title: How Much Can Embryo Selection Increase IQ?

video: https://www.youtube.com/watch?v=WxxMOi1Ph8I

venue: RiotIQ podcast

speakers: Russell Warne and Jonathan Anomaly

Video description

In this episode, Dr. Russell Warne talks with Dr. Jonathan Anomaly about embryo selection, polygenic scores, and how modern biotechnology can predict differences in traits like IQ, health risk, and longevity. They discuss how polygenic scores work, what kind of IQ gains are realistically possible when selecting embryos, and why most parents do not simply choose the embryo with the highest predicted IQ.

They also cover disease risk prediction, trade-offs between intelligence and health outcomes, personality traits, validation studies, CRISPR and gene editing, long term social effects, ethical concerns, and what embryo selection might mean across multiple generations. This is one of our most detailed episodes on the future of human genetics and intelligence. If you enjoyed this episode, subscribe for more conversations about IQ, intelligence, and human cognition.

Interested in embryo selection for your children? Check out Dr. Jonathan Anomaly's company Herasight: https://www.herasight.com/

IQ testing: https://riotiq.com/

https://x.com/RiotIQ

https://linktr.ee/riottest

Introduction and Guest Background

Hi, and welcome to another episode of the IQ and Human Intelligence podcast. Today, my guest is Jonathan Anomaly, Director of Research and Communications at Herasight. We're going to have a great talk today about embryo selection and how intelligence fits into the wider biotechnology landscape. Thanks for joining me on the program today, Jonathan.

Yeah, I appreciate your having me.

You're the first guest we've had on the podcast who is not primarily an intelligence researcher. So please tell me about your training and your background so that we can understand how you approach these issues.

Sure. My training is mostly in philosophy and economics, and I taught in PPE programs. So that's more familiar to British listeners than Americans, but PPE is philosophy, politics, and economics. And the point of that discipline, and it is its own kind of discipline, is to both be able to model human behavior, to sort of understand how people respond to incentives, whether they be economic incentives, psychological incentives, and so on. But also, in the end, to be able to do something with that.

So policymakers in England often study PPE because they want to understand the world and change it. But of course, it helps to be able, before you try to change the world, to sort of channel Marx's view. You know, he sort of criticizes philosophers as having only tried to study the world and not change it. The point is to change it, he says. But actually, the point is both, right? Because if you want to change it, you also want to get things right first, human motivations and so on.

And so, yeah, one of the sort of essential ways in which economists look at the world is people make rational decisions in some narrow sense of that word, in the sense that they understand their own preferences, their own budget constraints, their time constraints, and they choose accordingly. And we just define that as rational in economics, more or less efficiently pursuing your objectives.

But when they do that individually, sometimes the collective pattern that emerges, the emergent order, as we say, is good from a social standpoint. Usually markets lead to those outcomes, right? You're each pursuing your own interests, constrained by various things, and the collective upshot is usually better for everyone. But sometimes it's not. Sometimes it's worse for everyone. You can think of pollution as a byproduct of mutually beneficial market exchanges and so on.

Anyway, so I thought a lot about that, taught about that, wrote articles in that area in that intersection. And actually, to tie it together with what we're about to talk about, I began applying these kinds of basic principles in the realm of, to the problem of antibiotic resistance. It's a kind of collective action problem. I consume antibiotics when maybe I'm feeling a bit down, but actually it's probably a virus, not a bacterium. Who cares? You know, I have some antibiotics just in case. But of course, when we collectively do this, we overconsume them, produce antibiotic resistance as a side effect. Same thing for animal agriculture, where half of all antibiotics go probably wastefully. And the collective upshot is actually a world that's a lot worse, more resistant.

So fast forward, then I wrote a book and a few articles on what would happen when individuals had access to reproductive technologies. And I was thinking mainly about gene editing, thinking about the far future. But then I started realizing kind of halfway along the way, actually, you know, embryo selection with polygenic scores, which we'll talk about today, that's already kind of here. It's right around the corner. So I wrote a book on this and then I got tapped eventually in 2021 by a company, a nascent company that was doing polygenic scores, you know, for embryos and they asked me to join them and I did. So anyway, that's my academic background and how that ties in with what we're going to talk about today.

Polygenic Scores and Embryo Selection Explained

Now, you've thrown around some of these terms, polygenic scores, embryo selection. Explain to me like I'm 10 years old what polygenic scores are and how they're used in embryo selection and what embryo selection is in the first place.

Yeah, yeah, good. Exactly. We throw around these terms like embryo selection, as everyone knows. Yeah, so the first thing I guess to say is that women who are either infertile or a bit older, or they have hereditary diseases that they know about in their family, have been doing in vitro fertilization for almost exactly 50 years now. And yeah, so the first babies were born right around now, 50 years ago, first in England. I believe '78 was the first so-called test tube baby.

Yeah, that's right. So anyway, when you do in vitro fertilization, you basically stimulate the ovaries to produce eggs and combine the egg with sperm, and hopefully you get some embryos, enough to have a child with. And one of the reasons we associate in vitro fertilization with twins with having twins or triplets or even more is back then at the beginning they would just you know implant a bunch of embryos and see if any of them took you know and sometimes all of them did and it's like oh you've got triplets now or sometimes none of them did you know.

But what happens about 20 years after the advent of this is in the 90s the first genetic testing is done and that's for chromosomal abnormalities. This is called—it's like Down syndrome and—

Exactly. Karyotypes, okay.

Correct. So this is called PGT-A, preimplantation genetic testing for aneuploidy, which just refers to chromosomal abnormalities.

Now, you mentioned Down's because it's quite common. Well, it's commonly understood, but actually the majority of these are incompatible with life. And so the idea is this would, in the vast majority of cases, if you implanted one of these embryos, the body will actually reject it. It's a so-called spontaneous abortion.

And so one reason you don't see twinning anymore is this PGT-A test is pretty accurate. And so what they do is then test the embryos, stage one, right, to see if they have one of these terrible chromosomal abnormalities. And, you know, they're going to set those aside. So that we don't waste the implantation on an embryo that will never ever be viable.

That's right. And so instead, you're going to, of course, if you're going to implant an embryo and you have multiple, you're not going to deliberately implant the one with chromosomal abnormalities. So now you don't get much twinning because, generally speaking, you implant one and it'll work. And if it doesn't, you would implant another one. But, yeah, those are the first genetic tests.

And so you've already got, quote, embryo selection really right off the bat. There's the embryologist looking at it visually. It's not very accurate, but you can kind of look at the morphology. Morphology, it's a bit of a guessing game. And then you look at these chromosomal abnormalities.

