Manifold

Philippe Lemoine is a PhD candidate at Cornell University in philosophy and a widely-read public intellectual. We discuss philosophy, the scientific research used to justify COVID lockdowns, and the Russia-Ukraine war.

Philippe’s writing: philippelemoine.com
Philippe on X: https://twitter.com/phl43

(00:00) - Introduction and Guest Background
(00:54) - Philippe's Academic Journey
(08:29) - Philosophical Insights and Career Shift: Public Intellectual
(46:17) - Russian Energy and European Economy
(48:27) - Covid Epidemic Modeling: Bad Science
(56:22) - Critique of Scientific and Policy Incentives
(01:31:54) - The Messy Reality of Ukraine Maidan Uprising
(01:32:59) - Could Security Guarantees Have Prevented the Ukraine War?
(02:07:21) - Ukraine War: Long-Term Predictions

Music used with permission from Blade Runner Blues Livestream improvisation by State Azure.

--

Steve Hsu is Professor of Theoretical Physics and of Computational Mathematics, Science, and Engineering at Michigan State University. Previously, he was Senior Vice President for Research and Innovation at MSU and Director of the Institute of Theoretical Science at the University of Oregon. Hsu is a startup founder (SuperFocus.ai, SafeWeb, Genomic Prediction, Othram) and advisor to venture capital and other investment firms. He was educated at Caltech and Berkeley, was a Harvard Junior Fellow, and has held faculty positions at Yale, the University of Oregon, and MSU. Please send any questions or suggestions to manifold1podcast@gmail.com or Steve on X @hsu_steve.

Creators & Guests

Host

Stephen Hsu

Steve Hsu is Professor of Theoretical Physics and of Computational Mathematics, Science, and Engineering at Michigan State University.

What is Manifold?

Steve Hsu is Professor of Theoretical Physics and Computational Mathematics, Science, and Engineering at Michigan State University. Join him for wide-ranging conversations with leading writers, scientists, technologists, academics, entrepreneurs, investors, and more.

Steve Hsu: Welcome to Manifold. My guest today is Philippe Lemoine, who is a PhD candidate at Cornell in philosophy, and also a well-read public intellectual. He writes both, I think, for magazines and also on his own Substack. I had many suggestions from listeners to have Philippe on the podcast. Let me first ask you, Philippe, am I pronouncing your name correctly?

Philippe Lemoine: correctly enough.

Steve Hsu: I still have some remnants of high school French.

Philippe Lemoine: Yeah, yeah. Much, much better than most Americans, I will say.

Steve Hsu: Yes. My French teacher would be proud of me, I guess.

So let me start by asking you about your background. And as I just mentioned to you before we started recording, people in my generation, intellectuals and American intellectuals and academics from my generation, had a certain fascination with France and the idea that, oh, intellectualism still survives in France and things like this. So, tell us a little bit about your childhood and how you came to Cornell and, and how you view the, how you view the world as somebody who, you know, has one foot, maybe one foot in France and one foot in the United States.

Philippe Lemoine: okay. so, I have a somewhat unusual background at least in so far as my studies are concerned. I was, you know, like I remember being in high school, hesitating. So, I started programming when I was like 14 or something. So, I was really into programming as a teenager. so, like a typical nerdy kid in that respect.

but I also liked physics a lot when I was in high school. So, I remember hesitating between physics and computer science, you know, for university, in high school, I took the path of least resistance, which was computer science, because I already knew most of the stuff. so, I studied computers. I did a Bachelor of Computer Science initially.

then what happened before that, I don't know. I feel like in that respect, you know, I had the life of a nerdy kid, which would be similar in France to what it was in the U. S. Maybe we can talk about the role of intellectuals in public life in France and the U. S. This, I think, is different, but I don't think it really affected, My childhood, for instance.

So anyway, I went, did a bachelor’s in computer science by the end. I was, it's not that I stopped liking computer science. I still do, but I realized that this was probably not something I wanted to do, professionally. And so, it was more of a hobby. That's something I wanted to do as a job, you know, and I had this friend who was in this French school. Where, which is a well-known French school where you do political science, but a bit of everything, you know, economics, history, what I learned, what I discovered later is that you do a bit of everything because you do a bit of everything very poorly. And that's why you can. But, anyway, of course, I didn't know that at the time.

