3697 Will Genius Be Genetically Engineered? | Stephen Hsu and Stefan Molyneux
With significant advances in genetic analysis, genome sequencing and precision gene editing, it’s increasingly likely that embryo selection and genetic engineering will be used to prevent disease and increase the intelligence of future children. Dr. Stephen Hsu joins Stefan Molyneux to discuss human intelligence and the rapidly expanding technology which could soon enable embryo selection for both desirable and undesirable traits. Dr. Stephen Hsu was founder and CEO of SafeWeb, an information security startup acquired by Symantec and is the vice president for research and graduate studies at Michigan State University. Dr. Hsu serves as scientific adviser to BGI (formerly Beijing Genomics Institute), and as a member of its Cognitive Genomics Lab.Twitter: http://www.twitter.com/hsu_steveWebsite: http://www.infoproc.blogspot.comYour support is essential to Freedomain Radio, which is 100% funded by viewers like you. Please support the show by making a one time donation or signing up for a monthly recurring donation at: http://www.freedomainradio.com/donate
Stephen Su, who is the founder of two Silicon Valley companies.
The Real Keeners did three, but he only did two.
That's all right.
And the Michigan State University's Vice President for Research and Graduate Studies.
Stephen serves as a scientific advisor to BGI, formerly Beijing Genomics Institute, and as a member of its Cognitive Genomics Lab.
Dr.
Su, thank you so much for taking the time today.
My pleasure.
So, as we are wont to do in this show from time to time, we are dipping into the thorny and exciting realm of IQ research.
Now, I wonder if we can just start with a little bit of the 101s with regards to IQ, what it is, and dispelling some of the common myths that people have about IQ. Sure, I'll answer any questions you have about it.
All right.
So people understand that there's this thing called IQ which you can measure, but they think it often is measuring very specific skills that don't translate to life outcomes as a whole or and or that it's culturally biased, invented by white people to further the impression that white people are smart or something like that.
I wonder if you could give people just a brief overview of the length that IQ has been used and the amount of accumulated research that seems to show it's measuring something very real and not culturally specific.
You know, IQ has been under study for roughly a hundred years now.
And The actual construct is a little bit complicated.
What psychometricians noticed is that if I test people on a broad array of different things like memory, geometric visualization, vocabulary, reading comprehension, all kinds of things which you might think of as separate mental capabilities, because there is a positive correlation between all those things, one can extract a kind of maximally informative combination of the scores.
And that is the thing which is often referred to as the general factor of intelligence or G and popularly known as IQ. And there have been many, many uncountable studies testing the so-called validity of IQ. So in other words, if you take a population of people and you give them all IQ tests, There are a bunch of questions you want to know, like how stable is it?
If you rank order the people on their G-score, then you come back one year later or five years later or ten years later, does the rank ordering stay more or less the same?
Answer, yes.
If you know which kids have the high IQ scores and which kids have the low IQ scores, can you kind of guess who's going to be a kid who maybe drops out of school early and which kid is going to get a PhD?
Answer, pretty much yes.
Now, you can oversell IQ. You can say that, oh, if I know your IQ score, I know everything about you.
I know that you're going to make a lot of money when you're older.
Those things are not true.
And it just takes a moment to think about life and how complex it is to realize that no single score about one person is going to predict every aspect of their life.
And certainly we know people who are brilliant but lazy, brilliant but irascible and can't get along with others.
So there are all kinds of reasons why someone could be smart in a kind of analytical or abstract sense, but not successful in the real world.
And vice versa.
Lots of CEOs and politicians have very modest analytical or abstract intelligence, but because they have good people skills or they were just lucky, they become fabulously wealthy and powerful.
There is also this idea about cultural specificity as well that you can train for an IQ test, and I've seen pretty strong repudiations of both of those conjectures.
There's a lot of data on how trainable, for example, the SAT is or the ACT, and those are good proxies for IQ tests.
So there are huge studies with thousands of students in them in which they classified the population to one group who took maybe even very expensive SAT prep courses and another population didn't do any training.
And what's found is that the change in score, the increase in score on average that you get from really pretty intensive training, training that's quite expensive, is relatively modest.
It could be like something like a third of a standard deviation, something relatively minor.
So the training effect is relatively small.
Now, having said that, if I had a very specific kind of IQ test and I prepared you intensively for many hours a day for a year specifically for that test, I'm sure you would do better at it.
But in a way, you're kind of circumventing the purpose of that test by doing so.
Well, it's sort of like saying I'm a great athlete because I can use steroids or blood boosting or whatever.
All you're going to do is probably end up in a situation where you can't end up competing because you've trained for something without showing your natural ability.
Exactly.
And really, the fraction of the population that would do something like that is so small that you can sort of discard that in sort of the statistical application of IQ tests.
Well, as you point out, the SAT tests used for undergraduate and the GREs used for graduates, they actually correlate fairly well with the outcomes of people who take those tests and go in.
The people who score well in those tests generally go on to do quite well.
And the people who don't score as well generally, I mean, the tests are used for a reason and that they do actually predict real-world outcomes fairly reliably.
They absolutely do.
And one thing that's not understood by most people is that The College Board or Educational Testing Service that administer the SAT, they're one of the largest employers in the world of psychometricians.
Psychometricians are basically statisticians who study cognitive ability data and they do what is called item level analysis.
So before any single question gets added to the SAT to actually be scored, It has appeared on tests in previous years, but was not officially scored.
So in other words, they know how students did on that question, but they didn't include it in the score of that student.
So they know a lot about whether a particular question is biased toward one group of students or isn't a good reflection of capability.
It doesn't correlate with the overall score on the test.
So they've done all kinds of very high level scrutiny, which ordinary people can't even imagine.
So these tests are really kind of quite highly refined scientific instruments by now.
Well, and there's hundreds of millions of dollars riding on the efficacy of these tests, if not more, because you don't want to invite people into your university who flame out.
So it is really, really important that they get these tests right.
Otherwise, they're going to get a reputation of taking people's money without providing them the requisite degrees if they work hard enough.
The other place where there's just ginormous data about IQ is in the military.
So the U.S. military has been using IQ tests also for about 100 years.
And when you try to enter the military, they not only use your IQ scores, they call it something else like ASVAB or AFQT, but they use those scores to determine what track you're allowed to pursue in the military.
And they do this based on lots of experience.
So they know that a soldier who scores low should really not be assigned to a communications specialty, for example.
They also have basically a kind of absolute floor.
So depending a little bit on their recruiting goals and what the situation is, the floor floats kind of around 90.
So a little bit less than one standard deviation below average.
If you're below that, you basically can't get into the military.
And so this is not understood by the general population, but the military is in a way a very IQ-filtered population, and therefore, you know, you can have some confidence if you're an officer that your men meet some minimal level of capability, and that's very important for the military.
Right.
Now, there is, of course, a misnomer as well.
Whenever we publish these things, people write below, well, I took this online IQ test, and the actual IQ test is fairly intensive, needs to be administered by a trained professional, and measures a wide variety of things.
Now, as you pointed out in a couple of your interviews, Steve, there are a few proxies that you can use that are fairly well So,
as you said, the really sophisticated tests will have multiple components, which range from pencil and paper type things to Two, actually, the psychologist throws a bunch of pieces on the table and says, assemble this into a block and times you to see how fast you do it.
So a real IQ test takes a long time to administer.
It's quite expensive, and it's quite sophisticated.
However, what we know is that if you take a population, you give them that really exhaustive, expensive kind of test, but then you also give them some very specific narrow tests, like figure out what's going on with these little Bravens matrices, and There can be a very high correlation between the sort of narrow shortcut test and the more elaborate test.
And so that's why we often default to the shortcut test if we don't have the resources to administer a really sophisticated test to someone.
Well, and the GREs and the SATs are fairly decent proxies for IQ and, of course, collected from countless numbers of students around the world just to try and get into college.
That works fairly well as a proxy.
And, of course, then you have all of the associated information with grades and success and path to graduation.
And through there, if they succeed in their particular field, you get wonderful things like...
Number of scientific publications and so on.
So there is a lot of data regarding intelligence if you expand the basic tests beyond the formal IQ ones, if that's a fair way to put it.
Absolutely.
And the recent, you know, recent is maybe not the most precise way to describe it because these longitudinal studies can take decades.
But what's recently become very well established is that even if you look at the far tail So people who are not just in the top 1%, but in the top 0.1% or 0.01% of scores on these tests, you can actually see the returns to intelligence do not go away.
So if you compare people who score at the 1 in 10,000 level to people who score at the 1 in 100 level, the first group actually does better.
They're more likely to get a PhD in a hard subject.
They're more likely to have published scientific papers, more likely to have a patent, more likely to have published a novel.
So you can now map out the returns to intelligence even in the tail.
Well, and when I was growing up, there was a couple of myths around IQ. One, of course, was that after a certain point, it doesn't matter.
Well, as it turns out, it still does.