Then really shortly after that, you get monogenic testing. So for the thing that you mentioned, so for example, Tay-Sachs. Sickle cell anemia. Sickle cell, any of these. These one gene diseases you can inherit. Precisely. And that's been done since the 90s. And that's already embryo selection. You're in a forced choice situation. You know, nobody would deliberately implant the one with a really nasty disease.

So what we're doing now is just the next step in this in a way. IVF is old even embryo selection is 30 years old. But what we can now do and we get into the details I just give you a rough outline is you can more or less scan the whole genome of the embryo and you either need to sequence the embryo directly, or there's another method we can talk about, and run them through your polygenic scores. And you can gauge with those polygenic scores, what's the likelihood that embryo 3 will get type 1 diabetes or breast cancer? And of course, we're doing intelligence and we'll get into that.

But I suppose before we do, let me just say, since you asked, what is a polygenic score? How does it work? And the rough outline is that there are a number of countries, including the U.S., that have one or more biobanks. And what is a biobank? It's basically hundreds of thousands of people, typically, sometimes tens of thousands. And these are, of course, opt-in. You know, you can consent into these.

And what will happen is, whether it's a private or a public biobank, they scan your genome, you opt in, you know, and, and when you have access to enough genomes, let's say, let's say 100,000, half a million, and then you phenotype those people, meaning you measure the outward expression of those genes in the form of height, or it could be, these are simple examples, eye color, hair color, those are really easy to measure. Or an IQ score. Or an IQ score, which is much harder to measure.

Right. And similarly, schizophrenia, which would also be in a way harder to measure. You want to be sure you're not just asking people, hey, man, you got schizophrenia. You better get, you know, a psychiatrist on the scene. Are you hearing voices?

Yeah, that's right. Yeah. Sometimes. But usually it's just my wife, you know.

So, yeah, in order to really build a great biobank, like in Sweden or the UK or Japan and the US, you want to not only have accurate genotyping, but also accurate phenotyping. And some of these biobanks, for example, they do IQ test people, but it's not a very good IQ test. It's not Riot's IQ test. It's a five or ten minute one. And you can get some signal from that, but not great signal.

And then there are other ones that are maybe smaller biobanks, but they're, they're properly IQ tested and get a lot of signal from that. And so there are these trade-offs. And to be fair, if you need to collect data from 300,000 people, you don't want to do it on just one or two phenotypes.

And to be fair, if you need to collect data from 300,000 people, you don't want to do it on just one or two phenotypes. So now suddenly you're opening up everything and you're going to ask about health and mental health and all personality and all these other things. It might not be feasible to give an hour-long IQ test. So we do have to be fair. It's not from lack of interest. It's not from problems with the tests. If you're going to collect all this data, we need to be considerate of people's times. You're not going to find 300,000 people willing to sit through a week of medical and psychological tests. Time and money, exactly.

It's their time. It's the government usually paying for it or some private, you know, whatever foundation. But yeah, to sum it all up, if you have access to enough people and you have good enough phenotype data, again, height is the easy example. IQ is the hard example. Well, then you can scan the whole genome, look across the billions of base pairs that comprise our genome, and you'll see some common hits across all those people that are associated with diabetes or intelligence or height. And you sum those up and you get a polygenic score. There's just a variance more common for this disease or that disease.

Correct. Correct. So that's it. And then you can use those. I mean, if you have those polygenic scores, I mean, I like to say there's really three ways of using them. And, you know, a lot of enthusiasm over embryo selection. That's what we're talking about. But the truth is they're already used to explain the human past to some extent. David Reich at Harvard is explaining human evolution. And, you know, when did people migrate to different places by looking at ancient DNA? And then you can kind of, you know, you could either run them through polygenic scores to see how tall were they actually at different or what was their genetic potential? Because their height actually depends on nutrition and things, too, just like with intelligence.

But yeah, you can use these to help explain the human past, to help predict for yourself, like what's my likelihood of getting Alzheimer's disease, that sort of thing, and for embryo prediction. So, yeah, I mean, I took a 23andMe test before they went out of business. And one of the things I was interested in was my risk for health scores. And so that's based on my risk for health conditions. That's based on polygenic scores. I also wrote an article that came out about three years ago telling educators hey polygenic scores are coming. You better be ready. Your school district is coming.

And so, yeah, there are a lot of interesting things. David Reich's rewritten our prehistory almost single-handedly with DNA and these polygenic scores.

Predictive Gains and Validation for Traits Like IQ

And we've mentioned some of these traits that we can use to predict either in ourselves or in potential children in embryos. What is the degree of benefit you find for different traits?

Yeah, so it's going to vary based on how good your polygenic scores are. So that's point number one. And that in turn, apart from depending on, well, you need some talent, you need some really smart people, speaking of intelligence, to interpret the data because it's not going to interpret itself. It's not something you can just run through an AI.

But on top of that, it's going to depend on how good your phenotyping is and on data across different ancestry groups. So, for example, I mentioned, let's say, the UK Biobank. Well, you know, with that alone, we can't get a very good polygenic score for intelligence, but we can create a good one for something else like height. So we're going to have to look at other biobanks. And some of those are going to have better data on other ethnic groups or, you know, better phenotype data on intelligence than height or whatever.

And so, you know, there's no general answer to how good these scores are. It depends on the data to which you have access, how good your team is putting them together. And then, and this is the crucial bit, you know, how good your validation studies are.

And, you know, one of the people on our team, Alex Young, is, you know, someone who's pioneered this idea of family within family studies. And the really simple version of that is, you know, in the context of embryo selection, rather than explaining the human past, let's say, or individual prediction about your health, you know, you're talking about, you know, embryos that share half their DNA from each parent, and they're just as related to each other, more or less, right, as existing grown siblings are, right?

And so the gold standard for figuring out how good your polygenic scores are for embryos is to run the DNA of existing adult siblings through your polygenic scores and then predict and see how good is your predictor for height for eye color. Actually we not doing eye color right now but for diabetes whatever the case is, and they're going to vary, right? And so what our job is, and I think the really responsible thing that companies need to do is show exactly how much of the genetic variance that they can explain with their polygenic scores, and that's going to vary by trait.

I see. So for IQ, typical family comes in, they're doing this in vitro fertilization IVF procedure, and they come to you and your colleagues and say, we want to select the smartest possible baby out of these choices that we have in the Petri dish. What sort of gain could they expect for IQ?