I just could see that. And, you know, it sounded good for me because, for two reasons, basically. One, because I wasn't sure, like I said, I had become convinced that I didn't want to do computer science professionally, but I didn't know for sure what I did want to do. And the problem wasn't really that, I had no idea about what I wanted to do.

The problem was rather the opposite, that I had too many because I was interested in a lot of different things. And so, the appeal of this school was that precisely because you get to do a lot of different things, I figured, you know, and also, it's fairly prestigious, so I figured, you know, it would both give me the opportunity to try.

Variety of topics and also leave me a maximum of doors open after that, because it's fairly prestigious. and another appeal was that, you can get, you can enter like the competitive exams directly in third year. At least at the time you could, I don't know if it works now. so, I could just study on the side, you know.

For the exam while I was finishing my bachelor’s in computer science and then getting directly in for a year rather than start at the beginning. So that was another appeal of the school in question.

Steve Hsu: this schools, PO CS's PO or

Philippe Lemoine: yeah, yeah, science. So, okay. so, I did that. I got into the school, And then, like I said, I mean, you know, I liked it a lot, but I was disappointed by the, the classes, but I liked it a lot because I could study on the, you know, like, even if I thought the courses were, you know, very superficial in general, if you, if you're willing to study on your own, you can.

You know, you don't even need a school, really. but I guess, you know, so I did learn a lot of stuff and, I suppose you could say that it was prompted by the topics I had to study for classes, but a lot of it was like stuff I would. Read on my, on my own, not all of it, you know, I don't want to be like, there are some good classes, but, just like in general, like I remember, I remember taking a macroeconomics class, for instance, and to give you an example, like the, I remember the exam was a multiple choice questionnaire type of exam.

And so, what students do, so, you know, we'd go over the basic models of standard macroeconomics and, and so, you know, you, you would get asked questions such as suppose you're in model X, you know, manual Fleming or whatever. And, you know, with two countries and B and like, there is a, there is a shock in country B and like, okay, according to the model, what's going to be the effect in country A or something like that, you know?

And so, what's. Almost every synth, I say almost every, because It's not even that I know even one other person who didn't do that besides me, but I'm, I'm, I'm assuming there has to be some, like in over a hundred students and like in the class is, they would just like go over the exams from previous years and learn by heart all the possible question, all the possible answer.

And they are just. Answer the of course, you know, this gives you no understanding of the models that you're supposed to learn in the class. Whereas, you know, the, the, like the, the more economical and smarter way of doing this kind of exam is you, you, you learn the equation of the model and just, you, you derive the answer to the question from the equation.

And those models are not very complicated. It's not but, but almost nobody does it because it's not incentivized by the way it's taught and the way you're graded. Because also you have very limited time to do this, multiple choice questionnaire thing. So, if you learn by heart, you actually have an advantage over people, someone who has to derive the thing every time.

so, you know, just to give you an example of the way in that school, and I think it's fairly common, the way in which the teaching works and the exams work did not incentivize actually understanding the material. which I think is really bad. And so that's what I meant earlier, when I said that you do a lot of things, but very superficially, and so people get out of this school, they can talk very convincingly about pretty much anything for five minutes and when it gets beyond five minutes, then it starts like to collapse because you, you start to realize that they have no idea what they're talking about.

I mean, it's amazing.

Steve Hsu: it must be a great feeder for McKinsey and other consulting companies.

Philippe Lemoine: It's, it's, it actually. Actually, a lot of people go there after that. It's also a great feeder for French politics, which explains a lot if you ask me. But, so anyway, after this so, you know, it was useful to me, like, because like I said, I did get to, study like a lot of, a lot of things, you know, during those years.

And it also helped me.

So, you know, at the time I was considering economics a lot, you know, like, I figured while I was there, I figured I wanted to do a PhD. I just wasn't sure in what field and I remember history, economics were topics I was seriously considering, but in the end, I decided to go for philosophy after reading Aristotle totally by chance.

like I read the back cover of the physics of Aristotle and there was a quote by Heidegger, who I've since read and therefore hate with a deep passion. But at the time I had never read him and all I knew was that he was this famous German philosopher, and he was saying on the back cover, something like this is probably arguably the most important book in Western, Western metaphysics or something like this.

I was like, okay, so I should probably read this. And I thought it was, I thought it was going to. I was going to read it in two weeks, you know, and, and then move on to the next thing, but it was kind of a revelation. I was really impressed by it, and it was really challenging. Like I ended up spending all summer on it.

