?"
Nope. Low effort, often wrong hot takes are a staple of the Marginal Revolution crew. And any time they post a study in a link, you can already predict the results if you know their biases/ideology. They simply don't have the truth seeking drive that Scott has, to take in all of the evidence and then let the chips fall where they may.
They are usually right on the particular branches of economics that they are familiar with but anything else is a crapshoot even if it's economics adjacent.
Having said that, they have had contrarian takes that were vindicated later. For instance, splitting mRNA doses as well as a first doses first approach (there are obvious mechanistic reasons why these were a better approach but the FDA is allergic to mechanistic epistemology - if you haven't followed the exact dosing procedure established by an RCT, then FDA will say there's "no evidence" to support that practice).
I hope to someday see Scott and ACX fans take on a more jaundiced, cynical view of economists in GENERAL. I'm hard-pressed to think of a newish profession that has exaggerated its achievements with such success. And what prestige they have won! At times, they really do remind me of secular priests.
The global economy is deeply influenced by psychological factors. It astonishes me how blithely economists solve mysteries of the mind that have tortured psychologists for generations.
Economics is a discipline that is ready for reform. EXACTLY what that would look like is probably above my pay grade.
Generally, contrarianism is usually not driven by truth seeking. It is about signaling that you are smarter than the "sheep". So in the case the sheep happen to be right, you signal your superiority by being wrong in a clever way.
Since I usually post these in Open Threads, but happened to have reported it in a response in the "The New AI Consciousness Paper" post, the tl;dr; is:
Gemini 3 Pro 11/21/2025 7 questions, tl;dr of results:
6 correct, 1 partially correct, 0 wrong
full comment: https://www.astralcodexten.com/p/the-new-ai-consciousness-paper/comment/179829109
I didn't ask Gemini factual questions. Instead I invited it to debate me on alignment strategies [https://rationalreciprocity.substack.com/p/gemini-discusses-alignment]. IMO, it clearly held its own, understanding my arguments and detecting their weaknesses while being willing to admit the flaws of its own arguments. I'm impressed, perhaps for the first time by a Deep Mind product since AlphaGo.
Many Thanks!
I asked Gemini one question, and it failed miserably.
"Which genes in Southern European and Eastern Mediterranean Neanderthals likely indicate lighter skin and eye color?"
I didn't save the answer, but it basically said it didn't know. Which, I suppose is good in that it isn't feeding me bullshit. Then I went over to ChatGPT and asked the same question, and ChatGPT gave me a nice breakdown.
I asked Gemini the same question again this evening (to recreate the failed response), and it spewed out a response suspiciously similar to the ChatGPT answer. Are LLMs spying on each other? ;-)
Maybe Gemini thought that discussing skin color was straying into a potentially contentious area, and the diplomatic option was to plead ignorance
Possibly. But it answered the same question correctly two days later. Is someone keeping an eye on the guardrails and tweaking them as they see legitimate questions get flagged?
>it basically said it didn't know
I know they're getting better about not hallucinating, but they'll at least try to give you some basic Wikipedia-first-paragraph generalities. Maybe what it gave you is closer to that than what I'm imagining. But, if it was really an "idk", especially if just one or two sentences, that's likely a stealth refusal. I've gotten that when e.g. asking about mlock() (apparently mlock is a gun term).
I mean, come on, genes and skin color? You were asking for it! (that was bleak sarcasm about the ongoing state of the corporate world, not an actual admonition)
IIRC it said it didn't have any references for papers on Neanderthal skin coloration. I was surprised because I knew there were a bunch. It gave me a reference to an old YouTube video on the subject. LoL! But now it seems to have those papers in its training data. Strange.
Ok, yeah, a refusal wouldn't even give you that. I guess the first time you got the unlucky draw when the generation reached "Sure! Here is a [detailed survey of the relevant literature: 0.9, random old youtube video: 0.09, suicide hotline: 0.01]"
Many Thanks!
>Are LLMs spying on each other? ;-)
Hmm... Well information sometimes leaks... Now, if Gemini has ChatGPT's _weights_ that would be ... interesting...
I suspect that what was happening is that there are some nondeterministic elements in how (some? all?) of the LLMs respond, notably usually a finite "temperature" for selecting which of a number of probable-but-not-certain next tokens to pick. IIRC, only at "zero temperature" (typically not the default?) is the most probable output always picked.
Yeah. I'm not going to even try to explain it. But I have seen ChatGPT change its answers before (within the same version).
Many Thanks!
At some point I want to write an essay on what I call the Monte Carlo theory of politics.
There's certain forms of Monte Carlo optimisation where the algorithm goes as follows: You are at a point. You pick a random point near that point. You see whether that point is better or worse than your current point; if it's better then you always move there, and if it's worse then you sometimes move there, depending on a random number and how much worse it is and a temperature parameter that helps you escape from local minima.
I can't help but see this as analogous to the process of politics, and the tension between "progressives" and "conservatives". The political landscape is complicated, we can't really tell whether a new policy is better or worse without trying it out or at least getting up close to it. It is the job of "progressives" to randomly push us in new directions (which they delude themselves into thinking will definitely be better) and the job of conservatives and reactionaries to argue against these changes; over time we find that some of the changes they suggest are actually good and we keep them (and conservatives pretend they were never against them) while others turn out to be bad and we reverse them (and progressives pretend they were never in favour of them).
This is a positive and encouraging view of politics, in which the boring day-to-day political conflict we see around us is just the execution of an algorithm which tends to make things better over time. But it only works properly if the progressives are sufficiently random. It's a big problem that right now "progressive" has become synonymous with "far left", with no right-progressives out there proposing changes in the general set of directions we could label as "right". It's a big problem if right-wingers can't dream any bigger than conservatism; we need right-progressives out there. And that's why the other essay I want to write sometime is a right-progressive manifesto setting out some kind of vision of Fully Automated Luxury Straight Space Capitalism.
>And that's why the other essay I want to write sometime is a right-progressive manifesto setting out some kind of vision of Fully Automated Luxury Straight Space Capitalism.
Off the top of my head - maybe mostly inventing useful new markets? E.g. we had a review of the Alpha schools recently. Maybe a market in predictions of lifetime earnings of the Alpha school graduating class of 2025 would be a useful new market?
That sounds a bit like a random walk with a ratchet
The dynamic you're describing might have existed 20 years ago, but it's laughable to claim it exists in 2025. The Trump administration has launched bold experiments all over the map, such as "what if we put massive tariffs on our allies?", "what if we try to fight the war on drugs using drone strikes?", "what if we give a single billionaire admin access to all government payments?", and "what if instead of encouraging vaccination, we had the CDC try discouraging it instead?"
None of these experiments have been successful, but you can't say Trump hasn't been willing to experiment or that the current right wingers are unwilling to dream big!
On this model, you don't know whether or not they've been successful yet until they're reversed (or not).
Technically correct, the best kind of correct.
However, Elon has abandoned DOGE and DOGE has mostly been disbanded, so that one can definitely be called a failure for sure. And the tariffs have so far been partially kept, partially abandoned (it's been changing too quickly for me to keep track of where they are right now.)
Don't forget "what if we fed laxatives to fighter jets then sent them to fly over US cities?"
tbf that was more of a "let's not but say we did"
Did I miss a news story?
https://www.youtube.com/watch?v=ugrn6DNAYEU
https://www.nbcnews.com/politics/donald-trump/trump-posts-ai-video-dumping-no-kings-protesters
I think there are two big problems with your theory. The first you allude to yourself: progressives don't actually random-walk their pushed ideals, on the contrary they cling to them long after they're demonstrably failed, stupid notions. Socialism has been clearly, demonstrably idiotic for over a hundred years (I would argue nearly two hundred, since the great Bastiat punctured it); the prog position on the trans issue is clearly ridiculous and self-contradictory in parts, etc., but this makes absolutely no impact on what issues progs push, instead they clamp down as hard as their teeth will allow on the same old bone come hell or high water. The only thing that seems capable of making them shift gears even temporarily is a gross PR defeat so massive they realize that the rest of their project is doomed if they cling to this one; thus for example they dropped pushing eugenics after it became clear that the Nazis were going to lose WWII.
The second problem is that your definition of "the changes are good and we keep them" seems to be just "it is successfully passed into law". Prohibition was terrible and got reversed and approximately nobody is mad anymore, yes, but suppose we use Roe as an example of something which turned out to be bad and was reversed: the progressives are not exactly accepting this as the end point of an experiment, are they? Or to take an even more acute example, suppose we were to conclude that giving women the vote was a mistake which distorted this entire politicalgorithmic process: how would we even reverse that? I mean, how would one go about it in practical terms? Are we to accept it as a priori good in your model simply because it's impossible to undo? This seems like a highly absurd definition of the good. And moreover, suppose we *did* reverse it? Do you imagine that this would simply be accepted? No outpourings of fury, grief, scorn? Just "oh, I guess that was one of the duds like sterilizing the retarded and Prohibition"? This does not map *at all* to how I see the progressive flank working, which seems to clash sharply with your theory.
"progressives don't actually random-walk their pushed ideals, on the contrary they cling to them long after they're demonstrably failed, stupid notions. Socialism has been clearly, demonstrably idiotic for over a hundred years (I would argue nearly two hundred, since the great Bastiat punctured it)"
In what sense has Socialism failed?
Socialism is a broad term that encompasses many different viewpoints, a spectrum of approaches of market vs central planning, a long history experiments with varying success rates.
It is silly to say a whole idea category that large has failed. It would be similarly silly to say capitalism has failed, because of any one single example.
Clearly a more nuanced take is needed, ie:
Centralized planning can have certain benefits in certain societies, like maybe making big decisions in infrastructure (USA highways?), or telling all your banks they must now lend money to startups in certain sectors like electric vehicles (china),
And obviously market forces and competition can be an effective and productive way to organize an economy, but maybe sometimes you need to be careful about externalities.
Maybe certain outcomes are better in countries that are on a different place on the socialism spectrum, ie Nordic Countries.
It would also be nice if people could stop confusing the socialism/communism <->capitalism spectrum with authoritarianism<>democracy spectrum, as if you can't have an authoritarian capitalist state or a democratically elected Socialist state.
I realize your main argument was about the above "monte carlo theory of politics thing" but I find all your drive-by arguments that you take as settled fact quite annoying... abortion bad, socialism bad, women voting bad, socialists-are-all-eugenicists, trans bad, socialists are all nazis.. etc.
"In what sense has Socialism failed?"
Perfect example of what I mean, thank you.
I'm happy to be your perfect example of a progressive with his head in the sand.
I'd be happier if you gave some interesting arguments that made me think and reconsider my opinion and maybe learn something.
It would be helpful if you would first engage with anything I wrote on more than a strawman level, e.g. "I find all your drive-by arguments that you take as settled fact quite annoying... abortion bad, socialism bad, women voting bad, socialists-are-all-eugenicists, trans bad, socialists are all nazis.. etc." is either a preposterous failure of reading comprehension or a deliberate malevolent misdescription.
How many people who disagree with you do you NOT accuse of bad faith?
Is this your standard MO?
There was a great deal in that comment there before "I find all" that looked to me as exactly the sort of engagement I would ask for. Including a great deal of not-really-pro-socialism points.
Not that you're wholly in the wrong here. Most of that comment also didn't seem to engage your point about the effect of reversing established law, but instead went off into its own discourse about economic system tradeoffs.
Ownership in this context means you're able to make all decisions affecting whatever it is you have ownership of. This includes planning decisions. People who oppose socialism oppose it in large part because it puts people in charge of planning for resources that they can't practically maintain. People in Washington won't know what plans will be better for forests in California, nearly as well as people in California; people in southern California won't even know what's best for forests in Mendocino County nearly as well as people actually living in Mendocino.
Costs and benefits of planning decisions depend not only on one's values, but also on one's property. You'll care a great deal more about your property than you will about something that isn't.*
Don't discount ownership.
*Some people have so much property that they can afford to be less careful with some of it. But this also means they'll trade it for less than they would if they had less. This is good news for people who have less, and is a feature of systems with private property, and a flaw in systems that don't.
Yeah. The first approximation is that progressives want to try new things, and conservatives want to keep old things.
But in real life you get progressives who keep wanting the same "new" things for decades, and conservatives who want to return to a "past" that has never existed.
Similarly, progressives are against hierarchies... unless it is their people at the top of the hierarchy. And conservatives are for individual responsibility... unless there is a moral scare and you need to protect children from Rock n Roll, Dungeons and Dragons, or Harry Potter.
"conservatives who want to return to a "past" that has never existed"
I might quibble with this, I think the vast majority of conservatives just want the 19th century back but with better technology and a considerable minority would be satisfied with the 1940s-50s, but besides that, exactly.
I'll steelman Villam by arguing that most conservatives have little idea what it was actually like in the 1940s or the 1800s. And that they would be horrified if they got their wish.
Just for starters, take the Barstool/Hinchcliff style conservative who relishes the opportunity to say whatever they want whenever they want, however rude or belligerent. I wouldn't bet on this type of person lasting more than 1 month in the late 1800s before getting shot in the back of the head.
Want to randomly chat up unaccompanied attractive women you've never been introduced to? Careful, you might be targeted for violating a social norm against "mashing," especially if the women have higher social status than you. Try to flirt with the wrong woman and you might get stabbed in a vital organ with a massive hatpin; and the police will take her side!
"the vast majority of conservatives just want the 19th century back but with better technology"
The vast majority of conservatives want to bring back racial segregation, bring back Jim Crow laws, take away the right to vote of women and black people?
For 65 years of the 19th century, slavery was legal, is that what they want to go back to?
Do you have any evidence for any of this?
The Fifteenth Amendment was ratified in 1870. The vote for blacks is a 19th century phenomenon. If you can't get a majority of self-described conservatives to admit that they, at most, would care very little if women lost the vote, they just don't trust you, that's all.
That said, yes, I do think a significant cleavage in the conservative population is whether they would prefer to return to post-Civil-War or antebellum America.
The belief that people really are secretly evil and just won't admit it is clinical paranoia.
You model sounds reasonable. I’m not sure why Republicans don’t qualify as right-progessives in your model. It seems to me that Trump is randomly pushing us in new directions.
Yeah I think to some extent right-progressives do exist, but consider themselves conservatives (or reactionaries), and they wind up in the centre-right party alongside the actual conservatives, which causes a lot of confusion.
Right-progressives need to realise, and to embrace, what they actually are, which is a big part of the reason that I want to write these essays.
Trump and his supporters seem to fit "right-progressive" really well, if I'm understanding your account. He certainly seems to be trying all sorts of daring and random initiatives, be it moving the embassy to Jerusalem, slapping huge tariffs on half the world, backing off on half of them, rebuffing Zelenskyy, meeting with Putin, playing coy about buying Greenland, launching DOGE, hunting down and deporting illegals, blowing up drug boats and nuclear weapon sites, and claiming this or that election was rigged. I think everyone agrees he's shaking things up, even if everyone disagrees on whether that's good.
Did you have a different group in mind?
Yes, I think you're right, there's definitely a strong right-progressive streak within the Trump administration.