Thank you. Exactly. So we've got, thanks to Tobias Bjørnebo et al., but mainly Tobias, the best genetic IQ predictor in the world. And it's not even close. He spent a lot of time on this and a lot of time on carefully validating it across ancestries and so on. And what we've got now is about a 0.5 correlation. That's roughly, it's 0.47, I believe, within family. Wow. 0.54 in the general pop.

And what that means is that if you have, for example, three embryos and you're a European couple, we can talk about Asian, African, et cetera, but just to take that case, with three embryos, we can predict a spread of about eight points between the highest and lowest scoring embryo, which is about half of the total difference.

Now, with 10 embryos, we can predict about a standard deviation, about 15-point difference between the highest and the lowest scoring embryo, and that's about half of the actual genetic variance. And so the best way of explaining it is if you had 10 siblings, you plus nine, good Mormon family, and you've got 10 siblings. Sorry, I grew up in a family of five boys. Okay, well, it's pretty good. You're half. Half, yeah.

So, you know, the actual spread in intelligence on average for 10 siblings is going to be a little above 30 points, which is pretty wild when you think about it. Between the highest and the lowest. It's crazy. And we can explain about half of that. So it's really good at ranking. It's not going to get your exact IQ. You know it never will because IQ is not fully heritable and will never get every single genetic variant. And we had some episodes in the past talking about environmental influences as you can imagine. Absolutely. Nutritional.

So you're saying if between out of 10 embryos, the expected or predicted IQ range is 15 points. That means implant the predicted. We don't know for sure, but the predicted smartest embryo compared to the predicted least intelligent embryo will get about a 15 point gain. With 10 embryos and with an average compared to an average embryo select at random about seven and a half points and i don't know about you but i'm not going to turn down a seven and a half point IQ boost.

Yeah although you know hold your horses because the reality is even our you know many of our earliest customers we were in stealth for a few years you know really kind of trying to perfect the product and add new predictors and so on we've got a whole bunch of disease predictors of course. And so we had, you know, the usual kind of Silicon Valley types, people who really care probably more than average about IQ, probably know more than average about it, know some of the benefits which we can talk about, associate with it.

But even they, Russ, even they very rarely just selected on IQ. In fact, I don't know of a single case where they did. In fact, I can tell you two cases right now. Two of our most enthusiastic customers who really wanted to IQ max did not do so. Because in both cases, and I'm thinking, yeah, of two specific people here, one of them had their highest predicted IQ embryo was also the highest predicted for type 1 diabetes.

And she was training to be a physician. This is one of the few people who has authorized us to talk about her case. She's kind of been in the newspaper talking about us. So I can reveal her identity. And yes, she was training to be a surgeon, but she has T1D. And that's a pretty serious thing. She then knows firsthand how good type 1 diabetes can be on someone's life.

That's right. So there was no way she was going to just sacrifice T1D for a couple of points of IQ. That's crazy. And so she chose a different embryo. Another one, he didn't know that there was a risk of, they had one of the BRCA variants. They wanted a girl. For breast cancer. For breast cancer, sorry, yeah.

And one of the highest predicted numbers for IQ had one of the BRCA variants. So it's like, well, no way. I don't want her to die at 35 of breast cancer. So the reality is there's always trade-offs. And one of the highest predicted embryos for IQ had one of the BRCA variants. So it's like, well, no way. I don't want her to die at 35 of breast cancer. So the reality is there's always trade-offs.

People are looking at intelligence, among other things. And many people only have maybe four embryos. So you're not going to get... You only have so much selection then. That's right. So you're not always going to get some huge gain. But yeah, the gains are pretty big.

Validation Studies for Polygenic Scores

Now, you mentioned validation studies. And how, in a short explanation, how do you verify these when these oldest children who have been born from this procedure, this embryo selection, are probably still approaching kindergarten? How do you verify that?

Good. This is the most important question. And I would say anyone who's thinking about doing this, for example, whether you're getting a polygenic score for yourself or your embryo selection, whatever, or you're just interested, you should always ask of companies or academics who are generating polygenic scores, you know, how did you validate these?

And, you know, again, the gold standard is on existing adult siblings. What we can't do and can't expect people to do is wait for 100 years, you know, and see which of these hundreds of thousands of embryos that were implanted eventually developed Alzheimer's disease or whatever. I mean, it'll be nice when we have that data, when our grandchildren have that data.

Yeah, our grandkids can co-author on that study. That's right. Exactly. But that's not coming for a while.

The good news is, or maybe it's not good, the interesting news is you don't really need that. And in fact, it would be an unnecessarily sort of rigorous demand for data. I mean, it'd be nice to have that. But the truth is, if your DNA, as you know, it doesn't change your life. I mean, yes, there's some epigenetic things, but the basic DNA that you have as an embryo, you have as an old person. It's really not changing in fundamental ways other than some random epigenetic stuff.

We talked with Evelyn Lin, who's in the lab at King's College London, and she says that's one of the beautiful things about DNA and polygenic scores is it is unchanging. You only have to collect it once from a person, just get a little spit and you're done. You never had to go back for that variable.

Exactly. And so given that that's the case even though there are environmental effects and so on that we can talk about if you can predict from DNA alone an individual outcome or the differences between adult siblings because remember that the context of embryo selection an adult sibling has the same genetic relationship to another sibling as an embryo would to another embryo.

And so if you can predict the differences between adult siblings as well as the absolute numbers, say for height or whatever else, adult siblings, then there's just no reason to expect that it wouldn't work in embryos given that it's the same DNA. So you can come up with some exotic explanations and so on. But if you want to demand that excessive rigor for that, then you should also be demanding it for David Reich. You should be saying, well, who are you to claim to explain the human past with DNA alone? Resurrect these people. Build a time machine and go back to ancient Babylonia and test these people. You didn't go to ancient Babylonia.

Yeah. So, well, you know, and you could make the same rigor argument for environmental predictors. I come from educational psychology where the default environmental explanation for individual differences is always family socioeconomic status. And when you actually crunch the data, the correlations only about point three between socioeconomic status and educational outcomes.

Now, it can vary depending on exactly how you measure those things, etc. But you're saying that already within siblings, we're approaching a 0.5 predictor. I mean, that means that today, with these within family predictions that don't even consider between family variables and population stratification, all the other stuff that we spoke with Evelyn about in an earlier episode. you're already making better predictors than you can from education's default assumption.

And so if you want to make that, if you want to have that level of rigor for a genetic predictor, okay, go ahead. Let's also do it for an environmental predictor and say, okay, how are you validating this? Let's talk about what level of rigor you're going to apply to that because it weaker and explains far less and probably matters less than individual genes do. Or I should say the collective genes do. Yeah that right.