I wrote some kind of summary slash commentary of it that ran like well over a hundred pages just to reconstruct the arguments. I remember being very impressed by how clever he was. Like, I mean, I still think he was one of the smartest people who ever lived. I think it's very impressive.

and also thought it was fascinating to see, to put myself in the head of someone who hadn't, who didn't know about modern physics. And so, we had like a completely different way of understanding, you know, a theoretical scheme to understand the physical world. And, and that was really interesting because it really helped me appreciate better, Both the achievement that constituted by modern physics, but also how we got there, because, you know, we did build, if you look later, I learned, I read more stuff and how do you specialist of the issue, but I read more stuff about the history of, of physics. And it's really interesting to see how you have like, you know, nominalist physicists in the middle age who still feel like something that looks like it's the beginnings of like, what will become calculus and like they build on this.

And so, it's very gradual, you know, we have this image of the scientific revolution as this neat, clear break with the previous physics. But that's not at all how it actually happened. You know, in a sense, it was a break, but it wasn't nearly as neat as, and they built on previous stuff. and so, it's very interesting to look at, you know.

Aristotelian physics and, and then how it evolved over time because it wasn't, that's another like cliche that it was this fixed thing, but no, people like, you know, there was innovation within that framework, you know, paradigm as Kuhn would say. And, and he did like, you know, although, yeah, there was a change of paradigm, but there was some continuity too.

So yeah, that was, that was interesting for both of those reasons. So, I went to do philosophy. In France, I don't know if you're aware of this distinction in philosophy. There are two like schools in contemporary philosophy or tradition, I guess, one is called continental philosophy. The other is called analytic philosophy and analytic philosophy dominates in the Anglo-Saxon world, whereas continental philosophy dominates on the continent and especially in France, the names are actually funny because analytic philosophy was born in Austria, which you, it's hard to get more continental than that.

But, now it's more accurate, I suppose. And so, I was never really into continental philosophy because it always struck me as gibberish. And I don't know, it was just like, it didn't interest me, you know, like Aristotle was different, you know, like, I was like, okay, this guy is. is making real arguments.

It's hard to reconstruct them often, but you know, you can, you know, make plausible reconstructions. You know, I read Heidegger and I have no idea what's going on here. And I honestly, I don't believe anyone does, but so I wasn't really interested in continental philosophy, but I also didn't know analytic philosophy because especially at the time, you know, in France, it was almost nonexistent.

Not quite, but almost so, I was mostly doing history of philosophy and then by chance, basically because I heard people bash analytic philosophy saying that, you know, it was, they thought that you could reduce philosophy to logic. And I was like, I mean, sure, that sounds dumb. Like, surely, philosophy can't be reduced to logic, but I should probably read about this stuff to see for myself, you know, what this is about.

And so, I read a book called, Companion to Analytic Philosophy. I don't remember the name of the author. It was an overview of, like, contemporary analytic philosophy, and of course, I immediately loved it. So, in my case, criticizing incessantly analytic philosophy kind of backfired because it made me check, look into it, and I really loved it.

So that's when I decided to go to the U. S. basically because like I said, in France, it's not that there is no analytic philosophy whatsoever, but there is little of it, you know, there are a few places where they do that. if you want to get a scholarship, it's difficult when you don't know anyone, and I didn't know anyone, whereas in the U.

S., in that sense, at least it's more meritocratic. You could write, at least in philosophy, the way it works. The main element of your application is what they call a writing sample. So, I wrote this writing sample about some topics in the philosophy of language, specifically how the reference of terms is fixed.

And it's funny, I read it a few years ago, I read it again and it like, you can see that it's almost entirely, self-taught stuff, you know, I reinvent the wheel on several occasions or reinvent some concept that have been like invented before, but like giving them, giving them like different names, like, you know, presenting them in slightly different ways and stuff.

So, it's hardly ideal in that respect, you know, but, but still, I guess, it was still interesting enough that I got in at Cornell. So, I went to Cornell. I was mostly, even though the, well, I wasn't sure exactly what I wanted to do at the time, but I already was, inkling towards philosophy of science.

and, once I got there, I also got into logic. A lot more. So, I started reading more about like mathematical logic, which is really at the, this is one of those topics where the, there is no clear border between, you know, where philosophy stops, you know, and math thoughts and vice versa. because you can't, you know, I feel like I like statistics in that respect for similar reason, because you very quickly, when you go into statistics, you know, very quickly, you run into some epistemological question that are really philosophical, but you can't deal with them.

without understanding the math either. So, it's not, you know, there is no clear separation. And in fact, people who publish in those fields frequently publish, you know, and are in philosophy department or statistics department or physics department, et cetera, and publish in similar journals. And so.