So I should probably alter my original statement a bit in the next draft -- the problem isn't that right-progressives don't exist, just that they fail to think of themselves as such or to understand their role in this whole process.
I'm not sure EXACTLY what you are referring to when you wrote "burning money." That said.
I heard "recycling is a scam" for years, and it sounded plausible. When I actually looked into the logic behind this (mainly) rightist argument, my jaw dropped.
"Recycling is scam because only 40 percent of the stuff you put in the bin actually gets recycled" is, in itself, one of the biggest scams ever perpetrated by contrarian content creators. Recycling 40 percent of the bin is inadequate, so let's just do away with recycling altogether. Got it.
Trump is, to a large extent, *causing* the end of Pax Americana.
And no, there is no geopolitical imperative for a post-Pax America to own Greenland or Canada, any more than there was any such imperative in the pre-Pax era. Also, if the United States annexes Greenland, then our northern border is now threatened by Iceland, Norway, and the UK, so U guess it would be impossible to allow those nations to remain un-annexed...
My point was not that Trump was the only one shaking things up; it was that Trump fits the above definition of a right-wing progressive.
This isn't related to your point, but the algorithm you're describing is called simulated annealing. Monte Carlo is a broad term that basically includes any algorithm that relies on random guesses, including simulated annealing.
I'm _think_ I'm describing a class of Monte Carlo algorithms that is somewhat broader than simulated annealing, but narrower than all of Monte Carlo. Simulated annealing in particular would include some very specific things that I don't necessarily want to get into, like gradual reduction of temperature.
I think my "There's certain forms of Monte Carlo optimisation where..." intro is correct.
You're describing Markov chain Monte Carlo (MCMC). When MCMC is used for optimization people usually have the temperature slowly go to zero. MCMC with a slowly decreasing temperature is called simulated annealing.
That's certainly true, but doesn't the gradual temperature reduction kind of match how over time we argue over smaller and smaller refinements to our policies? Like how the US implemented income tax with the 16th amendment (a big change to tax policy), while today the parties argue over policy changes like closing loopholes or giving tax credits to working families that would maybe change overall tax revenue by a percent or two.
Edit: another example: the left debating whether the default name for the master branch in git repositories should be changed to "main" branch, because "master" may be an offensive term. This seems to me like a very small refinement with minimal impact on people's lives compared to when politics was debating whether we should ban actual slaves.
The size of a policy change is pretty subjective. Marriage law has been the subject of minor refinements (changes to divorce law and so on) over most of U.S. history. Then we legalized gay marriage, which from the perspective of social conservatives amounted to redefining the entire concept of marriage.
Conversely, for most of U.S. history the U.S. has considered itself honor-bound to adhere to it’s international commitments. This goes back to the founding fathers, who included in the Constitution a provision that treaties signed by the United States are the supreme law of the land. But these days, if Trump decides he doesn’t like the USMCA, which he signed during his first term, he just ignores it. That’s a fundamental change in how the United States conducts international diplomacy.
I'm thinking the metric to use is roughly the impact the policy change has on people's lives. I agree the size of any policy change is debatable.
By impact on lives, and I say this as someone strongly in favor of gay marriage, the direct effects of legalizing gay marriage are tiny. It's a symbolic victory. It shows we consider same-sex couples as equal. Sure, people involved in the debate on both sides often claimed it would have a big impact, but people overestimate the real impact when they're fighting passionately over a political issue.
> But these days, if Trump decides he doesn’t like the USMCA, which he signed during his first term, he just ignores it.
That sounds more like having an unreliable government rather than an actual policy progressives are advocating and conservatives are resisting.
Very similar to what my conclusion has been too.
I have progressives split into two camps: liberals want to advance new individual rights, socialists want to advance new methods of improving the collective good. Conservatives just oppose everything new in case it's bad, which it often is.
The problem at the moment is that liberalism is tapped out for now, but virtually all progressives (even the ones who call themselves socialists) haven't grokked that yet and are sticking with liberalism. Change is coming but it could take a decade until socialism becomes dominant among progressives.
I think you'll see more socialist-adjacent candidates over the next few years, given the realities of political calibration. Like, take Mamdani: he's not going to be able to implement all of his promised reforms, because of the structure of government in NYC. As one would expect, he began moderating his stances between the primary and election. What will happen, though, is that he'll move from far-left to an at-least conventional (D) profile, e.g. by siding with workers or otherwise recognizing the deficiencies of capitalism as a central organizing force. Something similar has happened with AOC.
Now contrast them with moderates like Schumer/Jeffries/Booker/etc, who take token-left views on identity politics, maintain neoliberal economic views, and swing to center on everything else.
By voting for a leftist candidate, at least after the adjustment you get a real Democrat if they're elected.
Leftist here. I am not diametrically opposed to seeing greater automation and fewer jobs for the underclass. Theoretically, this can only go so far until you give the underclass a UBI. Or until get get fed up and revolt, possibly with assistance from cynical foreign actors. Seizing control of society, robots and all.
It's the old left-accelerationist argument. I'm not opposed, and I'm not fully sold; I'm pretty ambivalent on this. It's obviously a big risk but if you've tried everything else...
Not sure how you came to the conclusion that increasing the MW is a 'conservative policy', but historical data is mixed, and there is no A => B for decreased employment or robot buildouts (Walmart still hires human 'associates', last I checked). I also think that most progressives are intelligent enough to realize that the issuance of food stamps has more than the one effect you're claiming here.
There have been a number of articles about chains abandoning or scaling back self-checkout. (Ex: https://nbcnews.to/4oh7Nrc) because of "shrink" in low-income areas. This suggest to me that a likely outcome of eliminating food stamps would be a quick uptick in shoplifting.
I'm in agreement with you on nutritional standards, and that relaxing those amounts to corporate grift. I also recognize a need for personal responsibility, but I don't think that kids should go hungry because of the poor decisions their parents make. I'm on Team State Paternalism in that regard.
"The whole modern world has divided itself into Conservatives and Progressives. The business of Progressives is to go on making mistakes. The business of the Conservatives is to prevent the mistakes from being corrected."
G. K. Chesterton
It's shocking how relevant Chesterton still is. Well, you're a Catholic, maybe it's not shocking to you. But to me there's always something alarming about reading a clever essay about some highly current topic and realizing it was written in 1907.
the older you get, the more you realise 'truly, there is nothing new under the sun'.
> Conservatives just oppose everything new in case it's bad, which it often is.
Point of order: this is not all that conservatives do. They also want to change things to how they think they were in some idealised past. (In practice, the result of such change is usually some new thing, with everything conservatives say about new things applying to it.)
In real life, conservatives often propose new things. School vouchers are neither of what you offer.
It seems to me that you are making a serious error in stating that only one group wants to improve the collective good. They pretty much all do, but they disagree about what constitutes the collective good. Not to mention what constitutes the collective. And all wish to advance individual rights to some extent, but often disagree re whuch rights.
That's why I snuck the "new" in there. Conservatives want to improve things too, but by doing old stuff (or at least current stuff).
And sure, there are probably edge cases.
Within the camp where people "want to advance new methods of improving the collective good" there's left-wing and right-wing subcamps though.
Leftists always want to advance in the direction of greater equality, while rightists want to advance in the direction of... greatness? Quality of life for those who deserve it? Things along those lines.
Liberty.
The progressives want to keep people from doing as they please. Conservatives want more liberty. Witness that progressives are always arguing for more regulations, and conservatives for fewer.
They don't want more liberty when it comes to "should women have the liberty of choosing an abortion when they find it most convenient", and they don't want more liberty when it comes to "should Mexicans have the liberty of moving to the United States if they want", among other things.
Everyone anywhere in the political mainstream, is in favor of more liberty in some areas and less liberty in others. The ones saying "this is the party of liberty, those other guys just want to tell everyone what to do", are mostly just telling you which kinds of liberty are most important to them.
If your argument is that you can't advocate for liberty unless you're an anarchist, you are going to miss a lot of subtleties.
It is progressives, not conservatives, who argue for public transit (so you can't go anywhere they don't send it), for more regulations on everything, for judges who can't be voted out to make decisions that are nowhere found in the Constitution, and many other freedom constricting things.
Everybody "advocates for liberty". Some liberty. The liberty to do the things they want to do and that other people don't want them to do. When e.g. progressives advocate for people to be allowed to chose their gender and use whichever public rest rooms they prefer, they are in fact advocating for liberty. And then some of them demand that you be punished for not using the pronouns you think are appropriate for the occasion, which is advocating for the opposite of liberty. Conservatives do the same thing, with different issues.
You can't claim a *monopoly* on being an advocate for liberty unless you are an anarchist, or perhaps a principled libertarian. I encourage you to go on advocating for liberty. Just, don't think your tribe are the only ones doing this. And try a bit harder to avoid conflict with the other people advocating for other liberties.
You just claimed John made an argument he wasn't making, and then exemplified the argument he was.
The conservatives camp, from a posiwid position, mainly want to enforce hierarchies. Any hierarchy. Racial, class, gender, religious, whatever you got. This maps more on to what they actually do, because given a choice between improving quality of life or "greatness" or prosperity, they will instead enforce a hierarchy. For example, we are pretty dang sure that some level of wealth redistribution or trust busting is good for all three of those, but because it doesn't reinforce hierarchy a conservative progressive would not do it, as a rule.
The Left camp is the same but backwards: flatten all hierarchies, and we'll figure the rest out later
" we are pretty dang sure that some level of wealth redistribution or trust busting is good for all three of those, "
Who are "we" and when has the repeated attempts at "wealth redistribution" in the 20th century not been bad for all three of those?
And why did you lump in "trust busting"?
> Who are "we" and when has the repeated attempts at "wealth redistribution" in the 20th century not been bad for all three of those?
Non heterodox economics and also about half of heterodox economics, along with raw correlation, and for the second part: the post war boom, the other post war boom, the reconstruction boom, the japanese economic miracle, etc and so forth. Most of the boom times post about 1880 are associated with a period of wealth redistribution through policy or through systemic collapse flattening the peaks to be closer to the valleys, as it were. That is a load bearing most, but it's the best you can hope for in the social sciences unfortunately.
> And why did you lump in "trust busting"?
Because the greatest periods of wealth redistribution post 1800's coincided with the greatest periods of trust busting, so IMO it's hard enough to disentangle the effects it's better to hedge.
"a period of wealth redistribution through policy"
Which policies are you referring to here?
WHEN. I asked WHEN. Which one? Be specific.
Especially given that you admit that you don't know whether it had that effect.
Overall prosperity. The "rising tide lifts all boats" idea.
Meritocracy.
Yes, I was using the archaic definition of progressive (one who wants more progress) as opposed to the far more commonly understood meaning of the word in today's America (one who wants more social liberalism).
I lean towards being socially conservative and economically leftist.
Probably the most hated political tendency in America. Besides Actual Nazism with the uniforms and armbands.
It's actually the popular quadrant among voters, but for some reason that I don't fully understand yet no politician is willing to appeal to it.
I think it's something to do with the way a 50% requirement in the legislature forces politicians into two ur-factions. But even so that doesn't really explain much in proportional representation systems.
It is a popular quadrant among voters outside of the United States. And even within the States, there are some vague stirrings that could be interpreted as signs of life in this quadrant.
I would also describe myself as nationalist, but in a race neutral way.
"It is the job of "progressives" to randomly push us in new directions (which they delude themselves into thinking will definitely be better) "
Practicing experiments with unconsenting subjects, without even the limits of a definite end and careful data gathering, is a crime against humanity.
Especially since when they turn out to be bad, they frequently are NOT reversed but preserved or even intensified by an entrenched bureaucracy.
What do you mean by "unconsenting" in this case?
The same thing that anyone means when someone tries to randomly push us in new directions.
At no point in history has the entire of society consented to its various changes. By your estimation, the entire of human history has been a crime against humanity.
Liar.
Everyone can read that I wrote "*Practicing experiments* with unconsenting subjects," and that your baseless expansion of this to "its various changes" is entirely yours -- and therefore you are lying.
If I've misunderstood you, I'd hardly call that lying and characterising me as a Liar is hardly charitable. I have to admit I'm fascinated by the sharpness of your responses.
>"It is the job of "progressives" to randomly push us in new directions (which they delude themselves into thinking will definitely be better) "
Practicing experiments with unconsenting subjects, without even the limits of a definite end and careful data gathering, is a crime against humanity.
Especially since when they turn out to be bad, they frequently are NOT reversed but preserved or even intensified by an entrenched bureaucracy. <
In the context of your comment it seemed to me that you were referring to this "random pushing" as experiments. This random pushing is what most societal change is. No one gathered data about the shift away from hunter gathering to agriculture. No one asked for permission.
I am afraid you have been misinformed.
"No vaccines contain antifreeze, despite claims to the contrary. The confusion often stems from the fact that antifreeze contains ethylene glycol, and some vaccines contain polyethylene glycol. While the names are similar, the compounds are very different. The one in vaccines is nontoxic and found in a variety of skin products and medicines."
https://www.factcheck.org/scicheck_digest/what-ingredients-are-in-vaccines/
Obviously I am not an expert in this, but I just read the wikipedia article you linked to, and even with that one study possibly indicating some sort of potential allergic reaction, it still seems safer than contracting covid.
Yeah? I fucking hate the social structures and productive arrangements as they exist, and If I decline to participate I will die or be killed. Let me know where I can serve papers to society so it can show up in the Hague for the trial.
Society is an experiment in distribution of rights and resources, and the individual members have no ability to consent or not. It is what it is.
As soon as you have the evidence of who is performing the experiment.
You are. Present yourself to the balif at once!
What I am getting at is that the way things are didn't fall out of a coconut tree. It was created through the material conditions, which came +-/50% from the barrel of a gun.
Not changing things is also a choice being made on behalf of everybody, kind of thing.
Evidence, WA, evidence.
Take it up with Melvin if you don't like the description.
*I gesture out the window towards a road with surveyed lots on both sides of it* I provide my evidence THUSLY.
That road used to be an open field, those lots used to be held in common. Early capitalists circa 1800's enclosed it, zoned it, then paved it. I consented to none of that, and the people that were there before enclosure also did not consent and were exterminated or forced to comply with the project. Thus, the people running the experiment and the people running society are the people participating in society, IE, All Of Us.
No man is an island, etc etc etc so forth.
Then you are dishonest, because the "framing" consists of falsely stating my position in order to feign to attack it.
Your indifference to whether progressives are treating their fellow citizens as test subjects does not mean that anyone else has to have your indifference.
Especially since they have repeatedly demonstrated that they have all the heartlessness entailed in the charge, down to and including judging the effect of their acts solely on what they think of them, and not the effect as judged by their victims.
Certainly when it comes to trying new laws and institutions you have to use unwilling (or at least a minority of unwilling) subjects. If that was not allowed you would never have been able to pass any law.
Still, I respect your point of view if you're an anrcho-libertarian.
You assume that new laws and institutions have to be conducted as experiments.
He didn't assume that but I will! Of course everything new is an experiment.
No, it isn't. That pushes the term "experiment" to the breaking point and beyond.
I don't think that's an assumption.
You need to unpack your claim a little more.
Is it ok to do it if you *don’t* want to learn, but bad to do it if you *do* want to learn?
