Deus Ex Machina: A Man, Machines, and God
Steve Hsu: Welcome to Manifold. Today's episode is an interview that I recorded with the podcast, Deus Ex Machina. The host of that podcast, Bohan Lou, grew up in China, attended Yale University, where he studied computer science and religion. And in our conversation, we get into topics ranging from geopolitics, US China competition, genomics, and we even get into questions like free will and my views on religion.
It's a unique conversation and I think you'll enjoy it. Thanks for coming to the show.
Bohan Lou: Super excited to have Professor Steven Hsu on today. Professor Steve is a professor of theoretical physics and I believe computational, mathematical science and engineering at Michigan State University. And Steve is also a co-founder of multiple companies, including Genomic Predictions, which uses preimplantation genetic screening devices for human embryos and Superfocus AI, which makes, you know, narrowly focused non hallucinatory AI that's actually useful for enterprises.
But you know, I think that's the official bio. I'm happy to touch on multiple subjects, I hope. but as I was going through Steve's extensive blog, and I think you've been blogging for over 20 years now, maybe one of the earlier ones on the internet, I basically went through, I think, most of your blogs to do some research on you.
And it is just all kinds of topics, you know, from China to, genius to, you know, obviously your fields of genetics. So it's quite fascinating. so I'm very excited for the conversation today, I thank you so much for coming on Steve.
Bohan Lou: Awesome.
So, starting off, I wanted to talk about something that kind of struck me, maybe visually, looking at your background and you know that, you know, you're an Asian American who You know, got a PhD in physics and is now a professor.
Okay, that's all. But then you actually, you're a pretty big guy. You played football, college football, at Caltech. Way back when they had a football team. Well, it was a D3 team, but still college football. And then you also, you know, did things like, jiu jitsu and martial arts. And, you know, they actually, you know, The only other football playing mathematician I know is John Urschel, who I believe is at MIT, which I'm sure you've heard
Steve Hsu: He's a little better player than I am.
Bohan Lou: right?
you did talk about at some point maybe realizing that the NFL maybe wasn't for you, though, you know, you showed some promise. But I was kind of curious about how you decided to get into sports and, you know, growing up as an Asian American, especially at a time when, you know, maybe the stereotype was even stronger when you're growing up about associating Asians with math and just STEM and not sports and athleticism.
Steve Hsu: Yeah, interesting question. I mean, there's several strands to this discussion. Part of it comes from my family background. It turns out my mother's side of the family is a military family. And so my grandfather, my mother's father attended the same military academy as Chiang Kai shek and was a general in the KMT.
And when they moved to Taiwan, he was actually briefly the, in charge of the Navy in Taiwan. so I guess at one point he was kind of an admiral. So anyway, that's, that's a very military family and that side of the family is very athletic. So my. My mother's little sister represented Taiwan in the Asian Games in swimming, and one of my two cousins actually is among the only Asian Americans to actually play D1 college basketball.
One of my cousins, Richard Chang, is about 6 '7 And he was a power forward at Cal and he was recruited pretty widely. Like he could have gone to UCLA or, you know, other schools as well. So a pretty athletic family on that side. And then. I guess I learned from my father's side, the Hsu side, which, you know, recently comes from Zhejiang mostly.
Actually, this is a weird thing. Like everybody knows about the Imperial examination system, but there was some kind of combined hybrid system for some period of time where to be selected as a candidate in that category, they combined physical activities as well as your ability to do well on the written tests.
And so there were even things like weightlifting, and I don't know, horseback riding, just all kinds of crazy stuff. So, for some reason, like, we have this weird, multi thousand year family history. I think I blogged about it at one point, you can look it up, but. My relatives in Zhejiang gave it to us, and it has the names of all these, the male shoes going back thousands of years, actually.
And some of the more recent guys from Zhejiang actually were selected, candidates based on this weird hybrid exam. So, I guess they were probably pretty. athletically talented on that side of the family too, back then. So anyway, it's a weird story. So that's one thread. so I think my mom basically, my brother and I, she encouraged us to do sports when we were pretty young.
I think basically because in her family, the military family, that was a standard thing.
Bohan Lou: Yeah.
Steve Hsu: my brother and I were competitive swimmers for a long time. And, so I don't know. I just gravitated toward that kind of stuff. I actually like contact, like I like hitting people in football and I like wrestling and jiu jitsu and judo.
I just like that kind of stuff. not everybody does. Like I had friends who are basketball players or track people or even other swimmers who, if you just grab them, they really didn't like it. And
Bohan Lou: Yeah.
Steve Hsu: getting hit in the face, let alone getting hit in the face. So, yeah, I think it's kind of just hardwired.
Bohan Lou: yeah, I, first of all, you know, extremely illustrious family history. So my great grandfather was also going to be down, but he was just a mere, I think, Deputy secretary to a county. No, not a general. That's also why our family history probably only goes back 150 years, I think to the late Qing dynasty, whereas he goes back thousands.
Cause you know, your family is more worthy of record keeping. Um, but. You know, I think you mentioned that, yeah, at some point the Chinese, maybe in pure exams, added a physical component and then, they probably got away, got rid of that at some point. Because growing up in China, I don't think it was that big of a deal.
There was some element of it. I remember I was forced to learn how to swim in Shanghai because I was in the second grade. The government at some point decided that there could be major flooding and we need to make sure the populace knows how to swim for safety. So everyone had to learn how to swim. And I did that, but I think actually recently they've started bringing some of it back in the effort to kind of rejuvenate the nation and You know, masculinized nation.
And so now, now it's actually a heavier, portion of the, I think both middle to high school and high school to college exam, which of course in Asian society means there's, there's a whole industry of test prep geared towards physical test prep now to just optimize it, to get those scores, you know, as, as our society does.
But I just, I think like, You know, you are an interesting embodiment, I think, of these ancient virtues that China has valued historically, which is the combination of the intellectual and the martial, which I think has kind of fallen to the wayside a bit more recently. And I know that ancient Greeks were also people that cared about that, too.
And I feel like, you know, in our society, we kind of see a bit more of this dichotomy, right? Imagine the scrawny nerd, and they're not supposed to be jocks, and the jocks are separated, they're supposed to be dumb. You know, yeah, why do you think we, you know, kind of like maybe lost some of that emphasis on this well rounded, both physical mens et manus kind of, kind of, integration of a full man and, you know, whether you think there is any, maybe because due to increased specialization, whether the world inherently has to split a bit more to focus on their body or the mind,
Steve Hsu: Yeah. It's a super interesting question.
And also to get back to what you asked me earlier about the stereotype of Asians being. not athletic. for the historians out there, I'll point you to an interesting fact. If you go back to the L. A. Olympics, which must have been like late twenties or early thirties or something.
So the earlier there was an L. A. Olympics also like in, I think, 84 or something, but, the previous one, which was early in the 20th century. If you go and do a little research, you'll find that the Japanese swimming team was the dominant, by far the most dominant, swim team in the world at that time. And they had a tradition where, I guess because the samurai were required to learn how to swim, they were required to be strong swimmers. And so they just had a tradition of it and also physical culture and stuff like that in judo and stuff like this. And so people just don't realize that these things are time and culture dependent.
And, you know, I think the modern kind of HBD people who want to think of Asians as non athletic would be shocked at, what people wrote in the newspaper about the Japanese dominating swimming, back then. And, so anyway, things are more complicated than I think you're, the simple minded modern approach.
The way things are today isn't necessarily the way they, they quote, have to be. now, in terms of like, I think there definitely is an idea in Chinese culture that, It does not particularly value the physical, so that's, I think, a Confucian idea that not only, you know, is, is, is the Confucian scholar, the ideal person supposed to be very cerebral and have very high self control and be a role model for other people doing mundane things like working in, you know, with soil, like, you know, with that, you know, like, you know, You know, the Chinese word Nongming.
So, all that is kind of denigrated, even being a merchant or a soldier is denigrated. And so I think that has a long term effect on what families tend to pour their energy into. And one thing I'll point out is that growing up in Iowa, which is really one of the most kind of all American places you can grow up, you know, all of my friend's dads, or at least predominantly, some of them were nerdly professors, but a lot of dads would take their kid out, their son out and throw the football with them or throw the baseball with them or play basketball with them.
And that was much less common among Asian American kids. And of course that sets you up, like if you're trying to make the like seventh grade basketball team, of course, if you spent a lot of hours with your dad showing you how to do it, you're going to have a huge advantage, whereas the Asian kids generally wouldn't have had that extra, you know, time.
So it's, it's all much more complicated, I think, than people think.
Bohan Lou: Yeah. And did your dad ever, you know, throw the, throw the ball with
you, you know, in a.
Steve Hsu: No, that's the funny part of it. My dad was actually literally a kind of Confucian scholar. So he didn't really do that with the kids, but luckily for us, my mom came from a very likable, athletic, active family. And so we were competitive swimmers from, I was a competitive swimmer since the age of like seven.
And,you know, I did other sports as well and took judo when I
I was growing up.
Bohan Lou: Yeah. And so, just cause you mentioned your grandfather, I believe was a general and, you know, you said went to the same military academy as Chiang Kai shek. is that the one in Japan where he was trained or are you talking about Whampoa Academy?
Steve Hsu: Exactly. No, in Japan. So both he and Chiang Kai shek studied, military, and were trained in military education in Japan. I think in Tokyo, actually.