Traits Selected For and Parental Decision-Making

Interesting. So you already said there are parents out there today selecting for IQ, but even in the most IQ maxing parents, they're pulling back for some things. What are some other traits people are selecting for? And how do parents decide trait X is more important than trait Y? What do these conversations look like?

Yeah, good. I'm actually not usually on them because we have a full-time certified genetic counselor that kind of helps people interpret their results. And the truth is, you know, even really bright early adopters of this, and we've had about 80 customers so far go through the process at one stage or another, not 80 babies born yet, maybe a dozen.

But so we've got some experience walking people through, the genetic counselor does. And, you know, from my understanding and talking to at least some of the clients, you know, they're looking at especially disease traits. Typically, the ones they focus on most are the ones that run in their family or that would have devastating early life impacts.

But, you know, sometimes it's a little bit arbitrary. We did have one case. They didn't end up being customers. So I guess I can just say it. One case of like an actress who just obsessively, you know, is like, I just want blue eyes or whatever. And we're not even doing eye color anyway. But, you know, there's occasionally people who are like really into an aesthetic trait or maybe they're very short and they just want like taller sons or something.

But it's typically, you know, priorities are minimizing inherited disease. We have about, you know, a couple dozen disease predictors, not quite that, but close to that. Autoimmune diseases, for example, those run in my family a bit on my grandmother's side. So Crohn's disease, rheumatoid arthritis.

Mental health is especially important, I think, to both men and women, especially women. Things like we don't have severe depression yet, but we'll be offering that soon. But we do offer, for example, bipolar and schizophrenia, which are partly correlated with those. And those matter a lot to people, right? I mean, they'll sacrifice some IQ to avoid schizophrenia for obvious reasons.

So that what it usually looks like. It sounds like most people are nuanced. They considering different things and that the monomaniacal oh i have to have my designer baby with blue eyes is much more rare and that when you get down to brass tacks with these couples um it no wonder you need a genetic counselor on staff because it sounds like it can be complex let's talk about the pluses and minuses of a dozen different traits.

Yep. And in the end, I mean, we kind of both accept it as a moral principle and a legal dictum, which is that, you know, informed choice is our maxim. So we don't report any, like, sorry, we don't recommend that you implant this or that embryo or prioritize it. We just try to make the numbers as clear as possible. And we don't want to inflate our predictors. Like, we just want to say, like, here's the best data we have for this or that trait.

You know, we have an embryo comparison tool where you can just see like embryo three, you know, like lower on, let's say IQ, but higher in terms of some other health score and go down the line. And then we have an overall longevity index where you can compress that data and you can kind of look along a bell curve of disability adjusted life years and then see like what's contributing to fewer life years in this embryo.

Oh, it's a high risk of schizophrenia or high risk of, you know, some autoimmune disease, which means that your body is going to be constantly attacking itself, low levels of inflammation your whole life, and that's going to remove life years or whatever. So yeah, people typically think about what runs in my family. Maybe there's one or two things that are bad. What do I prioritize?

You know, some do prioritize, I guess, you know, intelligence over length of life. Maybe, you know, that's fine. We'll let them do that. And then, yeah, what am I learning new here? Like maybe the way in which our, you know, me and my wife, our DNA is combining is introducing some new risk in an embryo that we didn't know we had. And well, look at that. That's elevated. I never thought about that, but I don't want to select in favor of it, you know?

What about personality traits? Are those, are those on the menu yet?

Not yet. It's partly because the genome-wide association studies from which you get these polygenic scores, those are still being done. As you know, they're less heritable than things like height or intelligence or some other diseases, but they're pretty heritable. They're usually right around 0.5, pretty moderately heritable.

So I assume that these will be on offer, and I think I often get asked, like, what would you not let people select for, or What would you personally not select for? Always select against? And I think that would be something like, you know, psychopathy. If we had, we don't have a predictor for it, but that would be one of those things where I actually want my child to be a psychopath.

So, you know, I assume that these will be on offer. And I think, you know, I often get asked, like, what would you not let people select for? What would you personally not select for? Always select against? And I think that would be something like, you know, psychopathy. If we had a predictor for it, but that would be one of those things where I actually want my child to be a psychopath.

Path that's right well i mean i i actually think it would be a good idea to offer it precisely because people would select against it i don't know if anyone who'd select for it well it's the same reason for schizophrenia i can't imagine someone's coming into your company and saying oh i want to implant the embryo that has the greatest chance of delusions hallucinations.

I can't imagine a parent, number one, because it wouldn't be ethical. Number two, you're responsible for this child for at least 18 years, and they're a part of your life until the end. Why would you purposely bring that into yourself?

Yeah, I think maybe I'll have a potential serial killer in the family. That's right. No one's going to do that. And you're going to subsidize it, spend your own time and money to have a serial killer.

Good point. Sense it's one of the only things that i think that we if we had a predictor for it and somebody some insane person wanted to select for it i think that would be one of the clear cases where we just say no like we're not we'll we'll turn that one down we'll turn that one down um yeah but otherwise personality is going to be interesting i think because you know i've talked to some experts on this read a bit about it but in terms of like you know the field of evolutionary psychopathology.

You know, it leads you to expect that generally speaking, somewhere in the range of the middle of the bell curve for most personality traits is where you want to be. You don't want to be at the extremes of either side. And it would present, I think it will present some challenges in how you explain the various kinds of life outcomes associated with these things to people.

But the truth is, generally, most of your embryos are going to be right around the middle, of course, of where you and your spouse is. Most traits, yeah. Yeah. Yeah. You're not, you're not creating whole cloth, you know, new people with radically new personality traits.

And that's a good point with embryo selection is that you can't select for what's not there. No, I'm pretty sure I'm five, seven. My wife is five, two. We not going to select for a future NBA player. It not the genes aren there. Exactly Exactly.

Future Uptake of Embryo Selection

So on the one hand embryo selection is new. And there a growing number of people who are using the technology. On the other hand, making a baby the old-fashioned way is much easier and cheaper. Probably much more fun, too.

So what do you think is the upper limit for uptake of embryo selection?

You know, it's unclear to me where that's going to be. And I think it'll depend in part on, on the one hand, people's objections, because there are reasonable religious objections. Some people simply will never want to use this. And that's fine in the same way Jehovah's Witnesses refuse blood transfusions and that sort of thing. I'm not going to tell them, hey, you know, I should force you to transfuse your blood or whatever.

So there's going to be some who never, never want to use it, presumably religious, social reasons. I fully respect that. and you know there's going to be some who maybe in 50 years this technology will be if not moot mostly irrelevant let's say if gene editing is is the game in town or something like that and i'm not even talking about anything radical but let's say like you could dramatically reduce the risk of schizophrenia more efficiently through gene editing than embryo selection well then probably you're going to do that although although i will say in order to get the embryo you probably will still have to do either IVF or what's called...