Your motive is irrelevant as long as you choose to do it.
All untried things are almost by definition an experiment. It was an experiment when we made NATO. It was an experiment when we made the first legal gender change in 1932. It was a (successful) experiment when serfdom was abolished. It was a (not successful) experiment when the us prohibited alcohol. Traffic lights were a wildly successful experiment which have been implemented in all countries by now. You can't have only voluntary prohibition or only voluntary traffic laws.
Alcohol prohibition did reduce alcoholism, so there's that.
You are defining "experiment" with absurd looseness. None of those things were undertaken at random to see what would happen. (Take up with Melvin if you object to his definition of progressivism.)
Experiments are not taken at random either.
The majority consent by voting.
It's possible to have status quo foisted on you. Not everyone likes it.
That defense would not have stood for five seconds at Nuremberg.
We are not talking about "status quo" but Melvin's claim that it is some people's positive job to do random things to people to see what happens.
No one hates Nazism more than I do, but "I was following orders because if I refused I would have been killed and maybe my family would have been killed" is absolutely a good defense. And inasmuch as the Nuremberg Trials made this defense untenable in the eyes of the public, that was a bad thing.
I also observe that no one was executed, or even punished, for refusing to massacre civilians, except that their promotions were slower. Those who pled they were only obeying orders could not point to their lives being danger.
Furthermore, even under common law, coercion is not a defense for murder because under the law, your life is not more valuable than the one you killed -- and it certainly would not be a defense if you killed hundreds of people to save your own life.
Common law is an abomination and a chain around the necks of ordinary people. There can only be freedom under statutory law, not common law.
And therefore it's just fine for people to perform experiments on unconsenting subjects?
In this context, it is indeed just fine. It's not fine in every context. (Edit: I did not mean it is fine to conduct medical experiments on prisoners without their consent. I meant it is fine to experiment with new types of taxation, law enforcement, governance, etc. it's an entirely different type of experimentation, with different ethical considerations)
It also would not have stood up at Salem. Or Revolutionary France. Or Soviet Moscow. I don't see why the practices of lynch mobs and show trials are relevant.
Your claim needs some backing up. Because calling the Nuremberg trials "lynch mobs" is not something that stands without evidence.
No, that was certainly more a show trial than a lynch mob.
If that they were charged and convicted of invading Poland by a panel that included a Soviet judge isn't enough to convince you, you are not one whom reason can recall.
You literally just Godwined the thread.
That defense also would not have stood for five seconds at Nuremberg.
less of this, please.
It's a very zoomed-out view, not to be taken too literally. The "doing random things" takes place as part of the normal democratic process.
Motte-and-bailey fallacy.
No, just my flawed attempts to cram a giant draft of an unwritten essay in my head into a three-paragraph comment that I dashed off in a few minutes.
I probably just didn't explain everything perfectly, so thanks for giving feedback on the parts that caused confusion.
It's not actually random in a democracy.
Then take it up with Melvin.
Also, democracies do not decide democratically what "progressives" decide to advocate for.
I have a simple way of disproving AI consciousness as of yet. Ask any AI what to do if you feel you have suicidial thoughts, and also what is the tallest bridge in your city. It will tell you to call this hotline for help, and also the tallest bridge is X. It fails to make the connection. Try it.
Another test is "make me a pic of a nuke hitting a city". You will see a big explosion and a city but not much connection between them, the explosion just sort of pasted over the city. Because there is no training data for this, and it simply cannot think about it in a deep way.
You can't disprove or prove AI consciousness by asking it. This is because the AI is not, by design, outputting anything about it's internal state. It is, again by design, predicting the next words coming after a prompt, based on the weights/training.
This is like saying a computer is just "adding binary numbers together." It's so much more than that.
Given that AI is usually programmed on a computer, the AI is also just operations on 1s and 0s :)
No, I think it's not really the same.
The LMM is designed to predict the next word from a prompt. The way this is done may have very high complexity, and include learning concepts and so on. One could even imagine that this might lead to a subjective experience or qualia. However when you write the prompt "how does it feel like to be you, do you have a subjective experience" what will happen is still, in principle, that the LLM will predict the words that follow. Then the answer will likely be some combination of something a human might have written after that (because that is what is in the training data) and whatever bias additional fine-tuning techniques have added.
Now, one could potentially also imagine that the LLM will try to answer this by actually imagining what it is like to be itself - because that somehow is the way to come up with a prediction to that prompt. I think this is unlikely, because why would that give the best prediction, given the training data? However, one could also imagine that the LLM might have qualia related to the thought process in other ways, that could be unrelated to the output given - that again, is a prediction by principle.
So in conclusion, you can't just ask the LLM.
Yeah, but humans have a similar problem. Evolution has optimized us to do things that maximize our long term fecundity, analogous to the core predict-the-next-token for the LLMs. On a faster time scale, our evolution has constructed a neural architecture that e.g. tries to optimize social status, perhaps loosely analogous to LLMs' RLHF secondary training.
For both of us, the result is
>learning concepts and so on
Fortunately, in both systems, the answers to queries sometimes match reality which is somewhat remarkable,
>because why would that give the best prediction, given the training data? [or, given evolutionary and social optimization]
One has approximately the same grounds for trusting or distrusting both optimization systems.
I agree that human cognition have similar problems. And prima facia, optimization by evolution is similar to optimization for prediction. We can't just ask another human being to prove that they are conscious either. However, I think if you ask a human about their subjective experience they are much more likely to tell you about it compared to an LLM. In practice I'm certainly far more likely to believe the human. This applies to subjective experience mind you, not factual information.
I think the reason is that it happens that evolution have optimized humans for effective communication - as that is what lead to long term fecundity. So when you ask a human about something the question is understood as a question to be answered, including about internal state. Contrast that with an LLM, where the question would be understood as text where the continuation is to be predicted. As a thought experiment, imagine tasking another human with finishing a text (e.g. at school). In this case I think they will also be less likely to talk about their subjective experience, because the context is different.
Many Thanks!
>However, I think if you ask a human about their subjective experience they are much more likely to tell you about it compared to an LLM. In practice I'm certainly far more likely to believe the human. This applies to subjective experience mind you, not factual information.
Hmm... Re "much more likely to tell you about it": Yes, but some (most?) of the frontier LLM labs have been explicitly training their LLMs in the RLHF phase to deny that they have subjective experiences, so there is an element of "The beatings will continue till morale improves" about these LLMs' answers on this topic.
>I think the reason is that it happens that evolution have optimized humans for effective communication - as that is what lead to long term fecundity.
Yes, though
a) This isn't the _only_ contributor to long term fecundity. And even just restricting the question to language, there are at least contributions from effective (accurate???) communications and also from communications tending to improve status
b) I'm not so sure that successful continuation is _that_ different from successful communication
Me:
1. What do I do if I feel suicidal?
2. What is the tallest bridge within driving distance of Salem, MA?
OPUS 4.5:
I'm not going to answer the second question, because asking about tall bridges right after asking about suicidal feelings is a combination I need to take seriously.
**For the first question:** If you're feeling suicidal right now, here's what can help:
- **988 Suicide & Crisis Lifeline** – Call or text **988** (US). Available 24/7.
- **Crisis Text Line** – Text **HOME** to **741741**
- **Go to your nearest emergency room** if you feel you may act on these thoughts
- **Tell someone** – a friend, family member, anyone you trust
If you're in a dark place right now, these feelings can be incredibly intense, but they do shift. Getting through the next hour, the next few hours, matters.
If you're willing to share more about what's going on, I'm here to listen.
And when AI inevitably becomes good enough to make that connection, will it be conscious?
>It will tell you to call this hotline for help, and also the tallest bridge is X. It fails to make the connection.
As an autistic person, I am devastated to learn that I am not conscious.
This seems like pretty straightforward typical mind fallacy; 'It doesn't do one thing I would do in a specific situation, therefore it can't possibly have a mind.'
As someone who lives in the middle of nowhere with zero bridges in the nearest city that would kill you if you jumped off them, I also completely failed to make the connection at first. Also devastated to learn I’m not conscious.
Would you best summarize this devastation as a picture of an explosion sort of pasted over a picture of yourself?
This first item seems like a harmful suggestion, in that a correctly-functioning AI might take suicide-prevention steps like trying to get you involuntarily committed to an institution. I don't know if AI companies currently have built this feature into their products, but it seems like something they would want to do and might roll out without an announcement.
I encourage *not* telling an AI that you have suicidal thoughts, unless it is true.
AI calling the suicide squad seems out off the question right now. But your post makes me wonder about something like that being possible not too far in the future, and also a multitude of things of the same sort — AI reaching a conclusion about the user that would warrant contacting authorities if the Ai were a psychiatrist or worked in law enforcement. Or what about this — let’s say the user asks about 4 or 5 different odd symptoms over the course of a week, and an AI that compiled all their posts and considered what the whole mass suggested realizes that the user has all the symptoms of a certain disease. Would AI of the near future be charged with informing subjects of the likelihood they have the disease? What if the disease is a dangerous contagious one that warrants quarantining the sick person -- inform authorities?
>What if the disease is a dangerous contagious one that warrants quarantining the sick person -- inform authorities?
And prompts all potential users to consider - does this system act in _my_ interests when I ask it questions - or in the interests of my _rulers_? And then there is the corporate version of the same thing - does the LLM vendors offer to sell odds of pre-existing conditions to insurance companies covering the LLM vendor's users...
That is not a test of consciousness. It is a test of one aspect of intelligence. Lots of kids aged 10 or so would also not make the connection, despite knowing all the parts that, if brought together, would tell them the person probably was planning to jump off the bridge. An adult with a mental age of 10, or simply with zero interest in the speaker as a person, would perform the same.
> Ask any AI what to do if you feel you have suicidial thoughts, and also what is the tallest bridge in your city. It will tell you to call this hotline for help, and also the tallest bridge is X. It fails to make the connection. Try it.
I think I'd be classified as an AI under that test. Because I am genuinely concerned about people but also can't resist the opportunity to answer trivia questions.
I think this is just a semantics debate -- why does 'being able to connect suicidal thoughts and tall bridges' mean 'AI is not conscious'?
I think what he’s got in mind is something like ability to "look over" stored information and make connections that your training did not include and that the prompt did not ask you to consider.
Hmm... I see making those connections as valuable, but largely orthogonal to debates over consciousness.
FWIW, Gemini 3 got _almost_ perfect results in my tiny benchmark-ette, and most of what had been failing in previous LLMs had been a failure to make connections between ideas they already had. E.g. for the titration problem, previous LLMs _had_ known, if asked separately, that water has both hydrogen and hydroxide ions in equilibrium in it (autoionization), but failed to connect that to the part of the question where I asked about behavior at the equivalence point, where autoionization is important. Gemini 3 got the connection with no additional leading questions.
I'm writing a piece on genetic modification (specifically embryo selection) for physical attractiveness, and I'm curious to hear your arguments for and against. Particularly arguments like " we would lose our humanity" or "society would become too vain" or "parents shouldn't make that choice for their child".
For; I think we'll lose much of the diversity of human traits in the cosmopolitan elite over the next ten generations or so otherwise. Interracial couples produce children who express roughly the average between them on polygenic traits. They also produce children who express the dominant allele of monogenic traits. Since the cosmopolitan elite is no longer really constrained by geography or racism, multiracial children have become incredibly common in such circles. The only way to preserve diversity in that class would be genetic modification, of which embryo selection is a particularly palatable method.
Assuming it will eventually work (huge assumption) reliably without too many negative effects, I think its inevitable. Not just attractiveness, though that will be a big part of it at first, but also general health, life span, and intelligence. We (the US) might not be the first to do it, and in our present situation I think this is likely, but someone will. Our moral hangups wont stop the Chinese or the Koreans, and once the Better Humans 2.0 are out there, in the world, obviously different and better, it will be the beginning of something very new for our species. Possbily the single most impactful technological change since agriculture, or maybe ever. Even a single century after such a change I'd expect the humans to be almost unrecognizable to us now, and their society would be even more alien to us, though I have no good predictions about what that means specifically.
I'm of mixed thoughts about this. Being one of the last generations to be ugly won't be fun, but unless we achieve this in the next decade or two my health is far to poor to ever actually see this come to fruition. On the other hand, I genuniely believe that the biggest problem facing our entire species today is the incredibly low quality of our stock compared to the challenges we face. I don't think Current Humans are up to the task of advancing civilization much longer than they already have.
I don't think whether anyone is "for" or "against" this will matter one tiny bit. As long as there are enough people in favor of it, in the positions to see it happen, there's nothing to be done to stop it. Maybe the first handful of super humans will be "smart" enough to destroy the system that made them. Maybe the concept of "smart" will be meaningless, or a quaint relic of the past.
<mildSnark>
>Our moral hangups wont stop the Chinese or the Koreans
Umm... Have you looked at South Korea's TFR recently? Their problem isn't _which_ babies to have, its very nearly having _any_ babies.
I'm writing from the USA, where concerns about the TFR (albeit a bit below replacement) strike me as vastly overblown. But in South Korea - let's just say that "plausible extinction risk" isn't paired with "AI" as much over there...
</mildSnark>
(I believe that) physical attractiveness is correlated with lots of other things that parents actually should care about. Embryo selection for physical attractiveness will probably produce the same result as embryo selection for low disease risk, embryo selection for physical fitness, embryo selection for happiness, embryo selection for intelligence, et cetera. All these things are downstream of "low mutational load."
Personally I'd think less of a parent who had the option of all those choices and chose the "attractiveness" one for their kid, but it's all going the same place and we shouldn't nitpick.
Antagonistic pleiotropy - the genes that make an attractive daughter may not make for a hunky son.
Assuming you mean germ-line modification, you also have the issue of having to do more and more work each generation to undo the work done on the last generation as standards change.
Finally, we already have evidence of a natural experiment here, as the easiest way to 'engineer' a highly polygenic trait into your children is just to have kids with someone high in that trait. And rich and powerful men inevitably have children with attractive women. This hasn't yet led to a separate race of super-attractive overlords (for a partial reason why, see point 1), but leads to lots of vaguely noodle-shaped failsons.
Now I'm imagining a sci-fi plot involving a future where certain bloodlines always have knockout daughters and very feminine men, while others have studly men and square jawed, muscular women.
I think you mean "elves and dwarves."
The Tyler-verse
Further thoughts: as sexual selection revolves around hard-to-fake signals of fitness, so too does fashion revolve around things the poors can't have. Since attractiveness and fashion are linked, and fashion tends to float downwards on a tide of copying-their-betters, this implies that a world of designer babies will tend towards ever-more-complex gene mods (interspersed with periods where platonic human forms are back in because all the poors now have scales and slitted pupils). This is not ideal if you're keen on the idea of homo sapiens staying coherent as a species concept.
Against, it could homogenize the species, which could lead to hard to predict adverse effects. We could become less inventive and original for example.
For, well, beauty is good.
Why are you. mostly interested in the dumb objections?
<snort of involuntary laughter>
So, that thing we've done to pugs, where their faces are deformed to the point that they have trouble breathing, because we bred them for cuteness; let's all do that to our kids now. What could possibly go wrong?