Bohan Lou: Right. Right. And I'm wondering, you know, growing up with that kind of family background, did you guys talk a lot about Chinese politics growing up and did that influence your family's views on the politics in mainland China and cross race relations?
Steve Hsu: You know, we actually, it's, again, it's this funny inversion that, although my dad was an officer, also in the KMT military, he went to college. During the war, he went to,he was in, in college in Kunming, this special university that was made out of four different Yeah, exactly. , they, so he spent the war there.
And then afterwards he was, actually, an instructor briefly at the Air Force Academy training pilots, because his field was aerodynamics. So anyway, he was also, he also had a military background, but the funny thing, or at least in that era, the World War II era, a lot of men had military backgrounds, but, the funny part of it is though, Although my mom's family was involved, obviously, at a high level with the KMT.
I didn't ever hear anything political, ever, from my mom. My mom just wasn't that kind of person. But my dad's family was getting basically screwed over in the Cultural Revolution because they were still in Zhejiang when I was a kid. So this would have been like late 60s, early 70s. And I heard my dad would get these letters from his relatives in China and they were written on the thinnest, cheapest, shittiest paper. That's all probably you could get in communist China, you know, at that time. And they would write very densely on these little letters. But those letters were priceless to my dad because they were his only contact with his family back home. And he would spend time telling me about how terrible the cultural revolution was and what his family was going through and how it made people into animals, like the brothers did terrible things to each other and to the father, to the parents, taking stuff from their house and just, just really terrible stuff.
And it's very funny because if you're an intellectual kid and you grew up in that era, a lot of intellectuals were pro communism, pro leftism. And there was a kind of romanticization of, you know, both the Soviet Union and Communist China by leftists here in the United States.
And so I would come home, like I, like some of my friends, parents who were leftists, like this is a university town, remember. I would be at their dinner, a dinner at their house and hearing all this stuff about how bad capitalism is and American imperialism. I come home and talk to my dad and my dad would like, be like, Those people don't know fuck all about what they're talking about.
Your relatives are being fucked over, right? Well, he, my dad would never use language like that, but, but your own relatives are suffering in China right now. And these people know nothing. So no, nothing about it. And so in the modern era, I am very, very able to discount. quote, expert opinion, because, you know, these professors that quote experts can be 100 percent wrong in, you know, very strongly held beliefs.
And it's only in a few technical fields where the way I like to say it is that if you say something that's wrong, the facts or the mathematics are going to punch you in the face right away. And it's only in those sub disciplines where if someone's, quote, an expert, that means something. In all of these other fields, sorry, people trained, all my friends with PhDs in history and, you know, other subjects, like, literally, you'll find people who are supposed to be the world's experts on a particular topic, and they're literally 180 degrees off.
Bohan Lou: Yeah, you know, the period you talked about, yeah, in the 60s, 70s, when all these leftist intellectuals were, you know, enamored with these, you know, platonic ideals of communism and historical materialism, how irrationally it just seemed makes so much sense that they had to support the Soviet Union.
And even on the right, the fascists had a lot of those intellectuals too. And, one of the, one of the guests I'm going to have on, Professor Mark Lilla at Columbia, he's writing, Well, he did write a book about, I think it's called Reckless Mind, basically this phenomenon of intellectuals who really support tyranny often because they, they have this delusional view that their tyrant is going to implement their views and it's got to work out and things like that.
Steve Hsu: You know, I, just another comment on this topic, on Twitter, I'm often pointing out, like, I can't help but notice how dumb, like, our intelligentsia is, well, pretty much everywhere in the world. Like, I mean, I'm arrogant enough to say, like, well, if you're not, like, a theoretical physicist or something, there's a good chance that you're sure you're right, but you're actually wrong.
And even physicists could be wrong about things, too. Like, one of the big things is people in the West are just totally wrong. They're totally miscalibrated about China in many respects. I'm often pointing out, like, simple, obvious things that are, you know, you can go and check them very quickly that they're true, but they go 180 degrees against what people say about China.
And so now I have people saying, like, Oh, Steve, you're a CPC shill.
Bohan Lou: You're, you're, you're a tanky, you're Wu Mao.
Yeah.
Steve Hsu: Yeah.
I'm, I'm look, remember my dad taught me, you know, when, before any of you guys were born, my dad was teaching me about how shitty the communists were and what Mao did to our family, you know, remember that. Okay. So if I tell you like, well, they do actually seem to have pretty good infrastructure and yeah, they actually did put a satellite up that can do X, Y, Z, you know, it's not because I'm pro communist is because like, I actually want to understand the world as it is not how your ideology wants it to be.
Bohan Lou: Yeah. no, I actually feel very similar to some, I'm a Chinese person living in the West, especially I came to the U S for college in 2016 and basically two months after I came down, Trump was elected. Right. And then China was just brought to the national forefront as this big demon, this geopolitical peer competitor and you know, this bipartisan threat.
And I often come off as, you know, a China apologist, even though, you know, you know, there's definitely parts of the government policy. I don't necessarily agree, but just because there's things that are just. You know, patently untrue that is said. And I, so I just correct the fact. And then I come off as, they think, Oh, classic, you know, butthurt Chinese person.
But it was interesting when you talked about these, these leftists, you know, what they were, you know, how, what they were glorifying, you know, communist China during the cultural revolution. and how you said, you know, you generally discount expert predictions unless, you know, it's like a really objective, you know, factually based truth.
In this case, it seemed like there was an objective fact, right? Like if they actually just ever showed up there, they would see, they would see, see the truth on the ground. Obviously, maybe if they're on an official tour, they'd be taken to a Potemkin village, but like the fact is they, they never went there on the ground. And I think that was different.
Steve Hsu: In those days, you could make excuses for the lack of information that a Harvard professor would have about what was going on in China, because they might not be able to go there, or if they went there, as you said, they might be tightly controlled what they could see. But, so you can apologize for their lack of knowledge.
You cannot apologize for the level of conviction that they have. conditional on their level of knowledge. So they, as always, like all these people in the humanities and, and people who even like people online who either are far left or far right or whatever, if they're just using what I call the calculus of words, word calculus, no data, no equations, no analysis, you'll find people that are way just incredibly overconfident based really mainly on their feels.
That's all the word calculus is. And that's all these people have. And so they're just constantly miscalibrated. So, now you don't have such a good excuse. You, you can actually, this, and people hate. me on Twitter for saying shit like this. And I'm always doing this. I'm like, look at this two minute YouTube video.
Do you think this was faked or do you think these electric cars are actually good? Like just a few years ago, I was putting up videos of guys in London or guys in Sydney, Australia. Or a guy in Israel who bought a Chinese car because he could, they could buy Chinese cars already at that time. And they're evaluating these cars.
It's like, this thing is awesome. It's as good as my Tesla. And it costs, you know, half as much. So I'm just asking you like, do you think all these people are fake? Do you think like, you know, what, what, cause people would come at me and say, Oh, Chinese technology, it's junk. These Chinese EVs are going to fall apart.
It's all low quality. Only Tesla can make this shit. And, and so I would, I would get into like, you know, Discussions like that. I'm like, well, at least do a little bit of work. And like, I understand if you think like all these white guys who know cars and are telling you these cars are pretty good and dudes in Tel Aviv, you know, sure, maybe it's all fake.
Maybe the communist party is paying off all these dudes in Tel Aviv, you know, but maybe actually I'm giving you a little glimpse of the real world and you could get it yourself if you just had an open mind.
Bohan Lou: Yeah. The very first internship I ever did was working at NIO in Shanghai. That was pre IPO, right? You know, that's China's premium electric car manufacturer. And that was, that was working on Thomas driving the time, which didn't really pan out for them, nor did it for any of the automakers.
But even that time you could see, like, it was real. Like I drove them. They were extremely good. And obviously in the U S you don't see them for geopolitical reasons, but recently I was in Europe and in the Netherlands. Like, all the Chinese DVs are, like, all over the streets, and, you know, the people there love them.
Unless, unless, of course, the tariffs kick in, and then, you know, really force it out of the market.
Steve Hsu: But even now you can find guys online. Who will trash talk about these cars and I'm like, what is your basis for this? Like by now you have crazy things like GM pays some famous. auto consultancy to tear down one of these cars and do like a full report on the quality and you can go and look at it and also like the the consortium of japanese automakers have done this too and by by now it's like totally beyond question that these are actually good cars and stuff and yet you can still find people who will tell you like oh it's all shit it's all fake it's you know so
Bohan Lou: So I'm just thinking from the point of, you know, strategic competition between the US and China. I think in the art of war, it talks about how you want the enemy to underestimate you, right? So in the sense like, you know, the Americans, maybe HR McMaster is called strategic narcissism where they just refuse to see the truth, you know, it's, it's good for China, right?
Plays into China's hands. Right. And obviously I know you're, you're an American, so you would want Americans to, you know, overcome these biases and see the competitor for what they are. But I'm kind of thinking about how. I guess from a Chinese perspective, I think, you know, during the era of Deng Xiaoping, it was very much like, bide your time, hide your strength, you know, you want, you want the other side to underestimate you.
But in the last few years, I think China's been coming out a bit more guns blazing, in a sense, maybe even over, overplaying the cards a little bit sometimes, you know, and it's causing the, the, the West and the world to really respond in such a forceful manner in a way where, you know, I think it's interesting.
