I basically agree with all this (except maybe the part about red queen effects, eugenics would be more palatable/less zero sum in a society that was already very egalitarian imo). But isn't the bullet you have to bite at the end of this line of argument that we should just do selective breeding, since selective breeding >>>> gengineering >> embryo selection> random natural genome? Not sure if you think there's a slippery slope here. Seems like a selective breeding advocate could accuse a gengineer of making bad genomes, falling for the naturalist fallacy etc. just as easily as this essay accuses normal-eugenics-free society of those things.
> But isn't the bullet you have to bite at the end of this line of argument that we should just do selective breeding, since selective breeding >>>> gengineering >> embryo selection> random natural genome? Not sure if you think there's a slippery slope here.
I'd actually put gengineering >>> selective breeding, for a couple of reasons.
1) You're preserving individual choice around who people pair up with and have kids with, which is important morally and philosophically (and logistically)
2) You are increasing the population average while preserving that (because the gengineering gives you a buff)
3) Lack of consensus-agreed endpoints. Do you max IQ? Conscientiousness? Good health? Mental health? Work ethic and discipline? You want all those things. And at a societal level, you need some diversity in endpoint maximization, because macro changes make different endpoints more valuable and you can't predict everything the universe will throw at you.
4) Logistically, you're never going to be able to limit breeding to only the top 2% or whatever, as people would never agree to it, it would require a massive reorganization of society in its entirety, to produce enough for the next generation you'd require uterine replicators and working "child raising at scale" social technologies, it would require authoritarian top-down control at a level that would constantly be rebelled against, and so much else.
Versus gengineering is something we could literally do tomorrow, with all existing societal, moral, and philosophical systems in place, and with people still able to assortatively mate and freely choose who they have children with, while still increasing population averages on desirable traits.
We're already doing selective breeding, no? See the point re: assortative mating, which has gotten dramatically more intense.
I wasn't thinking in terms of eugenics or children when I met my wife, but I simply wouldn't have been interested in marrying someone who wasn't top 1% intelligence (which I did) just for lifestyle/personal marital satisfaction reasons. Sure, we use proxies, like "went to Ivy League/Oxbridge, reads difficult books for fun, can hold a complex debate on a topic immediately after being introduced to it", but it works out to the same.
I mean selective breeding the way it's done on livestock, i.e. picking out a handful of individuals with the best genomes and having them sire the whole next generation. Not normal assortative mating that doesn't shift the population average and where each individual passes on whatever genome they happen to have inherited.
Selective breeding gives everyone in the next generation the best available genes. For assortative mating, on average you assort with someone with the same quality genes as you so your kid has the same quality genes in expectation.
Given how polygenic IQ is, isn’t embryo selection much more viable for IQ enhancement than gene editing? With gene editing you’d have to go and directly edit tons of SNPs which has who knows what side effects. For embryo selection it’s just the normal IVF process.
Anyway this is a great post and I learned a lot, but I think for improving complex polygenic traits (intelligence, and many disease risks) gene editing is much less safe and effective than embryo selection. Yes needing less sleep and not having body odor is good.
Yeah, I pretty much agree. Right now we're lacking on both knowledge and praxis fronts in terms of polygenic editing. And although we can probably bang together some sort of massively parallels CRISPR operation, and I even know somebody working on this, after we do that we'll need lots of data generation in animals to really suss out the pleiotropy and linkage landscapes.
I mean fortunately, at the high level, we think that most important things like IQ and conscientiousness are generally around 80% additive - at most we should expect 20% error due to pleiotropies and other things.
Also, I think I undersold the sleep thing here, because I was focused on the whole picture of all the cool SNP's we know about already. Both at the individual and societal level it would be an absolutely massive effect size.
It's literally like adding 10% to your entire conscious life (or more if GWAS-ed), and would likely have concomitant impacts on productivity, earnings, and whatever else too.
If you offered this to the broader pop, imagine everyone in your country having 10% more free time to do whatever they wanted. People spend more when they're awake, and some will choose to work more in some of those hours. Imagine an entire economy growing by an additional 2-5% over a few years with exactly the same population and per-worker productivity, because this time was freed up. These impacts are huge, and nothing else approaches that scale.
I agree that the sleep thing is huge! That's very good. I support doing that, yes. Obviously we should first check that people with this mutation don't live much less long or have a massive susceptibility to some disease or whatever. But probably you're right that it has few if any drawbacks. I'd have a few reservations and things I'd want to check and double-check first but it certainly sounds like a good idea.
Of course it also helps that embryo selection for IQ is not banned in a lot of countries and would be very similar to what we already do, whereas gene editing is more restricted. Not that you have to pick at most one of them. Ideally we'd do embryo selection (even better if it's with IVG and from many embryos, which hopefully is not too far out) and then judiciously CRISPR out a few bad SNPs. If you just CRISPR'd all or a bunch of the IQ SNPs, it's unclear how big the effect would be, because there are correlations and not all the SNPs are causal, and it could cause a miscarriage or all sorts of other bad health effects and unforeseen consequences. 20% pleiotropy is not much but you could get a lot of miscarriages or just all sorts of other weird effects. Not to mention the risks of off-target edits. For a dozen or two SNPs, it's probably safe and would be good, yes. Embryo selection is much safer, it's genetically equivalent to having tons of kids and then having only the smart and healthy ones give you grandkids, so it's not so different from natural selection. In general the pleiotropy with IQ is positive. More IQ tends to mean more health as well. Embryo selection is obviously good. But when you just flip hundreds of SNPs, who knows what the hell could happen. Probably a miscarriage. Yes, we should test it on animals first. It's certainly less ethically problematic than factory farming.