Some of the work that has been done points out that if you're like IQ 160 +, which is extraordinarily rare, in the sort of category of eminent scientists, well, you're going to be doing a lot more work, be a lot more successful, contribute a lot more, particularly to the essentials of the field, than your average PhD clocking in at 130%.
So there doesn't seem to be a bit where it peaks out.
It's almost like an extrapolation, like it just gets better and better the more IQ you get for particular fields.
I think that's sort of true, but it's not intuitively.
That observation is clear from the data now, but It's not intuitively obvious to the average person, because the average person can only sort of imagine, everybody can imagine people less intelligent than them, because they can't even how dogs see the world, right?
Well, and we were all younger, and we had smaller brains, and we all grew up.
I mean, I know IQ can be tested when you're young, but you do grow up in expertise and knowledge, and so you have a sense of what it's like to be less well-informed and less adept with your brain when you were younger.
You can always imagine, like, what would the 12-year-old me have thought of this?
But you can't imagine how the, unless you have a very high IQ, what the IQ 160 version of me or the IQ 180 version of me would think about this.
You can't really imagine that.
So it's very easy for people to discount the returns to very high capability.
Now, I personally, I went to a very good high school that was in a college town, so many of the kids were kids of college professors.
And so there were many kids in my high school class who were in the top, say, half of 1%, which was what you needed to become, say, a National Merit Scholar in the U.S. system.
But then I went to Caltech, and most of those kids who were in the top half of 1% in my high school class would have flunked out of Caltech.
They would not have been able to absorb the material at the pace that it was offered.
So there's already, you can see a gap there.
But then I would say in theoretical physics, which is my field of research, probably no more than 5 or 10% of the Caltech class could do theoretical physics.
So you'd have to probably be at the very top of a Caltech class to be a good theoretical physicist.
So you can see their layers.
And I think professionally, I know people who are much smarter than I am.
So, you know, I think there's just more.
As we develop AI, for example, we're going to see intelligences that eventually are way beyond anything humans can imagine.
Oh, yes.
We definitely want to talk about some of the genetic possibilities of this kind of stuff in a sec.
Actually, you just reminded me.
I had a friend who was very much a big fish in a little pond when I was in high school.
He was like the guy who would get more than 100% of his math tests and so on.
And then he went to a very tough university with a math and physics double major and hit the limit of his cognitive abilities, which is quite a shock to his system.
You know, when you keep moving up in these layers, at some point you're going to be among people who are babbling away in arcane language that's fairly incomprehensible to you.
Maybe you can reach where they are.
Maybe you can't, but it is kind of a shock when you keep moving up, when you finally get to that summit where there's still people above you and you're like, I don't think I can particularly reach that.
And there's a biological basis to all of that.
Now, before we get to that, though, let's talk a little bit about IQ and socioeconomic status because...
Because, of course, there is a very common perception in the world, which is that our potentials, if not our personality as a whole, are very much influenced by environmental factors.
And so the idea is that, well, as you point out, Steve, you went to high school with a So naturally they had all of the stimulation and they had the toys and they had all of this wonderful high socioeconomic status environment like probes getting the brain going.
And so people think, well, if we take everyone or anyone and put them into that kind of privileged environment, they're going to do as well.
So where does socioeconomic status lie with regards to IQ according to current research?
So certainly you can harm people by not giving the right nutrients, by not giving them access to school, etc.
But when you get to a certain level of good environments, we start to see that the genetic component that contributes to variation is larger.
And, you know, they're very simple studies you can do to test the hypothesis that the environment is everything.
So, for example, you can look at pairs data where you're looking at siblings.
So both siblings grew up in the same house.
They probably both had access to pretty good environments, similar environments.
But oftentimes you'll have a case where one sibling scores much better than the other sibling.
And then you can look and see how they do in life or how they do on scholastic subjects five years down the road.
To take an example, my brother is quite bright.
He was a national merit scholar.
He studied electrical engineering in college.
But, you know, I think there's somewhat of a gap between us in cognitive ability.
No offense, Mike.
We'll look forward to his email.
It comes in right after he sees this, but go on.
But I think he experienced, if anything, as good an environment as I did, maybe better because he's the older child.
And some people say the older child tends to get more attention from the parents.
So anyway, it's pretty clear environment is not all.
And I'm sure everybody is aware of all the twin studies and adoption studies that establish this fact.
Right.
Okay, so let's start talking about a topic.
I know that people sometimes get uncomfortable about it, but, you know, we must follow the facts wherever they lie.
And there is, I think, great good and great value to society in following these facts.
But when we start to talk about the genetic basis of intelligence, people, and it's funny because we understand that there's a genetic basis to how good looking you are, to how tall you are, to even your capacity for athletic endeavors or muscularity, or we understand all of these things.
And we also understand with regards to height, sure, you can harm someone's height, you know, don't give them enough nutrition, don't give them enough protein, and they'll grow up shorter.
But if you give them extra nutrition and extra protein, they're not going to end up way taller, they're just going to end up, well, wider.
So when it comes to intelligence, though, it goes against our grain to some degree because we have this egalitarian ideal, or at least a lot of people do.
And we don't like the bell curve of distributed intelligence because it seems to give some people an unfair advantage.
I don't know what fair or unfair means in biological terms, but let's start talking about Thank you.
Because it seems like every time I talk to some expert like yourself, there's new information that either because the cost of sequencing the genes has gone down or there's been some new progress.
Let's start talking about how we know the effect or the extent of the genetic influence on intelligence.
So the extent is well established from, say, twin and adoption studies studies.
So, for example, if you look at identical twins, so they share all the same genetic information in their DNA, but you adopt them into different families, so you effectively kind of, if you look at a sample of people, you've effectively randomized over the environment.
What you find is that as the twins get older, their IQs become more and more similar.
And when they're young, their parents control their environment.
And it does matter whether their father is a coal miner or their father is a university professor.
But as they get older, you find that that environmental effect fades.
And some of the estimates for IQ in late adulthood, so say when you're well into your 20s, are as high as something like 0.8 of the variants.
And so, you know, that means most of the difference between people is coming from the genes in the long run.
And this is, I think, frustrating for people a lot of times because what they do is they look at how groups are doing in school, junior high school and high school, and they say, oh, well, you know, somewhat similar, somewhat similar.
And then what happens is, of course, as the genetics exerts more and more influence over time – Then groups end up doing differently and then everyone says, ah, it must be because they got out into the adult world, they're facing bigotry, they're facing ostracism, the sexism and so on.
It's one of these curses of how IQ develops that it seems to throw the scent off towards socioeconomic or bigotry or racism or sexism factors when it is to some degree just how it plays out in terms of our brain development.
You know, it is a really complicated question.
So you can forgive people who are not specialists for really not knowing what the answer is.
People are intensely curious about this question.
So when I talk to them, they always want to ask me, well, how do we know this?
How do we know that?
What's the best study supporting this hypothesis?
They're intensely interested.
And furthermore, you have some experts who really kind of want to obfuscate the truth.
They don't really want to You know, they have ideological commitments which prevent them from endorsing what really is the scientific consensus.
So it's quite complicated and you can understand why people are confused about it.
Right, right.
So let's start talking about what is happening now with regards to being able to sequence genetics.
And of course, if you have people who have IQ proxies like GREs or SATs, and you have access to their genetics, you can begin to see a bell curve spread, even if we're just maybe a little bit more focused on the higher abilities, because of course, people of low to average ability are less likely to be doing GREs and so on.
You have a bell curve of the high abilities.
You have a significant amount of genetic information.
And of course, the cost of sequencing this stuff, I was reading an interview of yours from, I think it was 2013, where it was still a couple of hundred bucks to do a little bit of genetics work.
That price has come down to the point where I think you're able to start to sift through and tentatively, I mean, it's all very tentative, begin to see where genes may be showing up that influence intelligence.
So this is a very complex study.
And the field is moving fast on a timescale of literally years.
The situation changes significantly.
So just to jump into it, our DNA contains about 3 billion base pairs.
So just to start from the very beginning.
As a physicist, I can't help but start from the first principle.
Please, please do.
There's about 3 billion base pairs.
And they can vary from person to person.
But some regions of the genome are more likely to vary from person to person than others.
And in fact, if you were to just take a chunk, take a region of the genome from two people, say from you and me, probably about 99.9% of that chunk were the same.
We will differ at maybe one in a thousand base pairs.
If you were a Neanderthal and I was a modern human, We would differ at few per thousand base pairs.
If you were a monkey and I was a normal human, we would differ at about roughly 1%.
So you can see what, you know, as we become progressively more similar, we're eventually modern humans or even maybe from the same ethnic group or maybe siblings or, you know, identical twins.
But the rough level of variation is just those one per thousand base pairs out of three billion.
So if you have one per thousand out of three billion, you have a few million differences between two randomly chosen people.
Okay?
And all the variation between us is in those few million differences, right?