There's a, there's simultaneously some level of overestimation and underestimation going on. It seems like at the same time about China's.
Steve Hsu: Well, it's very easy in about obviously you're evaluating. I mean, forget about evaluating LLM on 10 different metrics to figure out which one is better. I mean, imagine evaluating another culture, another country across like, oh, the mil tech, the semiconductors, the social cohesion, the level of their education.
There's so many ways in which you could compare the strength of two different competing countries. So, of course, like you're gonna have things that are all over the map where, Sometimes you say they're 10 feet tall and actually this is not one of their stronger attributes and other times you're like, you don't think they can do X. In fact, they could do X, you know, 10 years ago and now they can do X better than you.
So, it's all over the map. And I think one of the most important things is for us analysts, whoever, whatever passes for the thought leadership in our military industrial complex and national security complex, they have to be well calibrated.
If you're not well calibrated, You could go to war overconfident thinking that's going to be easy and you get killed or you could be the other way where like you think the enemy is 10 feet tall and you, you know, you fold before you should fold. Right? So, all of this is super important. And I honestly just don't think the intellectual, I mean, of course you need several things.
The ideal analyst would be somebody who you. speaks fluent Chinese, has been there, has studied military history, technology, hard science, all of those things. And, and believe me, the intersection of those requirements that I just listed is, is, is very, very small. So basically, like, I, I kind of feel like Washington is flying blind at this.
Bohan Lou: yeah. Especially the going there and knowing the people part, right. If you look at the current generation of China analysts really, especially during the Trump administration. They had people like Michael Pillsbury who love to talk about how he read Chinese. And that was his competitive advantage where no one at the CIA could read Chinese besides him.
It was just like really fanciful interpretations of Chinese idioms.
Steve Hsu: One thing I just want to say, just to reinforce your point, is that when you think about the characters that you hear about, like Pillsbury, you hear about this guy, right? It means he is a, he is a policy entrepreneur, okay? He's not building a startup to sell a product, he's selling himself as a product.
Okay. So the way that guy wins is he gets invited to panels. He's published his stuff in foreign policy or other journals and he gets, you know, eventually he'll have a high position in national security or start his own consultancy, whatever it is. But he's an entrepreneur and he's selling, just like everybody else.
Okay. So just take that into account. And when you judge what these guys say. On the other hand, for example, I just recently was, I don't know if you've heard of the office of net assessment in the Pentagon. It's a very famous office, used to be run by Andy Marshall. And I was just talking to some analysts there and, you know, those guys are the official think tank within the Pentagon that advises the Secretary of Defense.
And they're very serious about trying to figure out what is going to happen, future trends in technology, drones, you know, autonomous weapons systems, AI, China, they're trying to figure all this stuff out. And you don't know who they are. They're, they're like, totally, you know, anonymous people, but they're the serious ones, right?
The guys you hear about are basically, yeah, they're basically grifters. Right. I mean, you know.
Bohan Lou: Yeah. I mean, I always lament, you know, the quality of maybe, and I know there are some, there are, there are definitely people in the ministry who would know their stuff, but like the average quality compared to maybe the area of the China hands, where, you know, during World War II and immediately after during the, you know, kind of the early cold war, that generation of China experts were basically all children of foreign missionaries who were born and raised in China.
These were like white people who spoke China as a native language, including the dialects. They knew Chinese people better than Chinese people even knew themselves. And that was like that level of understanding, who led the nation at that time. And that is obviously they were all persecuted later because of McCarthyism, but it's no longer.
Yeah, yeah.
Steve Hsu: A class, the people, the government, if I were advising president Trump and saying, who can give you good advice about what's happening in China and our competition with China, and even extending to the miltech side. I would talk to American entrepreneurs who had to get their device manufactured in China. Right. So, like I interviewed a guy recently who sells toys on Amazon, but all his stuff is made in China and he's been going there and visiting factories for over 10 years. He knows what's going on. He knows how the system works there. He knows how hard people work. He knows how competent the engineers are.
He knows he can't get that stuff made at that price anywhere else. And maybe even at twice the price, he can't get it made in the U. S., right? So, those are people who have the best insight. His Chinese is not perfect, but he certainly understands, you know, he's like 10x, 100x some dude in Washington who sits in a cubicle, right?
So, I would talk to those guys. The other point I want to make is that for students of history, if you look at what people wrote in the Cold War, what was happening during the Cold War? You had two very formidable superpowers, the Soviet Union and the U. S., Both of whom were pouring unprecedented amounts of money into developing satellites, thermonuclear weapons, and intercontinental ballistic missiles. These were all computers. These were all driven by military competition between the two superpowers and the people who were analyzing and planning this strategic competition were predominantly physicists. Okay? So, this was an era where, you know, you, you went and got your physics PhD and then you became a cold warrior.
Right? So that the level of analysis is so much higher than what we see today. Now you'll never find even the people who are, I find it so hilarious. I'll be reading some think tank reports. Like I'm one of the only people in the country who reads these things. Like I'm reading some think tank report about some weapon systems in China.
And I, there's like two authors and I go and look up the authors. And you're really lucky if one of those authors has a technical background, like an engineering degree.
Sometimes the guys who wrote the article are both, like, political science. One guy's, like, a PhD in political science. You don't know, you don't know anything.
Like, if you came and took my freshman physics class, you would probably fail. But here you are, like, telling me about some very intricate weapon system that, where our guys are gonna have to fight.
Bohan Lou: Yeah.
Steve Hsu: Right? And this is what we have. This is what passes for in the old days is not like this. You got your Ph.D. in physics at M.I.T. and then you went into the defense field and then you became an analyst, you know, in the office of net assessment or something. It's totally different situation than what we [are in].
Bohan Lou: Yeah. Like, I think your article is, you know, you were talking about many worlds hypothesis. Hugh, is Hugh Everett who came up with it. Right. And he, I think had to drop his PhD to go into the military industrial complex because of like issues
Steve Hsu: He left. He left right after he finished his PhD at Princeton. He, he basically, yeah, he basically worked at the office in that assessment. Actually, I think it was actually a slightly different office on analysis of systems. There was another office that was even more technical. And, but, but anyway, the, the, there's just night and day difference between the level of competency. Competency today. But I think China, I think in China, the people they have analyzing this shit are competent. That's the thing is I think they have a much more realistic sense of what we're capable of than vice versa.
Bohan Lou: Right. Well, I mean, I think China's definitely a society in general that values technical competence, I think more than the us. But I think, so growing up there, what I noticed is that basically developing countries, poor countries valve stem more than humanities because humanities in a sense, is a privilege is an aristocratic discipline, right?
To get to engage, that kind of thing. We always knew in college that the. The wealthiest on average, the wealthiest majors with our history in the classics majors, right? Because the immigrants and the poor kids study engineering to climb the social ladder. And I think growing up in China, it was always clear that, yeah, the smart kids with the highest gokko scores, they go into STEM.
The number one's still going to Tsinghua, but they go in as a philosophy and the literature
majors.
Steve Hsu: Let, me just give you two examples. You know, the current guy who's the national security advisor, Jake Sullivan. Is a law school grad, and I'm pretty sure his undergrad degree is non technical. So you have some guy who went to law, like, oh.
Bohan Lou: Well, I mean, that's the whole U.S. government, right? I mean, that's that in America, that is the prestigious. Your prayer out. Yeah.
Steve Hsu: What's your qualification for grand strategy and deciding like now is the moment to call an all out chip war. Oh, I guess it's your law degree. I mean, really? Come on. No wonder they made such a colossal mistake in the chip war.
Bohan Lou: So because Jake Sullivan is probably his background is like the golden poster boy of a U.S. policy person's background, right?
Steve Hsu: It's a joke, you know, you know, you the guy to come, you could compare him to, for example, Henry Kissinger, who, although he didn't have a technical background, was legitimately super smart. And if you read, you know, if you read his senior thesis at Harvard, you realize this guy and live through World War II.
So he understood a few things. but you know, you can find examples of secretaries of defense and such who are literally like PhDs in physics and stuff.
Bohan Lou: Like George Shultz, right? He was a professor at MIT, I think.
Steve Hsu: Yeah, he was an economics professor, but you have people like Harold Brown, who was a physics PhD and later president of Caltech, and he was a secretary of defense and compared to what we have now, it's just like, come on, everybody now, you basically can't find any technical people doing these jobs.
Bohan Lou: Well, so, you know, I wanted to get your thoughts on the value of the humanities, right? Like, you know, I actually am someone who does really value the humanities, and I didn't grow up with that exposure growing up in China. So, like, I always thought, you know, You know, arts were dumb, you know, it was all just about STEM.
And then I went to Yale, which is a very big liberal arts focused school. And I got, I fell really in love with religious studies and history and philosophy, and I do find a lot of value in them, but over time, I think I have been trying to wrestle with whether we are fundamentally in a world where the technical disciplines, I mean, they clearly do make a bigger impact on society and whether, you know, these things that we've studied for thousands of years, they may have still have some value, but they are actually diminishing in terms of, you know, how big, how, how much, yeah.
Steve Hsu: I have nothing against the humanities as a subject. I think. They're obviously, for example, just to take an example, history is a great thing to study if you want to think about great power competition or strategy, grand strategy. Of course you should study history. But is it enough to only study history that I think it's necessary, but not sufficient.