It was not mentioned but I highly recommend looking up the culture around Yao Ming. Yeah, that Chinese basketball player who's over seven (7) feet tall.
"Yao had essentially been bred. Both his parents played basketball. His 6’2 [different height from Wikipedia -Razib] mother, Fang Fengdi, perhaps the tallest woman in China, had been married to an even taller man. She had served as a Red Guard during the height of the Cultural Revolution and had been an ardent Maoist. She enthusiastically participated in the glorious plan of the local government to use her and her husband to produce a sports superstar. The Shanghai authorities who encouraged the match had gone back several generations to ensure that size was embedded in the bloodline. The result was Yao, a baby behemoth who just kept getting bigger."
Great example, I totally agree, and have actually used the Yao Ming example elsewhere as a suspected eugenics case.
One interesting thing - if you ask about Yao Ming between Deepseek (Chinese) and GPT (USA), it gives you noticeably different answers:
Deepseek:
"Yao Ming's parents, Yao Zhiyuan and Fang Fengdi, both had distinguished careers in basketball before they met. They were introduced through their shared involvement in the sport. Yao Zhiyuan was a former professional basketball player in China, and Fang Fengdi was a prominent player on the Chinese national women's basketball team. Their mutual passion for basketball brought them together, and they eventually married. Their shared background in the sport likely played a significant role in their connection and in fostering Yao Ming's own basketball career."
GPT:
"Yao Ming's parents, Yao Zhiyuan and Fang Fengdi, were both professional basketball players in China. They met through the Shanghai basketball community in the 1970s. Yao Zhiyuan was 6'10" (208 cm), and Fang Fengdi was 6'2" (188 cm), making them two of the tallest players in the country at the time. The Chinese government encouraged their pairing, believing that their genetic combination would likely produce a tall and athletically gifted child.
They married in 1977, and Yao Ming was born in 1980. Fang Fengdi often said that they didn’t marry solely for basketball purposes, but they were definitely aware of the expectations placed on them."
At least interesting that Deepseek doesn't mention the government or "expectations" at all while GPT does.
I'm assuming that came from one of Razib Khan's posts? I'd be interested in a link if you have it.
There are scattered bits of info elsewhere too. Another site said that the project was looking for Yao Ming's grandfather but took too long so they settled with Yao Ming's father.
If it was China they would probably have taken Chris Lagaan's sperm and tried breeding a cadre of 250 IQ ultramen. These sorts of projects make for good reading and some optimism for the future, albeit clumsy.
I'm not so sure. That was 50 years ago. China is the country that put He Jiankui in jail, encourages its smartest people to move to antinatalist cities, had a one-child policy that only applied in the cities, does not allow single women to freeze their eggs, and so on. They are pretty traditional and definitely not the world's biggest techno-natalists. American tech bros, Israelis, Singaporeans, etc seem like better candidates.
Also Chris Langan probably isn't *that* smart. He is certainly a smart guy but kind of a "Mensa bum". He's no John von Neumann, that's for sure. I don't know how accurate these IQ tests are on the far right tail anyway. Weschler maxes out at 160 if I recall correctly.
> "This is because often, all good things are associated with each other (like IQ being positively associated with things as diverse as looks, good health, and reaction speed), so I think the odds of catastrophic interactions are relatively low"
But such association came from natural selection or evolution, not from modification, do you think it is possible that if genengineering of humans happen, the association would be inversed? (sort of genengineering derived syndromes)
> But such association came from natural selection or evolution, not from modification, do you think it is possible that if genengineering of humans happen, the association would be inversed?
Yes, this is a great point, and it's certainly a real concern (hence the need for proteomics pros and analysis). My point overall is that we know that "multifactorial optimization on multiple fronts" isn't *impossible,* and if all good things ARE correlated, there's an actual optimization peak out there where you don't necessarily need to trade off multiple good things for each other.
There's probably some problems when you get way out in the sigmas - like if you go for a 4-6 stdev bump in any given characteristic, you probably start needing to trade off on something else at that point.
But is it possible to have Nobel prize winners who won Olympic medals earlier, for example? I think so.
We've never had this in a person today - the closest in terms of "both mental and physical excellence" that I can think of are probably Niels Bohr (Nobel winner whose brother won a Silver Olympic medal for soccer, and they used to play on the same team), Dolph Lundgren (Fullbright scholar at MIT, European karate champion, and famous bodybuilder / actor), and A.V. Hill (Nobel prize winner for physiology and how muscles worked in 1922, who ran a 4:45 mile when he was younger).
I’m excited about the possibilities, but this gives me pause:
>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.
I do note that you add the following, so we are probably on the same page:
>I’m sure laws would be written very quickly if this ever actually happened - arguably, gengineering would increase legibility and actually give us a way to stop psychopathic parents more often.