But this is, sorry to interrupt, because I can see where people's minds are going, having talked about this a bunch of times before.
Sorry to interrupt.
But people then think, well, we are 99.9% identical in all characteristics.
That is not how this stuff plays out.
I just want to reinforce, and of course, correct me wherever I go astray, Steve, but this doesn't mean that we're all 99.99% the same or whatever.
This is why I brought up the monkeys because our DNA is 99% roughly the same as theirs, but we're obviously quite different from them.
Now, we have 10 digits and two arms and two legs and our head sits here and our eyes are here.
So those are common features between us and the monkeys and that's probably due to the 99% DNA that we share with them.
You can't reason simply about mapping a phenotypic difference to number of base pairs.
For example, A single base pair change, there are known cases of single base pair mutations which can cause significant mental retardation, significant physical changes, etc., etc., lethal diseases.
So one base pair change could, in principle, cause a huge effect phenotypically.
For complex traits like height and IQ and Body mass and muscularity, all the complex traits.
Typically, we expect there are going to be thousands, perhaps 10,000 of those millions of places where we two individuals will vary from each other.
Maybe about...
1,000 or 10,000 of those will be the specific subset of genetic variants that control that particular complex phenotype.
And we would expect a lot, because there's a bell curve.
I mean, if it was just one switch, it would be binary, right?
And so because there's a bell curve, we would expect there to be a huge amount of complexity, and we've got genes leaning on other genes and dependent and so on.
And then this doesn't even count into the complexity of the environmental effects on genetics, but...
The complexity, I mean, it would be kind of cool if, in a way, one switch, but of course it wouldn't work out that way mathematically.
Absolutely.
And, you know, I mean, it always had to be the case that the human brain, which is literally the most complex known object in the universe, there's no known object that we can think of that is more complex than the human brain.
It had to be the case that the genetic program which makes that thing is quite complex and If two people are different in the way their brains grow or wire up as they grow up, you would expect there to be many, many different places in the genome where if I made a slight change, it would affect the function of the brain in some subtle way.
Right.
So it's like height.
I mean, if there was just one switch for height, you'd be, I don't know, five foot or six foot.
That's about it.
But because height is a bell curve, just like intelligence, we would expect there to be thousands of annoyingly complicated interactions to produce it.
That is exactly correct.
However, even for professional geneticists and genomicists, it has come as a surprise to them just how many variants are in play.
And this is because they have a cognitive bias toward thinking in terms of single gene conditions or single gene diseases, because A, that makes the science much easier.
B, it makes it much easier to, say, design a drug target.
And so lots of research is...
Focus on sort of single gene stuff.
And because things have turned out to be, I wouldn't say, predictably so complex, a lot of the field has not been really prepared for the amount of computation and mathematics and statistics that is required to make progress.
And so you can actually do an interview with somebody who has a PhD in genetics, but they actually don't have any of the cognitive skills to really understand complex traits because they really have not learned the mathematics necessary to do this.
Right.
Now, let's talk about some of the outcomes in terms of time in jail, divorce, even body mass index, things that you wouldn't necessarily expect to be significantly influenced by.
We'll just say G, because in general intelligence or whatever it is, the sort of conflagration of things that the IQ test measures.
And income as well, because this idea that we can sort of all end up roughly around the same, generally middle class, really does seem to be repudiated by the amount of influence that general intelligence has on something as complex as how much money you end up earning.
There have been studies which show, for example, that after it became known that tobacco smoking causes cancer, the probability that you're a smoker becomes correlated with G. Um, there are studies, uh, there's a famous Scottish study where they, uh, tested, I believe every kid of a certain age in Scotland and then looked at, now they can look back because many of these people are dead or very old, uh, and they can look at longevity.
And actually, um, IQ measured at age 12 or something like this turns out to be the best predictor of longevity.
Uh, it's better than adult BMI. It's better than adult, uh, cholesterol level.
It's fair.
So, um, There are all kinds of funny ways in which your sort of mental capability affects your life.
Well, and as far as income goes, I've read studies that show that for every standard deviation you are graced with, as far as intelligence for every standard deviation increase, you're getting about $30,000 a year in additional income.
And again, that's not particularly fair, whatever fair means in this kind of context, but it does seem to be a reality.
I don't think there's any sense in which the allocation of genes, which in a way you could just say is like the way that whoever God or evolution designed you as an individual, if there are differences between people, I can't imagine that they were assigned in a kind of fair way.
Unless you had a previous life and the people who were really good in the previous life got good genes in this life.
Other than that, it's just unfair.
It's a natural endowment.
It's effectively random.
And so some people got more and some people got less.
And the only plus that I can say is that because of this very evolution and mutation and randomness, that's why we've developed the brains big enough and moral sensibility sensitive enough to be bothered by this stuff.
So we can get mad at it, but it's the whole reason why we're here to talk about it.
So for me, I can't get too mad at it at all.
I think if you went back in time and you talked to a bunch of people, like in early days when most people were farmers, By looking at animals and crops and such, they had a feeling that genetics breeding would affect the way crops are, the way even the character of dogs or the character of their cow.
They would have had a very, I think, different view of human nature that was actually closer to the mark than what many moderns have.
Well, this is the whole point of domestication, right?
Is to take whatever animals you can find and breed the ones with the most neotenized elements or the ones who are the most friendly or nice.
And there's a guy in Russia who did this with foxes just in a couple of generations, produced feral, vicious foxes and really nice, child-friendly, cuddly foxes.
And this is, of course, why we have dogs and cats and other animals that don't try and kill us in our sleep.
Yes.
I have a set of experiments.
I mean, if you're willing to exert very strong selection pressure, you can move the mean of the species very far, very fast.
And that's remarkable too, because again, when I was growing up, the sense was that evolution was glacial and grindingly slow and you had to like have a giant time machine to see the differences.
But you see people in the heights of Tibet, you know, you take people from China, move them up to the mountaintops and they have miscarriages all the time.
The people at the top of the mountaintops are fine and it can happen extraordinarily rapidly.
Right.
So there are many examples known now of single gene conditions or things determined by small numbers of genes that clearly swept through populations on timescales of thousands of years.
So that's very rapid evolution.
What is a little bit harder to pin down, but there are cases now which are understood, are complex traits.
And so, for example, height is well enough understood now.
We now know of thousands, roughly, of genetic variants that influence height.
We can look at, for example, the European population where there's actually about a one standard deviation difference between the heights of Northern Europeans and Southern Europeans.
So one SD difference between populations.
You can now show that that is mostly due to different allele frequencies.
So the Northerners are a little bit more likely to have the plus version of the allele and the Southerners are a little bit less likely to have the plus version.
And furthermore, if you look at the evolutionary history of it, it appears that that difference probably arose in less than 10,000 years.
So you have a 1SD difference in a quantitative trait between two groups, and it arose due to evolution.
We don't know the exact causes, but it clearly was due to evolution because it's a systematic effect.
All the plus variants are elevated and the minus are depressed in the northern population.
And the timescale is only roughly about 10,000 years.
Which, of course, coincides with the development of agriculture and then the resultant need because you have a population growth, a population explosion dependent upon agriculture, which requires that you get really good at deferring gratification.
Because if you're not good at deferring gratification, you're going to eat your seed crop throughout the winter and not be able to survive until spring.
And it seems that the deferral of gratification is highly associated with intelligence.
I almost feel like in order to survive the winter, nature had to build bigger brains.
And, you know, intelligence is just a nice byproduct of the need to be able to defer gratification.
So in the case of height, you know, we can actually take large numbers of Northern Europeans and large numbers of Southern Europeans and actually look at the specific alleles to see that, you know, this height difference is at least partially genetic and it seems to have evolved relatively quickly.
The story you just gave for how cognitive abilities would be affected by, say, the advent of agriculture or maybe by climate, those are all very plausible, too.
We don't have...
Good enough control over the genetic architecture for cognitive ability to answer that question fully but we will probably very soon actually relatively soon we'll be able to do the same analysis that we did on height but instead on And how would you design a methodology or a process,
Steve, for trying to find out whether this sort of fire and ice theory as regards to cognitive development, as a big influencing factor, probably not the only one, but a pretty decisive one, how do you think that would be designed or what information do you think is necessary to try and make that conclusion more valid?
I think to go the full way to saying, oh, it was definitely the cold winters that did it, That's still going to be tricky.
That may take quite a while.
So if you ask, where is height right now?
We don't actually know the causal mechanisms for why the north and south of Europe diverged.
We just know that they did diverge.
We can kind of estimate the timescale.
We can tell it's partially genetic.
We can get to that level with IQ probably relatively soon.
Right.
Well, that would be interesting.
And to me, it's one of the very fascinating things as to why we've ended up with ethnicities who have different levels of IQ, and whether that's related to climate or agriculture or something to me would be quite fascinating.
So, what about...
Sorry, go ahead.
I should say that for very distant populations, there's a little bit of a problem.