You'll still have to genotype the embryo in the first place to figure out whether you need to do the gene editing anyway. That's right. So you'd still have to get it out of the body. So it'd still either be IVF or what's called IVG, which we probably don't have time to get into, but that's taking an adult cell, turning it into a pluripotent stem cell, then an egg, and then combining with sperm. Either way, that's...

But for now, I would say the next couple decades, the relevant timescale, you know we're at a point where you know something like two and a half to three percent of americans are already born through ivf and ten percent of danes danish people now it's partly because people are going to denmark for you know fertility tourism but but a lot of them are using it themselves natives and and some of this is infertility who knows this is speculative you know whether it's environmental stuff or people just waiting until they're older and we do know that when you're in your late 30s, the rates of aneuploidy dramatically increase, meaning chromosomal abnormalities.

And that means that even if you're technically fertile, you can have a kid, maybe your body is having quite a few quote spontaneous abortions because many of those embryos that you doing again naturally just through normal sex are just damaged. And so it harder to get pregnant. And if that the case many of these people then turn to IVF and then they screen those embryos for aneuploidy. Of course, we're doing a step further and giving more information, but look, it's already 10% in Denmark.

My view is as people learn that they can dramatically reduce certain risks, not all risks, but some risks that may run in their family, maybe boost IQ a bit, maybe even a little more in the future. I suspect, you know, it could triple, quadruple. And, you know, we could talk about the moral status of embryos. I have my own views on this, but I think the end of the day, a lot of people will want to do this electively because they want to, again, just increase the well-being of their children.

No, I think you're right. And you, you mentioned, um, opposition to, for example, the so-called test tube babies in vitro fertilization. Uh, you know, I looked back prepping for this and there were some wild news stories and ideas and opinion polls in the 1950s to seventies. Oh, a baby born through in vitro fertilization won't have a soul. Uh, I found a, uh, opinion poll from the 1960s that something like 60% of Britons and almost 70% of Americans were opposed to in vitro fertilization. It was playing God.

Historical Acceptance and Future Trajectory

And then the first test tube baby was born and within two, three years, poof, goodbye, all opposition. Do you think we'll follow a similar pathway in regards to embryo selection, or is this different?

Not only do I think we'll follow a similar pathway, but it'll be sped up by the internet in the same way that, you know, opposition can spread, you know, through memery, you know, whether it's on sort of Twitter or just through mass communication, you know, the news and so on. so too will reversals happen.

I mean, you've seen what's happened to our culture in the US, 2020 versus 2025. This is like a generational shift in five years, and it's kind of wild to see this happen. My view is I in North Carolina at our lab right now and right up the street I used to teach at Duke University. And in my department was a guy named Timur Kuran who wrote a book called Private Truths, Public Lies.

And the book was, you know, a bit of psychology. He's a political scientist, but it was really mainly focused on the end of communism. How is it that we went from mass support for communism in Eastern Germany to mass opposition, where a revolution happened just kind of overnight. And people danced on the wall. They were dancing on the wall like, hell yeah, and two years earlier they would have been executing people for this.

And part of the story is, to bring it back to IVF and embryo selection, you get some people, once they understand communism is bad in this case or you can indeed dramatically reduce disease risks or embryo selection, and that's good, people will privately think that. And maybe they'll privately use it. In fact, I know people who are religious conservatives who are publicly, let's say, ambivalent, if not outright against this, but have privately come to us to use it.

I think there's going to be more and more of that until there's a preference cascade. This is, again, Timur Kuran's view. This is what happened with communism. Privately, people were less and less happy. And then they kind of tell their close friends or family, like, hey, this is not working well. and enough people sort of understood below the surface that that was the case.

And then when some elites or some people who were not afraid to say what they believe, maybe because they're rich or they're famous or they just, you know, they're uncancellable, so to speak, you know, then what happens is more and more people realize publicly, oh, actually other people think the same way. And then suddenly everyone, you know, not everyone, but large numbers of people are like, oh yeah, IVF's fine, it's not a big deal, embryo selection's fine.

And I'll one-up you on your example, not only was there widespread opposition to IVF, this happened for lightning rods. You know, I love this example. We don't have polling data, but we do know there were some crusades against Benjamin Franklin's invention, because after all, isn't lightning God's punishment to sinners. And if you had that view, you had a moral obligation to tear down every lightning rod.

Instead, what happened is, you know, a few years on, people saw, you know what, you know, God hasn't... I don't like replacing my barn when it gets hit by lightning every couple of years. Yeah, Satan hasn't kind of come and destroyed me. Probably they're fine. And I think the same thing is going to happen here, so.

Gene Editing: CRISPR and Future Prospects

I see. Now, you mentioned gene editing earlier. There's one procedure called CRISPR that exists now. Now, don't ask me to say what CRISPR stands for. Clustered Regularly Interspaced Short Palindromic Repeats.

Whoa, I'm impressed that you could just rattle that off. There you go. But CRISPR is a reality now. From what I understand, it's only good for very narrow, short gene editing. So maybe for some of those single gene diseases. But is gene editing for more complex traits across hundreds or thousands of genes like IQ is, is that on the horizon?

So I don't think so. I have no expertise on this other than I casually read the literature. And the only thing that I can say is I do try to keep up with the relevant scholars in the area. So I talked to a few people at the Broad Institute and the Church Lab, you know, after George Church. And my understanding is that, you know, even for single gene disorders, it's not fully safe now. There's still, you know, these so-called off-target mutations and so on.

And you do have to edit really early on. Like right now, for example, if you have an embryo, you know, we wait for about five days. This is normal in IVF for the cells to divide maybe 150 times. they're still undifferentiated these are not blood cells or anything like that so it's not like a fetus it's just undifferentiated stem cells but you know you get 150 of them then you biopsy the outer layer to test it for for disease and so on.

The thing with crispr is you'd want to get it right off the bat right because you have to edit every cell right in the body and so first you'd have to get it early and you'd have to be reasonably sure there are no off-target effects. And, you know, you can probably get that for one for one gene.

It looks like the guy in China. He's a weird guy, by the way. He Jiankui. I'm not a fan of him at all. I'd watch him on Twitter and he seems insane. Well and I am skeptical because no one ever verified. I too Gene editing claims And editing claims. Correct In science we don hide our results. No.

