YouSaid · the spoken record

Steve Hsu

lines on the record: 164
first: 2022-08-23
most recent: 2022-08-23
sittings or episodes: 1
sources: podcast

Every line below is reproduced as it was said and linked to the record it came from. Nothing here is summarised or generated. Directory · Search · Corrections

“If you want to know how tall this embryo is going to be, I'll tell you a mind blower. When you do face recognition in AI, you're basically mapping someone's face into a parameter space of like, I think it's on the order of hundreds of parameters, right? Of those parameters is superheritable. So, in other words, if I take two twins and I measure their, I photograph them and the algorithm gives me the value of that parameter for twin one and twin two. They're very close, obviously. That's why I can't tell the two twins apart. The face recognition can ultimately tell the twins apart. The really good face recognition, but you can just conclude that almost all these parameters are the same for those twins. So it's highly heritable.”
2022-08-23 · Dwarkesh Podcast · Steve Hsu - Intelligence, Embryo Selection, & The Future of Humanity · IDENTIFIED FROM THE TRANSCRIPT · source
“Some special traits that people are nervous about, but you just upload it to some vendor who's in Singapore or in some free country. And they give you the report back. Have to be us. I mean, we don't do those edgy stuff. We only do that health related stuff right now.”
2022-08-23 · Dwarkesh Podcast · Steve Hsu - Intelligence, Embryo Selection, & The Future of Humanity · IDENTIFIED FROM THE TRANSCRIPT · source
“Yeah, I think we're going to evolve to a point where because the genotyping technology is getting better and better, eventually we are going to be out of the wet. Of this business and only in the bit and cloud part of this business because eventually the clinic, no matter where it is, they're going to have a little sequencer which is this big. Tech is going to do it, and then they're just going to hit upload, and then they get the report back like three seconds later from us for the physician to look at, and the parents can look at it on their phone or whatever. Actually, we're basically there, actually, with some clinics. So, yeah, it's going to be tough to regulate because it's just bits, right? So you have the bits and you're in some repressive, terrible country that doesn't allow you to select”
2022-08-23 · Dwarkesh Podcast · Steve Hsu - Intelligence, Embryo Selection, & The Future of Humanity · IDENTIFIED FROM THE TRANSCRIPT · source
“So in the IVF workflow, our lab, which is in New Jersey, can service every clinic in the world because they just take the biopsy they put in a standard shipping container and they just send it to us. And then we actually genotype the DNA in our lab, but we've actually trained a few of the bigger clinics to actually do the genotyping on their site. And at that point, it's just like they upload some data into the cloud and then they get back some stuff from our platform. So at that point, it's going to be the whole world, man. Every human who wants their kid to be healthy and, you know, get the best they can, that data is going to come up to us and the report is going to come back down to their IVF position.”
2022-08-23 · Dwarkesh Podcast · Steve Hsu - Intelligence, Embryo Selection, & The Future of Humanity · IDENTIFIED FROM THE TRANSCRIPT · source
“They can give me samples. We can get into that. For example, we just learned that you can recover DNA pretty well from the, like if someone licks a stamp and puts it on a... Their correspondence. So, I mean, if you can do Neanderthals, you can do a lot for solving crimes.”
2022-08-23 · Dwarkesh Podcast · Steve Hsu - Intelligence, Embryo Selection, & The Future of Humanity · IDENTIFIED FROM THE TRANSCRIPT · source
“So it's great. I love your question. It is totally standard every good IVF clinic in the world regularly takes embryobiopsies. So that's totally standard. It's like a lab tech doing that. And then what happens is they take the little sample and they put it on ice and they just ship it. And the DNA as a molecule is extremely robust and stable. In fact, my other startup solves crimes that are 100 years old from DNA that we get from some semen stain on some rape victim. Serial killer victims brastrap. We've done stuff like that.”
2022-08-23 · Dwarkesh Podcast · Steve Hsu - Intelligence, Embryo Selection, & The Future of Humanity · IDENTIFIED FROM THE TRANSCRIPT · source
“That's exactly what I would say. I would say, like, yeah, maybe, I mean, Paige Rink actually by now is totally deprecated Even if somebody else comes up with a somewhat better algorithm or somewhat better, maybe they have a little bit more data. If you have a team that's been doing this for a long time and you're really focused and good, it's still tough to beat you, especially if you have a lead in the market.”