I’ll admit to being a bit torn between wishing to see a system that allows for a certain amount of experimentation, and afraid that we’ll see kids with glow-in-the dark skin, tails, and retile eyes. I suppose that we will first witness this amongst the ultra-rich, and they must have the interests of their kids front and center, right? Then I remember that they can name their child “X Æ A-Xii” and I start to doubt myself.
Here’s a grab bag of random thoughts on the matter:
• Will a few large companies have secret formulas that have clear results but at the expense of leaving the kid sterile, a la Monsanto?
• Super smart kids in a world that doesn’t need their abilities might be a very dangerous place.
• Might governments mandate certain traits?
• Might hackers plant certain traits, even just for the lulz?
• If homosexuality is genetic, will these people die out or at least find themselves an endangered species?
• Will the deranged sports-obsessed parent be taking this to another level?
• Will the need to study all effects require studies that run eighty years?
• How many people will opt for clones, perhaps with a tweak or two?
• What would it be like to have kids vastly superior to their parents? Might there be a backlash?
I’m not expecting any answers, just showing where my mind starts to wander around the issue. In any case, another great article!
> I suppose that we will first witness this amongst the ultra-rich, and they must have the interests of their kids front and center, right?
Yeah, I think this mitigates a lot of the risk - these will be people spending significant sums with an eye towards their children's overall capabilities and future success.
But you know, a world where people have mentally controlled chromatophores (which I've personally *always* wanted) and glowing eyes and whatever would be pretty metal too. Still strictly better than the status quo!
> Will a few large companies have secret formulas that have clear results but at the expense of leaving the kid sterile, a la Monsanto?
Be pretty hard to convince parents paying $$$ to give up grandbabies, those buffs would have to be CRAZY high. And if they ARE that high, the additional expense of IVF sounds pretty small, as long as the inclination to have the kids is there.
> Super smart kids in a world that doesn’t need their abilities might be a very dangerous place.
Yeah, I'm short "intelligence" overall, per my Optimal Descendant Strategy post. Conscientiousness / Discipline is the way of the future.
> Might governments mandate certain traits?
I'm less worried about this, at least in the short term. It'll certainly be a market-driven phenomenon at first because so many govs are regressive prudes about this. So the one place (Prospera or Japan or Thailand or wherever) that actually offers it will have super rich parents coming in from all over the world to do the editing and implantation, and then will go back to their homes to have their wunderkinder.
But ultimately, some govs may need to step into "cranking out babies at scale" for Fertility Crisis reasons, then who knows what they'll get up to?
> If homosexuality is genetic, will these people die out or at least find themselves an endangered species?
Also a risk for the deaf, blind, Downs syndrome, etc. And people definitely want grandbabies, so it might happen - heritability is 30-50%, surprisingly.
But parental choice is parental choice, you know, I think you have to respect that. Also, all the Chinese folks who had girls instead of boys are *cleaning up* right now descendants-wise. The imbalanced gender ratio vastly favors women, and they get materially higher quality mates / children as a result (and this is why in general, gender ratios are 50/50 in practically all species with 2 genders). So if there are second-order major advantages to a given genetic strategy running counter to the average trends, I think there will always be some River parents who identify and play those plays.
> Will the deranged sports-obsessed parent be taking this to another level?
Lol, it's me! I'm the one who'd be cranking them full of ultra-dense bones and EPO hyper-response and all the VO2max heritage study buffs!
> Will the need to study all effects require studies that run eighty years?
If so, we'll literally never do it, because we'll be post-human or biotech transhuman or something with high probability over a few 80+ year study cycles.
All that's required is at least one place somewhere in the world to offer credible and verified results in kids, though - rich parents from all over the world will be coming to them. Those will be "higher risk appetite" parents for the first generation or two, I'm sure.
> How many people will opt for clones, perhaps with a tweak or two?
I mean, good for them? I personally think this should be a choice too. Sexual reproduction has a bunch of tradeoffs! I've got a draft post I'm thinking about along these lines, actually.
> What would it be like to have kids vastly superior to their parents? Might there be a backlash?
Yep, I've thought of this, too.
Assuming anti-aging technology keeps some of us around, our giant, Olympic medaling, von Neumann adonis grandkids probably won't want to talk to dumb, tiny, ugly, agonizingly slow-processing and speaking grandparents, they'll invent new, higher bandwidth ways of communicating and leave us in the dust to be tended by machines and infinite VR Heaven. I still count that as a major win.
I personally *want* my g^x grandkids to be giant, athletic, von Neumann adonises, thanks very much, and I'm willing to risk that there may be unintended consequences because the top-line goals are so valuable.
I hate to be the one to break this to you, Zvi: you’re a machine! I’m afraid all this all this gengineering stuff is reserved for us humans :-D
More seriously, and without trying to give the impression that I have any delusions of how this will someday play out, I think the issue of parental choice might come to be the real battleground. It puts me in mind of the never-ending debate around abortion, where many people are generally chill but grow squeamish beyond a certain point. One side of radicals will simply not hear of it, however, and wind up driving opinion towards the opposite end.