So anybody like, say, Harold Brown, you know, Secretary of Defense under, I think maybe Carter. And a former high energy physicist, you know, like, do you think he went into that job and didn't like to start reading a lot of books about Russia and the Soviet Union? And, and, you know, do you think he went in and just said, like, nah, I don't need to know anything about history.
I'm just going to do this job. No, of course he immersed himself and he probably reads twice as fast. As you know, your average history PhD. And, if you know anything about, you know, psychometrics, and, and so I'm sure he educated himself and said like, well, surely I need to understand a lot of this historical background before I can opine on anything related to competition with the Soviets.
But you can do that. You can earn your PhD in physics and then go back and say, Oh, I didn't have time to read all these books on, you know, Chinese history, but now I'm going to do it. whereas show me a Jake Sullivan who later on can have some intuition about exactly why hypersonic weapons are different from ordinary ballistic missiles.
Like what, like how much intuition can he really gain late in life about these really complex questions or semiconductor physics, you know, it's, you know, extreme ultraviolet lithography. No way, no way. I mean, they were there. These guys are afraid to even discuss the subject. If they open their mouths, they immediately make fools of themselves.
So, the, the, in physics we say, or mathematics, we say the two, the two operations don't commute. You can't do A and then B. It's not the same as doing B and then A. You better do it in the right order. You better get your deep education in the hard stuff. Otherwise you're not going to have it to rely on later in life.
Bohan Lou: I mean, that basically is a Chinese intuition, right? Like, you can go from one way to the other, but not bad. Because one is, yeah, more.
Steve Hsu: But here everything's so politically correct. And you know, like if a physicist, no physicist on campus other than like a guy like me would dare ever say stuff like this, because then you'd make an enemy in the history department and in campus politics, it would really be bad. It's a, so nobody says anything.
That's all, we live in a world where everybody's bullshitting all the time. Everybody's, it's all preference, all press preference falsification all the time.
Bohan Lou: Right. And, you know, I definitely agree that on average people who study theoretical physics are probably like, you know, two to three standard deviations of the average. So they do everything and they're basically successful. but I think one phenomenon I wanted to get your thoughts on was, Just the prodigious amounts of physicists go into finance, and maybe now, you know, more AI research and I know in one of your blog posts, you know, talking about Charlie Munger and, you know, his prescient views on like Ricardo's comparative advantage, with China, but, his, the, the final quote you put on the blog post was your favorite about, Charlie Munger lamenting the amount of hard science PhDs are going to finance how they're all getting brain drain and in China There's Shi Yigong if you're familiar, right?
He was he's I think dean of the Beida Bio, biology department is now the founding president of Westlake University, which is China's probably next generation research powerhouse University And he also had a big speech where he lamented how actually I believe Counter to most people's expectations, the highest Gaokao scores in China don't actually go into physics or computer science.
They go into finance. The most elite schools in Tsinghua Beida are their finance undergraduate majors, right? So it's also the top talent getting sucked into, you know, the highest paying, you know, money making jobs and, you know, obviously he laments that as, you know, for, for a state and a civilization.
I'm wondering about your thoughts on this.
Steve Hsu: Well, I agree with Munger. Munger said it's not good for society to attract your best talent into what is really a kind of zero sum, he used the word casino. And it's not quite that bad because it's not exactly zero sum. Like you can make arguments about whether having more liquidity in the markets is good.
Having well functioning markets is good. And it's a kind of underpinning technology for economic decision making and allocation of resources. It's all true to a degree. The extent to which a really talented person can add value in finance, I think, is dominated by people who do real positive things in the technical world.
And, it is kind of good that, you know, people can now, you know, make, you know, more money in Silicon Valley. You know, if you're a successful AI founder or something, it may dominate the career path. of somebody who goes to Jane Street and just does little stupid arbitrage tricks all day. So, you know, I have friends on all sides.
I have friends who were trained in theoretical physics or math and computer or computer science and went into finance and were the first early generation that introduced quant techniques like derivatives, pricing, options, pricing, all that stuff. High frequency trading. a guy I was, I was the TA in graduate school at Berkeley of a guy who built the whole H.F.T. set up at citadel and is a billionaire. Okay, so I know all these guys, right? And I also know guys who are tech founders. first author on the scaling paper from open A.I. Okay. high energy physics guy. So it gives me, it's such a rich tradition, physics.
It's the first science that became truly international and attracted people. The glamor around physics after the Manhattan project and Oppenheimer and all these things was so high that for generations, like the top people were sucked into physics. If you read Donald, do you know who Donald Knuth is?
The guy who wrote The Art of Computer Programming. So, he writes in his autobiography, he says like, well like every kid in my generation, Cold War generation, every kid in my Cold War generation, I was sucked into physics. And he was a, he started out at Caltech as a physics major. Okay. But of course, he just ultimately realized he was more interested in discrete math and, and, you know, computers and stuff.
So he ended up there. But, the point is that at one time everybody was sucked into this kind of education. And so coming from that tradition, I've seen people go in all directions, you know, to hedge funds, quant finance, trading, into startups. You name it, they're, they're all over the place. Right. And I feel that it is a misallocation of resources for too many people to go into quant finance.
And, and I think now it's not as glamorous. Like, I guess it kind of went like this, like For a while it was very glamorous and then it kind of came back down a little bit. But now with Jane Street and some of these two Sigma and stuff, it's kind of come back a little bit, but still there, if you're really creative, there are much greater things you can do like founding a company or, you know, building a really earth changing new technology.
And I think that's a healthier way to distribute talent in society.
Bohan Lou: Right, so then, you know, with the students you see then, are you doing anything to try to nudge them away from finance in a sense, right? You think it's misallocated? You think it's too many, you.
Steve Hsu: I, so I've been through this many times, either training a PhD student or a postdoc, in my group, and I, I tell them like, it's very tough to get a permanent position, a position where you're basically being paid to do research, on, you know, all the time in theoretical physics. It's very tough. And you should just assume you're not going to make it.
There was one period of time where they kept careful statistics. And even from the very top PhD programs like Princeton, Harvard, Berkeley, Stanford, Caltech, MIT, it was like maybe one in four, one in five PhDs from those schools would find a permanent position. So it's not very many if you think about the quality of the people in these PhD programs.
So I would just tell them, you're not, probably not going to make it, but luckily with me as your advisor, I have pretty broad interests. So if, if it looks like you're not going to make it, we'll spend six months, nine months at the end of your PhD or postdoc period where I'll train you up in something like options pricing theory or, you know, quant finance, or it could be in more machine learning type stuff so that you'll be more marketable on the job market.
And so I've been through that with, you know, at least, a half dozen, students and postdocs, maybe more. And, so I'm pretty familiar with that. Even that, like the psychology of that transition, between being really, really focused on having an academic career and wanting to do fundamental science to all of a sudden being like, Hmm, what's my net present value for doing that? You know, that totally different way of thinking that these.
Bohan Lou: Yeah. Oh, I see it a lot too. One of my best friends, he is about to finish a math PhD at Sony Brook and he is at the point where he's tired of being poor and he's going to quant finance. You know, he, you know, these realities.
Steve Hsu: Look, I don't really think against them, you know, of my really closest friends, like guys who I was best man at their wedding and stuff like this. A lot of them are in finance at hedge funds and stuff like this. I don't hold anything against them. You know, life took them where it took them.
And, you know, in the scheme of things, those were also fascinating careers too. I mean, they got to do a lot of interesting stuff. So in terms of base contribution to like the advancement of society or deep knowledge about the universe, of course, we don't want everybody to end up.
Bohan Lou: And what, what, do you think you were able to resist the temptation? Right. And you didn't go to quant finance.
Steve Hsu: You know, I thought about it and I actually spent a lot of time trying to understand financial markets, economics, you know, quant finance type stuff. I actually wrote the first paper, which was not published, but it was circulated around Wall Street, um, which used something called path integral.
So there's a thing called the path integral, which Feynman introduced in quantum mechanics. And, it's become like a big thing in quantum field theory. I used path integral methods to give closed form solutions for options, exotic options, pricing, values. And at the time that was actually faster than the just purely Monte Carlo methods people were using.
So that, that actually was used by at least one or two trading groups for a while, but, but that, that worked. So, I actually came pretty close to going down that route. I actually, if you talk to my girlfriends from that era, They would be like, yeah, Steve was always undecided, like whether he should stay in physics or he should just like, you know, go to Goldman Sachs or something.
And I think what happened for me was my career just went in a certain path because I was a junior fellow at Harvard, which is like a high prestige thing. And so I kind of knew I could stay in physics if I wanted to, I didn't have to go through quite the job difficulties that most people, even PhDs from good universities have to go through.
Cause I kind of knew I was going to get a good faculty position. So that's kind of what kept me in. Also, you know, it's funny because if the ground zero for quant finance had been in the Bay Area instead of New York, I probably would have gotten into Quant Finance because I just don't like the Northeast.
I don't like New York City. I never liked it. When I was a professor at Yale, I used to go down to New York City to hang out with my, you know, bond trader friends and shit, or to train at Henzo Gracie's school, jujitsu school. That was the closest, you know, BJJ school to New Haven at the time. So, but I used to hate New York. And, just the idea that I would have to live in Manhattan just never appealed to me. But if it had been in Berkeley, it could have been like, Oh, you can live in Berkeley, California. You know, if the roles had been a little bit reversed, I might've gone to finance part of the thing repelling me from finances. I just did not want to live in Manhattan.