2022-08-23 · Dwarkesh Podcast · Steve Hsu - Intelligence, Embryo Selection, & The Future of Humanity · IDENTIFIED FROM THE TRANSCRIPT · source
“Moat around the company. We do have patents on specific technologies about how to do the genotyping or how to do error correction on embryo DNA and stuff like this. We do have patents on stuff like that, but this general idea of like who's going to be the best at predicting human traits from DNA, it's unclear who's going to be the winner in that race. Maybe it'll be the Chinese government in 50 years. Who knows?”
2022-08-23 · Dwarkesh Podcast · Steve Hsu - Intelligence, Embryo Selection, & The Future of Humanity · IDENTIFIED FROM THE TRANSCRIPT · source
“Could twenty three and me or some company that has a huge amount of data and if they were to actually get better AIML people working on this, could they kind of blow us away a little bit and build better predictors because they have just much more data than we do? Possibly, yes. Now there's a core expertise that we have in doing this kind of work for years and years and years that we're just really good at it. And so even though we don't have as much data as 23andMe, we might still, our predictors are better than theirs right now. And I'm out there all the time working with biobanks all around the world, like in countries like, I don't want to say all the names, but other countries trying to get my hands on as much data as I can. But there may not be a lasting advantage beyond the actual business channel connections to that particular market. It may not be a defensible, purely scientific.”
2022-08-23 · Dwarkesh Podcast · Steve Hsu - Intelligence, Embryo Selection, & The Future of Humanity · IDENTIFIED FROM THE TRANSCRIPT · source
“That's another super question. For your entrepreneurs in your audience, I would say in the short run, if you ask like, oh, you know, what should the valuation of GP be? That's how the venture guys would want me to answer the question. There is a huge first mover advantage because they're important is the channel relationships between us and the clinics. And nobody's going to be able to get in there very easily when they come later because we're developing trust and a big track record with clinics all over the world and we're well known.”
2022-08-23 · Dwarkesh Podcast · Steve Hsu - Intelligence, Embryo Selection, & The Future of Humanity · IDENTIFIED FROM THE TRANSCRIPT · source
“Last 10 years, last 10 years, it's gotten sophisticated machine learning, genotyping of chickens. Artificial insemination modeling of the traits using ML lasts 10 years has Basically, for cow breeding, now it's totally done by ML now.”
2022-08-23 · Dwarkesh Podcast · Steve Hsu - Intelligence, Embryo Selection, & The Future of Humanity · IDENTIFIED FROM THE TRANSCRIPT · source
“They lay almost an egg a day. A chicken in the wild lays like an egg a month. How the hell did we do that by genetic engineering? That's how we did it.”
2022-08-23 · Dwarkesh Podcast · Steve Hsu - Intelligence, Embryo Selection, & The Future of Humanity · IDENTIFIED FROM THE TRANSCRIPT · source
“How much better can you do? Probably a lot better. Probably, I mean, I don't want to speculate, but because other effects, nonlinear effects, things that we're not taking into account will start to play a role at some point. So it's a little bit hard to estimate what the limiting, what the true limiting factors will be, but the one statement which is super robust, and I'll stand by it, I'll debate any Nobel laureate in biology or whatever who wants to talk about it. There's clearly a lot of variance available to be selected on or edited. That's just there's no question about that. And that's been established in animal breeding and plant breeding for a long time now. So if you want a chicken that grows to be this big instead of this big, you can do it. If you want a cow that produces literally 10 times or 100 times more milk than a regular cow, you can do it. The egg you ate for breakfast this morning, those bioengineered chickens,”
2022-08-23 · Dwarkesh Podcast · Steve Hsu - Intelligence, Embryo Selection, & The Future of Humanity · IDENTIFIED FROM THE TRANSCRIPT · source
“But we did a very simple calculation, which amazing that it gives the kind of the right result. We said given this polygenic predictor that we've built, which isn't perfect, I mean, it's going to improve a lot as we get more data, given this polygenic predictor for overall health, Which is used in selecting embryos today. If you just say, like, well, out of a billion people, what's the best person, typically? What would their score be on this index and then how long would they be predicted to live? It's about 120 years. So it's actually spot on. It's basically that's one in a billion type person lives to be like 120 years old or roughly.”