This is a very technical problem.
I don't know if you want me to get into it, but...
Please do.
The probes that we use to measure genetic variation in people, through a statistical analysis, we find that a particular SNP, so a particular single nucleotide polymorphism or a particular base pair in the genome, we find, oh, that's correlated with IQ.
And having one variant makes you a little bit smarter.
Having the other variant makes you a little bit dumber.
We don't know whether it's that specific variation that is causal or it's just correlated to something else, which is really the causal thing.
That's a tricky next step that you need.
And you then need to figure out, like, what exactly are these causal mechanisms?
Are they the same in different populations?
Has a predictor that I built on Europeans?
Is it not working on, say, Asians because of this different microstructure?
Those are tricky problems that I think are going to make it difficult to do comparisons of widely separated groups for a while.
That's going to take longer than people think.
Within the European population, say, just between, say, Italy and Sweden, it's actually relatively straightforward for height, say.
But if you want to compare, say, two widely separated ethnic groups on a trait like intelligence or even height, it's going to take a little bit longer, actually.
Oh, the cause and effect with genetics is completely insane.
Completely maddening.
It's like, oh, if I give an NBA outfit to a short guy, he's going to become tall because to tall guys wear NBA outfits.
So it is something that is a real challenge and something where we should always remember that conjecture is very tentative because data is so hard.
And as you say, causality is such a challenge to find out.
So fascinating though it is, I find myself pulling back on the horses of charging forward to conclusions pretty hard because the data is very hard.
It's also true that, you know, for what I consider the most interesting phenotype, which is cognitive ability, there's actually very little funding for it.
You can generally get funded to do things which are specifically related to disease.
And it's because of those disease studies in which they just happen to have gotten The height phenotype.
They just ask people, how tall are you?
And self-report is reasonably accurate.
So there is gobs and gobs of data on height attached to genotypes.
There's very little, actually comparatively, data on IQ scores attached to genotypes because there's basically no funding for it.
There's no multi-billion dollar project that tries to do this.
But if I were a government, it would be very smart to actually do this because what you can do with the output, the outcomes of these studies is extremely useful.
Oh, I would argue absolutely essential, even more so.
And I think this goes back to the Charles Murray and Arthur Jensen and other people who faced the challenges of trying to talk about racial IQ differences and so on.
So because it's so volatile for people and because it has so many political implications, people are like, well...
I'm not that masochistic.
Why would I want to get involved in that kind of stuff?
Although, I do think that it is something that really needs to be examined.
We need clear science behind it because in the absence of clear science explaining different outcomes between ethnicity, Steve, what we end up with is a bunch of social justice warriors trying to explain it using crazy things like, you know, I don't know, white privilege, systemic bigotry, massive racism, and so on.
We're going to have to have some kind of answer as to why ethnicities do differently overall, right?
Individuals are to be judged as individuals, but why do blacks and whites or Australian aboriginals or pygmies or Ashkenazi Jews or East Asians, why do they generally do so differently?
There is an answer which may have something to do with genetics, and people are afraid to go there, but then all that happens is you create a vacuum of explanation that gets filled in by people with no scientific training, and I I think quite a lot of ideological bias, to put it as mildly as I possibly can.
Yeah, I think all the scientists who work in this area have to tread extremely carefully, and I think fairly, because we have a very tough, very ugly history of race relations on the planet, shall we say.
But it does retard the science.
The science could be moving forward much, much faster if we're not so politically charged and also if it were funded at the level which is commensurate with the kind of breakthroughs that are possible.
I mean, after all, we're talking about understanding how the brain works.
And you would think that a smart government would be willing to spend the same amount of money on this kind of thing that they spend on, say, a nuclear particle accelerator.
Right.
And you yourself have faced some criticism, some opposition for the work that you've done in this area.
Is that fair to say?
Yeah, you know, I have never, I've always said that I'm agnostic on whether they're, so there are observed score, test score differences between groups.
That's clear.
I think you can't deny that.
The causality of that, whether it's due to, at least partially due to genetics, I've always been agnostic on because I think Not because I think it's impossible, but because I think that it's such a charged thing.
We should really make sure the science is solid before we speculate.
We shouldn't just randomly speculate on something that sensitive.
But just even not being willing to categorically rule out that God could have created us with average group differences has gotten me into trouble.
And I think that's just absurd, right?
So for someone to attack me for saying, we don't know the answer to this question...
Let's do the science first and then let's talk about it.
Even that position is actually not tenable in the current social justice warrior political climate.
It is a staggeringly anti-scientific.
I mean, I have to have a rant here about this because it's very frustrating.
It is a staggeringly anti-scientific perspective to say that human populations separated by tens of thousands of years in wildly different climates and environments, that this would have zero effect on the evolution of the brain.
When it is our most expensive organ by far, what is it, 3% of our body mass but takes more than a quarter of our energy to sustain it.
So the idea that the most expensive organ that human beings have remains completely unaffected by tens of thousands of years of evolution in wildly opposite climates is so ridiculously anti-scientific.
I don't even know where to start to talk with people who hold that kind of perspective.
It may not be decisive.
It may not be the end result.
But the idea that there's no effect on the brain for disparate climates and significant separation to me is so anti-scientific, it makes creationism look like physics.
So I think, you know, I like to be, I like to be fair and specific.
There's something called the Ideological Turing Test.
Are you familiar with this?
I have a feeling I might fail it, but let's go ahead anyway.
I think you're probably very good at this.
What Turing said is an AI would be able to mimic a human.
An AI would be able to fool you into thinking that We're good to go.
And give the strongest arguments that that other side can muster.
And only then can you say, I've fairly understood both sides of the argument.
So let me just give you the best version of the SJW argument that I can for why there are not genetically cause group differences between people.
I think the only scientifically defensible thing they could say is that, yes, people lived in different environments, but perhaps the cumulative effect of those different environments on genetic intelligence has been relatively small.
And so we don't rule it out.
It could have happened, but maybe it's not really very big and we can barely measure it with our crude tests.
And I think while maybe I don't necessarily believe that that's plausible, I think it's definitely not completely excluded scientifically.
So my attitude is let's wait and see what the science actually tells us before we form a very strong opinion about this.
Okay.
The counter-argument to that, and I know you were sort of playing a particular ideological perspective, which is great.
You know, I think that's very important.
It's the old argument from Aristotle that a truly educated person can inhabit another belief system and even advocate for it without necessarily accepting it, and it is a very, very important mental exercise.
Yes.
Right.
But the argument I think against that would be that, first of all, the tests do show significant differences among ethnicities, like all the way from Ashkenazi Jews down to Amazonian pygmies and so on.
There's, you know, from the low 60s to the 115s, 120s, there are averages in groups that are enormous.
And particularly when you have a free market environment where the Pareto distribution is going to end up with I think?
For instance, it is due to white racism that the Amazonian pygmies have such a low IQ. And that, to me, is a very great challenge to sustain.
And it's also, again, to say that the environment has had very little effect on our most expensive organ biologically.
And again, that to me seems incomprehensible.
Again, until the data comes in, I think we still need to keep exploring the challenge I have and I think the challenge that most people have who really want the answers to these things or not that there ever will be certain answers and the answers won't ever apply to individuals.
What bothers me is people who say I have a hypothesis and therefore you should not pursue the truth in this area.
On that, I totally agree.
I mean, the idea that you would try to suppress discussion of the topic or suppress actual research, which could eventually resolve the question, is just insane.
It's just the craziest thing.
It's like the church getting mad for us talking about the Earth not being at the center of the universe.
I mean, we should check to see whether the Earth is at the center of the universe.
We should not just tell people that it's a fact without doing the science.
It's so funny, don't you think?
I mean, when you talk to people about, you know, Copernicus and Galileo and so on, the question of sort of heliocentric view versus the Ptolemaic view of physics, you know, when they had the Earth at the center of the solar system and the rotation of the orbits of Mars got ridiculously complicated and you need like 12,000 different calculations to figure out where Mars was until you put the sun at the center of the solar system and then the retrograde motion maps out perfectly and it's all very simple.
And people look back and they're, Oh, how could they possibly have persecuted Galileo?
How crazy!
I mean, why would people be bothered by this?
It's just an exploration of the facts.
And then you start talking about race and IQ, and it's like, ah!
Ah!
Attack!
It's like, you understand, this is not a dissimilar situation.
One way that I... One sort of rhetorical tact that I take, which often sharpens the discussion if I'm talking to somebody who's, you know, from the far left, is I ask them, okay...
Neanderthals are definitely humans too, because we interbred with them.
We separated from them maybe 500,000, a million years ago.
If you look at the way their technology evolved, they were in Europe for 300,000 years or longer.
We've only been out of Africa for 50 or 100,000 years.
Their technology changed very slowly.
And there were lots of things that they never figured out, and we seem to have kind of displaced them fairly rapidly when we came on the scene.