Bohan Lou: Right, right. Yeah, I mean, yeah, to this day I think most of these big, you know, quant firms aren't really in California maybe due to non compete reasons, but you know, still isn't a big presence.
Steve Hsu: I don't know why. I don't know why. Yeah. But, it's, it's just inertia, but, yeah, that, that actually idiosyncratically paid, played a big .
Bohan Lou: Right. But I guess, you know, you are now a professor, and you've served in academic administration functions, but you still engage in private sector activities, right? You had Genomic Predictions and you now run Super Focus. Like what is, you know, your interest in these private sector pursuits?
And also, you know, one or honestly, do you feel like it detracts from research, you know, potentially?
Steve Hsu: So I got involved in tech startups, over actually about 25 years ago. So during the first internet bubble and my first, the first company I founded was with a PhD student who was studying with me at Yale. And we started the company, you know,when he was still a PhD student and that became a company called Safe Web, which was acquired by Symantec.
And, so once I had that experience, which was so exciting, it's so exciting to like, you know, actually the, the headquarters at first was in Berkeley and then, and then we moved it to Emeryville, which is right next to the Bay Bridge, a little town on the East Bay, right next to the Bay Bridge. And, so we had the classic startup experience.
And once you understand that world, once you understand how venture capital works, how entrepreneurship works, how leadership of a team works, all these things, it's like, it's kind of addictive because when you have ideas, and they can be ideas in a broader space than just like some narrow area of physics, you can actually do something about it.
So if we figure out that, like, wait a minute, there's enough data now to actually predict quantitative traits in genomics. using machine learning, you can not only do it as a research project, you can build a company around it. So I've always, since then, I've always had my sort of one foot in one world and one foot in the other. If you talk to my advisor, you know, at Berkeley or, you know, my colleagues, they'll be like, man, I wish Steve would quit fooling around with these stupid tech companies and just spend all his time doing physics.and then, then like, I've, there actually many times when like a particular venture fund was like, you know, because I can't be the CEO of the company if I'm not full time.
So if I'm only involved as a professor, like I'm on the board and I'm, I'm helping the company, but I'm not actually the CEO. There have been times when I've had venture funds say, Hey, we really want to put a big chunk of money into this company, but only if you're the CEO. And so on both sides, people have said, yeah, you should drop that other thing.
That's stupid. Why are you, why are you writing these papers on black holes? Like just cut it out, like come over here and help us make sure this product ships, you know?
Bohan Lou: But You, you know, so far haven't been willing to give up either, right?
Steve Hsu: I haven't. And part of it is my, I think it's just hardwired polymathic nature because there are plenty of people who are smart, very smart, but they're not polymathic like they're able to just focus on one thing. Like I'm doing algebra topology and I'm not gonna even think about the next election.
I'm just doing You know, i'm not gonna worry about you know my kids like school board. I'm just doing algebraic topology from now till I die. Right. So there are people like that. Hedgehogs, if you, if you want to call them that. And then there are foxes who always like sniffing around and finding interesting stuff. Or even better, like you could say like Hawks or Eagles that fly above it and see interesting stuff and come down and look at it.
So, I just have that personality where I'm doing lots of stuff and I can get bored easily and I can get interested in something and I'm able to kind of go deep into something. And like you could say, like, Steve, why are you worried about the U. S. China geopolitics? That has nothing to do with startups and has nothing to do with physics.
What the hell is that? Right? And it's like, yes, but I still spend a lot of time because I'm interested in it. And I end up talking to people from the office of net assessment in the Pentagon, you know, or something like that, or the Jason's or whatever. So, I think that's just a hardwired aspect of my psychology that I have.
Bohan Lou: Yeah.
Steve Hsu: Some of my colleagues would say your interests are too broad. You should, you should not do that. All this stuff.
Bohan Lou: No, I mean, I definitely am a believer in, you know, polymaths and renaissance men and how, you know, I think empirically the foxes outdo the hedgehogs, at least in predictions, you know, like the just being in the world is interdisciplinary. It's complex. It's interdependent.
Steve Hsu: You know, it might be true over some set of predictions. However, it might be that the ones who make the really deep time, you know, eternal contributions are the hedgehogs. So Einstein didn't do much other than physics. Right. So, so, so the thing is like, and maybe von Neumann, if he had focused more on one thing would have made some real progress.
Well, some mathematicians would say von Neumann, although he was very quick and smart, didn't make any one, like, really, really deep mathematical contribution the way that, you know, some other people did, right? Who were maybe not even as talented as him. So, so there, there is a trade off going on here for sure.
Bohan Lou: Yeah, and Just to your point about like you're interested in US China geopolitics and I am kind of curious, you know Give me your Chinese background in this day and age, you know be interested in this field and maybe doing certain things the US government Do you think you're actually curious about whether your Chinese background played a role in shaping your identity growing up or even now you face any additional resistance? You know potentially with increasing geopolitical tensions. You know.
Steve Hsu: Well it definitely played a role in shaping my view of the world because certain questions that I regarded as fundamental, I was always astonished that other people were not interested in these questions. So I would even ask my dad, like, well, wait, you guys keep saying you have 5, 000 years of civilization.
Well, what, what happened in the last few hundred years? Like, what, what, what was that? You know, like, how come you guys, you know, You know,you know, there's an untrue statement that like, Oh, they didn't know the earth was round until the Europeans told them it was round. But that's actually not true.
That's a widely stated thing, even by like, quote, historians of China or whatever. And, and, but you know, I would ask my dad, like, is this true? You guys with your great civilization didn't know the earth was round until the Europeans came, some Jesuit missionaries came and told you like, what, really?
Yeah. Yeah. Yeah. So there are really some fundamental, deep fundamental questions about developments of society, progress in technology, economics, you know, deep economic history. So there's stuff like that I've been interested in since literally I was like a five year old kid because it was kind of obvious.
Some of these things are obvious to me from a very early age. and so it definitely shaped my identity that I'm interested in these questions. And so I'll just give you an example. The first time I went to China. I think around 1991. I attended a conference in high energy physics in Japan. And so I was in Tokyo and I deliberately said, " I don't think you could go to China at that point, except the special economic zone across the border from Hong Kong. And so. I deliberately made a trip. I guess my girlfriend lived in Hong. I had an ex-girlfriend who lived in Hong Kong. Anyway, so, so I went, I deliberately added a leg to my trip and these are this. These are the days when international travel is much rarer than it is now. So anyway, in addition to that physics conference in Japan, I deliberately went to Hong Kong. Not to see this girl, not to see my buddy who was trading for Salomon Brothers, in, in, Hong Kong at the time. No, I went because I wanted to go across the border and see the special economic zone. Because I wanted to understand what big China, real China's trajectory, not Hong Kong, okay, or Taiwan, but real China's trajectory was likely to be during my lifetime. And I remember people in Hong Kong, like my girlfriend saying, like, ex girlfriend saying like, why are you going to Shenzhen? It wasn't even, wasn't even called that then it was probably called that then, but it was mostly known in the West as a special economic zone, SEZ.
And, she was like, you should go to Guilin. You can go to Guilin. You can get a one, a few day temporary passport and go to Guilin and see these oddly shaped limestone rocks.
Bohan Lou: Right.
Steve Hsu: And I'm like, what the fuck do I want to see these rocks for? I'm going to go, I'm going to take this horrible bus trip across the border to Shenzhen where they don't even have sidewalks, they just have mud.
Bohan Lou: Right.
Steve Hsu: They're putting up skyscrapers and mud and then huge factories, huge infernal factories that scream of the early industrial revolution, where people, you know, quality of life is horrible.
Bohan Lou: Yeah.
Steve Hsu: People don't know how to behave in an, you know, organized industrial setting. People are poor. People are like six inches shorter than me because they had no protein to eat when they were growing up.
No calcium. You know, just insa I saw things that, you know, I will never forget. I wanted to see those things. And so obviously I had a deep interest in what was gonna happen there. And so, yeah, I've had this interest for a long time now. To me, it's like, if you just say like, what's the greatest story of the world over the last 20 years?
Well, it's obviously the development of China, right? It's like, okay, one fifth of the human population goes from being desperately poor to being one of the most technologically advanced countries in the world. How many times has that happened in the history of humanity? Well, maybe never, right? And, so that's the greatest story.
And if you're like, historian or political scientist or whatever. And you missed it. Well, too bad for you because I was following it. And, so, but now we're finally at a point which I anticipated as a kid. Cause I knew what happened to the Japanese Americans and stuff like this. I mean, I'm sure that, you know, I'm not sure, but there's a very good chance that NSA or, or CIA are monitoring my communications and people are like, is this Steve guy on our side?
Or is he actually a spy? You know, and it's, it's very weird because as I've said, in other interviews, I'm regularly talking to U.S. national security people, intelligence people, and stuff like this. And, I mean, just as a civilian, I don't do any spying or any for anybody. But sometimes people come to me and ask me, like, what I think about something, right?
And, I'm regularly doing stuff like this. and yet, because I am ethnically Chinese, there are plenty of people who will, you know, be suspicious that I’m actually a spy or something.
Bohan Lou: Yeah, it is. And it's inherently true. yeah, I think in this climate that, you know, our ethnicity does matter and it's, it's noticeable, right? You know, versus like in the cold war, right? The Russians were white and, you know, they could blend in a very different way, but we are, we're noticeable.