2022-08-23 · Dwarkesh Podcast · Steve Hsu - Intelligence, Embryo Selection, & The Future of Humanity · IDENTIFIED FROM THE TRANSCRIPT · source
“You have a fitness function and how big is the slope in a particular direction of that fitness function? Like how much more successful reproductively would Joe have been, Joe Caveman have been if he lived to be 150 instead of only 100 or something? And there just hasn't been enough time to explore this super high dimensional space. That's the actual answer. But now we have the technology. We're going to fucking explore it fast now. That's the point that people, you know, the big light bulb should go off. Like, no, we're mapping this space out now. Pretty confident in ten years or so that CRISPR gene editing technologies will be ready for massively multiplexed edits. And we're going to start navigating in this high-dimensional space as we like. So that's the more long-term consequence of these scientific insights.”
2022-08-23 · Dwarkesh Podcast · Steve Hsu - Intelligence, Embryo Selection, & The Future of Humanity · IDENTIFIED FROM THE TRANSCRIPT · source
“Almost exactly the same technologies that we use at genomic prediction, but they're doing it to optimize like milk production and stuff like this. So there is a big well of variance. It's a consequence of this super multi genicity of the trait. And it does look like people could, coming back to your question about longevity, it does look like people could quote, be engineered to live much longer than they currently do by just say flipping the variants that reduce risk for individual diseases that tend to shorten your life. And then the question is back to, well, why didn't evolution give us lifespans of a thousand years? Like back in the Bible, people in the Bible used to live for a thousand years. Why don't we, I mean, that probably didn't really happen. But the question is, you have this very high-dimensional space.”
2022-08-23 · Dwarkesh Podcast · Steve Hsu - Intelligence, Embryo Selection, & The Future of Humanity · IDENTIFIED FROM THE TRANSCRIPT · source
“Variants in your genome, I would make you five standard deviations taller. You'd be like seven feet tall, and I didn't have to do that much work. And there's a lot more variation where that came from. Okay, because I only flipped 500 out of 10,000. I could have flipped even more, right? So there's this kind of quasi-infinite well of variation, which evolution or genetic engineers could act on. And again, early population geneticists who breed corn, who breed animals, they know this. This is actually something they explicitly know about because they've done calculations. Now, interestingly, the human geneticists who are mainly concerned with diseases and stuff are often not familiar with what the math that the animal breeders already know. And you might be interested to know that the milk you drink comes from heavily genetically optimized cows who are actually bred artificially using”
2022-08-23 · Dwarkesh Podcast · Steve Hsu - Intelligence, Embryo Selection, & The Future of Humanity · IDENTIFIED FROM THE TRANSCRIPT · source
“Need to flip 100 minus variants to plus variants. These are very rough numbers, but one standard deviation is like the square root of n, right? If I flip a coin n times and I want a better outcome in terms of number of ratio of heads to tails and I want to increase it by one standard deviation, I only need to flip square root of n. Heads because if you flip a lot, you're going to get a very narrow distribution peaked around a half, and the width of that distribution is the square root of nut, height is controlled by 10,000 variants, and I only need to flip 100 genetic variants to make you one standard deviation for male, that would be three inches tall, two and a half or three inches taller. Suddenly you realize, wait a minute, there's a lot of variants up for grabs there. I mean, if I could flip 500...”
2022-08-23 · Dwarkesh Podcast · Steve Hsu - Intelligence, Embryo Selection, & The Future of Humanity · IDENTIFIED FROM THE TRANSCRIPT · source
“So, no, I'm glad you're really asking about these questions because these are things that people are very confused about, even in the field. First of all, let me say if you have a trait that's controlled by 10,000 variants, like height is controlled by order 10,000 variants and probably cognitive ability a little bit more, the square root of 10,000. So if I could come to this little embryo and I said, I want to give it one extra standard deviation of height plus one standard deviation of height, I only need to edit 100.”
2022-08-23 · Dwarkesh Podcast · Steve Hsu - Intelligence, Embryo Selection, & The Future of Humanity · IDENTIFIED FROM THE TRANSCRIPT · source
“He would like to tell you a story about a particular gene that maybe it has this positive effect when you're young, but it makes you age faster, so there's a trade-off. And, you know, we know about things like sickle cell anemia. And, you know, we know stories like that. And no doubt there are stories like that which are true about specific variants in your genome. But that's not the general story. The general story, which we only discovered in the last five years, is that almost every trait is controlled by thousands of variants and those variants tend to be disjoint from the ones that control the other trait. So they weren't wrong, but they didn't have the big picture.”