It's, you can see the same thing at the institutional level too, even though, for practical reasons, you do have those semi artificial like divisions, but which are inevitable, I guess. So anyway, I got into this and I wrote a dissertation, which technically is still not finished because I have to, you know, I should make revisions to defend this stuff in two chapters.

It's one of those, three paper kinds of dissertations. So, you write three, so it's three different papers. I won't even try to describe what the first is about because it's the most boring and uninteresting topic you can imagine. To this day, I was going to say, to this day, I have no idea why I started to, I decided to work on this.

Actually, I do know it's because some people were wrong, and I was annoyed by it. That's a bad reason. That's a bad reason to write about something when the underlying. Topic is completely uninteresting. And this, this one is, so anyway, the, the second one is it's something about, it's, it's, it's about something called power, consistent logic, which, very roughly is a kind of logic that unlike classical logic can deal with, with inconsistent theories without everything going to shit basically. and so, what I do is like, define a new class of such logics, prove that they have such and such properties and argue that those properties make them philosophically interesting for such and such reasons. the,

Steve Hsu: The problem with logic doesn't immediately allow you to prove anything.

Philippe Lemoine: Well, I mean, you, you can prove something about logical systems, but, so that's, that's what I do, you know, in the. But then, you know, you know, I was much more enthusiastic about this paper when I started working on it than I am now. But, I mean, I still think it's kind of interesting, but not nearly as I initially thought.

I guess at this point, the one I prefer is a third paper, which is about something called the representational theory of measurements. And it's basically something that a bunch of mathematicians and psychologists mostly came up with, in the fifties, fifties, sixties, although it goes back to work done by physicists at the end of the 19th century.

And it's a theory that's supposed to explain how we can describe the world and have theories that make, Accurate predictions about it using numbers, which is kind of puzzling when you think about it, because the world doesn't seem to be numerical in any of your senses. and so, this theory explains that if you posit that the world is a certain, that certain structures exist, then, there is a homomorphism between those structures and certain numerical systems.

And in certain conditions, if some axioms are satisfied by the world, then you can represent it, you know, numerically, but those guys, they, they came up with the theory at the height of the empiricist movement. And so, they were convinced that they could show that they could interpret this, the, the, the axioms in question.

In a purely empirical way, and I mean that in a very strong sense that if they thought that those actions could be verified empirically in the sense that they were observable in a very strong sense, like, so they thought that you could literally open your eyes and like, verify those actions.

And then once you had. You could assume you were in a position to assume that those actions are true. Then you get the whole apparatus where you can represent the structure numerically. And, you know, it explains why you can assign numbers to stuff like temperature or length or et cetera, et cetera. and what I do in the paper is argue that no, actually, this interpretation is untenable, like you cannot interpret those actions in, in such strongly empiricists.

And so, the picture they had is as if theory would, was it like the, the, was like really foundational in the sense that all the rest of science rested on this. Because the rest of science uses numbers, you know, to talk about. And so, you had to first be able to establish empirically this theory before you could do the rest of science, at least logically.

And, and so the way it, that's one of the reasons why they would think they thought that, you know, you had to, Be able to interpret those actions in as, you know, describing, like, observe observable, like states of affairs, because that means you could, you could verify them without having needing any theoretical apparatus or other, you know, the laws of physics or whatever.

And,

Steve Hsu: need, would you need to do all this work if you're only interested in discrete quantities or would it be straightforward.

Philippe Lemoine: so, I mean, you know, there's a, like a lot of, like, if you look at Modern physics, it uses like continuous mathematics. Now there are some philosophers. This is actually fascinating. I don't know this stuff very well, but there is a fantastic book. If people want to read about Hartree field, the philosopher called Hartree field, that's called science without numbers. And what he does in this book, he argues that mathematics is actually dispensable because one of the main arguments in the philosophy of mathematics is that numbers exist because, mathematics, is and numbers are indispensable to do our best science. And so, we have this kind of inductive reason to wish to assume that they're true.