And I, my, my, interviewer for Yale, who was the U.S. consul general in Shanghai and was on the national security council under Clinton Bush. He's a, you know, white Jewish guy married to a Chinese woman, and was one of the first Americans to go to China. So, and he tells me that, you know, he married a Chinese woman and nowadays someone like him could never serve in the positions he did just because he married a Chinese person.
It's just too sensitive now. They just wouldn't, even though he's a white American, they just couldn't even place him there anymore. Yeah.
Steve Hsu: I think things are, things are going to happen in the next 20 years. I'm pretty sure because this competition is probably going to intensify. And so, yeah, it, the, the, the, the part where the unpleasant, you know, up till now, it's just been an interesting intellectual exercise for me to follow, like what's been happening, in China development, but now it could get kind of nasty. Who knows?
Bohan Lou: Yeah, I wanted to touch a bit about, you know, the genomic predictions work and you know your work with, you know, cognitive, you know, prediction, psychometrics. and so, you know, you've, you published research showing, talking about the genetic architecture of intelligence and you know, how we're getting to a point where, you know, We probably, you know, correct me if I'm wrong, maybe getting really close, we're already able to predict intelligence, you know, with, polygenic, you know, genetic markers.
Yeah, first I'm kind of curious, like where, where are we at in terms of the stage of getting a baby or an embryo and just figuring out what their IQ is going to be?
Steve Hsu: So we can get the embryo genotype, the embryo, my company genomic prediction has genotyped something like approaching a hundred thousand embryos, you know, as part of our business and we get a whole genome genotype. And from that we can compute all kinds of things. Like if you're. an outlier and risk for diabetes or breast cancer, et cetera.
We can determine those things. We deliberately don't try to make any estimate of cognitive ability just because it's too controversial. If we did try with the current best cognitive ability predictors that are, that, you know, we've built or other people have built, you could maybe get a correlation of about 0.4, 0.3.
Bohan Lou: Okay.
Steve Hsu: 0.4, between the cognizability of the, the kid and the, you know, the genetic, the eventual cognizability of the kid and, and the calculated genetic score.
0.3, 0.4 is not. You know, it's not really good enough for, well, I mean, okay, like Gwern has done these calculations and other people have done these calculations where like you could, you can get depending on how many embryos you're choosing from, you can get a few IQ points out of this, the expectation value, Just the delta and expectation value.
Expectation value is maybe not the most interesting, important thing. Like you could also like with pretty high probability excluding rare negative fluctuations. So in other words, you could significantly reduce the chance that your kid is in the bottom 5 percent or bottom 20%. Okay. You can do that. So it's funny because even sophisticated, even like, fairly sophisticated people who think about this, don't really think about it properly.
They only think about what the expected gain is. Okay. Which really, like, for a family, like, the variance around that is pretty big. So it's like, the expected gain, okay, it's important to know, but really what you'd like to know is can I, like, exclude some tail risk?
And it, I have the hardest time explaining this to people, that, you know, when you buy home insurance, you're paying some percentage of the value of your house every year for that home insurance.
And most people never benefit from that home insurance. Right. So why do people do that? Well, they do it to eliminate some tail risks. They really don't like it when my house burns down. Right. So, so in the same way, like 0.3, 0.4 is not great, but I can eliminate a significant amount of tail risks that you don't like.
And, just based on that, a rational, I mean, intelligent, highly intelligent, rational person would understand there's a lot of value in this, right? But now, the main limiting factor for improving these predictors is just the availability of data.
Bohan Lou: Mm hmm.
Steve Hsu: So the political situation is that if you go out and you try to, you're building a biobank or you're doing something, And you, you try to measure the cognitive scores of the people that you're inducting into your sample, you will just be attacked.
They'll call you a racist. They'll call you a eugenicist. NIH is so scared about this, they won't touch it with a 10 foot pole. Nobody can get funded to do this kind of research. Even worse, NIH is banning some researchers, like, for example, my, well, I don't want to say the names, but some people that I've worked with before who are really, like, world class computational genomicists, statistical geneticists, they are not allowed to access what data the NIH does have of this kind, right?
So, it's clearly a Sharia law
Bohan Lou: Hmm. Sure.
Steve Hsu: reaction to a certain branch of science. This is a branch of science. Because when we do, when we build these genomic predictors for cognitability, and we find like, oh, there's this gene, a variation in this gene has some particular impact, plus or minus on the expected. IQ of the person immediately, like you can go back into neuroscience and say, Oh, I see this gene codes for this protein, which is involved in this neurotransmitter or the growth of this dendrite or what, you know, this is real science.
It's not like bullshit. Right? So, so, you know, there's real stuff here, but we are not doing this science because some leftists have butthurt feelings. That's literally, that's literally how our society is organized. Right? So, some leftists have butthurt feelings. So there's a very serious important area of science which we could make progress on, but we're deliberately not making science, and we're persecuting the people who are interested in it.
Bohan Lou: So basically right now in the West, there is actually no progress in enlarging this dataset,
right? It's just stagnant. It's just not getting
Steve Hsu: Well, I don't want to comment too, in too much detail, because then I would shine the eye of Sauron
Bohan Lou: Right.
Steve Hsu: on some researchers that are just trying to do science. So, obviously, like, oh, there might be incidental collection genotypes, and maybe you can hoover up and aggregate. Right? Those genotypes.
I just can't resist telling you the story.
So, there was a longstanding collaboration, which used to be based at UC San Diego, and then it moved to USC. It's called Pediatric Imaging Neuroscience. It's a project where they inducted lots of children, and then they did a bunch of brain imaging on them. Pediatric imaging.
So they did a lot of brain imaging on these kids. And they were studying the neural development of these kids. And of course, like, even just old timey, old school neuroscientist people, you would IQ, you would measure the cognizability of these kids as part of the project, right? But that generated a data set that was very dangerous.
And so, like, lately there's been tons, if you look carefully in the scholarly literature, it's not really the scholarly literature, it's the butthurt, woke feels literature, there are people like, oh, this is really bad now, because wow, we actually, from that data, and, and, and, and, you know, by the way, like, they, they were very responsible researchers, so they have, like, all the different American ancestry groups are in this, population sample.
You have Asian kids, you have black kids, you have white kids, you have mixed race kids, you have everything in this sample, right? It's thousands of kids that were imaged this way over the years and, and they have lots of data, bio data on these kids. So you can just take that data and trivially compute, there's a 0.4 correlation between just brain size, And cognitive score. And it's not brain size measured by craniometry like some old timey 18th century white guy with a, you know, mustache is like putting calipers on my head. It's measured using MRI, you fucking idiots. Okay? It's measured by, if you think I can measure like, the size of little subfeatures of your brain and not measure the total volume?
Really, that's what you think? No, I can measure the total volume. I can measure the size of different features of your brain. And I can run tests to see whether in a statistically significant way, those gross physical measurements are correlated to the IQ, the cognitive scores of these kids. Lo and behold, yes, it's, it's, it's for real.
There is a correlation. So what? Well, very dangerous. Must shut this thing down immediately. You better shut this thing down, right? So that's the, that's the America that we live in.
Bohan Lou: Right. So then do you think this kind of research is, you know, gaining speed in countries like China where other places may not have these qualms? Or like Singapore, which, you know, used to have a eugenics program, basically.
Steve Hsu: There are many, there are many places where they don't have the feels reaction that you would have here. So in other words, if I, if I like, Oh, my, my wife's in the humanities. If I, we, I was at some cocktail party that her department threw, and I just told the little story that I told you. About the ping collaboration.
I mean, ping collaboration probably burned like a hundred million dollars of NIH money, like this study over a decade, right? I mean, they accumulate a lot of good data, right? But, ooh, it's dangerous, right? So, if I went to my wife's cocktail party and told them, like, English professors, history professors, political scientists, you know, they would get very frightened.
Like, they would, like, be shaking, and they'd be like, What? Brain size? We would. Stephen Jay Gould wrote a whole book on this, Steve. It's called The Mismeasure of Man. You should really read it. Right?That's how they would react. Now, the Asian professors in the room would probably be like, wow, this is really interesting, making a little mental note.
So, Stephen Jay Gould's book is bullshit, I guess. Is what, that's what, wow, we were forced to read that in grad school. It's, it's actually bullshit. Okay, anyway, so, they don't have that kind of visceral discomfort with the whole thing. But even the scientists in China and Taiwan, a lot of them were trained here.
They understand the sensitivities of their white colleagues. So they, they actually also have like a,a kind of stop short kind of instinct. Also like, they're like, Well, I understand there's a good scientific reason to do this project, but my white colleagues might find this a little controversial.
So unfortunately that's crept even in Asia. So it's still retarded somewhat, but to a lesser degree than it would be here.
Bohan Lou: That's unfortunate because, you know, I would have loved to test myself, you know, taking my current sample, and just to know, did I achieve my full genetic ability, or did
Steve Hsu: noisy. It's not it's not accurate enough for you to tell something like.
Bohan Lou: Well, you said right now it's 4, but like, like, you know, if we have more data, right, then maybe, you know, eventually.