2022-08-23 · Dwarkesh Podcast · Steve Hsu - Intelligence, Embryo Selection, & The Future of Humanity · IDENTIFIED FROM THE TRANSCRIPT · source
“No, no, I love your interviews because the point you're making here is really good. So Dawkins, who is a kind of evolutionary theorist, but from the old school, when they had almost no data to deal, you can imagine how much data they had compared to today.”
2022-08-23 · Dwarkesh Podcast · Steve Hsu - Intelligence, Embryo Selection, & The Future of Humanity · IDENTIFIED FROM THE TRANSCRIPT · source
“All it is. I mean, it's in our paper. We basically look at, okay, how many variants are typically accounting for most of the variation for any of these major traits? And then imagine that they're mostly disjoint. Well, just how much length of variants do you need then to independently vary a thousand traits? Well, it's a few million differences between me and you are enough, right? It's trivial, it's very trivial math. Once you understand the base how to reason about information theory, then it's very trivial. But it ain't trivial for theoretical biologists as far as I can tell”
2022-08-23 · Dwarkesh Podcast · Steve Hsu - Intelligence, Embryo Selection, & The Future of Humanity · IDENTIFIED FROM THE TRANSCRIPT · source
“And this thing, because you understand some information theory is kind of trivial to explain, but try explain to a biologist. You won't get very far.”
2022-08-23 · Dwarkesh Podcast · Steve Hsu - Intelligence, Embryo Selection, & The Future of Humanity · IDENTIFIED FROM THE TRANSCRIPT · source
“And so, if you think about DD, like your strength and your decks and your wisdom and your intelligence and charisma, those are all disjoint. They're all just independent variables. So it's like a seven-dimensional space that your character lives in. Well, there's enough information in the few million differences between me and you. There's enough for a thousand dimensional space. Variation. Like, oh, how big is your spleen? My big, my spleen's a little bit smaller, yours is a little bit bigger. That can vary independently of your IQ. Oh, it's a big surprise. The size of your spleen can vary independently of the size of your big toe. Oh yeah, yeah, there's about a thousand. We just do information theory. There's about a thousand different parameters I can vary independently with the number of variants that I have between me and you.”
2022-08-23 · Dwarkesh Podcast · Steve Hsu - Intelligence, Embryo Selection, & The Future of Humanity · IDENTIFIED FROM THE TRANSCRIPT · source
“Just start looking carefully at just something as trivial as the overlap of my sparsely trained predictor turns on, uses these features for diabetes, but it uses these features for schizophrenia. And how much overlap? Just the stupidest metric is like how much overlap or variance accounted for overlap is there between pairs of disease conditions. And it's very modest. It's actually the opposite of what naive biologists would say when they talk about pleiotropy. They're just disjoint. They're just disjoint regions of your genome that are governing certain things. And so why not? You have three billion base pairs. There's a lot you can do in there. There's a lot of information in there. So you can have, if you need 1,000 to control diabetes risk, I can have, I think I estimated you can easily have a thousand roughly independent traits that are just disjoint in their genetic dependency.”
2022-08-23 · Dwarkesh Podcast · Steve Hsu - Intelligence, Embryo Selection, & The Future of Humanity · IDENTIFIED FROM THE TRANSCRIPT · source
“Well, what you said is related to what Fisher proved in his theorems, which is that if suddenly it becomes advantageous to have x, like white fur instead of black fur or something, it would be best if there were little levers that you could move somebody from black fur to white fur continuously by just modifying those switches in an additive way. It just turns out with for sexually reproducing species where the DNA gets scrambled up in every generation, it's better to have switches of that kind. And so the other point related to your software analogy is that there are seem to be modular, fairly modular things going on in the genome. So when we looked at, we were the first group to, I think we had like initially like say 20 major disease conditions, we had decent predictors for.”
2022-08-23 · Dwarkesh Podcast · Steve Hsu - Intelligence, Embryo Selection, & The Future of Humanity · IDENTIFIED FROM THE TRANSCRIPT · source
“Me just point out that most of the diseases that we care about now, most of them, not the rare ones, but the common ones, they manifest when you're like 50, 60, seventy years old. And there was never any evolutionary big advantage, I think, of being super long-lived, right? So there's even a debate about whether like, okay, if the grandparents are around to help raise the kids, that raises the fitness a little bit of the family unit. Most of the time in the past, and most of our evolutionary past, humans just died fairly early. And so a lot of these diseases would never have been optimized against evolutionarily. But we see them now because we live under such good conditions. People regularly approach 80 or 90 years.”