And he says, no, actually. So, he's a nominalist who says, no, we actually, we don't, it's, it is not indispensable. And, you know, many philosophers have argued this before, but usually they give theoretical abstract arguments in favor of that view. But what, but he went further, he endeavored to show concretely that you could reconstruct, you know, at least a part of modern. So, in this case, it looked like it's classical gravitational theory in the potential form and you reconstructed it, you know, without assuming numbers. It's very clever. It's, it's, it's, it's really a great book. it's well worth reading, even if you're not a philosopher.

And so yeah, I mean, you know, there are some people who think that, you know, you don't need, As much mathematics as we use in like standard science to, to do the same thing, but like, there are very serious problems with this too, you know, like, so there's various arguments to the effect that, yeah, you could do this with like this theory, but like, if you start getting into like, say, special activities and you run into certain, like the strategy used to do this with like, Newtonian gravitational theory, it's not going to work or something like so.

So, you know, but maybe there are other tricks that could be, it's, it's an interesting topic, but so far, I don't think anybody has managed to do it. So certainly, going back to measurement theory so they want to, they have this idea that, you know, this was foundational in an epistemic sense. And like, in the sense that, you know, you need to have, be able to empirically, verify this theory of measurement.

So that you could go on to do the rest of science because the rest of science rests on assigning numbers to physical objects or, or, you know, or social relations or, or properties, etc. And so, and because of that, you know, they wanted it to be possible that you could, you would have to be able to verify empirically this theory without having to assume any of the, this like scientific baggage, you know, other physical or nonphysical theories.

Uh. And so that's why they were, they tried to interpret the actions of the theory in like purely observable terms in a very strong sense. And what I do in the paper is, argue that actually this interpretation is not tenable. You have, you have, you know, those actions, they are theoretical in the same way that actions of other theories, scientific theories are theoretical.

And in fact, epistemically, this theory of measurement is on a par with the other scientific theorists. And what this means is that you don't empirically verify the theory independently of the others. And then you can go on to verify the others. you, you verify everything in a package. so, it's much closer to the kind of view that someone like Quine was defending.

And, you know, I, I mean, I think I'm clearly right about this. And I think it's one of those cases where, I really think their interpretation is untenable and I think it's very clear. But the reason why they really wanted to have this interpretation is that again, this, they, they came up with this stuff, during the heyday of the empiricist movement, and it was kind of like a dogma that this is, this was possible, and, and so they, they just didn't see the problems because of that, but it doesn't mean the problems don't exist.

Steve Hsu: still people. Are there still people defending this perspective? Like, who will you, who would you convince, if you could convince?

Philippe Lemoine: I don't, I don't, I don't know. I mean, you know, I think for the most part, people have stopped caring. My impression, at least , is that people have stopped caring about this topic. They just like this site, this stuff, and they, they talk about the classical interpretation in passing, but they don't pose to consider whether this is even tenable, like, and it's, Yeah, I don't think it's, it's really a question that people, you know, they sort of assumed, you know, this has been settled.

And so, this is not really, as far as I can tell, anyway, a question that interests a lot of people. But personally, I do think it's interesting to think about because it's like this, I think it's important to, and it's, it's one small part of the answer to the question. the more general question of, you know, the Wigner unreasonable effectiveness of mathematics and science.

And, you know, it's, and so that's a fascinating question, I think, and this is just one small part of life, so it's, I think it's important to get it right. even though, again, it's just a small part of it. Um,

Steve Hsu: So

Philippe Lemoine: anyway,

Steve Hsu: Sorry to interrupt you, but in describing your dissertation to me, I could see a little enthusiasm lighting up your face. So, are you, are you still interested in academia and continuing as a philosopher or?

Philippe Lemoine: so no, that's, I mean, I am still interested in philosophy, and, and I guess, you know, so like, but I'm not interested in pursuing an academic career anymore, which is part of the reason why I still have defended, I guess it's hard to find the motivation to, I have to make revisions on those two last chapters, and especially the, especially the second one, it's hard to find the motivation.

Also, I constantly like to get into other projects, which appeal to me more in part because I don't want to pursue a career in academia. I mean, I should do it anymore at some point still, but and so, yeah, no, I'm not really interested in pursuing a career in academia. So while I was almost done, or so I thought with the dissertation, I went back to France, to work in a startup for, for a while which was, Kind of like, finance, but, more like about, helping people get, personalized financial advice, you know, in something which like is not very, doesn't really exist, in France at the moment.