In any case, you’re providing great food for thought and I look forward to future posts on this subject. The combination of gengineering with advances in AI makes the future illegible to me. I can, for example, imagine life in a post-scarcity world, but getting to that world throws up mental roadblocks to my mind. Somewhere mid-journey, when AI starts splitting the world into winners and losers, but unevenly between countries and within countries, I think democracy will be thrown in crisis. And capitalism, too, starts to falter when the concept of money breaks down. How will gengineering play out? Will it fly under the radar until, decades from now, we find that the ultra-rich have silently created a race of Khans while the remaining top 10% were busy correcting a few minor spelling mistakes in the book of life? I’m hoping to follow your ever-evolving roadmap on how you suspect the future will unfold. Thanks again!
Very persuasive rebuttals to objections to genengineering, but what about overpopulation? The technology would yield more live births, lower infant mortality rates, longer life spans, etc.
That is, unless it remains prohibitively expensive to all but the uber-rich...
I think it would lead to *fewer* births given the cost and expense, and the feeling among many that it would be wrong to give some children the treatment but not others.
We already see this among people who make objectively good incomes but feel like they can't have, say, a third child because they can't afford triple private school fees.
Yeah, I think Arbituram has nailed it - if anything, it's going to depress fertility rate / elite populations until it gets cheaper. But the only way through is early adoption and the natural technological advance that the economics will drive, to make it cheap and ubiquitous.
Makes sense. Maybe by the time it's cheap and widely available, there will be regulations designed to prevent overpopulation (a la China's now defunct one family-one child policy).
In theory a society of highly intelligent superbeings would have the self control to practice abstinence when they don't need kids and hyper sex with 1 session leading to pregnancy when they do need them.
1) Two people with Down's cannot, in fact, have children. The men are infertile (couple of odd exceptions here , but very much exceptions) and the women can have (normal) children but appear to be subfertile anyway.
2) You're overlooking the biggest barrier of all in my view, which is that IVF sucks, even if it was free. Any use of these technologies implies IVF, which is painful, time consuming, expensive, with remarkably poor success rates, even now, let alone the emotional rollercoaster of it all. I've been lucky enough to not need it despite starting fairly late with my kids, but I've seen people go through it and am not tempted. Don't think it would be worth a minor gain.
Ah, thanks - I didn't know that Down's syndrome people were infertile / subfertile.
I definitely agree that IVF sucks for everyone involved. Technologically, I think we actually could get to a point where we could alter embryos in-situ with lipid nanoparticles or AAV's, but you need the technological growth curve that will be driven by early adopters willing to undergo IVF to get there.
I think enough people will be willing to do it, though. If you look at the ridiculous Red Queen's Races people already run (we need to get on the waiting list for the right exclusive pre-school 3 months into the pregnancy, or precious Jaden's chances of getting into Harvard are *ruined!*), I think that there'll be plenty of early adopters willing to undergo the trouble of IVF. Not to mention that the elite folk who will be able to afford it at first are probably already older and lower fertility and may need IVF to begin with.
Nice - definitely a much easier (and cheaper) IVF experience if you don't have to do the crazy amounts of hormones to stimulate your eggs.
And it sounds like they're working on induced pluripotent stem cells ultimately turning into eggs, which would be the REAL game changer. Then you don't even need a surgery to harvest eggs at all.
>working out is mostly unpleasant and boring as hell as we conceive of it and we need to stop pretending otherwise. Once we agree that most exercise mostly bores most people who try it out of their minds, we can work on not doing that.
I'm mildly surprised to hear you say this, seeing as you’ve written about once being a regionally competitive strength athlete. Was there really no pleasure in the training? Or are you saying that is mostly the case for the typical gym goer?
Wow, did I say that? I don't actually remember saying that, which post / comment was it?
But I think I would agree at the population level - workout adherence is *really* poor in gen pop, and the very best interventions out of many hundreds tried generally increase physical activity something like 5 min a day in aggregate.
But yes, for myself, I'm one of those irritating people who loves working out and gets antsy when I haven't been able to lift or do sprints in a while. My favorite day of the week has always been Friday because a) the gym is empty except for hardcore regulars, so it's easy to get a platform, and b) it's deadlift day, and I'm gonna move some heavy iron!
Sorry, but I obviously replied to the wrong post. And to cap it off, I couldn't easily find the original. But you did set my mind at rest :-D
Oh, and a tip of the hat for your advice on treadmills. I was treadmill-curious for some time, and one of your articles made me finally give in to my dark desire. Brilliant! When I work from home, which is most days, I'm putting in 20-35km (the low end is surprisingly easy, but 35km--about 22 miles--is just crazy). And while this is more datum than data, I'm finding that burning 2-3k of active calories, coupled with 1k calories of intake, works wonders. So far, this is shaping up to be my easiest New Year's diet ever. Thanks!
> Oh, and a tip of the hat for your advice on treadmills. I was treadmill-curious for some time, and one of your articles made me finally give in to my dark desire.
Nice. I definitely feel like folk who WFH are the very best candidates for treadmill-desking.
Thanks for relating that it's been good for you, I'm really happy to hear it - that's the whole reason I write these things! Because somebody out there might hear it, make a change, and materially improve their health and quality of life. :-)
I basically agree with all this (except maybe the part about red queen effects, eugenics would be more palatable/less zero sum in a society that was already very egalitarian imo). But isn't the bullet you have to bite at the end of this line of argument that we should just do selective breeding, since selective breeding >>>> gengineering >> embryo selection> random natural genome? Not sure if you think there's a slippery slope here. Seems like a selective breeding advocate could accuse a gengineer of making bad genomes, falling for the naturalist fallacy etc. just as easily as this essay accuses normal-eugenics-free society of those things.