So I think it's fair to guess we're probably somewhat smarter than they are.
So you ask them, well, so now imagine that I could resurrect, like in Jurassic Park, I could resurrect a bunch of Neanderthals.
Do you think that they would produce physicists and poets and...
Um, uh, medical doctors at the same rate in their population as modern humans, or do you think that maybe they would have some cognitive difficulties adjusting to modern life?
And I think most reasonable people would say, yeah, there's a chance.
And if they, if they won't admit it for Neanderthals, you could always go back to Homo erectus or you could go back to things which have really tiny brains, right?
Eventually they have to admit it, right?
So then they have to say, well, okay, so there, there definitely were human.
And in the case of Neanderthals, they really were humans because we could interbreed with them, right?
So they were in the same species as us.
So if you ask questions like, well, race is socially constructed, right?
Well, is the difference between Neanderthal and Neanderthal socially constructed?
I don't think it is, right?
Because they look very different, right?
And it's not just a social construction, right?
So at some point, by using that slightly more distant example, you can explore a lot of these issues without having quite the same sort of moral or ethical charge to it because nobody's really worried about discrimination against Neanderthals.
So you can get them to kind of admit that, yeah, they might have been dumber than us.
They're definitely different from us.
We could interbreed with them.
I'm not sure I want my daughter to marry a Neanderthal.
So you can have that conversation because it's just that little bit of extra distance.
It's not talking about somebody who lives down the block from you.
It's talking about a now extinct version of human.
And then you can start to narrow the gap.
Well, so a few hundred thousand years was probably enough for our brains to evolve somewhat differently than theirs, right?
Maybe a hundred thousand years, is that enough?
Fifty thousand years?
So you can force them to admit it's somewhere in there.
Somewhere in there is enough evolutionary time to cause these differences.
Until they do unfreeze and regenerate Neanderthals, in which case you're going to be a racist again.
I mean, until then, it's a great argument.
I just want to point that out.
Sorry, you were just about to make another point.
Right.
Once we give all of them green cards just for being Neanderthals or citizenship, yeah, then it'll be a real problem.
But, you know.
Right.
Neanderthal right citizenship.
And this also comes to questions of inferiority and superiority, and I've steadfastly rejected any question of ethnic or racial superiority or inferiority because it's merely adaptation to local circumstances.
Blacks in Africa are eminently adapted to the environment of Africa, and Inuit or the Eskimos in the north are Are eminently evolved, and you swap those two, and they're not going to do that well in each other's environment.
So there's no such thing as superiority or inferiority, because that, again, would be a repudiation of evolution, which allows or encourages, I guess is another way of putting it, encourages the best possible adaptation to local environments.
So I think that's where people get hung up as well.
If there are different ethnic characteristics, people somehow think it's superior versus inferior, when I don't see it that way at all.
If you look at these, the well-reconstructed Neanderthal skeletons, they're enormously big and powerful humans.
Every one of them would be a superstar professional athlete if they were resurrected today in the modern world.
And so they're definitely superior to us in those ways, right?
They'd all be playing in the NFL or professional rugby or whatever, or in the Olympics.
So yeah, they're not...
Maybe our brains are better than theirs, but it's not a...
The definition of superiority is rather subjective, actually.
Right.
So when it comes to social mobility, IQ is something that is very important.
And I want to sort of talk about how people look at a snapshot of the world and say it's unfair.
But I think if you want to look at sort of the cycle of how people move up and down, these are constantly moving escalators or elevators of social mobility.
And a lot of this has to do with regression to the mean.
So you might look at some brilliant person who's doing fantastically well and say, well, that's totally unfair.
But if you look at regression to the mean, particularly if he is very smart and chose a very smart wife, the snapback or the constant churning of this kind of stuff, you know, if you want equality, just be patient, I think is one way of putting it.
I wonder if you could help step people through how that works.
I totally agree with you.
I mean, just as an observed fact, the phrase that's often used is shirt sleeves to shirt sleeves in three generations.
So, you know, some exceptional first generation creates a big family fortune and Maybe the second generation sort of hangs on, but doesn't create more value.
And by the time it's the third generation, they've squandered it.
And that's a very common pattern.
And what's really happening here is that it's really quite complicated to understand how regression works.
So if someone is well above average on a trait, say IQ or something, perhaps most of their advantage comes from genes.
But a chunk of it probably also came from being lucky, maybe lucky in the developmental process or having a good environment that is matched to their particular learning styles, etc.
So there's a piece of that that can't be passed on.
Right.
So imagine that the superior guy marries an equally superior gal, but some chunk of their superiorness is coming from luck.
And that luck cannot be passed on to their kids.
And so that is why you see people regressing back to the mean.
So Shaq's kid is going to be very big and tall and athletic, but probably not as big and tall and athletic as Shaq, even if Shaq had married his female counterpart.
So...
It's hard for people to understand that because there's a sort of conditional probability aspect to it.
But it's definitely a well-verified natural phenomenon.
And so, yes, there are very few dynasties, family dynasties, that can last multiple generations.
They do happen.
And there is actually some evidence that you can have classes of people, even in Western societies, that are kind of mating mainly with each other.
And so you can have strata of societies An academic economist at UC Davis named Greg Clark has analyzed this, and he's looked at, for example, surnames.
So there are some surnames that are associated with, say, the upper classes in England, and they've been associated with the upper classes for a very long time.
So the regression back to the mean is a bit slower than you might think.
If you talk about a whole class of people, it's happening more slowly than this sort of three-generation thing that I described, which would apply to maybe a particular billionaire or something.
But if you look at a whole category of people, it can take place over much longer periods of time.
And that's a very interesting aspect of research.
It's not fully understand what's going on there.
Well, I think the sort of rags to riches to rags situation that you describe was much more common when the government had less control and involvement in the economy as a whole because, of course, now what people do is they take their good fortune.
And this is important to understand as well.
This happens in science, of course, as well, as you know.
That you just happen to have the right paper in the right hands at the right time.
That gets you your first big break.
You use that to leverage your next big break.
And that first one is kind of lucky.
But after that, I mean, everyone works as hard or everyone can work as hard.
Some people do end up kind of fortunate and that fortune tends to swell.
And then what happens in the business world is you say, well, you know, we've got to put up tariffs against foreign competition, and we've got to get special licenses for our guys now, and we've got to get special tax breaks, and we'd like it if the regulatory environment was more complex so smaller companies don't have the army of lawyers and accountants to conform to all of these incomprehensible rules.
I think if you want equality, you want as free a market as possible because it's going to chip away at the privilege that you can't use political machinations to defend.
I agree with you.
Big corporations can persist long after they've become very mediocre and horrible and are not creating much value, mainly because they can interact with government and carve out special privileges or protections.
That's a little bit different than the CEO. So the CEO of that company, say the guy who founded the company, his son and daughter might be a little bit mediocre and spoiled by their trust fund, and then their kids are not so special.
So that sort of Up-down cycle for individuals is sort of still there, I think, because even if dad did everything possible to protect you, by the time it's the grandkids or the great-grandkids, you know, there's nothing left to protect you.
The corporation is a different thing.
You could have a corporation which is milking the system for quite a long time, even after it's stopped producing a lot of value.
Right.
And of course, the government forces a lot of people to invest in the stock market who don't really want to be there but would rather invest in the stock market than pay taxes on their income, you know, if it's retirement benefits or whatever.
And that artificially feeds these entities long beyond, I think, what the free market would do.
Now, as far as the question goes, and it's an interesting question I think most people have thought of at one time or another in their life.
Tell me what you think.
Would you rather be born smart but broke or...
Rich, but average.
What do you think will give you the best outcome in the long run?
Because a lot of people look at the trappings of wealth and think it just sort of photocopies intergenerationally almost automatically.
Well, obviously, I have to quantify that a little more to give the answer.
But for a certain choice of parameters, it's better to be born smart, but of modest means.
Of course, it depends a lot on what your utility function is, what things you value in life.
I think one of the things about...
Really smart people is often they don't want, they don't focus on trying to make money.
They focus on trying to learn more things about the world and self-actualize.
And I can sort of feel that you're kind of this kind of person as well.
So it is interesting that you might think, oh, these super, super smart people are all going to become billionaires.
But actually, most of them don't necessarily want to become billionaires.
Right.
And there are, I think, some studies that show fairly conclusively that IQ is a much stronger, but significantly stronger predictor than race for escaping the bottom quintile of income.
If you want to test theories of how discrimination works, actually, I had arguments with a sociology colleague of mine many years ago about this.
I said, look, if you look at the, there's something called NLSY, which is the National Longitudinal Study of Youth.
And it's been going since, I think, Patrick Moynihan.
So probably, like, more than 50 years now.
So you can track individual kids.
You know their IQ scores.
And you can ask, like, for a kid born in the bottom quintile of family social economic status, what is the probability that they escape the bottom quintile or they make it to the top quintile or the middle quintile?