It's not, it exists, but it's not very common, even though a lot of people could benefit from it, but it's not part of the culture, I guess. And so, the. Main goal of the startup was to help develop that, which was interesting. But I guess, I left after a while in part because the initial strategy didn't work.

So, we changed the strategy. And so there wasn't really any role for me in the new strategy after this. And in part, because I was more interested in more technical stuff. I was, you know, like doing data science type of things, which especially after the change of strategy was. Was not really something that we needed to do.

And then, you know, during all this time, you know, starting from when I was like, at, in Ithaca at Cornell, was writing on the side, you know, on a blog, just in my spare time about various topics. and I got on Twitter. It's funny. I didn't think, I used to think that Twitter was like, I couldn't understand the appeal of Twitter because it was at the time it was like one of the 40 characters and was like, to me, in my, in my head, Twitter was just like a place where like dumb celebrities were having like fights and I, I got convinced, To join it because I was dating at the time I was dating a girl who was in communication and she was like, you need to get on because I had the blog already.

And it's like, you need to get on Twitter to promote the blog. And I was like, Twitter seems dumb, but eventually she convinced me. So, I,

Steve Hsu: I travel the exact same path as you because I have my blog first and I only use Twitter to advertise my blog, but then gradually you get sucked in because pretty much everybody is on Twitter. So, yeah.

Philippe Lemoine: found out, you know, I found out Twitter was great. I met so many interesting people whose existence I wouldn't even have suspected were not for Twitter. I mean, it can be, you know, it's tricky. Like you have to find the right balance. Which I'm not even saying I'm particularly good at finding, but where you can end up wasting a lot of time on Twitter, having debates with stupid people or people arguing in a bad face, which really drives me nuts.

but you can also get, you know, I, I, I think my writing benefits a lot from, from the debates I have on Twitter, because if you know who to argue with. Which this is the hard part, but if you do then, you know, it's a great way of testing your arguments of like getting like objections you wouldn't necessarily have thought about, or getting like hints, about like possible solutions to problems you're considering, getting like recommendations of papers, or, you know, like if you, if you follow the right people, you, you will like find, you will instantly know about, you know, All the newspapers in the, on the topics that interest you.

it's like such a great way for, you know, conveying information. It's like, it's, it's great in a number of ways. Like, of course I'm aware of the drawbacks to painfully aware of them. But overall, I'd say it's, it's been really good. And also, it got me a lot of likes, it just like my Twitter just took off very rapidly.

and so, it got a lot more attention. So, like, I guess my ex-girlfriend was right. You know, like it did. Help the blog a lot. and so, yeah, you know, so the thing started to be read like a lot more, I guess, and which. Again, he wasn't part of a plan or anything. I never imagined that this thing would work that well and that it would give me the opportunities it did.

And, but, but it did. And so, I ended up, so I had this blog for a while. Then I was recruited by Richard and Anya, I think, you know. I mean, I've seen you guys talk, so I know you, you know, each other for CSPI, you know, so I, I, I wrote a number of things for them. I still write for them. I have the series; we're probably going to talk about this, which is kind of like my interpretation of the origins of the Russo Ukrainian war.

So, the first, first part was published already on CSPI. And like, I'm working on finishing the second part. So, this is really huge, but in fact, I'm probably going to publish the second part in several installments because it's too big. And also, the different parts are. Pretty self-contained.

So that's, that's good. anyway, so, yeah, it was I've been like, working for, CSPI and, I created a sub stack more recently, so I'm also writing, I'm writing like, you know, on that sub stack where I wrote about, a number of things, I guess one thing that's also gave me like a lot of eyeballs, or the blog, a lot of eyes on my various writings, is COVID.

Like I started.

Steve Hsu: Mmm.

Philippe Lemoine: I wrote a lot about COVID, mostly criticizing some of the methods that were used, like the literature about particularly the effect of non-pharmaceutical interventions where, which were just awful. because yeah, I guess I should have mentioned, I, I mean, I suppose it started because I was interested in scientific modeling in general from a more philosophical perspective.

And so, I started to learn statistics on the side, because Statistical modeling is like a prominent example of modeling in science and it's useful and used in a lot of different fields. and so, and also because I have always been interested in social science. So, I figured, you know, if I want to understand that stuff, I need to learn more statistics.