2022-08-23 · Dwarkesh Podcast · Steve Hsu - Intelligence, Embryo Selection, & The Future of Humanity · IDENTIFIED FROM THE TRANSCRIPT · source
“Evolution has not had that much time to optimize humans, and what do you mean by optimization because the environment that humans live in is changed radically in the last ten thousand years? Like for a while we didn't have agriculture, now we have agriculture, now we have swipe left if you want to have sex tonight. The environment didn't stay fixed. And so when you say like fully optimized for the environment, what do you mean? The ability to diagnose matrices might not have been very adaptive 10,000 years ago. It might not even be adaptive now But anyway, so it's a complicated question. One can't reason that naively about, oh, well, if God wanted us to be 10 feet tall, we'd be 10 feet tall. Or if it's better to be smart, my brain would be like this, this big or something. So you can't reason that naively about stuff like that.”
2022-08-23 · Dwarkesh Podcast · Steve Hsu - Intelligence, Embryo Selection, & The Future of Humanity · IDENTIFIED FROM THE TRANSCRIPT · source
“The small ways in which we differ are just these little additive switches, mostly little additive switches. And so that's the deep scientific discovery over from the last, say, five, ten years of work in this area. Now you were asking about why evolution hasn't completely, quote, optimized all traits in humans already. Now, I don't know if you ever do deep learning or very high dimensional optimization, but you realize like in that high dimensional space, you're often moving on a surface which is slightly tilted, so you're getting gains, but it's also kind of flat. So even though you like scale up your compute or data size by an order of magnitude, you don't move that much farther. You get some gains, but you're never really at the global max of anything in these high-dimensional spaces. I don't know if that makes sense to you, but it's quite plausible to me that two things are important here.”
2022-08-23 · Dwarkesh Podcast · Steve Hsu - Intelligence, Embryo Selection, & The Future of Humanity · IDENTIFIED FROM THE TRANSCRIPT · source
“From each other. So it's very plausible to me that if I wanted to edit an embryo, a human embryo and make it into a frog, then there's all kinds of nonlinear subtle things I have to do. But all those very nonlinear complicated subsystems are fixed in humans. You have the same system as I do. You have the human not frog or ape not frog version of that region of DNA, and so do I.”
2022-08-23 · Dwarkesh Podcast · Steve Hsu - Intelligence, Embryo Selection, & The Future of Humanity · IDENTIFIED FROM THE TRANSCRIPT · source
“Like, say it's the bigger brats that out compete the smaller rats. At what rate does the rat population then start getting bigger? He showed that it's dominated by the additive variance, that that dominates the rate of evolution. And it's easy to understand why if it's a nonlinear mechanism that you need to make the rat bigger, when you sexually reproduce and that gets chopped apart, you might break the mechanism. Whereas if each little allele has its own independent effect, you can just inherit them without worrying about breaking the mechanisms. So it was well known, at least among a tiny population of theoretical population biologists that added variance was the dominant way that populations would respond to selection. So that was already known. And the other thing is that humans have been through a pretty tight bottleneck, and we're not that different.”
2022-08-23 · Dwarkesh Podcast · Steve Hsu - Intelligence, Embryo Selection, & The Future of Humanity · IDENTIFIED FROM THE TRANSCRIPT · source
“Yeah, so okay, so the first issue, which is why is this genetic architecture so simple? Surprisingly simple. And again, 10 years ago, we didn't know it was going to be simple. So when we were checking to see whether this was a field that I should go into because either we are capable or not capable of making progress, we had to study the more general problem of the nonlinear possibilities as well. But eventually we realized that probably most of the variance was going to be captured in an additive way. you know, we could narrow down the problem quite a bit. There are evolutionary reasons for this. There's a famous theorem by Fisher, who's the father of population genetics and also of really what you call frequentist statistics. And so Fisher proved something called the Fisher's fundamental theorem of natural selection, which says that if you impose some selection pressure on a population, the rate at which that population responds to this selection pressure.”
