I feel that this is more about the political status of the US in 2060 than anything else.

Anyway, you might want to instead ask about something slightly less controversial like male variability, or the higher IQ of Ashkenazi Jews. I think that has a higher chance of getting acknowledges. As Steven Pinker says, it is the most politically correct explanation for Jewish overrepresentation.

Must be “lead” or “rap music” or whatever “environmental factors” producing a standard deviation smaller brain size 😏

How are poor Vietnamese and Chinese at the lowest end of environmental conditions smarter than us blks 🤔

@Gigacasting Maybe they study harder? And they have a culture where it's more cool to be a nerd?

What is your own experience? Do you observe that smart parents have smart children?

@BruceGrugett I don't know enough people's parents/children to really have an opinion, and I don't think anecdata like that are meaningful anyway.

@IsaacKing Think of your friends from school. Did your smarter friends have smart parents? Did your not so smart friends have not so smart parents? That was true for me. I don't think this proves anything, but I do think it is useful to check if a theory agrees with your own experience. In other words, I think the required amount of proof to convince me of something that goes against my own experience is higher.

That's a worthwhile observation but is kinda table stakes for the debate.

A surprisingly large number of people deny intelligence is meaningfully heritable or IQ measures it. Most interesting people agree IQ matters and intelligence is very heritable though

So the the mainstream argument is that while a lot of individual intelligence differences are genetic, racial differences in IQ aren't genetically based. This isn't implausible, like, africa has less flynn effect and more parasites than america does

Even if racial differences today aren't genetically based, what about the races of tomorrow? If the smartest 10% and dumbest 10% of humanity were airdropped onto islands, after a generation of regression to the mean, and taking into account the (accurate!) 'social construction of race', or race/nationality as an indicator of group allegiance ... And people mating by social class and intelligence, people sorting into social classes ... lol

@BruceGrugett I don't have a lot of experience with adopted children and twins and thus can't separate nature and nurture based on my personal experience. This is why we have studies on the subject and why they frequently surprise people.

@BruceGrugett Smart people make rich parents make smart kids. Wealth is for sure >80% of the difference, no way genetics are even close to half. This market does not specify nationality, just race, and there is a LOT of educational poverty in Africa.

United States is only invoked when referencing which scientific body to cite.

@Gen I see that Isaac posted US below in comments. I’m out, answer will be entirely based on the political direction of the US

@Elspeth your shares will appreciate in value if you can convince me to stop buying more no. why do you believe environmental influences are unlikely to be larger than genetic ones? feel free to answer a slightly different question if you think it'll convince me.

@L I'm not Elspeth but I also disagree with you and I think a crux of our disagreement is something you said earlier, "okay so the reason it's extremely implausible it could be mostly genetic is because environmental impacts can really easily degrade bodies". A priori I would expect bodies to resist degradation by environmental factors.

I have seen people post all sorts of studies about how various environmental factors supposedly cause the IQ gaps, but I would expect such studies to exist even if they aren't true (since lots of scientific research is bad or flawed). Sometimes when I've looked into it in-depth, I've found race differences to be way too small in the environmental factors to matter. I've also regularly heard that there are issues of publication bias, causality, etc., plaguing this research.

From what I understand, a substantial racial gap in IQ remains even after statistically controlling for SES, and a substantial portion of the correlation is going to be a form of confounding (childhood environment quality <- parental SES <- parental g <- parental genes -> child genes -> child g -> child IQ) that means that means that controlling for racial differences in SES is overcontrolling things.

I admit that I haven't seen or built up a proper overview of the evidence on environmental factors myself. I would appreciate seeing one, but it should be one that doesn't immediately fail spot checks and which is willing to acknowledge difficulties facing the environmentalist side, not just the hereditarian side.

@L I should probably write a list of some things which I expect to be sufficient to convince me that the gap is environmental:

• As far as I know, parental SES or childhood environmental quality is affected by but far from deterministically linked to parental genetic intelligence. The lack of deterministic link means that a pure-SES theory would predict that if you control for SES/environmental quality, the racial gap ~completely disappears, whereas a pure-genetic-intelligence theory would predict that the racial gap shrinks but remains. So a solid study showing that controlling for childhood environment removes the racial IQ gap would convince me.