And you can ask, well, conditional on IQ, what's the probability?
And you can ask conditional on race.
And it turns out that if you take black kids and white kids born in the bottom quintile, but you match them for IQ, they have almost identical probabilities of getting out.
So my sociology colleague, when I showed him this data, was just his jaw just hit the floor.
And he kind of refused to believe it.
But then he sort of looked more carefully at it, and he couldn't refute it.
And I said, doesn't this at least put some kind of bound on the level of racism that you think is existing in society?
Because wouldn't we be able to detect it very easily this way?
If, in fact, life was so hard for the black kid, how could it be that just by matching IQ scores with a white kid from a similar family environment, they have equal probability of getting out?
So...
And I just want to give you this quote.
It's from the great infoproc.blogspot.ca, which we'll put a link to below.
Once one accounts for the AFQT score, the entire racial gap in mobility is eliminated for a broad portion of the distribution.
I wonder if you could just step people through that statement.
Right, so that's sort of what I was saying.
So imagine that you, so you have this huge data set where you've followed a huge 100,000 kids for 50 years, something like this.
It's a U.S. government funded study, but still on and going.
So pick out all the black kids who scored 100 on their IQ, effectively at exactly 100 plus or minus one point on their IQ test, and then take all the white kids who scored 100 plus or minus one point on their IQ test.
And then also only look at the white kids who are in the bottom 20% in the population for family income and education.
And also take only the black kids who are also in that lowest quintile for family SES. And then just ask, what's the probability that That a given white kid with those characteristics got out.
And let's say the probability is like, actually, if you have average IQ, almost everybody got out.
It turns out it's kind of funny.
If you're a libertarian, you love this data because it shows like, you know, if you're stuck down there, actually, there's some problem, right?
So almost all the white kids who had IQ of 100 got out of the bottom quintile.
And actually, almost exactly the same probability, the same fraction of Black kids who were born in that bottom quintile got out if their IQs were also 100.
And if you do this for different IQ scores, the two curves are almost sitting right on top of each other.
So it's almost as if IQ is the number one driver for much bigger than all the other effects like race for when they got out of the bottom quintile.
And this is not disputed data.
This is extremely solid data with huge sample size.
And it's been around for a long time.
And this is, you know, given that blacks in America are an average of about a give or take standard deviation below whites and a little bit further than that below East Asians, they do less well.
But if you take a slice of East Asians or Ashkenazi Jews or Inuit or whites with an average of 85 IQ, they all do about the same.
And I think what the market is fundamentally measuring is not skin color or hair texture or anything like that.
It's measuring G.
The market measures G. and that is very difficult for people to say.
It is something that I've said on this show for years.
The world is scarily more fair than you think it is.
And that is something that's very challenging for a lot of people to accept.
Now, I want to mention a couple of things.
One, the graph that I posted on my blog that shows this result, it was computed by a researcher at the Pew Trust, Pew Charitable Trust.
It was not computed by some libertarian right-wing, alt-right, evil person.
It was a progressive guy, an economist who sat down and did this calculation.
So I think we can trust it.
Now, in the spirit of Aristotle and Alan Turing, let me give you the counter-argument which my clever sociology professor friend gave me.
Well, actually, he didn't, but I sort of proposed it to him as maybe he would want to latch on to that.
Well, society is so unfair and there's so much institutional racism against black kids that for a black kid to get a decent enough education to score 100 on the IQ test at age, whatever, 17%, I, you know, all these hassles, he really is somehow in some hidden way a superior kid.
And so actually, you're not really fully matching, right?
The black kid really is better.
And that's what explains how in this racist society, he's able to earn the same income as though the white kid later in life.
And these kind of really subtle hypothesized invisible effects are really hard to get at, right?
I mean, where is the null hypothesis for this?
Yeah.
Design a study that actually falsifies that, right?
So at some level, you can just say there's some invisible miasma, which is pushing Asian Americans up and it's pushing African Americans down.
Maybe it's true, but I find it very ascientific to adopt that as your primary hypothesis.
in ether theories myself.
It's exactly the ether of, I guess they call it institutional racism.
Not that there isn't any racism in the world.
There certainly is.
But I'm sure I've experienced it and other people have as well.
But I don't think it's as strong as the typical SJW thinks.
Well, I guess another counter argument would be to the hypothesis would be to say, well, you know, the welfare state transfers a lot of wealth to the black communities.
And we have government schools, which are attempting to up the educational capacities.
And we have affirmative action, both for universities and for hiring as a whole.
And And that fills in the gap to the point where if we didn't have all of that stuff, then the racism would show up and the blacks wouldn't be doing as well, even with the same IQ. So we want to get as many dimensions as possible into this argument.
And again, that would be something which you could probably find some way to design a test to figure out.
The part where I really draw the line, which I really don't like, is that because there are these unpleasant facts that different groups have different averages when they take these tests, People just want to throw the whole test in the garbage.
When, in fact, if you look at all the things that psychology as a science has invented, cognitive score measurement is by far the best thing they've ever invented.
And it's used by the US military, every major university, every major university in the world is doing some kind of cognitive testing to filter students, figure out who gets in, who gets the scarce educational resources.
So to try to throw that in the trash just because you don't like the picture of the world that it paints is just the craziest thing.
It's like saying we're going to throw out gravity because I fell and skinned my knee.
Well, it is the one thing.
I shouldn't say the one thing, but it's the one thing that comes to mind for me at least when I think about objective value that the field of psychology has provided.
You know, sure beats Freud's rape theories and incest theories and so on.
Yeah.
Now, let's move on to, just in case there's one single solitary soul out there in the multiverse that we have not upset or offended yet, let's move on to genetic designer babies, because this is something that I find enormously fascinating as a possibility.
So...
As we begin to map these kinds of things and we begin to look at the levers that may be able to affect human intelligence down the road, the inevitable question comes up.
And eugenics, of course, is one of these ugly terms, but, you know, we do eugenics all the time.
We do eugenics when we choose who we marry.
We do eugenics when we're farmers and we figure out what seeds, as you point out.
Corn has had a 30-standard deviation alteration from its original form.
And if you've ever tried to live in the wild without any food around you grown by farmers— You'll know it's been quite a lot of...
So I think we need to sort of step out of that particular scare-word scenario.
But let's talk about some of the possibilities that exist for changing the possibilities for human intelligence from conception onwards.
Yeah, so let me describe the present and near future.
So everything I'm about to say is 100% not speculative.
It is either happening now, as I'll point out, or it will happen very soon.
So about a million times a year worldwide, a couple that is having fertility problems, it's almost always couples that are having fertility problems, except for one very small subcategory, which is super rich billionaire Silicon Valley venture capitalists who are also getting into this, even if they don't need it.
And I'll explain that in a second.
They decide they want to use something called in vitro fertilization, which means that they hormonally stimulate the mother-to-be, she overproduces eggs, they extract those eggs, they fertilize the eggs outside of her body, and then they re-implant them for it to become a baby.
Now, it's quite common for the couple to end up in a situation where they have more fertilized embryos than they want to use.
And in fact, the technology is so good now that you can freeze these embryos in liquid nitrogen, but as far as we can tell, no ill effect.
So you can have fertilized embryos and thaw them out when you're ready to use them.
That's totally there as a technology.
It's best practice now in IVF clinics to do that, to let the mother's body recover from the hormone shots.
So it's actually best practice to freeze the embryos for about a month before you implant them.
Now, the technology is completely there to...
When the embryo has grown to about 100 cells or something, pull a few cells off.
Usually they pull them off the region that is meant to become the placenta, not the actual being, but has the same DNA. And then genotype.
And the results of that genotyping then can be reported back to the parents, and then they can make a decision about which of the embryos is best to implant.
Now, I submit that this is what...
This does fall into the term eugenics, which literally just means good genes or improving better genes, improving genes.
But I don't see anything evil in this.
So what exactly is evil?
Let's suppose that at the moment, all we can do is screen against diseases.
So I can tell you which embryo is going to have Down syndrome.
I can tell you which embryo is going to have cystic fibrosis.
And the parents just make the decision, I'm not going to implant The embryo that is going to have cystic fibrosis because it's just horrible.
I think it's very hard to ethically criticize the availability of this technology or the decision-making of the parents.
I think even though it is, strictly speaking, eugenics, I challenge anybody to come out of the woodwork and tell this parent that, no, you should implant the cystic fibrosis baby that's going to die a horrible death at age two.
Well, and prevention is better than cure.
And if there was a genetic treatment for cystic fibrosis, people would take it.
And so if you have a way of having a baby born without cystic fibrosis versus one that is, I think that's in the realm of prevention rather than cure, which I think most of us would think is better.
Right.
Now, the probability that a A child that's conceived in the completely old-fashioned way without using any of this stuff has a kind of serious problem is about 1% or 2%.
And 1% of that is just Down syndrome.
But there's another 1% or 2%, which is sort of just rare genetic diseases that we're carrying around.