> But isn't the bullet you have to bite at the end of this line of argument that we should just do selective breeding, since selective breeding >>>> gengineering >> embryo selection> random natural genome? Not sure if you think there's a slippery slope here.
I'd actually put gengineering >>> selective breeding, for a couple of reasons.
1) You're preserving individual choice around who people pair up with and have kids with, which is important morally and philosophically (and logistically)
2) You are increasing the population average while preserving that (because the gengineering gives you a buff)
3) Lack of consensus-agreed endpoints. Do you max IQ? Conscientiousness? Good health? Mental health? Work ethic and discipline? You want all those things. And at a societal level, you need some diversity in endpoint maximization, because macro changes make different endpoints more valuable and you can't predict everything the universe will throw at you.
4) Logistically, you're never going to be able to limit breeding to only the top 2% or whatever, as people would never agree to it, it would require a massive reorganization of society in its entirety, to produce enough for the next generation you'd require uterine replicators and working "child raising at scale" social technologies, it would require authoritarian top-down control at a level that would constantly be rebelled against, and so much else.
Versus gengineering is something we could literally do tomorrow, with all existing societal, moral, and philosophical systems in place, and with people still able to assortatively mate and freely choose who they have children with, while still increasing population averages on desirable traits.
We're already doing selective breeding, no? See the point re: assortative mating, which has gotten dramatically more intense.
I wasn't thinking in terms of eugenics or children when I met my wife, but I simply wouldn't have been interested in marrying someone who wasn't top 1% intelligence (which I did) just for lifestyle/personal marital satisfaction reasons. Sure, we use proxies, like "went to Ivy League/Oxbridge, reads difficult books for fun, can hold a complex debate on a topic immediately after being introduced to it", but it works out to the same.
I mean selective breeding the way it's done on livestock, i.e. picking out a handful of individuals with the best genomes and having them sire the whole next generation. Not normal assortative mating that doesn't shift the population average and where each individual passes on whatever genome they happen to have inherited.
Selective breeding gives everyone in the next generation the best available genes. For assortative mating, on average you assort with someone with the same quality genes as you so your kid has the same quality genes in expectation.
In that case , agreed they are different, but selective breeding is, almost by definiion, non consensual.
Given how polygenic IQ is, isn’t embryo selection much more viable for IQ enhancement than gene editing? With gene editing you’d have to go and directly edit tons of SNPs which has who knows what side effects. For embryo selection it’s just the normal IVF process.
Anyway this is a great post and I learned a lot, but I think for improving complex polygenic traits (intelligence, and many disease risks) gene editing is much less safe and effective than embryo selection. Yes needing less sleep and not having body odor is good.
Yeah, I pretty much agree. Right now we're lacking on both knowledge and praxis fronts in terms of polygenic editing. And although we can probably bang together some sort of massively parallels CRISPR operation, and I even know somebody working on this, after we do that we'll need lots of data generation in animals to really suss out the pleiotropy and linkage landscapes.
I mean fortunately, at the high level, we think that most important things like IQ and conscientiousness are generally around 80% additive - at most we should expect 20% error due to pleiotropies and other things.
Also, I think I undersold the sleep thing here, because I was focused on the whole picture of all the cool SNP's we know about already. Both at the individual and societal level it would be an absolutely massive effect size.
It's literally like adding 10% to your entire conscious life (or more if GWAS-ed), and would likely have concomitant impacts on productivity, earnings, and whatever else too.
If you offered this to the broader pop, imagine everyone in your country having 10% more free time to do whatever they wanted. People spend more when they're awake, and some will choose to work more in some of those hours. Imagine an entire economy growing by an additional 2-5% over a few years with exactly the same population and per-worker productivity, because this time was freed up. These impacts are huge, and nothing else approaches that scale.
I agree that the sleep thing is huge! That's very good. I support doing that, yes. Obviously we should first check that people with this mutation don't live much less long or have a massive susceptibility to some disease or whatever. But probably you're right that it has few if any drawbacks. I'd have a few reservations and things I'd want to check and double-check first but it certainly sounds like a good idea.
Of course it also helps that embryo selection for IQ is not banned in a lot of countries and would be very similar to what we already do, whereas gene editing is more restricted. Not that you have to pick at most one of them. Ideally we'd do embryo selection (even better if it's with IVG and from many embryos, which hopefully is not too far out) and then judiciously CRISPR out a few bad SNPs. If you just CRISPR'd all or a bunch of the IQ SNPs, it's unclear how big the effect would be, because there are correlations and not all the SNPs are causal, and it could cause a miscarriage or all sorts of other bad health effects and unforeseen consequences. 20% pleiotropy is not much but you could get a lot of miscarriages or just all sorts of other weird effects. Not to mention the risks of off-target edits. For a dozen or two SNPs, it's probably safe and would be good, yes. Embryo selection is much safer, it's genetically equivalent to having tons of kids and then having only the smart and healthy ones give you grandkids, so it's not so different from natural selection. In general the pleiotropy with IQ is positive. More IQ tends to mean more health as well. Embryo selection is obviously good. But when you just flip hundreds of SNPs, who knows what the hell could happen. Probably a miscarriage. Yes, we should test it on animals first. It's certainly less ethically problematic than factory farming.