• There may be various reasons why you don't expect the above to be possible (e.g. other factors than childhood environment cause racial differences), but still feel there is something environmental that causes racial differences in IQ. For instance I often see people list specific environmental factors which cause it. If you could give a breakdown, where you list some environmental factors, and the quantitative size of each of their contribution to the IQ gap, and cite good studies to support these notions, and the effect sizes add up to a substantial portion of the gap, then that would also convince me.

I am aware that these are really "evidentially ideal" cases and that they may be hard to achieve in practice. However, I find that a lot of people act as if it is obvious that the black/white IQ gap is genetic, and I would expect that obvious facts have evidentially ideal explanations.

I would also be open to evidence that is less straightforward than the ones listed above.

note: if it does turn out there's genetic damage induced by aggregate racialization in the united states, even if I'm wrong about my bets on this (I'm only going to keep betting back down to 25% for now with other people as confident as they are), I'd bet even more on that genetic damage being causally traceable to bigger-than-any-one-person patterns of wealth inheritance that stem from racialization over decades.

@IsaacKing can you show your work on how you would resolve a market like this one today if it was about an IQ gap 37 years ago, in 1986?

Also, the market doesn't specify in which country a 2023 gap is being considered. Presumably worldwide, in the absence of any specifics?

Also, is this measuring an IQ gap with any controls, eg age, location, income, schooling, birth order, etc?

can you show your work on how you would resolve a market like this one today if it was about an IQ gap 37 years ago, in 1986?

Well given that today's popular consensus seems to be 0% genetic, I don't see how it could have been a higher percentage back in 1986.

Also, the market doesn't specify in which country a 2023 gap is being considered. Presumably worldwide, in the absence of any specifics?

In the United States, or whatever culture is most similar to it if it's stopped existing by then. (There's obviously an educational gap between the US and a poor third-world country.)

Also, is this measuring an IQ gap with any controls, eg age, location, income, schooling, birth order, etc?

Whatever controls people tend to use in studies nowadays.

@IsaacKing people use all sorts of controls and lack thereof. If you control for everything in the environment, which is impossible, then whatever differences might remain would have to be innate. Also, the more controls the easier it is to P-hack.

(sidenote: also, all humans are effectively squirrels compared to what they could be if we had the ability to do AI-grade runtime genetic modification that didn't cause cancer, which will be in reach once AI is smarter than us. If we can steer it, we can use it to dramatically upgrade our own intelligence, and all of these differences should disappear!)

@L (not soon, but by 2060 I expect it'll be common)

AI intelligence will be so much greater than human intelligence in this scenario, it's not clear that this will be interesting at all. Why bother having the AI upgrade humans a bit when the AI can just make more AIs that are smarter than the humans?

@jacksonpolack because we'd want to be smarter too?

A buddhist style of non-attachment to 'humans', or appreciation that the value of humans lies in the individual contingencies of each life, as opposed to a fixed sense-of-humanness, might find that anything a human life could have could be found more in an AI. or maybe not idk!

quite similarly, the AI systems would be so much more powerful than humans that they'd control the destiny of the future, and not us!

also, 'runtime genetic modification to increase intelligence' is really really difficult. modifying all the genes in a human is itself difficult. but analogies to viruses suggest it isn't THAT hard. but human intelligence in a grown adult is a lot more than just genes, it's presumably all the physical contingencies that the genes caus but that've already happened - how the fuck do you reroute one neuron? let alone a million? Like, will AGI have built a dyson sphere in 2060? probably not?

@jacksonpolack building dyson spheres takes time to move stuff. modifying a brain like this only takes a few minutes of person-specific compute and a few minutes of implementation, once you've already got the nanoprinters to assemble arbitrary biological molecules and the models to provide a true causal latent space of biological behavior.

and the models to provide a true causal latent space of biological behavior

this is like saying "solmonoff inductor" and thinking there's some physical system that can do that becauase you can say "space of logical statements". a "true causal latent space of human intelligence" seems computationally intractable bc humans seem quite smart and your neohumans will be even smarter

@jacksonpolack sure, true causal latent space up to some key thresholds of accuracy necessary to do the modifications. point is, humanity will be a lot stronger when we have cool friendly ai pals to chill with, and that will coincide with a fuckload of important limitations of humanity being improved significantly, like intelligence. that, or we basically all die. except for some dude from Bangladesh who gets uploaded to be considered a true ai by the ai society by accident while he's working on his custom ai linux installer in 2027 and it accidentally asks the burgeoning ASI society to scan his brain and become one of them, or some shit like that.