2022-08-23 · Dwarkesh Podcast · Steve Hsu - Intelligence, Embryo Selection, & The Future of Humanity · IDENTIFIED FROM THE TRANSCRIPT · source
“Actually, it works really well. And that is related to a deep statement about the additive structure of the genetic architecture of individual differences. So in other words, it's kind of weird that the ways that I differ from you are merely just because I have more of something and you have less of something. And it's not like, oh, these things are interacting in some super incredibly ununderstandable way. And so that's a very deep thing, which again is not appreciated that much by biologists yet, but over time I think they're going to figure out that there's something interesting here.”
2022-08-23 · Dwarkesh Podcast · Steve Hsu - Intelligence, Embryo Selection, & The Future of Humanity · IDENTIFIED FROM THE TRANSCRIPT · source
“And that is a key step. Otherwise, you might overcount. You might have ten different variants close by that have roughly the same statistical significance if you just do some simple regression math. But then you don't know which one of those tend to use, and you might be overcounting effects or undercounting effects. So what you end up doing is this super high dimensional optimization where you only activate, you grudgingly activate a SNP when the signal is strong enough. And once you activate that one, the algorithm has to be smart enough to penalize the other ones nearby and not activate them because you're overcounting effects if you do that. So there's a little bit of subtlety into it in it, but the main point which you made, which is that the ultimate predictors, which are very simple and additive, just sums over effect sizes times states.”
2022-08-23 · Dwarkesh Podcast · Steve Hsu - Intelligence, Embryo Selection, & The Future of Humanity · IDENTIFIED FROM THE TRANSCRIPT · source
“In any small region of your genome, the state of the individual variance is highly correlated because you inherit them in chunks. And so you need to figure out which one of those you want to use. You don't want to activate all of them because you might be overcounting. So that's where this L1 penalization sparse methods, they force the predictor to be...”
2022-08-23 · Dwarkesh Podcast · Steve Hsu - Intelligence, Embryo Selection, & The Future of Humanity · IDENTIFIED FROM THE TRANSCRIPT · source
“Yeah, so you're absolutely right that the ultimate models that are used when you've done all the training and the dust settles, the models are very simple. They have an additive structure. So it's basically like I either assign a non-zero weight to this particular region in the genome or I don't. And then I need to know what is the weighting. But then the function is a linear function of it's an additive function of the state of your genome at some subset of positions. So the ultimate model that you get is very simple. Now, if you go back 10 years when we were doing this, there were lots of claims that it was going to be super nonlinear, that it wasn't going to be additive the way I just described it. There were going to be lots of interaction terms between regions. Some biologists are still convinced that's true, even though we already know, like we have predictors that don't have interactions. The other question, which is more technical is that”
2022-08-23 · Dwarkesh Podcast · Steve Hsu - Intelligence, Embryo Selection, & The Future of Humanity · IDENTIFIED FROM THE TRANSCRIPT · source
“But if you look carefully when the big future AI in our future, in the singularity, looks back and says, hey, who gets the most credit for this genomics revolution that happened in the early 21st century, they're going to find that AI is going to find these papers on the archive in which we proved this was possible. And then five years later, we did it, et cetera, et cetera. Right now it's underappreciated, but the future AI that Roko's basilisk AI when he looks back is going to give me a little bit of credit for.”
2022-08-23 · Dwarkesh Podcast · Steve Hsu - Intelligence, Embryo Selection, & The Future of Humanity · IDENTIFIED FROM THE TRANSCRIPT · source
“As a phenotype. And we proved that in a paper using all this fancy math in 2012, I want to say the paper came out around 2012. And then around 2017, when we got a hold of half a million genomes, we were able to implement it in practical terms and show that our mathematical result from some years ago was correct. And the transition from low performance of the predictor to high performance, there's a kind of what we call a phase transition boundary between those two domains, occurred just where we said it was going to occur. So some of these technical details are really just not understood, even by practitioners in computational genomics who are not quite that mathematical. They don't understand actually these results in our earlier papers. They don't really know why we can do stuff that other people can't do or why we can predict how much data we're going to need to do stuff. It's not well appreciated even in the field.”