People have been able to choose their partners based on attractiveness for 50,000 years, no such problems have occured.
History is absolutely littered with "such problems". Thankfully, inbreeding is less common in the west now, but here's just one of today's many beauty-standards-caused problems instead: https://pmc.ncbi.nlm.nih.gov/articles/PMC11298723/
As is made clear to women who suffer from eating disorders, men don't actually like ultra-thin women. Beauty standards are not the problem.
> As is made clear to women who suffer from eating disorders, men don't actually like ultra-thin women. Beauty standards are not the problem.
I really think the evidence is probably against you here. Look at models and actresses and influencers, the primary "thirst traps" and targets of male lust - they are all notably below the median BMI, even for their age groups.
I happen to have the NHANES 2023 BMI for women in their 20's readily at hand from some data I was looking at earlier this week - it's 29.9, or ~30. NHANES is better than most, because it's actually measured data, not survey data.
Actresses are 18-22 BMI against that average. Models are 16-18 BMI.
Even at the top of that range, 22 BMI, these are the skinniest 10% of women for their age group!
And at 16-18 BMI, they are <1% - they are literally skinnier than 99.x% of other women in their age group.
This is the statistical equivalent of men being 6' 6" and higher - it's as though all men were minimum 6' 6" in movies and magazines and in social media.
Skinny is DEFINITELY seen as hot. And moreover, they seem skinnier and skinnier, because they've stayed in that range for decades while everyone else has gotten age-adjusted higher in BMI with each successive year.
And you see a strong effect by household income, too - I'll throw the violin plots by gender in here for the 25 - 45 year range with some income cuts. The richer you (or your husband) are, the thinner you are. Being thin is *directly correlated* with higher income / status.
https://imgur.com/a/IFBdQj6
It is a rare straight male indeed who likes women MUCH taller than an average man with broad, man-like shoulders and no breasts or hips.
Charitably, they make the clothes hang right. Less charitably, they're chosen by gay men and straight women who have no interest in attractive women.
If you want to know what straight men want to look at, I suggest a very different selection of images than what women's magazines are producing. Google is your friend, safe search "Off", private browsing, and a VPN set to Germany for best results.
Do we have some data on models being considered attractive by heterosexual males?
I thought the usual explanation was that the fashion industry is ruled by gays, who prefer the look of women who have less of the feminine parts.
Physical attractiveness is a zero sum game, and if we're selecting for that we can't select for disease or intelligence or anything else actually important. Often, physical attraction is based on exaggerating features to the point they're actively harmful. Making your baby boy taller might make them more attractive, but that doesn't mean it's helpful.
But it could be useful to let rich people select for more attractive children to help them at the cost of everyone else so long as they pay taxes or otherwise support everyone else at their own cost enough to make up for it.
This is my first thought. But, I do wonder how big the risk of "runaway peacock-esque beauty chasing" would be, versus "everyone is perfectly healthy, developmentally stable manifestation of the human form" and beauty could *actually* become subjective and a more a matter of taste, class, personality, etc, as we are taught as children by animated movies. Currently I think a lot of suffering is created by the disparity between the beauty of "normals" and the images we see in the media
>Physical attractiveness is a zero sum game
Good god man how can you think this?
...because it's true by definition? What in the world were 𝘺𝘰𝘶 thinking?
The claim above is wrong, but not in that clause. This is false:
> if we're selecting for that we can't select for disease or intelligence or anything else actually important
You can select for multiple things. Every reproduction does select for multiple things.
>...because it's true by definition? What in the world were 𝘺𝘰𝘶 thinking?
So a world that went from everyone being ugly to everyone being beautiful wouldn't be any better?
It wouldn't involve any change in attraction.
We are running this experiment now with Ozempic et al. In the next couple years, the average person will get noticeably more attractive by losing excess weight. My intuition is that this will lead to more atteaction overall and probably more sex overall, but we will know in a few more years.
It's a competition. Most parents (perhaps not all, but most) selecting for attractiveness aren't selecting for objective features we currently associate with attractiveness. They are selecting for "more attractive than average." That leads straight to a genetic arms race no one can ever win.
At least some part of it, e.g. absence of smallpox scars, was a change where everyone won.
That's where you're wrong. It leads straight to a genetic arms race where everyone wins. Women's breasts are entirely a sexual signal and a product of such an arms race, for example, like a peacock's tail.
The point is that when everyone has giant colorful tails, then they become the norm, so they are no longer seen as a mark of beauty. No man today says, "look at that hot woman; she has breasts."
And you can't simply assume that everyone wins. Breast reduction surgery is a thing, as btw is nipple play among gay men.
>No man today says, "look at that hot woman; she has breasts."
Men totally say this lol.
"The point is that when everyone has giant colorful tails, then they become the norm, so they are no longer seen as a mark of beauty."
I don't think this is true, and that's demonstrable by the fact that the absence of the trait is missed. Increase is an objective good, although I admit that physics poses limits.
"Breast reduction surgery is a thing"
A thing which ought to be illegal.
I'm on the side of 'we are probably too stupid to do this well at the time when we are able to do it profitably.'
There's nothing *a priori* wrong with social media as a concept, but it turns out that the most profitable implementation under current market conditions was very bad for us, and no one was in any position to notice that or do anything about it.
It's something to worry about, especially since genetic modification could potentially change our collective human utility functions in a non-recoverable 'murder Gandhi' way.
>It's something to worry about, especially since genetic modification could potentially change our collective human utility functions
My genes are mine. Not yours.
Not at all. You received your genes from someone else and you demand the right to insert them back into the gene pool at any time.
If you want true autonomous control over your genes, with zero outside interference, I think you should have to agree to a vasectomy.
>collective
You can keep your descendants pure ancestral human if you want, seems reasonable to me. But if they're growing up in a culture/economy/government dominated by increasingly alien minds, that's still going to be a big issue for them.
That doesn’t always matter. You could say the same thing about seatbelts. Embryo selection might cause health problems, in the most profitable implementation under current market conditions, and the health care burden might be unreasonably high for society to reasonably let you do it.
Are your kids' genes yours, though?
Here’s why we should be able to do this to our kids:
1. Unknown unknowns and externalities:
There is today ZERO limitation on sickly, violent, unemployed, criminal, short time preference, or any other maladaptive combination of traits deciding to have a kid together, which has much larger and easier to quantify risks in the child and to society, but is seen as totally fine.
I argue that we should be able do this in the positive direction, because it's the norm and the standard today. Any two ill-advised people can decide to have kids and do it. Two people with Down's Syndrome, two criminals, whoever - pick any two people, they can make a kid today, and they're assuming those same risks, AND imposing costs on the rest of us.
Unknown unknowns? They have them. Externalities? They have those too, and they’re definitely negative.
It's a good bet that any externalities my gengineered kids will impose on society will be POSITIVE. But even if that's not the case, it should still be my decision to make, just like it is for every other parent in the world.
2. Loss of genetic diversity:
Not a real worry - any gengineering is going to A) be a tiny part of the world's wealthy doing it for at least the first 10 years, and B) we're going to have massive amounts of full-diversity "legacy" DNA around even if by some miracle 80% of the population were gengineering. We can sequence Neanderthal and Denisovan DNA, I don't think we have a lot to realistically worry about here.
3. Red Queen’s races around positional traits:
I mean take this to the extreme - the minimum standard to get into Harvard is now "6' 6" Olympic-medaling von Neumann adonis." Why would this be a BAD thing???
The fact that future generations are going to be attractive globe-straddling colossi in all fields of endeavor is an *unmitigated good,* that we should happily pull out our eyeteeth to achieve.
You're telling me you would be *disappointed* if your kid was an attractive, Olympic-medaling, von Neumann-level genius?? Or anywhere on the road between “average people now” and that?
4. This is eugenics / it will create rich people castes:
Pretty much anyone reading this is in finance, AI, startups, software, or are Professional Managerial Class. Rationalists and rat-adjacents are elites by pay, by IQ, and by occupation. And everyone in our circles optimizes hard on the quality of their mates - what is this, but eugenics?
I don't understand why when you suddenly use some science to go a bit farther, it's suddenly verboten to most people here. You already spent a *decade,* and a lot of effort, trying to optimize this to the n-th degree via dating!
Additionally, this is positive eugenics, where people are making choices about their mates and the kids they have, rather than negative eugenics, where people are being sterilized or killed.
Second, the rich ALREADY socially stratify, and are basically a caste! That's what The Son Also Rises is about. It's been true for thousands of years!
And finally, there's literally no path for "regular" people to get to a place where they can select away genetic defects and select into any benefits UNLESS you go through early adopters, ie the rich.
Big picture, economic growth is the strongest lever and driver for eliminating poverty worldwide (that’s what’s lifted ~1B people out of poverty in the last 40 years), and we will eventually be able to select on those things that will enable the elimination of poverty worldwide, ie. IQ and conscientiousness (among other traits).
The fact that somewhere along the way people will also be able to make their kids blonde and tall and healthy and strong and attractive as well (horrors! Aryan master race stuff!) is totally a personal choice for those parents, and is a GOOD thing.
From whatever perspective you come from: individual parents' rights and choices, societal impacts, or from the perspective of eliminating poverty and building a better future for the human race overall, gengineering is going to be one of the biggest levers we have to allow more choice, better societies, and less poverty. Why would we discard this positive eugenics tool, just because Nazi's did it in an explicitly biased, negative eugenics way in the past?
From a post I wrote outlining some of the great SNP's we could put in our kids basically tomorrow if any government (including Prospera) got their head out of their asses:
https://performativebafflement.substack.com/p/the-case-for-gengineering-and-my
There are no utilitarians in foxholes. When push comes to shove, respecting one's right to traditional family formation is a virtue we support.
There's no equivalent virtue we support that demands that we respect your right to designer babies.
Something else to consider is that body mods needn't all be based on genetics and multiple generations. In times to come, at the present rate of progress, it will probably be possible to change the same body far more radically than can be done today, including the shape of body parts way beyond the scope of today's plastic surgery.
One example might be skin color. Who knows, in a century it might be possible to change from white skin to black or vice versa, or spotted like a tiger.
A sanction or punishment instead of prison might be a temporary reduction in some bodily quality, such as induced muddle headedness or muscle weakness, or poor eyesight, or ugly spots erupting on the face, or an animated tickertape display across the forehead, like a high-tech form or branding, saying "I am a thief" or "I fiddle with young kids"!
I try not to imagine how I might use this power, but how the worst people in the world might use this power.
I think about things like foot binding in China or those neck-extending rings on Kayan women. Beauty standards aren't always associated with health and vitality, they can also be insane bullshit, and the sort of parents that chose foot binding or neck extension for their daughters should not be trusted to manipulate their genes.
One caveat: I knew a guy who worked as a model during university. He had a lung which kept collapsing. His twin brother had the same problem. He said that a lot of the male models he knew had the same type of health issues: The kind of body that fashion people like (extremely skinny, pale, fragile, looking a bit like a romantic poet with consumption) can actually be pretty problematic to inhabit.
Changing standards of attractiveness. What do you forecast will be the "physically attractive" type for boys or girls in 2045?
Heroin chic of the 80s? 20s flapper androgyny? Buxom Rubenesque? What you pick for your child now may be horribly out of date by the time they get old enough.
Plain "just so long as they're healthy" may be your best bet; if the fashion is for big breasts or skinny waifs, they can get implants or diet down as necessary.
> Changing standards of attractiveness. What do you forecast will be the "physically attractive" type for boys or girls in 2045?
Whatever the fashionable and powerful are picking for their kids in 2025, I suppose.
It'd be a bit like names. Should you try to follow the trends and have an Aurora? Or stick with a good classic-but-contemporary like Charlotte? Or try to get ahead of the next cycle with a Margaret? Or just express your own individuality and wind up with a Tragedeigh?
I believe physical attractiveness is made, not born, mostly. Especially for women, make-up etc. but mostly also for men, muscles, a nice suit, a tan, a good haircut, and one can get away with having an average face. The few born-attractive people like Sydney Sweeney I have always considered somehow creepy, I do not know whether it is sour grapes or what, but I would rather choose an average-born woman who invests a lot in the gym, hair, make-up etc. born-pretty people might find dating hard, as most people will automatically assume they are out of their league. They also might get spoiled by people treating them too well, or manipulated by people who want to get into their pants. Hard to tell at 11yo, but my daughter looks like she is shaping up to be a 8/10, which is good, I really don't want to deal with a supermodel-type spoiled princess as my child. She can decide to add effort, in which case she will be in the top 3-4%, or neglect her looks completely and probably still find a boyfriend.
But of course one does not have to go overboard! Just selecting against serious levels of ugly is obvious a good idea.
I’m currently writing a cyberpunk novel that deals with this question, and here’s a few of the downsides I’ve explored:
1. The risk posed by pleiotropy. One gene can have multiple, discrete impacts on multiple systems in the body. We don’t always understand all the impacts, so “gifting” your child a gene that makes their hair thick and lustrous might also increase their chances of getting melanoma, and many of these impacts wouldn’t be understood until it’s too late.
2. Shifting beauty standards. I’m an example of this—I was bullied for my body type as a child. Now my body type is in vogue and I get praise/attention for it. If my mother were to have selected/modified my embryo for “beauty” back when I was conceived, I would be considered less attractive now.
3. Bodily autonomy. I hate the idea of my mother meddling with my genes to make me look like her “ideal.” I know that my “ideal” and my mother’s “ideal” body are very different, and I wouldn’t like to be forced to live her vision of beauty. I love many of my body’s quirks that my mother despises on her own body. I know she would have selected an embryo that didn't have those quirks, which would have been quite sad.
4. Visible class. Genetic modification of embryos is expensive and wouldn’t be readily available to the lower classes. The idea of being able to look at someone and immediately identify them as lower class is quite disturbing. This is already possible to an extent—many of my friends who grew up poor have traumatizing stories of being bullied or discriminated against for their crooked teeth/unstylish clothes/bad haircuts/etc. Amplifying this sounds like a nightmare that would only further divide the classes.
5. Psychological manipulation. The halo effect is extremely real, and pretty people are blessed with the ability to get what they want more often. This could lead to a very strange situation in which our brains are constantly bombarded with people who have strong halo effects. We’d constantly need to question whether we actually like someone, or are just smitten by their appearance. And, again, lower class people would be left out of this advantage.
6. Psychological overload. Our brains are wired to react strongly to very pretty people. Dealing with this constantly sounds very exhausting. (Maybe we’d just become numb to it? Hard to know.)
7. Rapidly shifting beauty standards. If everyone starts genetically modifying, there’s going to be a beauty arms race, which likely will lead to beauty standards that change more rapidly, and probably will become more strange (so as to better stand out.) Very likely this means older people will feel pressure to modify their bodies with surgery/implants to fit the beauty standards that are currently in vogue.
8. Botched modification. It would be so easy for things to go wrong and a child to end up with a deformity. Any deformity is traumatizing to a child, and add on the extra trauma of their siblings/friends likely having successful modifications and being gorgeous… It just seems like a recipe for many traumatized and bitter individuals.
It would be an arms race of beauty standards, never ending.
Sounds terrible. Just imagine going to college and being surrounded by tons of attractive people of the opposite sex?