And so if you just say, hey, by supplying this technology, I can eliminate 2% risk to you of a really nasty outcome in your family...
I think that's all net good, right?
Who could criticize that?
Okay.
But now, suppose instead of just single-gene diseases or problems with chromosome counts that lead to Down syndrome, suppose I can make quantitative predictions about the embryo.
Suppose I could tell you embryo 4 is going to be the tallest, embryo 3 is going to be the smartest, embryo 2 is going to be very antisocial.
Suppose that level is Of granularity and accuracy was available in the reports.
Yeah, we go from a dice roll to a buffet.
Yes.
And that is very close.
We're just, depending on, so I could be more specific depending on which complex trait we want to talk about.
But for some complex traits, it's there right now.
For other complex traits, five years at most before we'll be able to Make some decent quantitative predictions about them, like IQ, for example.
Well, if as a parent...
I'm sorry, you were just about to say something else.
Let me hold my thought.
I was just going to say, so that is the near-term future that we as a civilization face.
This is going to be here.
And I predict, actually, that as the delta in outcomes, positive delta in outcomes, gets better from IVF coupled to these new technologies, as that gets bigger...
More and more people are going to do, like certain billionaires and Silicon Valley types, are going to do IVF even if they don't have a fertility problem.
They just want to get rid of that tail risk and they want to have a better outcome.
And it's going to become normative.
I think probably within a generation it will become normative to do this as the way that you reproduce.
Right.
Now, that starts to sound just a little bit unfair.
And I can understand how people have ethical issues with that, which is to say, aren't we going to create a superclass of geniuses that aren't going to be subject to the regression to the mean?
And the technology to maintain their high genetics is only available because of the previous accidental generation of high genetics.
And I think that's, unless the price gets down to the point where you can do it by backing into a vending machine in some horrible manner, I don't see how people could not have some ethical questions about all of that.
So if you're a country like England, which has national health system or Canada or Singapore, the cost benefit for this is very positive.
So in other words, it costs at slightly higher economies of scale, it would only be probably less than $10,000 to do a cycle.
And that $10,000 pays off because you're eliminating these very expensive tail risk conditions and getting a generally healthier population.
So, if you do the cost-benefit analysis… Well, except for the old age pensions, because these high IQ people, they're like vampires.
You can't kill them.
I mean, it takes forever.
But, you know, that's down the road for politicians.
They don't think that far.
They raise your GDP. That's true.
That's very true.
And of course, I've talked to a bunch of criminologists on the show as well, Steve, and they talk about how the sweet spot for criminality is around IQ 85.
And we'll get to sort of the IQ thing, but if you had fewer children born around that sweet spot of criminality, I mean, the social cost, the trauma cost, the emotional cost, the incarceration cost, I mean, the massive cost that crime has in society would enormously diminish.
I would not be surprised if That cost was just as big as the medical savings from getting rid of, you know, rare genetic diseases, etc., etc.
Right.
Well, and of course, I mean, the trauma aspect, the fear aspect, all of the associated costs, you know, security systems and all of that, people living in fear, people moving out of neighborhoods and kids getting beaten up in school by crazy criminal kids and so on.
I think the tertiary effects of crime would be even more widespread and the sort of associated shadow costs.
Most Westerners are shocked that a woman can walk across Seoul or across Beijing in the middle of the night with almost no fear that anything's going to happen to her.
And in fact, actually, this is a thing that is even less well-known.
In Tokyo, which has...
I don't know if you've ever been there, but they have this tremendous metro system.
It is common for parents to send kids, young kids, like seven-year-old kids...
By themselves to a school which is several metro stops and transfers away.
And the kid goes to school, has their school day, and comes back by themselves.
In Japan, they're 7 years old, 10 years old.
If I did that to my kid in beautiful East Lansing, Michigan, which is a college town, I would be locked up.
But somehow, this is normal behavior in Japan.
So that's a different world.
Well, I mean, Mexico City and Seoul are not dissimilar in size.
And yet the crime rates are extraordinarily different.
And of course, we've had Jason Richwine on the show talking about Mestizo IQ. And of course, if you look at the average IQ of Mexico, it tends to hover a little bit higher than the sweet spot for criminality.
But of course, in East Asian countries, as you point out, Beijing and Seoul, you get an average IQ of 104, 105, 106, and so on.
And therefore, criminality is just enormously lower, except in the political class, where corruption remains a problem across the world.
But this is why, you know, when people like Sadiq Khan say, well, you know, terrorism is just part and parcel of living in a big city.
It's like, not all big cities are not in Poland and not in East Asia and so on, because I think there are IQ differences in the populations that are important.
This idea that they're trying to normalize the terrorist threat is just absurd.
It's like saying, like, I have raw feces.
My sewer doesn't work.
I have raw feces flowing down my street.
Oh, well, that's okay.
Just get used to it.
That's what modernity is like.
It's good for you.
I think those politicians are insane.
Well, or power-hungry to the point of ignoring basic reality, which, you know, potato, potato when it comes to that kind of stuff.
Because here we get to the challenging, to me, one of the most challenging questions, and I sort of put this out for the audience to debate and discuss, and I look forward to your feedback.
You can go at it in the comments section below.
But this question of...
I think most people will be more or less comfortable with, let's avoid obvious deficiencies if we can do that.
You know, as you say, Down syndrome or cystic fibrosis or other things which are going to cause significant problems and result in a short and sort of painful life for parent and child.
But then there's a question of...
The lower end of the scale, you know, certainly mental retardation, significant mental retardation, I think everyone could say, well, if you have the option to avoid that, so much the better.
IQ 87?
You know, if you give parents the choice and you say, well, you can have a kid who's got an IQ of 85 or 90, or you can have a kid who's got an IQ of 120, well, I think most parents are going to say, I think we're going to go to the higher end of that.
But at what point do we say, I'm not fixing a problem, I'm reducing or eliminating part of the natural variance of humanity?
Yes.
Well, so I think you're reasoning about this in really the perfect way.
So this is how one should reason about it.
So for diseases, there's not going to be much dissent.
I mean, there'll be a little bit of dissent because there are people who don't, for example, want to abort Down syndrome babies.
And, you know, I don't want to judge anybody's choice.
I actually think the parents are the ones who should choose.
You know, if you adopt the opinion that these embryos are not yet people, Which is, you know, if you're Catholic, you may not agree with that.
But for most Americans, I would say, most people in the world, they would say, okay, embryos don't have full human rights.
Once you say that, then I think the people who are in the driver's seat should be the parents.
Now, I think there is this slippery slope.
So what will happen is that, as you say, okay, we can warn you if the IQ is going to be under 85.
And then you just kind of don't thaw that one out.
Okay.
Okay, well, we'll warn you if the IQ is going to be below 100.
Okay, well, we'll not thaw that one.
Eventually, people will get very uncomfortable.
I anticipate that you'll have countries where the laws are a certain way, diseases only, nothing more.
And then countries, for example, in Asia, where anything goes, parents just get to decide what they want to do in the clinic.
And that's going to give those countries an evolutionary advantage over time and rapidly too, because we're eliminating the, you know, the intergenerational slowdown of this kind of advancement.
And if this is going on rapidly, those countries are going to escalate up in terms of GDP, they're going to escalate down or de-escalate in terms of criminality.
And there's going to be...
It's almost like an arms race, but based upon positive genetics, if that makes sense.
Remember that the extra cost of having the IVF kid now...
And now, when it's only a million people a year who are doing it worldwide, it's a million couples.
It's not real true economies of scale yet.
It's still a boutique business with small fertility clinics doing it.
The total cost is significantly less than, say, a couple years at a very good private kindergarten, K-12.
So a couple years tuition at Dalton, In Manhattan will more than cover multiple cycles of IVF for you and all the genetic screening you want.
And let me just point out something that if you have the choice between genetic screening as far as resulting IQ versus paying a lot of money for a really, really great private school, you want to pick the genetics because schooling has not been found to be statistically associated with resulting IQ in any way, shape or form.
Absolutely.
You'd be much better off taking the IVF cycles than paying for a year of expensive kindergarten.
And so the value proposition is very clear.
And so I think you're going to see, you know, first wealthy people doing it, then slightly less wealthy people doing it, then pretty much anybody who can afford it eventually will be doing it.
And if you are not allowed to do it because of the local laws, you can fly to, say, Singapore, for example, and have your IVF cycle done there.
Which means that if you put restrictions in your own country, you're simply excluding people who otherwise may have been able to afford it and reducing it to the number of people who can afford to go to Singapore and have the contacts and have the expertise and the knowledge and the time off and the leisure and blah blah blah blah blah.
So it makes it even more exclusionary to deny it to the people in your country.
And then taking it even further, there will be some very organized countries that make it free, maybe mandatory.
Maybe make it mandatory to get rid of the really expensive diseases, for example.
Maybe make it mandatory to get rid of IQs less than 80.