It was not mentioned but I highly recommend looking up the culture around Yao Ming. Yeah, that Chinese basketball player who's over seven (7) feet tall.
"Yao had essentially been bred. Both his parents played basketball. His 6’2 [different height from Wikipedia -Razib] mother, Fang Fengdi, perhaps the tallest woman in China, had been married to an even taller man. She had served as a Red Guard during the height of the Cultural Revolution and had been an ardent Maoist. She enthusiastically participated in the glorious plan of the local government to use her and her husband to produce a sports superstar. The Shanghai authorities who encouraged the match had gone back several generations to ensure that size was embedded in the bloodline. The result was Yao, a baby behemoth who just kept getting bigger."
Ethics in China? ha. Trying out cool stuff? yes!
Great example, I totally agree, and have actually used the Yao Ming example elsewhere as a suspected eugenics case.
One interesting thing - if you ask about Yao Ming between Deepseek (Chinese) and GPT (USA), it gives you noticeably different answers:
Deepseek:
"Yao Ming's parents, Yao Zhiyuan and Fang Fengdi, both had distinguished careers in basketball before they met. They were introduced through their shared involvement in the sport. Yao Zhiyuan was a former professional basketball player in China, and Fang Fengdi was a prominent player on the Chinese national women's basketball team. Their mutual passion for basketball brought them together, and they eventually married. Their shared background in the sport likely played a significant role in their connection and in fostering Yao Ming's own basketball career."
GPT:
"Yao Ming's parents, Yao Zhiyuan and Fang Fengdi, were both professional basketball players in China. They met through the Shanghai basketball community in the 1970s. Yao Zhiyuan was 6'10" (208 cm), and Fang Fengdi was 6'2" (188 cm), making them two of the tallest players in the country at the time. The Chinese government encouraged their pairing, believing that their genetic combination would likely produce a tall and athletically gifted child.
They married in 1977, and Yao Ming was born in 1980. Fang Fengdi often said that they didn’t marry solely for basketball purposes, but they were definitely aware of the expectations placed on them."
At least interesting that Deepseek doesn't mention the government or "expectations" at all while GPT does.
I'm assuming that came from one of Razib Khan's posts? I'd be interested in a link if you have it.
Yeah, it was Razib Khan.
https://www.discovermagazine.com/the-sciences/was-yao-ming-bred
There are scattered bits of info elsewhere too. Another site said that the project was looking for Yao Ming's grandfather but took too long so they settled with Yao Ming's father.
https://mainstand.co.th/en/features/5/article/3252
If it was China they would probably have taken Chris Lagaan's sperm and tried breeding a cadre of 250 IQ ultramen. These sorts of projects make for good reading and some optimism for the future, albeit clumsy.
I'm not so sure. That was 50 years ago. China is the country that put He Jiankui in jail, encourages its smartest people to move to antinatalist cities, had a one-child policy that only applied in the cities, does not allow single women to freeze their eggs, and so on. They are pretty traditional and definitely not the world's biggest techno-natalists. American tech bros, Israelis, Singaporeans, etc seem like better candidates.
Also Chris Langan probably isn't *that* smart. He is certainly a smart guy but kind of a "Mensa bum". He's no John von Neumann, that's for sure. I don't know how accurate these IQ tests are on the far right tail anyway. Weschler maxes out at 160 if I recall correctly.
> "This is because often, all good things are associated with each other (like IQ being positively associated with things as diverse as looks, good health, and reaction speed), so I think the odds of catastrophic interactions are relatively low"
But such association came from natural selection or evolution, not from modification, do you think it is possible that if genengineering of humans happen, the association would be inversed? (sort of genengineering derived syndromes)
> But such association came from natural selection or evolution, not from modification, do you think it is possible that if genengineering of humans happen, the association would be inversed?
Yes, this is a great point, and it's certainly a real concern (hence the need for proteomics pros and analysis). My point overall is that we know that "multifactorial optimization on multiple fronts" isn't *impossible,* and if all good things ARE correlated, there's an actual optimization peak out there where you don't necessarily need to trade off multiple good things for each other.
There's probably some problems when you get way out in the sigmas - like if you go for a 4-6 stdev bump in any given characteristic, you probably start needing to trade off on something else at that point.
But is it possible to have Nobel prize winners who won Olympic medals earlier, for example? I think so.
We've never had this in a person today - the closest in terms of "both mental and physical excellence" that I can think of are probably Niels Bohr (Nobel winner whose brother won a Silver Olympic medal for soccer, and they used to play on the same team), Dolph Lundgren (Fullbright scholar at MIT, European karate champion, and famous bodybuilder / actor), and A.V. Hill (Nobel prize winner for physiology and how muscles worked in 1922, who ran a 4:45 mile when he was younger).
I’m excited about the possibilities, but this gives me pause:
>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.
I do note that you add the following, so we are probably on the same page:
>I’m sure laws would be written very quickly if this ever actually happened - arguably, gengineering would increase legibility and actually give us a way to stop psychopathic parents more often.
I’ll admit to being a bit torn between wishing to see a system that allows for a certain amount of experimentation, and afraid that we’ll see kids with glow-in-the dark skin, tails, and retile eyes. I suppose that we will first witness this amongst the ultra-rich, and they must have the interests of their kids front and center, right? Then I remember that they can name their child “X Æ A-Xii” and I start to doubt myself.