Steve Hsu: in the, in the fullness of time, maybe you could get it like the correlation could be 0. 7 or something, right? Or. Or whatever. But you, you, yeah, you still, there's still also clearly some variance or just, it could just be pure randomness, developmental randomness that is not controlled just by the DNA. So you'll never get rid of
Bohan Lou: Right. Cause I was just always curious whether environmental factors like lead exposure, microplastics, whatever. Did I, am I like stupider than I needed to be? You know, I'm
always just like, oh, you know.
Steve Hsu: You'll never know. And yeah, anyway.
Bohan Lou: But let's say another factor, cause it's not, I'm not the tallest guy.
Height is something that's less controversial and you guys have pretty good ability
Steve Hsu: We pretty much solved height. So, our uncertainty in the height predictor is a few centimeters.
Bohan Lou: So I could go get tested and know whether I was malnourished as a kid and was deprived of my full potential in height.
Steve Hsu: Okay, I think in the extreme case where you were very malnourished, then we could say that in a quantitative way, but you would probably already know it.
Bohan Lou: Right, right, yeah, most likely not extremely malnourished.
Steve Hsu: But you know, I'll tell you something, this is again something that dumb, like kind of dumb HOLDers don't understand, is that like If you go to Japan or China until recently or Korea, people really didn't drink any dairy and they had very limited sources of calcium in their diet. And, also like the average amount of protein consumed per day was way lower than what, say, Americans would eat.
So, so, you know, like if you look at Korea, the kids are so much bigger now. than just one generation, one and a half generations ago. Clearly, there's a lot of environmental stuff. And I don't think we really know the asymptotic kind of group average height for East Asians, because even now, there aren't that many populations that are eating a kind of completely comparable diet to what they would eat in, you know, the Netherlands or something like this, right?
Or, or, or Iowa, right? So, there's still some residual uncertainty like that.
Bohan Lou: Well, there's that and I feel like there's also You know social selection factors of you know, at least in East Asia I'm just reading reports coming out of China of like Obviously the East Asians compete on everything that can be competed on, and now height is like the new battleground.
Our parents are so anxious about, you know, injecting their kids with human growth hormones, just like doing whatever they can because they don't, like, lose out on the height competition. And I think that's also going to push the average higher, you know, marriage selection, things like that. yeah. I'm kind of curious.
So, So, you know, you wrote a lot of the many, many worlds interpretation of quantum mechanics and I'm wondering, you know, do you have any views on determinism and free will, you know, especially after the, the discovery of quantum mechanics or in a way that, or what does it even mean, you know, for, for you?
Steve Hsu: Yeah. I have written on this topic. I don't know if you saw my blog post on this, but so there's, there's sort of two, separate issues here. So, one is like, what we know about physics, what can we say about free will? And first of all, of course, it's very tough to define what you mean by free will, right?
I can define, to me at least, some philosophers don't like this way of viewing it. Um,I think there's a, there's a, there's an area, there's a school of thought called philosophers who think about free will, some of them are compatibilists. I think that's all pretty stupid myself. But if you think in physics that, forget about just your brain making decisions, just take a slice, a time slice of the universe, and ask, if you have full knowledge of the position and momentum of every particle on that slice, can you predict the next slice, the state of the universe on the next time slice and the next time.
So can you, basically, if you think about partial differential equations, is there an initial value problem that you can integrate up? And if the answer is yes, then what are you talking about? Free will? What do you mean by free? You know, because the whole universe is deterministic. right. Then you can say like, well, but what about quantum mechanics?
I mean, are there random events? Are there random outcomes due to quantum mechanics? And the answer might be yes, there's occasionally dice throwing going on, but it turns out the operation of your brain is not intrinsically very quantum mechanical. And furthermore, if these are truly random dice rolls, it In what sense does that give you free will?
Like if I have a robot that occasionally is a deterministic robot, but every now and then I perturb it using a random throw up the dice. How is that free will? So, there isn't really any way to accommodate. I think the traditional idea of free will, which is that I have a soul. And I quote, make decisions in the ether and this machine, meat machine, then acts on my decisions.
One way to realize this would be like, we live in a simulation. You are not an NPC. You're actually a player character. And in some higher dimensional other universe, there is a player operating you, and in that sense, you have free
Bohan Lou: Right, but that player may be determined, right, because they're in a.
Steve Hsu: Well, that's, that's okay. There's a, who knows what the laws of physics for that player are. Right. But, but, but like, that's a scenario where you could say like, Okay, the PCs have free will the NPCs don't have free will and that's the end of the story, right? So there's a situation where you could say at least Majula what is going on in this other universe? The PCs could have at least more free will than the NPCs so you can address it that way There's no evidence in physics that I'm more than just a machine That's described by this entire slice of the time slice of the universe. And so, therefore, if there's a good initial value description of the universe, there's a good initial value description of me plus my environment, and therefore I, quote, don't have free will.
And some philosopher guy's gonna go like, Well, this is a compatibilist situation. It could be compatible. I don't really care. I think that's a, my way of thinking about it is more clear, I think, than.
Bohan Lou: Your, position is basically my position. You know, since learning about physics, I find it difficult to believe in free will. And I'm kind of curious, like, you know, how does this impact your day to day life? Right? Do you, you know, are bothered by it that way? Or is it just the illusion of your agency is enough? You
Steve Hsu: Well, I, I, yeah, I, I think ultimately when you think about it, we are an evolved species and it's useful for evolutionary purposes for us to have the illusion of free will, the illusion of self,
Bohan Lou: Yeah.
Steve Hsu: The illusion of making decisions, even though they're determined by some lower level machinery in our neural net, those are all still useful illusions of self and agency that we have, but possibly they're just illusions, right? So it is just, it is what I mean, that, so maybe that's a very like spirit for some people, for some humanists, for example, it's a very dispiriting perspective, but that could be true.
It's also possible. We're in a simulation and I'm an NPC and. You're not, right? So, so, so that's also possible.
Bohan Lou: now. Yeah.
Steve Hsu: Yeah, I, I, don't, I, I enjoy discussing this stuff, but I can't imagine spending a significant amount of my entire career, like a philosopher would, working on this stuff. It's kind of laughable.
But because the basic things are not known, like, you got to figure out the actual physics. By the way, I just want to say one other thing, which is that another deep aspect of physics, which is even more deep in a way that it's not as, it's not as personal for humans, but it's even more deep is, is there actually true randomness in physics? So that's a really fundamental question, which actually I have to say most physicists are, most physicists kind of punted because of this stupid Copenhagen interpretation of quantum mechanics. They just punted and said, yes, there is true randomness. I accept it.
There's, there's true randomness, you know, but actually it's not clear whatever it showed us is that in fact, it could be that the wave function as a whole is evolving purely deterministically and there's only apparent randomness to the individual. on the branches of the wave function. So, so, but that is the, in a way, the most fundamental question in physics.
And yet 95 percent of physicists that you talk to don't really understand this point because they just sort of learned something about the Copenhagen interpretation when they took quantum mechanics and they haven't thought deeply enough about this.
Bohan Lou: Right, right. Or maybe they're too scared to know the truth, you know?
Steve Hsu: No, I well, I think it's just sad I mean if you look at the sociology of physics this Copenhagen interpretation thing is just the most convenient way.
Bohan Lou: Right.
Steve Hsu: to formulate it for students so we can just get on with calculating, calculating cross sections and reaction rates because that's much more important. That's kind of the way the field went. And then it's like only a small number of deep thinking people went back and looked at these things more carefully, including Everett. but then like that knowledge is very esoteric, even within physics.
Bohan Lou: Interesting. So right now the Copenhagen interpretation is the consensus view that's really taught.
Steve Hsu: Well, it's, Well, the funny thing is it's it's a kind of consensus view among the bulk of the physics population that has not thought deeply about this question, which is most people, and among the people who have thought deeply about it, I would say like, not that many believe in Copenhagen. I can actually almost, I actually document this on my blog.
I've gone back and looked at all the biggest brains, all the top theorists, you know, going back a couple of generations and tried to see how many of them actually thought about this and wrote. down their thoughts about, and, and overwhelmingly you can see that, you know, whether it's Julian Schwinger or Richard Feynman or Murray Gell Mann or, you know, you name it.
David Deutsch, well, I wouldn't necessarily put Deutsch in that category with those guys, but, but, every one of them, if you look at their, if you look carefully at what they wrote, because there was a very big social stigma about talking about Everett or, deviating from Copenhagen for a long time, nevertheless, you can find in, in their writings, I document this on my blog, for you, you know, huge number of these people, they were very specific about this point that like Copenhagen doesn't make any sense ever. It does seem to be self consistent, although we don't know for sure that it's correct, but it does seem to be self consistent and boy, my colleagues are dumb and I can't believe they didn't think harder about this.
Bohan Lou: Yeah, so why hasn't this trickled down to the standard curriculum yet? I'm sure there's lag but like, you know, why is it still not the case? That this is what's being taught in schools yet.
Steve Hsu: Because it's off the track of the way you advance your career in physics. So it's, it's, kind of irrelevant to the actual experiments that people do in physics to learn a little bit more about the fundamental laws or the mass of some particle or the properties of some semiconductor. It's all kind of irrelevant to that.
It's a much deeper foundational question. It's sort of hived off into an area called quantum foundations. But if you realize, if you have a slight philosophical bent or whatever, you realize, wait a minute, this is like the most, one of the most foundational questions, in physics. And it's kind of amazing that if you just pull like a hundred random physics professors, most of them have not thought deeply about it.