So I'm skeptical if even been done. But I do think it's perfectly feasible. I think it would be pretty straightforward to do it for one gene. You're probably going to get, if you did it multiple times, you're going to make mistakes. But in many cases, you won't. But this is problem number one is these off-target effects. So that needs to be solved first. Nobody wants to screw around with their embryos. And it's important to say we're definitely not doing that. Not now, anyway.

When you're talking about selection, you're just revealing information, which is we've already been doing that for 50 years. No, there's no genetic engineering going on right now. No, there's no prudine. No, but I do suspect, and this is, again, channeling friends that work in the area, in 20 to 30 years, it's a lot more likely to happen, so-called multiplex editing, where you're 50 or 100 or more genes.

But again, you've got to get it early. You've got to have a really good system. And then you want to be able to potentially scan the genome to be sure you haven't introduced any devastating mutations. So I think that's a ways off. And don't implant that embryo that you made mistakes with.

I think a lot of people, in fact, I'm glad you mentioned it and that you're skeptical. I'm skeptical, too, of He Jiankui's claims in China that he did this. And I'm even more skeptical that we're going to, you know, people will be like, oh, you know, the Chinese are already doing this, right? And it's like, no. If they are, they've got big, big problems. I don't think they're doing anything.

The Chinese government, which is not adverse to technological innovation, as we know, put him in prison for a few years. Yeah. I don't know if they proved that he did what he did, let alone whether it was successful. But, you know, clearly the Chinese government is not open to this. I can't imagine there's large numbers of people following in his footsteps.

No, that's right. And also a lot of the experts, as you know, come to the U.S. and other places to work. Some stay in China, but the experts in this area, even He Jiankui, as you know, was being monitored and mentored by his Stanford professors. And we have the email exchanges, right? He didn't really know what he was doing. And so, yeah, they really don't have the expertise to be massively editing embryos now. And I think it's a ways away.

And in fact I just say the real bridge technology is editing what I you know happen with our distant descendants is actually creating chromosomes from scratch and in some ways it might be easier than editing where you're micro editing across the genome it might be easier to just string amino acids together i mean nature does it yeah craig venter has done it for bacteria there's no reason in principle but when you consider that even the smallest chromosomes in humans you know the y chromosome for example are tens upon millions of base pairs that sounds like a harder job than coming in you know pulling out a defective piece of dna and replacing a new one but no when you when you explain well you got to do this really fast you have to make sure there's no off-target mutations, suddenly, oh, yeah, building a chromosome from scratch does actually sound easier and maybe in the long run safer. Who knows?

That's right. And moreover, I guess we should finally say you'd better have a good causal story in terms of the genetic architecture of what's causing traits. So like right now, we do have a partial causal story for some of the things we're selecting for, but mainly we're relying on associations. Like if you have these genes, you'll probably get these outcomes. And that's pretty much good enough for the context of selection.

But if you're going to edit genes, you probably want to know all of the downstream effects before you do that. So, yeah, that's a good point. And, and, you know, we talked about genetic prediction with this episode with, with Evelyn Lin of when you make prediction, if your sole purpose is prediction, you often don't care what the causal story is.

But yeah, with gene editing, you better darn well understand how this segment of DNA biologically has effects. And I could see that being much easier for diabetes and much harder for IQ, let alone extroversion.

That's right. I mean, one of the things especially interesting for you and your readers or listeners rather is, you know, we can have a kind of partial causal story, but not a full one. And that'll be good enough for us. Like if you see that this, this genetic variant or this set of variants is typically expressed in brain behavior and not in like your liver or something well you know we we have reason to believe then it probably got something to do with intelligence We thought it did anyway.

Now it really looks like it but we still not might not know the exact you know function of it but we we kind of yeah it good enough for selection there there a difference between i know that this gene functions in such and such organ especially based on my mouse models and oh well yes we know exactly what this series of base pairs does when it tells your body to make an amino acid, which then forms a protein, which your body then does something with the, that's a much harder story to tell.

Correct. Um, but Hey, knowing where the genes express is a first step and it's actually pretty impressive how much progress has been made on that. Yeah. And it helps you build your, your model is better so yeah.

Long-Term Societal Impacts

Um let's talk about long-term effects hopefully in the first generation these babies these toddlers your company's already seen be born grow up to be healthier well-adjusted children with long happy lives um what do you think would happen after two generations if 1% of the population is born through this procedure or 5% or 25%? What do you imagine, you know, if we put on our sci-fi hat, what's going to happen in two, three, five generations?

Yeah, good. Well, you know, one of our mutual friends, maybe he's mutual friends, certainly colleague give yours, Garrett Jones has done a lot of work on estimating how IQ helps predict certain kinds of outcomes in groups. And that's the context that it's relevant for your question. Let's say a country, like let's just call it Singapore or Israel, it's easier to imagine these small countries that are already pretty high IQ and wealthy and that sort of thing.

And they're already, in the case of Israel, they're subsidizing IVF and genetic testing. So this is just, even if they never subsidize polygenic testing, well, it's going to be pretty cheap for them. It's just an add-on.

So what happens if you, you know, 5% of the, 10% of the country uses this and selects for IQ, it maybe boosts, you know, the average IQ of the country by three points, or you could talk about reduces the average rate of certain disease by the same amount. It's going to have economic consequences.

But, you know, in Garrett Jones's story, again, since your podcast is about intelligence, you know, one of the most interesting things he found, which you're well aware of, is not just that on average, smarter people tend to be more innovative or make more money. If anything, I actually think some people overestimate how big those effects are on an individual level, you have an accurate view of this, but a lot of people think it's just going to massively blow up your income, which I think personality and stuff matters a lot, some luck.

But from a group level, it matters a lot. So if you raise the average IQ of a group, what you get are these really interesting interaction effects where you make each person within the group a little bit more cooperative, having a little bit longer time horizons, and those two are related.

The way to show this, as Jones did, and this is the stuff that I used to do in my academic world, you know, game theory and ethics, he would have his students play a prisoner's dilemma game or a public goods game. Basically, I'll just say the rough way of describing that is where there are short-term incentives to defect from mutually beneficial agreements, and then you get paid in the the short term more, but long term incentives, getting bigger payoffs if you cooperate in the short term and forego some big short term payoffs.

In situations like that, smarter people tend to cooperate more for mutual gains in the long run. And what that translates to is remarkable. It translates things to things like higher savings rates in your country, which is good for the economy, higher appreciation for the power of markets and prices to achieve their desired goal rather than command and control.

So if you have a democracy, people who are a bit brighter tend to support markets a little bit more. I don't want to get political, but it's just... That's just what the correlation is. That's what the correlation is. And that does tend to... IQ and free market support are positively correlated. Yeah. And there are obvious examples against us, North and South Korea, right? You've got the same population and as soon as you have communism, everything's ruined and they're brainwashed and so on.