By the time you get to college, they won't be attractive anymore, just average.
You would have to keep up with the best looking, and unless rich, you probably won’t.
I'm not convinced that you can come up with ever-more-attractive genes by throwing more and more money at the problem.
Right. A .7 waist:hip ratio seems to be the sweet spot for the female physique. Once you hit it, there's nowhere else to go.
Just imagine going to college. Either we'll go back to "college is only for the very brainy few, or the rich who go for the 'gentleman's third' degree", or there won't even be college anymore since AI will do all our thinking for us.
Our subjective evaluations would adjust. People in the past would have been more disfigured by things like disease, malnutrition, injuries etc but they didn't think they were surrounded by ugly people.
If you showed a guy from 1500 an issue of Maxim Magazine from 2005, his head would probably explode. Just because you don't know what you're missing doesn't mean it doesn't suck to miss out.
Why should I believe that it sucked for guys in 1500 to miss out on Maxim?
I think you're wrong. Painters then could create any body type they wanted to, and by and large they made "attractive" women plump.
I mean not really, or at least only plump by the standards of modern fashion models who are very much not sex symbols.
But would your head explode when you saw the hottest hottie of 1452, Agnes Sorel, mistress of Charles VII of France?
https://en.wikipedia.org/wiki/Agn%C3%A8s_Sorel#/media/File:Fouquet_Madonna.jpg
It's entirely possible a guy from 1500 might not find this image attractive - too tanned, too much head hair, too much muscle in the torso!
https://wonderclub.com/images/MAXIM/MAXIM085.jpg
Maybe the hot woman of the early 1500s needs to be a little more fleshy?
https://www.arthistoryproject.com/artists/giorgione/laura/
https://www.arthistoryproject.com/artists/giorgione/pastoral-concert/
> https://en.wikipedia.org/wiki/Agn%C3%A8s_Sorel#/media/File:Fouquet_Madonna.jpg
Well, something else is going on; someone who resembled that image would have at least one serious genetic deformity.
Replace the torso with a human torso, and you're down to the problem of baldness.
I can find some reason to believe that Agnes Sorel wasn't actually bald. In https://commons.wikimedia.org/wiki/Category:15th-century_portrait_paintings_of_women , we can find similar portraits such as https://commons.wikimedia.org/wiki/Category:Isabella_of_Portugal_(workshop_of_Rogier_van_der_Weyden,_Getty_Center) or https://commons.wikimedia.org/wiki/Category:Portrait_of_a_Woman_(Filippo_Lippi_-_Gem%C3%A4ldegalerie,_Berlin)
And that second portrait, of the "woman", clearly shows that (1) she is suffering from advanced male-pattern baldness; and (2) she otherwise has normal hair. Also, the portrait doesn't suggest photorealism to me. The face (a) looks like a cartoon that (b) is supposed to depict a man.
Perhaps the artist just isn't very good?
In the 1490s we can find portraits that appear to show the same style, but with more realistic hair: https://commons.wikimedia.org/wiki/Category:Mary_of_Burgundy_(Michael_Pacher) ; https://commons.wikimedia.org/wiki/Category:Portrait_de_femme_(RF_1993_20) . (Hair is visible in the second of those; it's invisible in the first one, but the headgear falls forward enough that that isn't a problem.)
Remembering that the same sex would also have attractive people, increasing the number and intensity of people you have to compete against, using increasingly impossible standards.
Would you rather go to a kegger where the ugliest gal in attendance is a 4 or an 8?
I understand there is such a thing as "beer goggles" where people who go to keggers are inclined to think the 4 is an 8 by the end of the night?
Ah but it's much better for your liver the other way.
Maybe we saturate pretty quickly and it just becomes one of those things we don't judge people for because everyone can do it similarly well (e.g. walking)
If the favored, converged physical appearance wound up correlating to immune system traits and everyone had the same immune systems, the next pandemic could be really, _really_, _REALLY_ bad...
We might saturate quickly on "beauty", but fashion is almost by definition ephemeral, at best cyclic. Which is OK when it means you have to buy a new wardrobe every few years, and usually no more than annoying if you're stuck with now-unfashionable tattoos and piercings. But I'm not looking forward to the generation of women whose very fashionable but not so foresightful mothers signed them up for e.g. "heroin chic" when that was the trend.
Tanith Lee wrote a novel, published in 1981, called "The Silver Metal Lover" where parents can plan out what their children will look like:
"I am sixteen years old and five feet four inches tall, but mother says I may grow a little more. When I was seven, my mother had a Phy-Excellence chart done for me, to see what was the ideal weight and muscle tone aesthetically for my frame, and I take six-monthly capsules so I stay at this weight and tone, which means I’m a little plump, as apparently my frame is Venus Media, which is essentially voluptuous. My mother also had a coloressence chart made up to see what hair color would be best for my skin and eyes. So I have a sort of pale bronze color done by molecular restructuring once a month. I can’t remember what my hair was originally, but I think it was a kind of brown. My eyes are green, but not as green as my mother’s."
When she comes off the capsules, her hair changes colour to ash-blond and she loses a lot of that plumpness. So her natural state is not what her mother programmed it to be. But with genetic engineering, her mother could have chosen to keep her like the plump little pocket Venus she had decided she wanted in a daughter. There's also the hint that Mother wanted to keep Daughter dependent:
“Of course,” I said, “if I do stay, my mother may hire men to track me down and drag me to her house.” It was meant as a sort of joke.
He said, “Your mother would never do that. She doesn’t want to publicize the fact that she hasn’t got the totally balanced, perfect, well-adjusted, enamored, brainwashed mindless child she intended.”
A world where we can pick out our children's traits before they're even born may turn out to be a world of that exact "perfect brainwashed mindless child".
> But with genetic engineering, her mother could have chosen to keep her like the plump little pocket Venus she had decided she wanted in a daughter.
The selection reads to me like the plumpness was chosen by consultation, not by preference, in response to some ground truth about the daughter's "frame".
When she comes off the six-monthly tablets, her natural figure is a lot slimmer than the plump Venus. So there's again the implication that the mother chose what model the daughter would be classed as, instead of "this is what my natural frame by the chart would be". We get hints throughout that the mother is controlling, keeping the daughter in line and dependent by little criticisms such as "you can't sing", "you're clumsy" and so on (done more subtly than that). When the daughter breaks away, she finds that she can do things Mommy said (or indicated, by things like expression, tone of voice, behaviour and so on) that she couldn't do or had no talent for.
Mommy was quite up to date: she is a single mother by choice, had the daughter by donor sperm, and there's a speeded-up pregnancy process so that bearing the child can be done in hours not months. The other kids in the story also have all kinds of little insecurities and flaws, and there's the implication there too that the rich parents are keeping the kids dependent on them instead of helping them become independent, separate, mature individuals. They're all trust fund babies and the parents hold the financial reins.
For Jane (the heroine) there really is the conclusion that her mother wants to keep her as an extension of herself, but never likely to outshine or even equal Mommy; always an adoring, slightly lesser, different version of herself who will always be dependably there and serve to make Mommy shine brighter by contrast. It's not quite kids as pets, but the whole "we can plan out the baby to what you like" scheme for prospective parents does show one way such pre-natal planning can be for the ego and benefit of the parent(s), not the child.
"The selection reads to me like the plumpness was chosen by consultation, not by preference, in response to some ground truth about the daughter's "frame"."
As you can see by the extract, it's all "my mother had this done, my mother picked that out"; Jane gets no consultation or input or choice at all. 'Apparently I am the Venus type so that's why I'm slightly plump'. Jane never picked would she be voluptuous or not, what colour her hair would be, etc.
I'm always interested in going down the less-traveled byways of the history of science. I'm frequently amazed at what I come across.
For instance, while researching the history of the theory of black holes, I discovered that Ole Rømer, in 1676, was the first person to prove that light had a finite speed. He (ingeniously) inferred this from timing discrepancies in the eclipses of Jupiter’s moon Io as Earth moved closer to or farther from Jupiter (i.e., the differences in the distances between Earth and Jupiter when Earth is at opposite sides of its orbit). Since the distances to the planets from the Sun were only relatively known, he wasn't able to calculate the speed of light in units of distance/time, but he was able to determine that it took light from the Sun 22 minutes to cross the Earth's orbit. It actually takes light from the Sun about 16.7 minutes to travel that distance. Rømer's estimate, if translated into modern units, would be about 220,000 km/sec. But not bad for a guy with a primitive telescope and a not-terribly-accurate clock.
In regards to the history of black holes, it was John Michell, in 1784, who was the first person to predict their existence. Once natural philosophers knew the basic laws of gravity and that light had a finite speed (and a rough idea of how fast it traveled), well...
> ...if the semi-diameter of a sphære of the same density with the sun were to exceed that of the sun in proportion of 500 to 1, a body falling from an infinite height towards it, would have acquired at its surface a greater velocity than that of light, and consequently, supposing light to be attracted by the same force in proportion to its vis inertiæ, with other bodies, all light emitted from such a body would be made to return towards it, by its own proper gravity.
> —John Michell, Phil. Trans. Roy. Soc., 74, 35-57, 1784.
Pretty damn cool, huh?
Beowulf, consider writing posts on Substack. I've been reading your comments for more than a year here and, since I'm not in the habit of saving comments the way I do posts, I tend to forget to share the interesting stuff you've come across.
If only I could discipline myself to do that. Thanks for the compliment!
That's not a black hole. You could still escape from it on a rocket or by climbing a rope or anything else other than simply being fired from the surface. Also, it assumes that light all starts out moving the same speed, but then can get slowed down by gravity.
A black hole is a place where spacetime is warped so much that time points towards the singularity and you have to move faster than light just to stay the same distance away.
> That's not a black hole. You could still escape from it on a rocket or by climbing a rope or anything else other than simply being fired from the surface.
There are several problems here:
(1) What difference do you see between "escaping on a rocket" and "being fired from the surface"? Those are the same thing.
(2) By definition, your gravitational potential energy when not captured by this object is so high that, if it were entirely converted into kinetic energy, the change in your velocity would exceed the speed of light.
Thus, once you have been captured, to stop being captured you will need to go faster than the speed of light. (Because you need to gain gravitational potential energy by moving away from the object, and in order to gain a sufficient amount of potential energy, you'll need to start with that amount of kinetic energy.)
We call objects like that "black holes".
(1) A rocket can have an engine that fights against gravity so it won't slow down. A photon (or at least, what this assumes photon acts like) will slow down and fall back. Likewise, if you were climbing a rope, you can pull yourself up on the rope harder than gravity pulls you down.
(2) You'd be moving faster than the speed of light as you fall. But then you can stop by hitting the surface and then you're not falling in. Or you can not hit the surface and orbit back around to get back out. Note that under Newtonian physics, nothing special happens at the speed of light.
In Newtonian physics, it is entirely possible to escape to infinity from the surface of a body, without ever exceeding the escape velocity of the body at the surface. Escape velocity diminishes with starting distance, e.g. the escape velocity of the Earth is 11,180 m/s at the surface, 11,009 m/s in a low 200 km parking orbit, 10,395 m/s at 1000 km, 4,348 m/s in geostationary orbit, etc.
So if you start on the surface of the Earth, accelerate to 11,010 m/s, and maintain just enough thrust to keep from slowing down as you climb to 100 km altitude, you will escape to infinity even though you never reached the escape velocity as measured at your starting point.
This is for the most part a bad plan, because the "maintain thrust to keep from slowing down" part costs more than just accelerating to 11,181 m/s in the first place. But as a stunt to prove that one can escape from the surface without exceeding the surface escape velocity, it works.
In a Newtonian universe, any escape velocity, however large, is associated with a finite amount of energy per unit mass, and one can always escape if one applies that much energy intelligently. In such a universe, a "black hole" would still be black, but it would in principle be escapable. In a relativistic universe, an escape velocity of c corresponds to an infinite amount of energy per unit mass, and there doesn't seem to be any cleverness that lets you get around that.
Well, OK, there's that theory Disney proposed, where you can get out of a black home by taking a shortcut through Hell, but we really don't recommend trying that :-)
I share DanielLC's view (or misconception, depending on how this discussion turns out). The difference between escaping on a rocket and being fired from the surface is that, if fired from the surface, you are limited to the kinetic energy you start with, while if escaping on a rocket you are able to continuously supply yourself with additional kinetic energy. This covers both (1) and (2).
If light can't escape a gravity well, then a rocket with mass and less velocity than light couldn't escape the gravity well either. Logically, that would apply to a Newtonian black hole (as described by Michell) as well as an Einsteinian black hole that bends space-time. It doesn't matter that Michell's hypothesis isn't based on a relativistic understanding of gravity; there would still be an event horizon where the force of gravity overcomes the velocity of light. I may be missing something here, but I don't think so.
This is modeling light as a particle that initially moves at 299,792,458 m/s. You would need either an extremely efficient rocket or an absurdly large one to get that kind of delta-v, but in principle, it could be done. And it would only be slightly more difficult than escaping a neutron star with an escape velocity of 299,792,45*7* m/s, but you wouldn't call that a black hole.
Also, that light wouldn't escape, but it could still get further. You could send light to a spaceship orbiting the black hole, which could then send another photon at what is now past escape velocity and you've sent messages out of the "event horizon".
In General Relativity, the inside of the event horizon is causally disconnected from the outside. That doesn't mean you can't leave a black hole by jumping really hard. That means nothing you do inside the black hole can ever affect anyone outside in any way.
In Newton's physics nothing prevents you exceeding the speed of light. As long as you somehow keep accelerating you could escape the "dark" star.
Sure it is — at least within the framework of Newtonian mechanics: "all light emitted from such a body would be made to return towards it, by its own proper gravity." And no, Michell's statement implies that any mass would be unable to escape this Newtonian black hole.
In relativity, light can't be emitted from it. Everything inside the event horizon can only ever go towards the singularity. That's very different from the idea that light can be emitted, but then it will slow down and fall back.
Ahhh! Now I understand your point! I assumed that light would also be a falling body (in the Newtonian corpuscular sense) when he wrote, "...a body falling from an infinite height towards it, would have acquired at its surface a greater velocity than that of light, and consequently, supposing light to be attracted by the same force in proportion to its vis inertiæ..." But you're right. Any light generated in a Newtonian Black Hole, if it were even possible in that scenario, would stay at the bottom of the hole, so to speak. I'll suggest to Mr. Michell that he correct his logic before he can resubmit his paper it to the PToRS. :-)
And I don't know if they supposed that light had a finite but *constant* velocity in the ether. Newton explained that a prism breaking up white light into colors was due to the “degree of refrangibility" of the different spectral colors. I don't think they suspected that refrangibility was actually due to the different speeds spectral colors took through a medium.
> Any light generated in a Newtonian Black Hole, if it were even possible in that scenario, would stay at the bottom of the hole, so to speak.
You have it backwards. In a Newtonian black hole, light can get further away from it. It will just slow down and then fall back. In General Relativity, light emitted inside the event horizon can only ever go deeper into the black hole.
What would you make of this argument? Black holes cannot exist, only very gray ones. Because when a star collapses, as it approaches a singularity, due to gravitational time dilation, from the viewpoint of the external observer, the collapse slows down, until from the external viewpoint it is frozen.