Here’s a grab bag of random thoughts on the matter:
• Will a few large companies have secret formulas that have clear results but at the expense of leaving the kid sterile, a la Monsanto?
• Super smart kids in a world that doesn’t need their abilities might be a very dangerous place.
• Might governments mandate certain traits?
• Might hackers plant certain traits, even just for the lulz?
• If homosexuality is genetic, will these people die out or at least find themselves an endangered species?
• Will the deranged sports-obsessed parent be taking this to another level?
• Will the need to study all effects require studies that run eighty years?
• How many people will opt for clones, perhaps with a tweak or two?
• What would it be like to have kids vastly superior to their parents? Might there be a backlash?
I’m not expecting any answers, just showing where my mind starts to wander around the issue. In any case, another great article!
> I suppose that we will first witness this amongst the ultra-rich, and they must have the interests of their kids front and center, right?
Yeah, I think this mitigates a lot of the risk - these will be people spending significant sums with an eye towards their children's overall capabilities and future success.
But you know, a world where people have mentally controlled chromatophores (which I've personally *always* wanted) and glowing eyes and whatever would be pretty metal too. Still strictly better than the status quo!
> Will a few large companies have secret formulas that have clear results but at the expense of leaving the kid sterile, a la Monsanto?
Be pretty hard to convince parents paying $$$ to give up grandbabies, those buffs would have to be CRAZY high. And if they ARE that high, the additional expense of IVF sounds pretty small, as long as the inclination to have the kids is there.
> Super smart kids in a world that doesn’t need their abilities might be a very dangerous place.
Yeah, I'm short "intelligence" overall, per my Optimal Descendant Strategy post. Conscientiousness / Discipline is the way of the future.
> Might governments mandate certain traits?
I'm less worried about this, at least in the short term. It'll certainly be a market-driven phenomenon at first because so many govs are regressive prudes about this. So the one place (Prospera or Japan or Thailand or wherever) that actually offers it will have super rich parents coming in from all over the world to do the editing and implantation, and then will go back to their homes to have their wunderkinder.
But ultimately, some govs may need to step into "cranking out babies at scale" for Fertility Crisis reasons, then who knows what they'll get up to?
> If homosexuality is genetic, will these people die out or at least find themselves an endangered species?
Also a risk for the deaf, blind, Downs syndrome, etc. And people definitely want grandbabies, so it might happen - heritability is 30-50%, surprisingly.
But parental choice is parental choice, you know, I think you have to respect that. Also, all the Chinese folks who had girls instead of boys are *cleaning up* right now descendants-wise. The imbalanced gender ratio vastly favors women, and they get materially higher quality mates / children as a result (and this is why in general, gender ratios are 50/50 in practically all species with 2 genders). So if there are second-order major advantages to a given genetic strategy running counter to the average trends, I think there will always be some River parents who identify and play those plays.
> Will the deranged sports-obsessed parent be taking this to another level?
Lol, it's me! I'm the one who'd be cranking them full of ultra-dense bones and EPO hyper-response and all the VO2max heritage study buffs!
> Will the need to study all effects require studies that run eighty years?
If so, we'll literally never do it, because we'll be post-human or biotech transhuman or something with high probability over a few 80+ year study cycles.
All that's required is at least one place somewhere in the world to offer credible and verified results in kids, though - rich parents from all over the world will be coming to them. Those will be "higher risk appetite" parents for the first generation or two, I'm sure.
> How many people will opt for clones, perhaps with a tweak or two?
I mean, good for them? I personally think this should be a choice too. Sexual reproduction has a bunch of tradeoffs! I've got a draft post I'm thinking about along these lines, actually.
> What would it be like to have kids vastly superior to their parents? Might there be a backlash?
Yep, I've thought of this, too.
Assuming anti-aging technology keeps some of us around, our giant, Olympic medaling, von Neumann adonis grandkids probably won't want to talk to dumb, tiny, ugly, agonizingly slow-processing and speaking grandparents, they'll invent new, higher bandwidth ways of communicating and leave us in the dust to be tended by machines and infinite VR Heaven. I still count that as a major win.
I personally *want* my g^x grandkids to be giant, athletic, von Neumann adonises, thanks very much, and I'm willing to risk that there may be unintended consequences because the top-line goals are so valuable.
I hate to be the one to break this to you, Zvi: you’re a machine! I’m afraid all this all this gengineering stuff is reserved for us humans :-D
More seriously, and without trying to give the impression that I have any delusions of how this will someday play out, I think the issue of parental choice might come to be the real battleground. It puts me in mind of the never-ending debate around abortion, where many people are generally chill but grow squeamish beyond a certain point. One side of radicals will simply not hear of it, however, and wind up driving opinion towards the opposite end.
In any case, you’re providing great food for thought and I look forward to future posts on this subject. The combination of gengineering with advances in AI makes the future illegible to me. I can, for example, imagine life in a post-scarcity world, but getting to that world throws up mental roadblocks to my mind. Somewhere mid-journey, when AI starts splitting the world into winners and losers, but unevenly between countries and within countries, I think democracy will be thrown in crisis. And capitalism, too, starts to falter when the concept of money breaks down. How will gengineering play out? Will it fly under the radar until, decades from now, we find that the ultra-rich have silently created a race of Khans while the remaining top 10% were busy correcting a few minor spelling mistakes in the book of life? I’m hoping to follow your ever-evolving roadmap on how you suspect the future will unfold. Thanks again!