But when you look at these really interesting correlations if you just boost 5 of the population by a certain amount you probably going to get more cooperation less political corruption which is an interesting point that Garrett Jones found. Because again, you know, I could hire my uncle or my brother now for this job that he's not qualified for, or, you know, I could have this more meritocratic system, which is better over the long run for my country, for me, for my own prospects.

And so what you get is all these really interesting effects, which will translate to overall economic effects, technological effects, and so on. And even things like, you know, violent criminality is negatively correlated with IQ. And so as IQ goes up, criminality goes down. And that's true even if you control for socioeconomic status.

Yes. I'm telling you things that I've learned actually from you and other researchers, but just for the benefit of the listeners.

And so actually, let me ask you, turn your own question on yourself. One of the things that the answer to your question depends on is, does it matter for, let's say, a nation or a group's overall outcome what the mere average of that group is? Or do the smart fractions matter? Does it matter more like what the top 3% are or something? Do you have a view on that?

You know, I think that it's so difficult to get causal evidence about average group IQ and what happens when that changes. So I don't know, but I do believe in Jones's thesis that benefits that accrue from either bright individuals or bright groups get distributed out throughout society.

And so I do find it plausible that, you know, maybe one day, 3% of babies, you know, just like I would now bore with IVF, well, you know, the parents also tack on embryo selection and two thirds of them select a embryo that would have higher IQ than, than than randomly um and they have children and grandchildren some of those are selected with ivf and this this occurs i could foresee a situation where even if my grandchildren don do embryo selection they still benefit from other people grandchildren.

And I would rather maybe this is my time horizon I would rather have my grandchildren be a well-paid middle manager at a Google type company than a CEO at a small company being paid less with worse benefits because that Google type company never got invented by someone else smarter.

Right. This is kind of like, do we benefit from having, you know, some more von Neumann's in the population? You know, it certainly increases inequality and it marginally makes me less likely to get that. Well, I would never get my grandchild. I want that CEO corner office, But Grandpa Russell is saying, I think we're all better off having someone else out there.

Yeah, exactly. But, you know, who knows? Can I quote Darwin on this? I was just looking for this. Well, you could never go wrong quoting Darwin. No, never.

So he, you know, intuitively understood this. I mean, with no with no numbers. So, you know, take it for what it's worth. But he says in The Descent of Man, it is most difficult to say why one civilized nation rises, becomes more powerful and spreads more widely than another, or why some nation progresses more at one time than another.

We can only say that it depends on an increase in the actual number of the population. Adam Smith agrees with that. You have to have enough people to have markets and so on. on the number of men endowed with high intellectual and moral faculties, personality, morality, as well as on their standard of excellence.

And I love this passage because it shows you both. He kind of intuitively understood IQ matters, but it's not the only thing. Virtue matters, personality matters, and the standards of excellence, both the social norms around you and the government policies and so on. And they all interact. So culture and political systems matter too. They really matter.

And so I do think other things equal. Sure. Boosting the IQ of a society by a few points, probably a good thing, given Garrett Jones's study. But I don't think it's everything. I think these other things matter a lot.

Potential Downsides and Eugenics Concerns

So last question before we wrap up let try to keep this one succinct But I think of the other flip side Are there downsides with multi selection for things like IQ In the book Brave New World which as much as people talk about it and not enough people read it there is a section where Huxley describes an experiment that this future society engaged in where they had an island If I remember right, Cyprus, where they put all of the high IQ people there and the society there falls apart because they realize we do need people of all IQ ranges for a functional society.

So do you think that's a legitimate concern if maybe one day most babies are born with embryo selection and every parent is in an arms race to get the smartest baby possible? Do you think that there's a problem where we realize too late we need your average Joe salt of the earth who mows his lawn and keeps society going, you know, goes to the power plant and runs it? What do you think?

Yeah. So I think, first of all, it's very unlikely that we're ever going to get a situation where we just have two kinds of people. You know, it's stratified. like the selected and the unselected or enhanced and unenhanced.

There's already a huge distribution, as you know, of personality traits, intelligence, and so on. And all you can do is, first of all, select your partner, which is by far the most important thing, and then kind of around the edges, you know, a few points.

Even if you had a perfect IQ predictor, you're not going to be able to get anything out of range of the random assortment of the genes of you and your partner, unless you're talking about massive gene editing and so on. So there's that point to be made.

Secondly, I think that if everyone moved to the right a little bit, if the average IQ, well, it's always going to be indexed at 100, but if everyone just got a little bit smarter, both those at the left side and on the right side, we all just shifted the goalpost, I don't think that's necessarily bad. It's probably good.

And if you think, you meaning the hypothetical person out there, thinks like we're at the optimal point right now, why? Why would you think that where we are exactly at this moment and not 100 years ago or 1000 years ago is somehow optimal? I mean, traits and distributions of traits are always changing. Evolution doesn't stop and culture just speeds it up.

Moment and not 100 years ago or 1,000 years ago is somehow optimal. I mean, traits and distributions of traits are always changing. Evolution doesn't stop, and culture just speeds it up, whether it's driving our IQ down or up or sideways and selecting for weird personality traits. This is happening and shall always happen through cultural norms and sexual selection.

The only question is, you know, are we going to deliberately do it just a little bit beyond the kind of non-deliberate bit. And, you know, my view is, yeah, if everyone was just a bit smarter, it'd probably be good. I don't have strong views on it.

But what I think Huxley's point is, is actually a little different than what he himself thought it was. I wrote a paper with Julian Savulescu and Chris Gyngell, two philosophers who support, as they call it, genetic enhancement, some years ago, called Great Minds Think Different.

And in our paper what we were doing is saying well when people can select for personality it turns out we i think as a society we we benefit from a variety of personality traits and what we concluded there is probably first of all people will not want to select in only one direction yeah they're going to select in different ones and if anything the problem is we're going to get too much diversity not not enough diversity and so i'm not worried at all about that as an introvert who married an extrovert. I could see that being an issue.

Exactly. I think the worry that Huxley is right to express, and by the way, Aldous Huxley and his brother Julian Huxley were lifelong supporters of what we would now call genetic enhancement. They would have called it eugenics, but they always meant individual eugenics. They always meant using genetic knowledge to whether choose a partner, children, et cetera. Julian was more—

The Brave New World is not, it's supposed to be dystopian. He is warning us in many ways. It not a warning About this compulsory society situation. That right So it a warning against compulsory government eugenics But his last book explores this topic again and he in favor of individual selection and even sperm donors and this sort of thing.