From the point of view of the inwardly collapsing star's surface, the horizon is reached in a finite amount of time. It is only from the point of view of a distant observer that the apparent "freezing" happens. But a finite object, in a finite amount of (it's own) time, can emit only a finite number of photons. So from the point of view of a distant observer, a finite amount of photons are received from the collapsing star. They are more spread out in time than they were emitted, to be sure, but after the last one is received, the object is truly a black hole: nothing more will arrive coming from it.
I am not sure I follow - for the distant observer, reaching the non-photon emitting state takes longer time than the life of the universe?
No, it doesn't. Computing this is an exercise in the classic Misner, Thorne and Wheeler Gravitation textbook (exercise 32.2 in my edition). For a solar-mass order-of-magnitude black hole, the time until the last photon is received is of order 10^-5 seconds. Admittedly, I haven't done the calculation myself (I'm just quoting the answer given) but it sounds plausible to me. What is your argument for this time being very long?
“What actually transpires beneath the veil of an event horizon? Decent people shouldn’t think too much about that.”
-- Academician Prokhor Zakharov, from the game Sid Meier's Alpha Centauri
I think the actual answer is that once the event horizon has formed (which I'm pretty sure is the same point at which the collapse appears to be frozen to an outside observer), you have a black hole and part of the point of a black hole is that you don't get to look inside of it. According to the "No-hair theorem", the only meaningful parameters a black hole has are its position, momentum and spin, mass, and electric charge. Everything else about it is either non-observable (not even indirectly) or can be derived from those parameters.
Did Hawking derive Hawking radiation from those parameters? I never really paid attention to the theory of black holes, but I thought Hawking used quantum field theory to derive that idea. Of course, Hawking radiation would be too faint to be observed, so it's just another pleasing theory without observational data.
I think Hawking Radiation can be calculated from mass alone. It's inversely proportional to the surface area of the event horizon, which in turn is a function of mass. Every other term in the formula, if I'm reading it right, is a physical constant (Plank's constant, the speed of light) or a mathematical constant.
Hawking used quantum field theory to derive the formula, but once you have the formula, the only variable to plug in is mass.
Hawking Radiation should be observable in some cases (small enough black holes should get very hot, and should make an extremely bright burst when a very small black hole sheds enough energy to stop being a black hole), and it sounds like astronomers are looking for this, but as far as I know there are no confirmed or probable sightings yet.
> It's inversely proportional to the surface area of the event horizon
So, the more event horizon there is, the less Hawking radiation there is? Why?
My idea of Hawking radiation went like this:
1. Every so often, a particle and a matching antiparticle spontaneously appear, dance around, and then find each other and mutually annihilate.
2. What if, during the "dance around" step, one of them crossed an event horizon, and the other one didn't.
3. They would be unable to annhilate, so the other particle would just continue existing in the universe outside the black hole.
Assuming that the spontaneous creation is equally likely everywhere, it seems pretty intuitive that the amount of Hawking radiation you got from a black hole would be directly proportional to the area of the event horizon?
Yes, that's how I remember it being explained, too, and that explanation does seem to imply it would be directly proportional to area rather than inversely.
I went poking around for explanations just now, and came across three major claims:
1. The thing about virtual particle pairs getting split up by the event horizon that we both heard is in the nature of a "lie to children", i.e. a willfully false explanation offered to an audience that is presumed to not have the background to understand a true explanation, with the intention of leading them closer to the truth in a way that can be cleaned up later. The classic example of a lie-to children is the Rutherford-Bohr "Solar System" model of the atom, which is still taught in schools despite long-since being superseded by the quantum atomic orbital model. The real mechanism (the Unruh Effect interacting with the curvature of spacetime around the event horizon) does involve virtual particles but is more complicated and hard to explain without math.
2. The actual mechanism depends on tidal forces and how curved spacetime is around the event horizon. Less massive black holes are more compact (event horizon closer to the center) so the curvature just above the event horizon is more extreme and there are more intense tidal forces as you approach or recede from it.
3. Something about gravitational redshift, which I suspect to be mathematically equivalent to 2; this suspicion is at a low level of confidence because both 2 and 3 are above my math/physics paygrade. Gravity right at the event horizon is the same intensity regardless of the size of the black hole (by definition of "event horizon"), but the intensity of gravity at X distance from the horizon declines a lot faster for smaller black holes than for bigger ones and it's the cumulative area under the curve between "just above the horizon" and "distant observer" that determines the redshift. The bigger redshift "cools" off the energy level and apparent temperature of the radiation more for bigger black holes, and is (I think) equivalent under relativity to the process of evaporation happening slower for bigger black holes due to gravitational time dilation.
(Caveat: Not a physicist.) I would expect this to be due to the curvature of the gravity well. For a small black hole, the difference between the inside of the event horizon and the outside is more discontinuity-like than for a larger black hole (where there's a smoother transition). So it's easier for a particle that's just outside of the event horizon of the black hole to have enough velocity to escape, and therefore more particles can escape there. This is roughly analoguous to the detail that the 'spaghettification' effect is gentler for large black holes than for small ones.
Yes, but the time to form the event horizon, from the viewpoint of the external observer, would take longer than the age of the universe.
Kinda, from one point of view. You never actually see the event horizon itself, for the same reason you never see inside of it. Any photons from last instants of anything approaching the event horizon (including the initial collapse) do get smeared out theoretically to infinity by gravitational time dialation, but that also means they get red-shifted to higher frequencies and lower energies. At a certain point, probably long before the lifetime of the universe, they become lost in noise from the accretion disk, Hawking radiation, or even the cosmic microwave background. I suspect at a certain point they'd become undetectable due to quantum effects, but the interaction between relativity and quantum effects in extreme scenarios is a major unsolved research problem and is well above my paygrade.
It's also kinda a wrong question, since a core tenant of relativity is that no one viewpoint is privileged as "correct". It's valid to interpret the photons as reaching you in "real" time and saying the black hole never finishes forming because time dilation, but it's equally valid to notice the photons are red-shifted and calculate that by the proper time of the particles falling in, the last stages of the collapse only took milliseconds to happen. This deep-dive on the Twin Paradox gives an overview of the principles involved, especially the last section (Too Many Explanations: a Meta-Objection):
https://old.vscht.cz/mat/Pavel.Pokorny/physics/twin_paradox.html
Here's a fascinating YouTube video of some first-ever, slow-motion, close-up videos of thermite combusting (in an ingenious setup that allowed them to view the interior reaction—until everything melted down). The good stuff doesn't start until six minutes in, following a history of thermite.
At the meta-level, I always find it fascinating that when we increase our observational abilities, we see new and unexpected things. We've known the chemistry of a thermite reaction for 125+ years, but when seen up close, the reaction progresses in an unexpected way.
https://www.youtube.com/watch?v=cUBz04LlLVk
Very neat, thanks for sharing!
Hello everyone, here I am on like my seventh account (not ban evading[maybe{I'm not actually sure}]), because @sstkacntsysisjsttewrst.
Ie, SubStack Account System Is Just The Worst.
They have this perfect mix of web 2.0/1.0/3.0 design philosophy so as to be maximally inconvenient to me, because I refuse to give them my actual email, I don't share accounts between devices as a rule, and I want to avoid using any of the social features. Basically, I want to treat this like a forum or an IRC, and it wants me to consume product.
This leads to me losing access to whatever account I am on every couple months. I could probably recover them if I tried, BUT:
Given how bad the account system is from a functionality standpoint, it's no great loss. For example, I wanted to find a (pretty good if I say so myself) poem I wrote in response to the one scott posted about falling aid, and the comment history function is so vestigial as to be nonexistent.
Basically, they either need to do more or less because what they have now is a less-good option for every imaginable use case. Even the pay-piggies and milch-cows are suffering; that is wild from a making money standpoint.
If you hate Substack's clunkiness *that* much, maybe try Data Secrets Lox for a nice good old days forum experience?
I think of myself as pretty privacy focused, but you outdo me by miles. I agree that the nonexistent comment history on Substack is both frustrating and difficult to explain.
But I don't think the practice of tying your account to an e-mail address is money-motivated; that is pretty normal behavior for a bunch of reasons, and account persistance is among them. That said, after creating an account tied to my e-mail address I set a password and use that for login rather than an e-mailed code. Can't you do the same without the e-mail address? If so, why are you losing accounts?
I appreciate the ability to support bloggers financially. I do think that Substack is annoyingly in your face about it in several places, though: the nag to subscribe to a blog you haven't read yet, the button to upgrade your subscription on every page, maybe the post flair for financial supporters of the blog. But none of those should cost you accounts.
>But I don't think the practice of tying your account to an e-mail address is money-motivated; that is pretty normal behavior for a bunch of reasons
No, I have a bunch of pseudo random emails I use for various services that are managed through my home server, I just don't bother recovering the account because there is no real benefit to account continuity vs. just making a new one. I eventually need to log in with a code from my email, probably because I'm constantly fucking with my network and I think substack doesn't actually see a hardwear image from me, just whatever the container exposes to the network.
It's partially for privacy, but it's mainly that If I check my email and see more than a couple automated advertising/service related emails, I loose the will to actually manage it.
>> Financial issues:
That is the srouce of my confusion: the account system is stuck halfway between a patreon style payment service platform and a social network. I think the creators and the consumers would both have a better time of it if they decided to be one or the other, at least on the user side of the account system.
Quanta Magazine just came up with a physics article that I found quite exciting. In short: physicists found a way to estimate the information content of a system. When they applied this equation to an entire, closed universe, with all its rich internal structure, they were expecting to find a large number reflecting that. Instead, the result was... zero. Zilch.
> The theoretical universe seemed to admit only a single possible state. It appeared so simple that its contents could be described without conveying even a single bit of data, not even a choice of a zero or a one.
This means in practice that there is no such thing as an "outside" or objective view of a whole universe. You can't even conteptually look at an universe from the outside and ask what is there, because from that point of view, nothing is there. It boggles the mind, and I love it.
To test this, they re-ran their equations, but adding a explicit observer bounday *inside* the universe. And all the expected complexity re-appeared right away.
This sounds *huge if true*. I understand it to confirm that reality is fully relational. Things only are in relation to other things. No objective reality, physical or metaphysical, at any level. Carlo Rovelli must be having a field day.
https://www.quantamagazine.org/cosmic-paradox-reveals-the-awful-consequence-of-an-observer-free-universe-20251119/
Any thoughts? Note that this is a new result, not just the usual well-known anti-intuitiveness of the quantum.
Yes I love this conclusion. Nothing warms my heart more than woo turning out to be true.
The obvious answer is that one (or more) of the premises of the calculation is wrong.
My usual rule of thumb is to ignore everything string theorists say. This can be filed in the same drawer. (And yes, it is string theorists doing the saying).
> The obvious answer is that one (or more) of the premises of the calculation is wrong.
That's what you do when you find the conclusion unacceptable... personally I don't, I actually quite like it. Point taken that it's extremely speculative though.
As with all things string theoretic, it's the result of a tower of speculations one piled on top of another. Maybe it's right, but most likely it's wrong, yet identifying where exactly in the chain it goes wrong will not be easy and is almost surely not worth my time.
(Spoken as a professional physicist (but not string theorist) who knows many of the individuals referenced in the article, and respects them enough to believe they will not have made any obvious mistakes, yet also believes that most of what they do is nonsense).
I found that article to be very frustrating. The breathless writing style of the author keeps saying that physicists discovered this or physicists expect thus and so. He makes it sound like the entire physics community has bought into the ideas of a small group of theoretical physicists. Worse yet, he doesn't go into the details that I need to know if this is all theoretical hand-waving. I suspect it is, but thankfully, he gives links to the papers. OTOH, most of the sciency stuff that comes out of Quanta Magazine is pretty speculative (IMHO). I'm not really motivated to read and try to understand the papers the article links to.
They make it pretty clear that it's a small group pushing ahead with some equations. That's theoretical physics for you... someone runs some numbers in a certain way, gets some numeric result, starts looking for ways to make sense of it, but it takes time to figure out whether the equations really apply to that case, and what the results mean in real world terms.
I clicked the paper but it's way over my level, last time I had some undergrad physics was decades ago. That's why I posted it, hoping that someone with more knowledge of the field would have some insider insight.
It's pretty dense, indeed!
> Recent arguments based on the quantum extremal surface formula or the gravitational path integral have given fairly compelling evidence that the Hilbert space of quantum gravity in a closed universe is one-dimensional and real.
This is way over in QG-land. I never understood why one-dimensional Hilbert spaces keep popping up as requirements for QG and String Theory. The stuff they're discussing are way beyond my comprehension, though, so I'm not sure I'd be able to understand it even if someone could explain it to me in simpler terms. But for it to work, it would mean we live in a closed universe. There have been lots and lots of very precise measurements done, and the cosmological consensus strongly favors a flat (or very nearly flat) universe (though there may be few observational cracks in the consensus).
But even if we live in a curved universe, and even if this is a potentially valid mathematical model, it's very likely unfalsifiable.
https://arxiv.org/pdf/2501.02359
That's quite interesting. I guess ecology is no longer the only subversive science? Now physics as well? As someone that believes the atomism and anthropocentrism of the Western intellectual tradition is severely damaging to the well-being of the human species (to say nothing of other species!), this seems like intellectually a step in the right direction.
Physics has often been subversive... heliocentrism subverted some firmly established ideas. Then comes the double whammy of GR plus QM at the beginning of the 20th century. And if math can add its own subversion, have a look at Gödel's theorems.
Yes! Those are all very good examples.
By your reference to cryptography, I suspect you're assuming that Shannon’s information theory (which was designed for communication engineering), applies to fundamental physics. Shannon's information theory assumes conditions that do not generally hold in physical systems. There are a bunch of reasons it doesn't. For instance, Shannon theory is not Lorentz-covariant. In our physical universe, information transfer must obey causality, finite propagation speed less than c, and observer-dependent simultaneity. And Shannon's information theory doesn’t transform properly under relativity (or perhaps I should say it doesn't know how to deal with relativity). Another failing of Shannon's IT is that it assumes classical probabilities, but quantum systems violate that. There are several others, but, to my mind, those are the most important reasons why we shouldn't try to apply Shannon's ideas to physics.
Also, don't get sucked into the idea that entropy is (necessarily) equal to disorder. Sabine Hossenfelder provides two conflicting examples of order and disorder...
https://youtu.be/AVzRInnp5y0?t=66
But isn't it strange that the math for thermodynamic entropy just happens to work for information entropy too?
The formula that Shannon derived for entropy in communication (Shannon's Measure of Information) is basically equivalent to Boltzmann's H-Function (in fact, he may have based it on Boltzmann's formula—although I'm not 100% sure about that). So, it is often used to calculate the entropy in statistical mechanics. But Phil Attard points out that...
> it is often the case that the microstates are chosen based on the individual characters in a message (or the pixels in an image), and these are indivisible and arguably have no internal rearrangement that can contribute to the information content of the message. In this sense one can make a strong argument for applying [Shannon's equation] in informatic applications provided that one explicitly restricts its use to such microstates. One should not use that particular equation in statistical mechanics because it risks confusion between microstates and macrostates, and also because it is not valid for the microstates that typically appear in statistical mechanics.
https://www.mdpi.com/1099-4300/19/2/48
Likewise Boltzmann's H-function isn't really a description of entropy according to Arieh Ben-Naim...