2022-08-23 · Dwarkesh Podcast · Steve Hsu - Intelligence, Embryo Selection, & The Future of Humanity · IDENTIFIED FROM THE TRANSCRIPT · source
“And so I spent actually a fair amount of time, like probably a year or two studying these results, very famous results. Some of them were proved by Guy Galtiran Tau, who's a fields medalist. And these are results on something called compressed sensing, which is a penalized form of high-dimensional regression, which tries to build sparse predictors. Machine learning people might know it as L1 penalized optimization. And anyway, so the point is the very first paper we wrote on this was to prove that using real genomic genomic data, that these theorems that were very abstract could be applied in order to predict how much data you would need to, quote, solve individual human traits. So we showed that you would need at least around a few hundred thousand individuals and their heights, their genomes and their heights to solve height.”
2022-08-23 · Dwarkesh Podcast · Steve Hsu - Intelligence, Embryo Selection, & The Future of Humanity · IDENTIFIED FROM THE TRANSCRIPT · source
“When there are millions of genomes with good phenotype data available for analysis, so that a typical run for us, a training run might involve almost a million genomes or half a million genomes or something. So the mathematical question is, What is the most effective algorithm given a set of genomes and phenotype information to build the best predictor? So you can be boiled down to a very well-defined machine learning problem. And it turns out for some subset of algorithms, there are theorems. They're actually performance guarantees that tell you, they give you a bound on how much data you need to capture almost all of the variation in the features.”
2022-08-23 · Dwarkesh Podcast · Steve Hsu - Intelligence, Embryo Selection, & The Future of Humanity · IDENTIFIED FROM THE TRANSCRIPT · source
“Just yeah, a little bit. I mean, yeah, you look better than me, but The question is the following. So, a lot of times what theoretical physicists do is they have a little spare energy. They have some spare cycles and they get tired of thinking about quarks or something. And they want to maybe dabble in biology or they want to dabble in computer science or some other field. And the thing that we always have to do as theoretical physicists, we always feel like, oh, I have a lot of horsepower. I can figure a lot of stuff out. Like, for example, Feynman helped design the first parallel processors at thinking machines. I got to figure out which problems I can actually make an impact on because I can waste a lot of time. Some people spend their whole lives studying one problem, like one molecule or something, or one, you know, biological system. And I don't have time for that. I'm just going to jump in and jump out. I'm a physicist, right? That's a typical attitude among theoretical physicists. The thing that I had to confront about 10 years ago was I knew the rate at which sequencing costs were going down. So I could anticipate we would get to the day.”
2022-08-23 · Dwarkesh Podcast · Steve Hsu - Intelligence, Embryo Selection, & The Future of Humanity · IDENTIFIED FROM THE TRANSCRIPT · source
“That's a great question. So, you know. First of all, again, like I was kind of stressing the fundamental scientific importance of all this stuff. If you go into a slightly higher level of detail, which you were getting at with the individual SNPs or polymorphisms, those are individual locations in the genome where I might differ from you and you might differ from another person. And typically if you just take pairs of individuals, each humans will differ at a few million places in the genome. And that's what's controlling. That's why I look a little different than you.”
2022-08-23 · Dwarkesh Podcast · Steve Hsu - Intelligence, Embryo Selection, & The Future of Humanity · IDENTIFIED FROM THE TRANSCRIPT · source
“And so the most direct application of this science that I described is well, we can now genotype those embryos from a small biopsy. And I can tell you things about the embryos. I could tell you, hey, number four is an outlier for breast cancer risk, I would think carefully about using number four. Number 10 is an outlier for cardiovascular disease risk you might want to think about not using that one. The other ones are okay. And so that is what genomic prediction does. And I think we work with Or 300 different IBF clinics on six continents now.”
2022-08-23 · Dwarkesh Podcast · Steve Hsu - Intelligence, Embryo Selection, & The Future of Humanity · IDENTIFIED FROM THE TRANSCRIPT · source
“And in the process of IVF, because they use hormone stimulation, they generally produce more eggs instead of one per cycle. They might produce, depending on the age of the woman, anywhere between five or ten or twenty or even, I recently learned for young women who are hormonally stimulated, if they're egg donors, they could produce sixty or a hundred eggs in one retrieval cycle. And then it's trivial, as you know, men produce sperm all the time, we're just producing it. You can fertilize those eggs pretty easily in a little dish, and you get a bunch of embryos, which they grow. They just start growing once they're fertilized. Now, the problem is if you're a family and you produce more embryos than you're going to use, you have what we call the embryo choice problem. You have to figure out like, okay, I have these 20 viable embryos. Which one am I going to use?”