He's imagining all kinds of kind of strange utopias. It's a novel, right? It's not a blueprint for society. But I think Huxley and his brother, Julian, they drew exactly the right lessons from World War II, which is maybe we were a little too enthusiastic about what the state could do in terms of like improving the human population.

But we were not overly enthusiastic when it comes to thinking, well, the traits that make us human are heritable. The traits that really, you know, predict a good life, like having a decent level of empathy, intelligence, health. These are heritable. They're good. It's good to encourage these things. It's just that the state should not be the agent doing it.

And so Julian, sorry, Aldous Huxley, even as he writes Brave New World, which is a parable against state control, still supported, you know, thinking about, you know, insights from genetics to help guide our reproductive choices. So I think he drew exactly the right lesson.

Interesting. I will say, I, just like the Huxleys, am strongly against compulsory decisions about reproduction, which spans a wide range of possible interventions. We haven't even scratched the surface of a lot of them, because state compulsion to such private decisions rarely ends well.

But, um, I see that as also a lesson from, from history from the early and mid 20th century. Um, I will say that where we're going with this, there is a huge difference between yes, that that brave new world state mandated compulsory genetic engineering And oh if you want to you can select your embryo or not There a huge difference It the difference between freedom and not having freedom.

Yeah. So I think it's important, as I've always said to when people say, well, X is eugenics, therefore it's bad. Like this practice, whether it's embryo selection or editing or whatever, or even assortative mating, you choosing a smart wife or something, that's eugenics, so it's bad.

I always like to say, like, you can't settle a moral or political dispute by just defining away the dispute, you know, like just labeling something. Because the truth is, you know, when you think of laws that ban incest, those are not only eugenic laws, they're state-sponsored, coercive eugenic laws, and they're good, you know, but obviously that's not what we mean when we're talking about state-sponsored, coercive eugenics being bad.

We're talking about the state controlling your reproductive choices that are otherwise, you know, voluntary and reasonable. I mean, it's not reasonable to have children with your sibling or possibly even a cousin. You know, this is a real problem in the Middle East right now with cousin marriage.

And so the truth is, you know, we can't just sort of, I don't know, have a stick a turn out there. Here's a dirty word. And say this is bad or whatever. Forget the term. What really matters is, should we be allowing the state to control these really intimate choices? And the answer, I think we both agree, is no.

And is it better if people are allowed to have information about their genetics and their embryos, or should the government control that information? I think, again, the answer is no.

You know, we have a long and bad history, the history, of course, of eugenics from the early 20th century that we can learn from. And so I think it's clear, you know, what we're doing, obviously, I think is fine and even good. But the key is that people should be able to opt into this. I do not want a world in which they're forced to use our company or we put pressure on them.

All we want to do is to people who are interested in this supply them with the best information we can and let them make their choices.

You know the whole idea of if we make society a couple IQ points smarter that can benefit many people That's been the justification in the education world for things like gifted education for over 100 years. We need to nurture these talents, let these kids grow up, be our physicians, be our engineers, be our scientists so that we can all benefit.

So the same logic works for why we need gifted education, why we need elite universities as, well, embryo selection should be permitted also. Yeah, and it's even compulsory in the case of education. Yeah, yeah, yeah. We force kids to go to school. boosting abilities you know yeah and funny that they and i think that society is better off because we force people to go to school for 10 plus years depending on the country uh and so you know if we've seen those knockoff benefits from increased literacy maybe there are knockoff benefits that will accrue in one two three generations from embryo selection for iq who knows the difference being that we just allow it rather than make it compulsory exactly i i wholeheartedly agree.

Conclusion

Well thank you jonathan anomaly for for joining us for the iq human intelligence podcast if you enjoyed this episode please do watch eugene lynn's episode i've already mentioned it a couple times we also have a great episode about the education side from karen rambo hernandez talking about the variability and cognitive achievement in kids which is surprisingly large just like it is within families. Please like and subscribe to the podcast and visit the links to Herasight. You can find those beneath the video in the description. Thanks for coming with us, Jonathan.

Thanks for having me. Great discussion.

Insights

• Polygenic Scores (PGS) Mechanics: PGS aggregate GWAS-derived effect sizes across ~millions of SNPs to predict variance in polygenic traits (e.g., IQ h² ≈ 0.5–0.8); key insight: ranking efficacy in siblings/embryos (within-family r ≈ 0.47–0.54) outperforms absolute prediction due to shared environment/ancestry confounding minimization.
• Embryo Selection Gains: With n embryos, expected max PGS gain ≈ √(n) * (h²/2) SD (e.g., ~0.5 SD IQ/15 points for n=10, half genetic variance explained); trick: family-within-family validation on adult siblings mirrors embryo relatedness (r=0.5), bypassing long-term longitudinal needs.
• Trade-offs and Tools: Multi-trait optimization via comparison tables/longevity indices (DALY-weighted); intuition: parents prioritize family-history diseases (e.g., T1D, BRCA, schizophrenia) over IQ, revealing non-monomania.
• Validation Rigor: Gold standard: out-of-sample sibling PGS ranking; compares favorably to SES (r≈0.3) environmental predictors; unchanging germline DNA enables lifelong stability sans rephenotyping.
• Societal Dynamics: Preference cascades (per Timur Kuran) accelerate acceptance (IVF: 70% opposition → norm); multigenerational IQ shifts yield group-level benefits (Jones: cooperation in PD/PGG, lower corruption, market support via longer horizons).
• Gene Editing Limits: CRISPR-Cas9 viable for monogenic (off-target/mosaicism risks); multiplex/polygenic infeasible short-term (needs causal pathways, early zygotic editing); future: de novo chromosome synthesis (Venter-inspired) potentially safer than iterative edits.
• Eugenics Distinction: Voluntary/libertarian (partner choice, info access) vs. coercive/state; counters Brave New World via trait diversity preservation (assortative mating bounds shifts) and optimal non-extremes for personality.

Transcription errors?

• Names: "Tobias Wolfram" → likely Tobias Bjørnebo (Genomic Prediction IQ lead; common speech-to-text error); "Tamar Karan" → Timur Kuran; "Eugene Lin" → Evelyn Lin (King's College; gender/pronoun mismatch, but context fits); "Chris Gingel" → Chris Gyngell.
• Technical terms: "Klein filters" → karyotypes; ""prudrine" → unclear (possibly "pruning" or discarded); "string amino acids together" → nucleotides/base pairs (minor; context synthetic genomes).

See also

• intelligence
• polygenics