So, I learned statistics on the side. And so, when COVID came I started reading those papers and I was just appalled by how terrible they were. And, and so I wrote about this, so I like one paper that was pretty popular that I wrote is about this, paper, which I think to this day remains the most The most widely cited paper on the effect of non-pharmaceutical interventions during the pandemic, which was published in June, 2020, I think, and so that claims that lockdown and other, mostly lockdowns, in, and other non-pharmaceutical interventions in Europe had saved in 11, 11 or 14, I don't remember, European countries alone, like over 2 million people during the first wave.

Steve Hsu: Mmm.

Philippe Lemoine: And so. I read that paper, and I was like, how could this thing even get published? It got published in nature. How could it even get published? You know, it was just, it was just preposterous, like the way, because basically it just assumed a conclusion in a way, you know, like it's, they basically defeated a model that assumed that the only thing that could affect the rate of transmission was, was those non, you know, government interventions.

Steve Hsu: Mmm.

Philippe Lemoine: And then and then concluded, you know, the, that, those interventions because the thing, you know, that was at a huge effect that had a huge impact, hence the 3 million lives saved because of them, thanks to them, et cetera. But this is just like a, purely mathematical. Consequence of the fact that incidents stopped rising long before anything like herd immunity was rich.

And, but it doesn't really tell you why, you know, this happened. Of course, the model assumes that this is going to be because of government interventions. But if you look at, Countries that didn't do that or places I didn't like, like Sweden, especially during the first wave. Sweden is the thing that really changed my view about this stuff.

Because going into the pandemics, I read about those, you know, compartmental models. and, and I was like, okay, so I, we, we do need to, in retrospect, I think even that was a dumb argument, but like, we do need to lock down because otherwise we're going to get, like health services are going to be overwhelmed so fast because it's going to keep climbing until we reach herd immunity.

And like, that was the basis of the arguments. But after the first wave, I thought, I mean, in, in practice, this was more gradual than that, but, but, but, you know, starting from April, basically, even end of March, April, and then it, it took me until like the next fall until I was fully convinced that those models were like, just bad.

and in part because of Sweden, I was like, look, I mean, in Sweden, that much lighter, it's not that they had no intervention, but they were very, by comparison to France where I lived, you know, it was almost nothing. And yet you know, incidence also stopped rising well before herd immunity was rich.

so, you know, I mean, of course there are other ways, you know, I remember for a time I was like, okay, maybe the, maybe the fatality rate of the disease is not as high as we thought. And that would be one explanation and, you know, and the prevalence of the infection that was higher than we thought in Sweden.

In fact, they did have reached herd immunity and another possibility is that in certain models where you introduce some complications in a basic serum model, you can get like a lower herd immunity, you can reach herd immunity at like a lower prevalence of infection. So, but, but eventually, you know, I realized that no, it's just that the way in which those models model the spread is just wrong.

And, and so one, there are, again, there are various explanations, you know, one thing I, and so one possibility, for instance, is that the people like voluntary behavior, the behavioral response, even in the absence of any kind of government intervention is such that it will be enough to drive the transmission rate down, even in the absence of government interventions.

And I do think, I do, I do think that there is something to this, like, you part of the explanation for why in Sweden it also, didn't get all the way to herd immunity during the first wave. And, but in another paper argued, I argued, you know, that if you look at the available evidence, it's not plausible that this is the only explanation.

Because there are various cases where, if you look at the data we have, which admittedly is very imperfect about people's behavior, stuff like, you know, mobility data, restaurant frequentation, you know, that sort of thing. There are many cases where you don't see any obvious change in people's behavior, and yet, you know, incidence starts falling all the same.

And this is very puzzling and, and so I was very annoyed by the fact that it's not as if I had an explanation for why, you know, this happened, but I was annoyed by the fact that nobody was bothered by that. There was a clear anomaly, you know, in the data and, and nobody is certainly not, you know, the professional, epidemic models were happy to publish like crap based on those like really simplistic models. Was really bothered by that and looking for an explanation for this. Um,

Steve Hsu: So, if I

Philippe Lemoine: know, they would.

Steve Hsu: If I could just comment a little bit based on my experience, you know, many, many years as a scientist and having some somewhat broad interests, just including outside of physics there are fields where things are messy. And so, someone who's a little bit rigorous minded looking from the outside says, hey, you guys can't conclude anything.

Right. And the dynamics I've found for those fields is that people self-select out. So, if you're not satisfied, like, like, let's suppose Philippe, I said, like, hey, you know, the government of France is determined. You are going to become a biostatistician. You're going to, you're going to do, you're going to focus on epidemiology.