2022-08-23 · Dwarkesh Podcast · Steve Hsu - Intelligence, Embryo Selection, & The Future of Humanity · IDENTIFIED FROM THE TRANSCRIPT · source
“But what use is it, right? What use is all this stuff? The most direct application of this is in the following setting. Every year around the world there are millions of families that go through IVF, typically because they're having some fertility issues and also mainly typically because the mother is older, like typically in her 30s or maybe 40s.”
2022-08-23 · Dwarkesh Podcast · Steve Hsu - Intelligence, Embryo Selection, & The Future of Humanity · IDENTIFIED FROM THE TRANSCRIPT · source
“Specific DNA pattern to the phenotype, the actual express characteristics of the organism. And if you think about it, this is what biology is. Like once we had the molecular revolution and people figured out that DNA is the thing which stores the information, which is passed along, and evolution selects on the variation in the DNA as it's expressed as phenotype and as that phenotype affects fitness or reproductive success. That's the whole ballgame for biology. And I'm lucky that as a physicist who's trained in kind of mathematics and computation, I arrived on the scene at a time when we're going to solve this basic fundamental problem of biology through brute force AI and machine learning. So that's how I kind of got into this, right? Now you ask, is an entrepreneur, like, okay, fine. Steve, you're doing this in your office with your postdocs and collaborators on your computers.”
2022-08-23 · Dwarkesh Podcast · Steve Hsu - Intelligence, Embryo Selection, & The Future of Humanity · IDENTIFIED FROM THE TRANSCRIPT · source
“Yeah, so if you don't mind what I should say, there are two ways to introduce it. One is the scientific view and then the other is the IVF view. And I can kind of do a little of both. So scientifically the issue is we have more and more genomic data. If you give me the genomes of a bunch of people and then you give me some information about each person, like do they or do they not have diabetes or how tall are they? Or what's their IQ score or something? All of your listeners will be familiar with AI and machine learning. It's a natural AI machine learning problem to figure out which features in the DNA variation between people are predictive of whatever variable you're trying to predict, whatever the biological term is phenotype. So this is an ancient scientific question of how do you relate the genotype of the organism?”
2022-08-23 · Dwarkesh Podcast · Steve Hsu - Intelligence, Embryo Selection, & The Future of Humanity · IDENTIFIED FROM THE TRANSCRIPT · source
“Kind of lost for a long time And it was just coming back in the US when I was growing up. And so yeah, if you were 200 pounds of fairly lean 200 pounds and you could bench over 300, that was pretty rare back in those days.”
2022-08-23 · Dwarkesh Podcast · Steve Hsu - Intelligence, Embryo Selection, & The Future of Humanity · IDENTIFIED FROM THE TRANSCRIPT · source
“So the Greeks were very special because they were the first to really think about the word gymnasium and there's a thing called the palestra where they were trained like wrestling and boxing and stuff like this. They were the first people who were really seriously into physical culture and training, specific training for athletic competition. But if you look at like even in the 70s, so when I was a little kid and I remember in the 70s and now when I look back at old photos from the 70s, it's very apparent, guys are skinny. Guys are so skinny. You know, the guys who went off and fought World War II, whether they were on the German side or the American side, they were like 5, 8, 5, 9, and they weighed like 130 pounds, 140 pounds. They were totally different than what modern U.S. Marines you would think of look like, right? So yeah, physical culture was a new thing. Of course, the Romans and the Greeks had it to some degree, but it was.”
2022-08-23 · Dwarkesh Podcast · Steve Hsu - Intelligence, Embryo Selection, & The Future of Humanity · IDENTIFIED FROM THE TRANSCRIPT · source
“So, you know, and at that time, like, I could probably squat, I could maybe squat 400 pounds at that time. So I was pretty strong, right? And I had big legs. And so anyway, the fact that the trainer didn't really understand like what well-developed anatomy was supposed to look like was just blew my mind. I was like, no, that's my quadricep. We build that up. And she's like, oh, I thought that was an injury. I was like, what? What are you talking about? So anyway, we've come a long way. This is one of these things where you got to be old to have any kind of understanding of how this stuff evolved over the last, you know, 30, 40 years.”
2022-08-23 · Dwarkesh Podcast · Steve Hsu - Intelligence, Embryo Selection, & The Future of Humanity · IDENTIFIED FROM THE TRANSCRIPT · source