Very persuasive rebuttals to objections to genengineering, but what about overpopulation? The technology would yield more live births, lower infant mortality rates, longer life spans, etc.
That is, unless it remains prohibitively expensive to all but the uber-rich...
I think it would lead to *fewer* births given the cost and expense, and the feeling among many that it would be wrong to give some children the treatment but not others.
We already see this among people who make objectively good incomes but feel like they can't have, say, a third child because they can't afford triple private school fees.
Yeah, I think Arbituram has nailed it - if anything, it's going to depress fertility rate / elite populations until it gets cheaper. But the only way through is early adoption and the natural technological advance that the economics will drive, to make it cheap and ubiquitous.
Makes sense. Maybe by the time it's cheap and widely available, there will be regulations designed to prevent overpopulation (a la China's now defunct one family-one child policy).
In theory a society of highly intelligent superbeings would have the self control to practice abstinence when they don't need kids and hyper sex with 1 session leading to pregnancy when they do need them.
Two points:
1) Two people with Down's cannot, in fact, have children. The men are infertile (couple of odd exceptions here , but very much exceptions) and the women can have (normal) children but appear to be subfertile anyway.
2) You're overlooking the biggest barrier of all in my view, which is that IVF sucks, even if it was free. Any use of these technologies implies IVF, which is painful, time consuming, expensive, with remarkably poor success rates, even now, let alone the emotional rollercoaster of it all. I've been lucky enough to not need it despite starting fairly late with my kids, but I've seen people go through it and am not tempted. Don't think it would be worth a minor gain.
Ah, thanks - I didn't know that Down's syndrome people were infertile / subfertile.
I definitely agree that IVF sucks for everyone involved. Technologically, I think we actually could get to a point where we could alter embryos in-situ with lipid nanoparticles or AAV's, but you need the technological growth curve that will be driven by early adopters willing to undergo IVF to get there.
I think enough people will be willing to do it, though. If you look at the ridiculous Red Queen's Races people already run (we need to get on the waiting list for the right exclusive pre-school 3 months into the pregnancy, or precious Jaden's chances of getting into Harvard are *ruined!*), I think that there'll be plenty of early adopters willing to undergo the trouble of IVF. Not to mention that the elite folk who will be able to afford it at first are probably already older and lower fertility and may need IVF to begin with.
https://www.businesswire.com/news/home/20241216400051/en/Gameto-Announces-World%E2%80%99s-First-Live-Birth-Using-Fertilo-Procedure-that-Matures-Eggs-Outside-the-Body#:~:text=Fertilo%2C%20Gameto%27s%20lead%20investigational%20program,vitro%20fertilization%20and%20egg%20freezing.
Just saw this. If this scales, consider me updated upwards on the plausibility of mass adoption of gene editing.
Nice - definitely a much easier (and cheaper) IVF experience if you don't have to do the crazy amounts of hormones to stimulate your eggs.
And it sounds like they're working on induced pluripotent stem cells ultimately turning into eggs, which would be the REAL game changer. Then you don't even need a surgery to harvest eggs at all.
Maybe not poisoning kids from day one is a better approach 🤷🏻
>working out is mostly unpleasant and boring as hell as we conceive of it and we need to stop pretending otherwise. Once we agree that most exercise mostly bores most people who try it out of their minds, we can work on not doing that.
I'm mildly surprised to hear you say this, seeing as you’ve written about once being a regionally competitive strength athlete. Was there really no pleasure in the training? Or are you saying that is mostly the case for the typical gym goer?
Wow, did I say that? I don't actually remember saying that, which post / comment was it?
But I think I would agree at the population level - workout adherence is *really* poor in gen pop, and the very best interventions out of many hundreds tried generally increase physical activity something like 5 min a day in aggregate.
But yes, for myself, I'm one of those irritating people who loves working out and gets antsy when I haven't been able to lift or do sprints in a while. My favorite day of the week has always been Friday because a) the gym is empty except for hardcore regulars, so it's easy to get a platform, and b) it's deadlift day, and I'm gonna move some heavy iron!
Sorry, but I obviously replied to the wrong post. And to cap it off, I couldn't easily find the original. But you did set my mind at rest :-D
Oh, and a tip of the hat for your advice on treadmills. I was treadmill-curious for some time, and one of your articles made me finally give in to my dark desire. Brilliant! When I work from home, which is most days, I'm putting in 20-35km (the low end is surprisingly easy, but 35km--about 22 miles--is just crazy). And while this is more datum than data, I'm finding that burning 2-3k of active calories, coupled with 1k calories of intake, works wonders. So far, this is shaping up to be my easiest New Year's diet ever. Thanks!
> Oh, and a tip of the hat for your advice on treadmills. I was treadmill-curious for some time, and one of your articles made me finally give in to my dark desire.
Nice. I definitely feel like folk who WFH are the very best candidates for treadmill-desking.
Thanks for relating that it's been good for you, I'm really happy to hear it - that's the whole reason I write these things! Because somebody out there might hear it, make a change, and materially improve their health and quality of life. :-)