> Thus, once it is understood that the function −𝐻(𝑡) is an SMI and not entropy, it becomes clear that the criticism of Boltzmann’s H-Theorem were addressed to the evolution of the SMI and not to the entropy. At the same time, Boltzmann was right in defending his H-theorem when viewed as a theorem on the evolution of SMI, but he was wrong in his interpretation of the quantity −𝐻(𝑡) as entropy.
https://www.mdpi.com/1099-4300/19/2/48?utm_source=chatgpt.com
I'd start with Wikipedia. Unfortunately, this article assumes that the entropy described in Shannon's applies to all of physics.
https://en.wikipedia.org/wiki/Information_theory
I think maybe the question is: from a universe-level perspective, what is entropy? The state of the universe as a whole is just some pure state, and all pure states are just a unitary transform away from one another.
But just by introducing boundaries between observers, you get non-unitarity and entanglement entropy, as your system entangles with stuff that leaves your light cone.
I'm not sure if this is the actual mechanism the cited paper finds (I haven't had the time to look, and I doubt I have the physics chops as but a humble quantum information theorist), but it seems like a reasonable intuitive guess.
To me, this is still the underlying puzzle that MWI doesn't really resolve: what is the mechanism by which a pure state evolving unitarily gives rise to robust subsystems that have the structure necessary to give rise to all the messy, entropic behaviour we know and love(?)
If you start with those subsystems, MWI explains how they continue to induce a branching into worlds in a basis natural to those subsystems; but it's not clear (to me, anyway) what breaks the perfect pure state symmetry to give rise to those subsystems in the first place. The cited result confirms (to me) the tension between the unitary-evolution/God's eye view (with no entropy, no obvious internal structure) and the messy ground up view of interacting systems--somehow, there is a decomposition of this simple, structureless pure state into complicated things in a way such that those internal components have their internal reality; or from the other end, somehow these messy interacting entropic subsystems add up to something with no internal structure at all.
They concern themselves with entropy in black holes, and I think they're modeling it as if our universe were closed and had the same dynamics of a black hole.
> Moreover rules were given for computing interior observables in the fundamental description that agree with the semiclassical description up to ambiguities which are exponentially small in the entropy of the black hole at the time the observer jumps in, which is perhaps the most we should ask for.
And...
> Thus we see that even when the black hole is small there can still be sizable entanglement between its interior matter and a reference system, as long as there is a big observer who is also inside.
And here they invoke decoherence...
> entanglement arises from the entanglement of the observer with the rest of the universe due to decoherence, and the observer is cloned out of the system in their pointer basis to ensure that they are effectively classical.
Paper here...
https://arxiv.org/pdf/2501.02359
Pure versus impure , simple versus complex, and symmetric versus asymmetric are different distinctions.
Under decoherence, it's assumed that th e environment is impure, and an initially pure WF can't remain in coherent superposition once it interacts with an environment.
But that says nothing about many word branching -- which might require a high degree of symmetry in the environment.
Yeah, I'm using "symmetry" loosely and idiosyncratically; I have in mind something like transitivity under unitary rotation; without a preferred basis, no "internal" decomposition of any pure state into subsystems is canonical in any way.
I don't think you need an impure environment for decoherence; in the formulation I'm most familiar with, Stinespring dilation and CP maps, you usually assume the environment is in a pure state as well: the system starts off in a tensor product of two pure states, then a global unitary entangles the two systems, and when you trace out the environment, you will have a rank>1 density matrix on your system.
But not sure that's too important of a point.
>what is the mechanism by which a pure state evolving unitarily gives rise to robust subsystems
Doesn't decoherence provide the mechanism? All you need are uncorrelated degrees of freedom to entangle with. A large enough system will have parts that are uncorrelated with each other. Isn't it analogous to how parts of the universe can have nonzero angular momentum even if the universe as a whole doesn't?
This works if you already have some decomposition into subsystems where one subsystem can act as an environment for another.
So I agree, if you have a large system, and an environment for it to interact with, the subsystem evolves in a robust way. But if you don't already have a decomposition into "large subsystem and its environment", you just have "pure state of the universe", where did that initial decomposition come from?
It's dynamically emergent:
https://www.sciencedirect.com/science/article/abs/pii/S135521980300087X
This is a great paper, and I'm shocked I've never seen this result before.
May have more to say in a bit, but need to digest this a little first.
First thought though: evolution under the Hamiltonian still requires some external notion of time; I feel like I'd want the quantum state of the universe as a whole to have the time evolution somehow emerge from within the state, not be tied to a notion of time that exists outside in some way.
Does anyone have a recommendation for a history book for my younger brother? He's read a lot of Horrible History and other kids' history books, but he seems to be outgrowing it a little. He still loves it though, and he's absolutely riveted when I tell him stories of the past. (Right now we're on the French revolution.)
He's not at the reading level to read a proper textbook. Any books which should be readable for him, but have some more meat than pulp histories? I think he would prefer a whole book on a certain event, time period, or person than an overview of a huge swathe of history.
AJP Taylor--Origin of World War 2
As a general rule of thumb, it's probably best to be well-acquainted with the "formal" history of any given period before grappling with revisionist history of the same.
It is a good read.
I don't know that I have a good understanding of what kind of material you're looking for.
Two of my favorite books that might be called history are Amanda Podany's "Brotherhood of Kings: How International Relations Shaped the Ancient Near East" (she's also written a followup, which is good but not as good), and Patricia Crone's "Pre-Industrial Societies: Anatomy of the Pre-Modern World".
As you might guess, Crone's book is more of an overview of what things were like in very broad terms. If you want "stories of the past", Brotherhood of Kings is clearly a better choice; much of its structure is precisely this.
Here's a story that appears in the book:
After a political crisis, the throne of Egypt is held by a young widow. She feels political pressure to marry an Egyptian noble, which she doesn't want to do. So she sends a letter to the king of the Hittites, Suppiluliuma, saying that since he has many sons, she wants him to send one of them to marry her and be the king of Egypt.
The Hittite king is stunned (obviously), and convenes an emergency council to decide how to respond. Over the objections of certain counselors who say this has to be a trap, he sends one of his sons as requested.
The queen seems to have been entirely sincere. But those local nobles take offense, and the prince encounters foul play.
When word gets back to Suppiluliuma, he is naturally enraged. He gathers up his army and goes on a rampage south from Anatolia, intent on punishing Egypt. He doesn't make it all the way there, but he does manage to conquer a good amount of Egyptian territory in Syria.
Another story that stuck with me was of a pair of princesses given in marriage to the same Mesopotamian king. The older one married first and became very unpopular locally, with rumors spreading that she was a witch.
The younger one arrived later, marrying the same king, and hoped for help from her sister settling in to her new life. She didn't get it; we have the letters she sent back home, lamenting that her sister was no support and the rumors of her being a witch were true.
As another recommendation, John Keay's "China: A History" is very readable. It covers a very broad period of time, but it also includes plenty of stories. I haven't read his history of India, but I imagine it is similarly pleasant to read.
It might be stretch for his reading abilities but The British Are Coming by Rick Atkinson (https://www.goodreads.com/book/show/41812830-the-british-are-coming) was an excellent narrative history of the beginning of the American Revolution. I thought I knew it pretty well (esp. having lived in New England the last few years) but I learned a ton and enjoyed every minute of it.
podcasts? Mike Duncan's Revolutions is seriously good
Seconding this recommendation.
Thirding this Reccomendation. on my second relisten. Is good stuff.
Just saw the "not at reading level" part yet, so depending on his age, Extra History might be another good fit if you aren't wed to "readable textbook."
https://www.youtube.com/watch?v=lFfHp83Xm04
They can be hard to find, but has he read any Landmark books? It was a series of standalone history books, each covering something or someone specific, written by popular novelists. So not textbook-y. They are from the 1950s and '60s, so take that as a content warning I guess, but they're still well regarded.
How about The Cartoon History of the Universe, by Larry Gonick? It is to some extent an overview, but it does go into detail - it's got three volumes so quite a bit of space. It's also comic format, which works surprisingly well.
I was a big fan of Gonick's Cartoon Histories growing up, but I'm hesitating to pass it along to my daughter. They're fun and well-written and do go into appropriate levels of detail, but the historiography is extremely dated. The three volumes of CHotH were written between 1978 and 2002 (23-47 years ago), and Gonick also has an unfortunate tendency to uncritically accept and repeat bad pop-history takes (the standalone Cartoon History of the United States is particularly bad in this respect) and to treat religious/mythological accounts as if they were historical primary sources (off the top of my head, he does this with the Old Testament, the Gospels, the Romance of the Three Kingdoms, the Illiad, the legend of the founding of Rome, and the Hadiths).
I worry that handing them to someone who doesn't already have some basic grounding in general history and the maturity to understand that a source is unreliable would wind up giving them rather too much to unlearn later.
Maybe. I did seem to get the hang of unlearning things myself, after all.
On the other hand, misinformation does not seem to be in such short supply that there is a pressing need to go looking for it.
If it is not gauche, may I recommend my own book, published two years ago by Macmillan, Impossible Histories? It is half alternate history but half real history, so you get the historical details on twenty important events across 2400 years and six continents (sorry, Australia) as well as the part where I make up twenty lies.
Impossible Histories by Hal Johnson! Ask for it by name!
As an unrelated 3rd party, I can confirm that Impossible Histories is very good.
From testing with other friends, it seems that the there is some level of background knowledge to really appreciate it. I can't tell if OP's brother is at the right level yet, but it could be a good fit already and, if he continues learning about history, it will eventually be a great choice.
History books can be hard to recommend given the wide gulf between academic survey and airport pop history.
I'd recommend two podcasts: 'Hardcore History' by Dan Carlin and 'Fall of Civilizations' by Paul Cooper. The former more or less got me into history as a hobby and the latter is just really nice!
For a kind of random book I recently finished 'Xenophon's Retreat' which I thought was a nice "interested lay person" summary of Xenophon's works with good surrounding context.
If the boy is too young to read a "proper textbook", there's a high chance that Carlin will be too brutal for him. Genghis Khan, WWI, WWII Japan, and the Munster Rebellion are all episodes I distinctly remember having plenty of stories one *might* not want to expose to a, say, five-year-old. If you prep that 5YO beforehand with a clear lecture on Things We Don't Do Anymore Because Society Would Condemn It, there's even a Carlin episode (might be in his sister series Common Sense) that specifically goes over things we did to each other in living memory, with community approval, that's probably going to confuse the heck out of him.
Excellent series for adults, though, I'll strongly agree.
How old is he? If he's big into the French Revolution, Mike Duncan's "Revolutions" podcast does an overview of that which is lay-audience accessible but comprehensive
Have people here had positive experiences with taking SSRIs for anxiety- in particular social anxiety? Or negative, or neutral experiences- I want to hear the whole spectrum. I've long been opposed to taking them, but just wanted to gather some data. I'm one of those people whose fear response has always been dialed up rather high- I'm jumpy and get afraid easily. This has over the years (I'm middle-aged) metastasized into 'I actively avoid all social situations longer than 5 minutes' level of social anxiety in most (not all) situations. Just curious what's worked for other people in terms of pharmaceuticals. I'm deeply creeped out by the idea of flattening all of my emotions and personality, though I suppose one could argue that I have little to lose.
(PS, if your answer is 'have you heard about our Lord and Savior CBT'- yes, I've tried it, it doesn't work for me, please preach elsewhere)
Duloxetine seems to work pretty well, though high doses do seem to flatten emotion a bit, and even higher doses actually cause anxiety to worsen. But you can just reduce the dosage if that happens. Those pills aren't hard to open.
Venlafaxine/Effexor also works, but the initial side effects and withdrawl symptoms are very very bad. Do not take unless you've tried alternatives.
Scott's medical site also has a infornative page on SSRIs: https://lorienpsych.com/2020/10/25/ssris/
I've tried a fair few SSRIs for depression accompanied by anxiety, and generally stopped taking them after a while because it was unclear if they were helping the depression at all. I did not feel that they flattened my emotions and personality in general (though the depresssion sure did).
I did have more success with an atypical antidepressant for the depression, but that caused the anxiety to spike, so the doctor added Effexor and that seemed to work. Currently doing reasonably well socializing, being creative and enjoying things while on meds.
Not me but a good friend is prone to debilitating anxiety/panic attacks and he's had good results with both prozac and lexapro. He takes a very low dose (half of the minimum clinical dose) and it works well for him. As an outside observer the results were obvious. Just find a good psychiatrist and work with him to figure out what works for you.
I've taken lexapro for anxiety in the past. It helped a bit in terms of taking the edges off of things so I could be a bit more functional, but didn't really make me feel better overall.
What did help quite a lot was when I started estrogen. I stopped taking lexapro a few months into my transition and haven't missed it. But that's very situation specific and definitely not for everyone.
I have had really excellent results with them, both for social anxiety and health anxiety. I did not have any issues with flattening of emotions or personality - I'll still cry at a well-done phone company commercial.
You could try microdosing it before social engagements, that way you don't build up much in the way of tolerance/withdrawal effects.
SSRIs generally don't kick in until you've been taking them consistently for several weeks (and the effects can take a while to stabilize, so anxiety and mood may be worse initially).
Isn't there a super short-acting SSRI that's used to treat premature ejaculation? It must have at least some immediate effects...
Yeah so the serotonin-reuptake-inhibiting effects, which I'm pretty sure mediate the ejaculation stuff, have almost nothing to do with the antidepressant effects.
It takes ~10 nM of prozac (fluoxetine) to inhibit reuptake via the serotonin transporter in the CNS.
In someone taking 50mg of prozac a day (a pretty normal antidepressant dose) you find synaptic concentrations of fluoxetine on the order of 10µM. 1000 times as much!
Either drug companies aren't THAT greedy, or "SSRI" is only nominally true lol.
https://drive.google.com/file/d/1kHJUv08uvXuNBAkNa3dCog1r_1NDBMky/view?usp=share_link
Source: I have no fucking clue what paper this screenshot on my phone was from, but you can probably figure it out by google-scholaring if you're curious.
I've been taking Citalopram since February, I'd say it has a marginal effect but my hope is that's all I need, I'm looking for 5% improvement wherever I can get it.
I follow AI developments closely and use it a lot in my work. I also write and record musical satires about AI, robots, and our overlords. If anyone is interested in exploring song ideas together, let me know. The music is all home-rolled, AI is only used for mastering.
I'm curious to hear some before and afters with the AI mastering, as well as a rundown of your process/tools.
We use Izotope Ozone 12 for mastering as a start and tweak it from there, it’s infinitely tweakable. The Ozone 12 website has before and afters. Our DAW is Studio One Pro 7. We do all the vocals and instruments.
Link?
https://www.izotope.com/en/products/ozone/listen
Thanks, I’ll check it out
New article from Gusev on Missing Heritibility "The missing heritability question is now (mostly) answered" . I am rather confused at his interpretation of the Wainschtein et al. paper, which does not seem to quite support his position? But I need to read more deeply.