It's very frustrating to me. I mean, I've written many papers on this kind of stuff, but, but, It's a sociological phenomenon.
Bohan Lou: right, right. And, I wanted to touch on something you mentioned earlier as we were talking about free will, you know, about the role of religion. I'm kind of curious, are you religious? And do you have any views on, you know, religion?
Steve Hsu: Yeah, this is a great question. So my mother, her family, actually this military family, were converted to Christianity in the 1800s by missionaries.
Bohan Lou: Wow.
Steve Hsu: So it's very unusual that I come from this long line of Christians. My mother is a devout Christian. I was raised as a Methodist. So at the age of like, I don't know, nine, I pleaded with my dad to accept Jesus Christ.
So that his soul, he would not, his soul could go to heaven and he would not burn in hell, right? So, I would say I have a very deep religious sensibility. I can very easily, if I walk into a church and I listen to them singing, or I walk into a beautiful cathedral in Europe and look up, I can easily attain that mindset where I think there is more than just this visible physical world.
There's a depth to the universe that humans will never fully grasp. I can empathize with all of that. However, I also have the smart ass physicist's perspective that like, yeah, this is just a kind of weird delusion that apes have, you know, they can't accept death or that they're a machine or that the free will is an illusion.
So they punt and say, oh, the universe is mysterious and God loves me. So I can kind of have both perspectives. It's kind of
Bohan Lou: Yeah, that's interesting. I understand what you're saying. You know, just rationally, right? Like just how you talked about if you will it's hard to believe in this non materialist soul ish thing going on. It just doesn't really make sense But just the fact that you know when you walk into a church and you do feel something tugging your heartstreams I don't know.
Maybe that's a sign of you know, something going on there where obviously there's material ways to explain it. But there's you know.
Steve Hsu: I also feel that I'm the same person. I grew up in Ames, Iowa, you know, 40, 50 years ago, even though every molecule has been replaced now in my body, right? But that's a useful illusion from the perspective of evolution that I feel that way, right?
Bohan Lou: Yeah, and you know as we get close to wrapping up I'm kind of curious. Are there any things you thought really intentionally about when it comes to educating your kids, you know?
Steve Hsu: I did. I was very high, my kids are in college now, so it's kind of over. It's really a weird milestone in your life to think like I invested so much energy and time. I didn't necessarily do the right thing all the time, you know, but I tried. And it's, it's an emotional thing to like to realize it's over now.
Like of course I still have some influence on my kids, but you know, it's decaying exponentially, you know, once they move out of your house. So I tried to be like, I tried to teach my kids stuff. Um. This is a common thing, actually, among physicists. It was funny because at my, at our A. I. company, Superfocus, where my son is working this summer, there's another kid there who's also the son of another physicist, a colleague of mine, and they're both interns there.
I was asking them about it, and they're like, Yeah, both in both cases, like Dad tried to teach me all this stuff when I was a kid and like, you know, probably ruined my interest in physics or something. So, or math or whatever. So I tried. I think what I would advise parents is to expose your kids to lots of stuff. A little bit of what you're doing is providing the kid the willpower that he doesn't have because his brain is still maturing. So don't let him quit immediately, like expose him to something and make him try. Sometimes you have to just make them do stuff.
Bohan Lou: Yeah.
Steve Hsu: It's good for them. You don't want to push it too far and create a nonlinear back reaction. A lot of the time you're just thinking like, it's good if my kid learns jiu jitsu. It's good if he learns how to swim well. And for a while it's going to be me supplying the willpower because the kid doesn't want to do it. The kid would rather just be watching TV or TikTok or whatever, but I'm taking them to the pool and I'm making them swim.
And, you know, so you can do some of that, but you can't push it too far. Or maybe you're going to have the opposite effect on your kid. And, and, and so that's, that's all I can say to people is that.
Bohan Lou: Right. So are your kids interested in physics at all? Right,
Steve Hsu: my, my daughter is more on the artistic side, so she's not. And, my son is actually pretty talented in math and stuff like that, but he's a computer science major and I think it's actually unfortunate that he's not studying more physics in college. He took some pretty advanced math and physics when he was in high school.
He had already taken, like, the college calculus based physics class. He had taken linear algebra, multivariable calculus in high school. and he's taking, he's, he's definitely going to take more advanced math, but maybe more along computer science lines and less,
Bohan Lou: Right, right.
Steve Hsu: So in a way, it's, it's kind of tragic. Like, he may be one of those people who will never really understand Maxwell's equations very well.
Bohan Lou: You got it. I mean, but like you mentioned, right, oftentimes, You know, kids have some level of rebellion, to, you know, what their parents vocation is, maybe. And, you know, sometimes also when they get older, they come around. I think about how, like, my dad is a professor of pharmaceutical oncology, the Chinese Academy of Sciences, and probably the thing I've always been least interested in and know the least is anything related to life science or biology.
You know, everything else I'll learn, but just not biology. And there's some maybe inherent thing there about trying to leave your parents' shadow, right? Given you're such an eminent physicist. Yeah.
Steve Hsu: Yeah, so that could be an unfortunate consequence of it. So parenting is one of these weird things where you just have to acknowledge at the end, you really can't evaluate the counterfactual very well. Like it, my wife will say things like, oh my God, we should have kept them in piano for another year, or we should have forced them, my daughter, to learn violin longer or whatever. You know, I'm like, well, how do you know what, you know, the consequence of that would have been. Like, you know, you'll never know.
Bohan Lou: I mean so many things I think it's quite hard to evaluate the counterfactual. As someone who's interested in Chinese history, I'm always thinking about the counterfactuals, right? What if the KMT won? What are this, this, this? Who knows? You know, there's romanticized versions of a different character, but it's really hard to say.
Steve Hsu: I certainly couldn't tell you how my life would have turned out if I had, you know, taken the offer from, you know, Morgan Stanley to join their quant trading group, you know, when I was 25, you know, or 24 or whatever. I didn't do it, but. I can't tell you what my life would've been like if I had done it.
Bohan Lou: Right, right. Maybe the final question for you. I'm kind of, you know, you did write some topics about happiness, your blog, and I am curious to hear, you know, what is your main approach to happiness?
Steve Hsu: Again, it's a very hardwired thing. I'm just a pretty happy person. Like me, I'm very lucky. I, I don't have, I'm not high strung or neurotic. I don't have problems with depression. I'm pretty much happy all the time. I'm so happy. I'm probably annoying to people. Like, I wake up in the morning and I'm just like, Hey, what are we doing today?
You know, what's, what's up with that? You know? And then they're like, oh my God, leave me alone. You know? So. I don't have much to say other than it's hardwired. I will say one thing. My dad was old when he had me, and my brother. And so the whole time growing up, I thought being a smart kid and not understanding death, like having thought a lot about death and infinity and stuff like this, I was like, wow, there's going to be a time when my dad is not around.
And since he's older, I can kind of calculate, like, I'm not going to be potentially, I'm not going to be that old when he goes. And so when my dad died and I was in my. I think it really affected me because I have friends who are my age and both parents are still alive. It's not that uncommon, whereas my dad's been dead for a long time. And when that happens, you then can consider someone's life in its entirety, someone you knew really well. You can think about, like, what was important to my dad? Like he had an academic career. He published books. He wrote scientific articles. But what was really important to him? It turns out it was family. And so, like, I always have this balance, like, some of my physicist friends who are very focused on their physics careers or whatever, guys who are trying to make money, I've always been like, you know what, there's a good chance at the end, like, all this stuff is, you're going to regard this all as kind of, like, vanity bullshit, like, you know, vanity, vanity, all this, you know, from the Bible.
You're going to realize, like, It's really your family and your human connections that matter. And in that way, actually, the experience of having kids and raising them has been such a joyful and rewarding experience that in a way, like, you know, some people might say, Oh, this Steve guy, he's such an elitist.
He's talking about cognitive ability and Richard Feynman and, you know, whatever. I actually am glad that almost any ordinary human who can have a family and raise their children really has experienced maybe 90 percent of the great stuff. You know, even if they were working as a gardener the whole time, they didn't have some highfalutin career, they still got to experience really great stuff.
And I feel that's, you know, it's a job in a way that's contributing a little bit to the justice of.
Bohan Lou: Yeah, I mean.
Steve Hsu: After going through that myself, I feel much better. Things are not as unequal as they could be.
Bohan Lou: Right. Right. It's almost a religious view, you know, like on the, on the blessing of children and how important it's like, I mean, you've accomplished so much and yet you say all of that is squeezed into maybe 10 percent of the happiness and value, you know, of life.
Steve Hsu: In a way, in a way. Yeah. I mean, of course you could go into some mode where you're like, Oh, well, what are my historical contributions to, you know, blah, blah, blah. But at the field level, at some deep level, it's more like, no, that, that, that other stuff is just vanity, vanity, all this vanity. But anyway.
Bohan Lou: Yeah. Well, thank you so much for your time, Steve. And I'm wondering, you know, just the audience, you know, you do write prolifically on Twitter and blogs and, you know, so where can people find you? And I'm happy to link them too.
Steve Hsu: Well, I moved my blog to Substack.
Bohan Lou: Right.
Steve Hsu: So I just found myself there. But the best place to start is just Twitter and my podcast, which is called Manifold.
Bohan Lou: Yeah. awesome. Well, thank you so much for your time, Steve.
Steve Hsu: Thank you.