You know, it can creep up that way as well.
Well, as long as the government's paying for healthcare, they do have a rather unholy right to interfere in some way as to other taxpayers.
Now, let's go even further.
Oh, yes, we're going further, people.
Because now we're talking about not just reducing some of the lower end of the bell curve.
And again, this is around prevention.
It's not killing anyone and so on, right?
But we can also, as you've talked about a number of times, we can hit the gas at the high end to areas almost...
Unimaginable.
We're talking about IQs in the hundreds and so on and, you know, the sort of once in a hundred years geniuses becoming something that we might be able to crank out on a regular basis from the old between the thigh factory line, so to speak.
Let's talk about what could be possible just throwing ethical caution to the wind and just talking about what's technically possible because when it's technically possible, it's going to come up sooner or later in our ethical questions.
Just to frame it in a fun way, imagine that you have some kind of civilizational techno-economic competition between big nation states.
And one government just decides, hey, we're going to plow billions of dollars into producing really great grains to accelerate our advantages over other countries.
It turns out my best estimate of how many Genetic variants are affecting intelligence or are responsible for making two typical individuals different from each other.
It's on the order of about 10,000 variants.
Now, when you say 10,000, it seems like a huge number.
But what's interesting is that if you flip a coin 10,000 times, you'll get about 5,000 heads.
But the exact number of heads you get will form a Bell distribution, Gaussian distribution.
And the width of that distribution only grows like the square root of 10,000, the number of flips.
So it's sort of 500 plus or minus 50 or 500 plus or minus 100 is getting you already to the edge of the distribution because it's very narrow.
If you flip a lot of times, you're going to get something very close to 50% heads.
So in order to make someone one standard deviation, say 15 IQ points, smarter, You only need to flip something like 50 or 100 variants.
So 50 or 100 places where somebody has the bad minus variant, you flip it to a plus.
So it's sort of just so people understand, if you want to go from, let's say, New York to Europe, you know, one or two degree difference at the beginning ends you up in completely different countries.
And so we're starting things very early, and you can end up with a sort of big arc from there.
And so you don't need to be able to manipulate all 10,000, if I can understand that correctly.
That's right.
In fact, if you took a very smart person and you said, well, how many more plus variants does that very smart person have than the average person?
It might only be 100 or 200.
And so you can map one to the other by flipping 100 or 200 variants.
So we haven't talked about CRISPR. There's a new technology which was just discovered in the last few years and it's being perfected at a very fast rate, which allows very accurate gene editing.
And it's already in clinical trials for a number of different applications.
And it's been done, for example, to monkeys.
In a lab in China, they deliberately made autistic monkeys by editing a particular gene that was known to be related to autism.
So it's moving very fast.
And I would guess, I would be surprised if in 10 or 20 years, we could not easily do hundreds of edits on a human embryo.
I'm sorry, you just have to...
How do they edit...
Geez, that just blows my mind.
So what's interesting about this is that a lot of the most powerful technologies in molecular biology and genetics, they come from systems that nature itself produced, in this case in bacteria.
So when these bacteria are attacked by viruses, they actually are able to attack the virus itself and chop up its DNA. And so they developed certain molecules which were very good at chopping and editing DNA. And some scientists figured out that these systems existed, and they figured out how to take these systems and generalize them and make whatever arbitrary edits you want to make.
And this has been tested now widely.
In fact, at my university, we have a gene editing core, which is using CRISPR. We're making genetically edited mice every day, genetically edited plants.
It's a widely used technology after just a few years.
And again, I will say that I would be surprised if in 10 or 20 years we were not able to make a few hundred edits on a human embryo with very low off-target error rate.
And so then, theoretically, we could push people to the maximum historical human intelligence or beyond?
Beyond, right?
If you go beyond.
Right.
So when I said, like, super smart guy like, you know, Einstein or Hawking or somebody has a few hundred more of the plus variants than average guy.
Yeah, you could make a baby with 500 more plus variants than the average guy.
Maybe only one person has ever existed in history like that.
And we might as well go to 700.
Well, no one in history has existed like that.
So basically, we get an 80-foot-tall basketball player competing with everybody else.
Let me give you the Shaq analogy.
I don't know if you follow the NBA, but so if you remember when I was growing up, most centers in the NBA were tall, skinny guys like Kareem.
They would be seven feet tall, but they would not weigh more than an NFL linebacker.
You know, they were only giraffes.
Yeah.
Yeah.
Now, no one at that time ever thought you would get a guy who was built like an NFL cornerback or running back, but seven feet tall.
No one ever thought that was possible until Shaq came along.
Shaq, if you scale Shaq down to six feet tall, he's built like a running back.
And if you remember when he first came in the league, he was destroying rims.
He would dunk with so much power, he would actually bend the rim.
The rim was not built to deal with him.
And they actually had to change the rules because the Lakers were winning so much that they couldn't stop him.
Anytime he got the ball in the paint, nobody could stop him.
He would just go to the hoop and dunk it.
But he had a combination of things which people didn't think was possible.
But Now there's Shaq and we'll see another Shaq eventually.
We'll see people who are like, we'll see a guy who's 7'5 and still athletic and explosive, you know?
And in the same way for people to say that like, well, there's actually not going to be anybody smarter than Einstein.
There's not going to be anybody smarter than Gauss or Stephen Hawking.
That's absurd.
It's totally absurd.
And it's contradicted by everything we've seen in domesticated animals, domesticated plants, breeding for agriculture.
We're able to, when we want to push things, we're able to push them Way beyond anything nature does randomly.
And I also wanted to point out when people say the possibility that governments might spend money to make their citizens smarter, this has already happened.
It's failed, right?
but the thought popped into my head earlier, Steve, was the Head Start program, which I think was started under Bush and was an attempt in many ways to close the racial IQ or performance differences in America.
They blew past $100 billion plus, and although there were some initial successes, there tends to be with these kinds of things that evened out and dissipated and vanished over time.
And so governments do spend enormous amounts of money attempting to influence the resultant intelligence of their citizens.
It's just that genetics seems to point out that doing it post-birth almost never works.
But if there's some way to do it pre-implantation, again, not even pre-birth, we're not talking about changing fetuses in the womb and so on, I'm trying to decreep it in a way because there is always this, you know, we're on a mountainside calling down the fires of Satan to alter the soul of humanity and so on.
But we do this all the time and governments have definitely invested countless amounts of money to try and affect the intelligence outcome of their citizens.
Yeah, so again, I think your intuitions are very good.
So there is what Steve Pinker refers to when I talk to him about this as the ick factor.
And it's going to take a while for people to get over the ick factor, right?
The idea that, oh, we'll put an extra student teacher in the classroom, we'll get you really nice books or an iPad to learn from, that's very easy to throw money at, even if the data show that it's of limited effect.
Now, this has a big ick factor, but what's going to drive it is the following.
It's more and more common among the elites to end up in an IVF clinic because the Women are very high-powered.
They delay childbirth because of their careers.
It's increasingly common.
And among billionaires and hedge fund managers and even university professors, it's not uncommon at all to know people who have had IVF kids, et cetera, et cetera.
So once the elites become accustomed to it, it will be easier for them to say, oh, well, gee, let's make this part of our National Health Service because it's not that icky.
In fact, my granddaughter was born that way.
So, Steve and I, Pinker and I actually disagree on this.
Actually, we were on a panel at the 92nd Y talking about genetics and genius, and I think that's one place, for example, where I think he uses the ick factor terminology.
I think we'll get over the ick factor faster than people think, and largely because elites actually already in most countries have been exposed to this technology.
Well, I think it's funny how to almost harness the social justice warrior momentum, because if there are people out there who can be offended by you being upset about something or you bringing some subject up, it tends to squelch that debate.
And as you point out, if we have millions of people who are only alive because of IVF technology, and people say, well, IVF technology is bad, are you saying I shouldn't be here?
Do you want to say that to my face?
Do you want to step up to me and tell me I shouldn't be here?
And then people will back down from that because, you know, this cult of not giving offense seems to have taken over the planet.
We might as well use it for good in some Right.
Well, you can define them as a protected class.
Babies and genetically engineered super babies are a protected class.
Yep, that's right.
Opposition to it is now hate speech.
Well, I feel we could chat for a while.
I want to make sure that we give you a chance to have something else with your evening.
I want to remind people to follow Steve's great stuff on Twitter, twitter.com forward slash HSU underbar Steve.
His website, one of the websites is infoproc, I-N-F-O-P-R-O-C dot blogspot dot com.
Dr.
Su, fantastic chatting with you today.
I really appreciate the time that you take to break down these imminent and explosive topics to something very digestible by the educated layperson.
It's a huge public service, and I hugely appreciate you taking the time today.
It's been a pleasure, and I'd love to come back on the show sometime.
Next time, we'll talk about AI. Oh, please, please, let's.
We're both software guys, so we can probably rub geeks together that way very, very well.