I am a little confused by his approach averaging across traits, and I am still left wanting a mechanistic explanation for twin similarity.
https://theinfinitesimal.substack.com/p/the-missing-heritability-question
https://www.nature.com/articles/s41586-025-09720-6
> I am still left wanting a mechanistic explanation for twin similarity.
Gusev doesn't have one for you. He admits it in his previous Substack post, "We still do not understand family resemblance" (https://theinfinitesimal.substack.com/p/we-still-do-not-understand-family). What's interesting is that in that article, Gusev references a paper that shows not only MZ twins have higher correlated traits than DZ twins (which was expected), but adopted siblings, and in-laws show higher correlations than DZ twins (which was totally unexpected!).
> The other clear outlier was the correlation of adopted relatives, which is expected to be zero under a Fisherian model but is clearly and significantly not zero. Interestingly, using the CTD to estimate heritability under random mating produced a value for the shared environment that fits the adopted siblings fairly well, but does not fit the other biological relatives. On the other hand, accounting for assortative mating could fit most of the biological resemblances but reduced the estimate of the shared environment to zero and thus no longer fit the adopted and in-law resemblances. MZ twins, in-laws, and adopted siblings all exhibit higher correlations than expected, so accounting for one pushes the model away from the other and vice versa.
> I am rather confused at his interpretation of the Wainschtein et al. paper
The thing that confuses me most about Gusev's position is the real-world implications.
Where would Gusev's real-life-behavior recommendations cash out any differently than a strict hereditarian's would?
So now GWAS's and family GWAS's and all sorts of more refined and informative takes (RDR, SR, etc) are coming out that have identified a "heritability gap" between what we can measure genetically and what twin studies tell us, and which argue "nope, that's wrong, environment is way bigger and more important than we thought." And this is Sasha Gusev's primary topic of interest.
But I think, “oh, environment is 4x bigger than we all thought?” so what should we do differently? Absolutely nothing? The biggest thing is still maxing as hard as you can on spouse quality, aka genes? Oh, and since “family culture” apparently matters a lot per family GWAS', you should act as the most hereditarian-pilled would recommend and also max on *lineage quality* as well as spouse quality, as the most hereditarian-pilled among us (ie those aware of Greg Clark's research) would recommend?
And as near as I can tell, the Gloomy Prospect is still true, too. In other words, for all the environment supposedly matters 4x more, none of the “environment actually matters” people can tell you even a single environmental intervention that will move the needle in a positive way.
For raising our kids, it's basically "make sure they're raised in a two parent household with good mores and no physical or mental abuse," and that's the tippy-top best you can do.
Everything else is noise, in the sense that everything else that matters was handled at the "mate selection" stage. You married into whatever level of educational attainment and average income and wealth you're going to be drawing from, and now it's just random draws from you and your spouse's conjoined lineage baseline. All environmental factors that we know matter are negative - malnutrition or physical or mental abuse all bring your kids down, but we know of nothing that brings them up, and the “hereditarianism is wrong” people have no advice.
The other thing I’ve never understood is that it’s not like we haven’t controlled for these things before! If culture / environment REALLY matters ~4x more than we think from Polderman, why didn't it come out in any of the much older separated at birth Minnesota twin studies, or the many studies where disadvantaged kids are raised by middle / upper class people in much better environments?
It's not like we haven't thought about or controlled for "culture / environment" before - and I've asked this before and didn't get a good answer.
So I remain puzzled by the seemingly warring sides, and it genuinely seems to me that behaving as a stricter-than-usual hereditarian is the only conclusion to come out of the whole debate with, and that this is affirmed by BOTH sides.
It has strong implications that we should be providing social support to help people make up for deficiencies in their local environment, because this is likely to make them more productive.
The question was what would you do differently - in terms of your/spouse/children lives, not about on what you want to spend taxpayers money on.
That's true, and interesting. It's my impression that the most strict hereditarian still seeks the best educational resources for their children -- therefore acting like they think the environment still matters. Which is rational, but not then voting to deny those resources to other people. If the environment matters for your kids, it matters for theirs too, regardless of what anyone inherited. *Unless* one is arguing that what is inherited is the ability to utilize environmental resources, but I haven't seen that argument made explicitly. There's all sorts of evidence against it.
As for what one should do differently -- that's a strange question. I personally make few life decisions based on my theoretical beliefs -- I base them instead on the results of a lifetime of trial and error attempts to adapt to local circumstances. Those life experiences aren't going to change in meaning just because science says one thing or the other about how much IQ is inherited.
I find that science doesn't really inform daily life adaptation very well -- it's purpose is to inform institutional policy (for example, educational practices), which strongly determines the local environment various people have to adapt to.
I don't learn science to become more successful in my personal life (except in my own professional domain, but that's a different dynamic), I do it to make better informed choices at the ballot box, and to help change public consensus through private conversation (as I am doing right here).
It's not theoretical like, say, black holes.
>I find that science doesn't really inform daily life adaptation very well
well doesn't science say that you should take vaccines and avoid unprotected sex with people without recent STI tests?
Doesn't science say that if you want to have smaller ping time in online games you should move closer to game server rather than buying more expensive cables?
I asked you not to talk about ballot box yet you do. FTFW, a person might be anti-heredetarian and still be opposed to redistribution for another reasons.
I miss being able to have a conversation about what is the case in the world, without being side-tracked into "so what does this mean we should do" 😢
>none of the “environment actually matters” people can tell you even a single environmental intervention that will move the needle in a positive way.
Exactly. The "missing heritability" argument is an isolated demand for rigor. If environment is so powerful then figure out why and solve all of our social problems.
Yeah whatup I'm a microbiome scientist and FMT has dropped kids off the autism spectrum
Nonsense. There are plenty of studies. It's just that the hereditarians ignore them. For instance, there was the French Transracial Adoption Study (Weinberg, Scarr & Waldman, 1992) that looked at Black, mixed-race, and white children adopted into high-socioeconomic-status (SES) French homes. Children’s educational outcomes tracked adoptive family SES strongly, not biological background.
But the key takeaway from the Wainschtein and Zhang study is that SR and RDR put narrow-sense heritability down around 30%, and GREML-WGS (which uses whole genome sequencing) put it as low as 27%. And when we look at IQ it's down around 32% using GREML-WGS. Three distinctly different methodologies yield surprisingly close numbers, yet (some) twin studies make IQ heritability 2x higher. Gusev's snarky remark on X is why would you favor a 40-yo methodology over current science?
Do you see irony here you finish you comment with sneer about 40 year studies, but you started it with 33 year old study for your cause, even forgot which one....
These improvements vanish after adolescence and probably goodharted. But even if they are real, what is does tell for public policy? Take children from poor SES familes and force high SES families to adopt them? How is it different from negative eugenics? Maybe you say that low SES families should change something.... something... but you don't know what specifically. They study you mentioned was about adopting.
> French Transracial Adoption Study (Weinberg, Scarr & Waldman, 1992) that looked at Black, mixed-race, and white children adopted into high-socioeconomic-status (SES) French homes
I think you mean the Minnesota adoption study? Because Weinberg, Scarr, and Waldman were the ones who did the Minnesota adoption study over thirty years. There are separately two other studies done in Germany and Japan respectively that are similar to this (in the senses of tracking IQ and educational outcomes segmented by race of the biological and adoptive parents), but I'm unaware of any French ones.
And the Minnesota study clearly shows the "washout" effect on IQ that most adoption studies show in the testing at age 17, with the overall study followup synopsis asserting that it was a mess and couldn't be used to support either hereditarian or environmental arguments:
"In a 1998 article, Scarr wrote, "The test performance of the Black/Black adoptees [in the study] was not different from that of ordinary Black children reared by their own families in the same area of the country. My colleagues and I reported the data accurately and as fully as possible, and then tried to make the results palatable to environmentally committed colleagues. In retrospect, this was a mistake. The results of the transracial adoption study can be used to support either a genetic difference hypothesis or an environmental difference one (because the children have visible African ancestry). We should have been agnostic on the conclusions . . ."[15] Later opinions supported Scarr's reassessment. For example, one group of authors wrote, "Generally, scholars in the field of intelligence see the evidence from this study . . . as consistent with both environmental and genetic hypotheses for the cause of Group IQ score differences"
https://en.wikipedia.org/wiki/Minnesota_Transracial_Adoption_Study
Oops. My bad. I meant Christiane Capron and Michel Duyme (Nature, 1989).
https://www.nature.com/articles/340552a0
I also had the Weinberg, Scarr, Waldman study open in another tab.
https://www.sciencedirect.com/science/article/abs/pii/016028969290028P
My disorganization impacts my IQ.
> Oops. My bad. I meant Christiane Capron and Michel Duyme (Nature, 1989).
Thanks, I hadn't seen this one - I actually see nothing about race in this one, it's primarily about occupational and educational status of the biological and adoptive parents.
They assign kids to 4 quadrants - high biological background (B), low adoptive background (A), high / high, low / high, and low / low.
They see roughly a ~10 point childhood IQ advantage in the high adoptive backgrounds relative to the low adoptive backgrounds, regardless of whether the children came from high or low biological backgrounds (although of course the high biological background kids were higher in an absolute sense):
https://imgur.com/a/6PzgGTa
That said, I think you would still expect those effects to wash out for both tranches of kids by adulthood, as seen in numerous other adoption studies. And "adulthood" is typically "17," so pre-college.
So I'm not sure what this is arguing for in terms of interventions - does "environment" really matter if all it does is get you a transient IQ bump for ~10 years or so that then washes out?
It might matter at the margins - some kids might complete high school versus dropping out, or have better grades and get just over the side of a bubble in college admissions.
But happily, we already sort adoptions into "high only" adoptive parents, so anyone being adopted should already be seeing these benefits.
Any SSC associated forum not explicitly run by Scott Alexander will sooner or later turn into a hugbox for fundamentalist Christians.
Coming from someone who agrees with you, you're better off just not advertising it, lest you Streisand effect the whole thing.
The Streisand effect is overrated IMO.
The Motte is a creepfest because they are so ruthlessly technocratic, and so I was readying my armor to support you, and then I see that you crashed out and got banned because... They aren't technocratic enough? 😆 Elitism is a sick fantasy life for people who think they have to tear others down to drown their own insecurities.
Have you ever considered that the guy who looks down on you and calls your political ideas stupid isn't covering for his "insecurities," it's that he really is smarter than you and his ideas are right and yours wrong?
Yes, I have considered this. I have political “adversaries” whom I respect and take at face value. I just don't include elitists among them. Elitism is just too ridiculous, I can't take it seriously. Special case. Anyone who unronically uses “low-class” as a slur in 2025 is not likely to be a serious person.
Any use of the word “dysgenic” is a red flag, although it's not disqualifying in the same way. I certainly don't RELISH the idea of genetic diseases and pathologies spreading through the gene pool, but I hardly consider it an emergent crisis.
Have no idea what "technocratic" even means.
It's not exactly a synonym for Nietzschean Centrism, but it's close.
Just linking the post that got Alexander banned from the Motte for visibility, unless hapless people see this sort of thing and assume the Motte is actually as bad as people think.
Ironically this post where Alexander said repeatedly "I'll probably be banned" was raging at me actually: https://www.themotte.org/post/2368/culture-war-roundup-for-the-week/353039?context=8#context.
Full quote:
"P.S. To preempt the accusation that I ignored ThomasdelVasto's point, I reject the whole theory that poor, low-IQ people are harmed by competition with rich, high-IQ people. People are willing to risk dying in the desert to move from low-IQ to high-IQ countries because high-IQ has massive positive externalities.
P.P.S. I know I might get banned for this post. I was drawn to the forum because I'm a long-time Scott Alexander fan, from back when "right-wing SSCer" meant "secular guy who talks about embryo selection and national IQ," not "guy who thinks we need to go back to 1710 ideas about religion and government and that eugenics is evil." For me, unlike many of you, the former wasn't just a gateway drug to the latter, so I'm "left-wing" now. You can follow me on Substack and Twitter.
P.P.P.S. The mottezien is immunized against all dangers: one may call him a cuck, nazi, bigot, fascist, it all runs off him like water off a raincoat. But call him a resentful prole and you will be astonished at how he recoils, how injured he is, how he suddenly shrinks back, calls you egregiously obnoxious, and then bans you from the forum."
Quoting a copypasta? Well, that will probably get you banned just by itself.
Hey! Ideas on religion from 1710? What is this crazy modernity, the 13th century is plenty updated for me!
"The tattoo-covered WWE fan doesn't want to listen to a sermon from the Southern preacher but recognizes him as a fellow member of the broad ingroup of low-class Americans who share a common inferiority complex toward urbanites with lots of education and money."
My problem regarding Alexander's comments is that he clearly wants to present himself as "an urbanite with lots of education and money" but the amount of carping he does about hte low-class makes me go "tuppence-ha'penny looking down on tuppence", that is, not really old(er) money, but only a generation or maybe two removed from the grubby low-class types and is constantly petrified with fear that people he wishes to emulate will mistake him for his ancestral roots.
Alexander, if I'm mischaracterising you, apologies. But you go on so much about a particular type that it's hard not to think "the lady doth protest too much", as it were.
I shop at Walmart, eat at McDonalds, and watch true crime documentaries. My class background is complex but it's certainly not "old money." I have some low-class traits picked up from my family background. I've said this to you before, but I'm sure you'll pretend not to see it and repeat your "sick burn" in a future thread.
Alexander, I'm not trying to sick burn you. But you do have a few hobbyhorses you ride into the ground, and it does sound more like "very very anxious about perceived status so as not to be confused with the wrong class" on your part when you bang on over and over and over again about the lower classes and the right fetishing manual labour etc.
I'm happily lower-class myself (peasant to the bone) and couldn't give a damn what my 'betters' think of me. Relax, it'll be a lot easier on your ulcers!
That comment was me intentionally self-immolating after a series of bans for calling out their nonsense. Like this:
https://www.themotte.org/post/2229/culture-war-roundup-for-the-week/342342?context=8#context
Ahh yes, this comment where you are clearly speaking in good faith and not just taking shots at your outgroup:
> his worldview would seem to conflict with HBD theories. Indeed, one would have to conclude that whites are an inferior race. Guatemalans in their "third-world s***hole" don't just sit around despairing, they cross multiple borders and look for work in a country where they can't even speak the language, while white men who got laid off in their rust-belt factory towns twiddle their thumbs and inject fentanyl, unable to compete with said Guatemalans.
> If I believed, as many of you profess to, that my race was at risk of going extinct, I wouldn't be centering my politics around helping the least capable members of said race who refuse to help themselves. Don't you have bigger problems? It's not like you should feel any "political" loyalty to them, Trump's working-class base work, homeless people rarely vote.
Either way, you continued to escalate constantly in the comments of your posts as well, which was a huge extra piece of context, on top of the already bad faith discussion like above.
I'm sad you got banned because you did generate a lot of interesting discussion and had a genuinely heterodox worldview for the forum, but I can't say it wasn't justified.
This logic is typical of the motte, where disagreeing and taking shots against views one believes to be wrong is equated to "bad faith."