You might get very irritated because you might say, like, your, your counter argument might be like, I can't conclude anything from what I have available for me. It's too messy. no one should have high confidence in the conclusions of any of these papers.

Philippe Lemoine: That's interesting.

Steve Hsu: Get me out of this field, get me out, right? And then, so the field becomes an accumulation of people who accept shoddy reasoning.

They like their model. They like to publish the paper on their model, even though, like, maybe secretly they know it's not very reliable, or maybe they're just the type of person who doesn't question very much, right? So, I fight whole fields of academia like this, and I knew immediately when the pandemic started because I had done a little tiny bit of work in epidemiology back when the servers back when the STARS, thing happened, and I just realized, like, there's so complicated that's getting to, really, you know, high confidence, high conviction predictions would be extremely hard.

Like, like the thing you mentioned, like, maybe people don't change the number of times they go to the restaurant, but maybe they're very careful, like, covering their faces or doing something differently the way they, the way they grab the, the, the handle of the bathroom door, you know, we're never going to have good models for exactly how this all works.

Right? So, but anyway, the funny thing, like, even like, history, like, I always argue with my wife about this. yeah. I think historians, having known a lot of historians, are people who are selected to be overconfident in what can be determined about the past from the kinds of sources that are available to historians.

Whereas someone else like me might come along and say, like, we don't even really know who killed Kennedy, and, and you're telling me you knew what happened, you're telling me some really complicated thing about what happened in the 15th century in Germany. I don't even know who killed Kennedy, right? So, so but the field is self-selected, so if you like that game, the game is to push your theory of What happened during covid or your theory of what happened in the 15th century in Germany?

Do you like that game? The other people that are more skeptical and careful self-selected out. They don't want to play that game with you. They're doing something else. And then you end up with these communities where they're kind of a little bit impervious to someone like you. I mean, sure, you could easily come along and just grab the most at any given moment during the covid epidemic.

Okay. You could grab the most prominent paper, the most influential people and just rip, and just rip it to shreds, which probably you did. Right. But then you got frustrated because nobody, nobody cared. Right. The epidemiologist. Yeah. So, sorry. I, I just had,

Philippe Lemoine: So yeah, it's interesting. This is very similar to how I think about science. You know, I think people have this idea of scientists as, people who like, you know, which is completely wrong, you know, people are like working assumptions were like assumptions free, very free spirit and stuff like, but you know, I think this is, I think really scientists are much more like plumbers than people realize.

Basically, you go to grad school, you learn what I call those scripts, you learn a bunch of scripts.

Steve Hsu: yeah,

Philippe Lemoine: And those tell you how to write a paper. And there are a number of scripts. And what you do in grad school is that you learn a bunch of scripts. They will tell you how to write different kinds of paper.

And then if you want to have a career, all you have to do is follow the script. And you turn up papers that fall. And as long as they follow the script, you're going to get to publish them, you know, somewhere and, and you will have a career. Um,

Steve Hsu: the

Philippe Lemoine: so,

Steve Hsu: Sorry to interrupt you. But the difference between fields is, I agree with your characterization, but the difference is that in some fields, if you're following the script and nothing much is happening, like, you're going to get punched in the face. So, in other words, if you're doing some.

Hmm. Experiment with superconductors and you say the wrong thing or some, some other lab immediately shows that you're wrong. You know, they're, they're, the incentives are different and the ability of other people to check you is stronger. Whereas I could publish some very influential, but ultimately pretty crappy epidemiology model and it could get thousands of citations, but nobody can really push back on me because there isn't a clean way to test my paper, right?

Philippe Lemoine: yeah. You're never gonna bump into reality in, in many fields or in some fields you will bump into reality and that will act as a check, you know, on this kind of stuff, but, but I do, I think, you know, the script thing is true of every field is just that in some fields. Yeah, you will bump into reality and that will like, they'll provide a correction mechanism that, that

Steve Hsu: These scripts actually work. So, in other words, like, if I say like, oh, genotype all these people, then do this machine learning process, then publish this predictor, and then we'll take the predictor to Australia, and it works, right? So, some of these scripts actually work, some of the scripts don't work. And

Philippe Lemoine: Yeah. And, but what's, but the point you were making is, is that in many fields, we have no way to know whether it works and that's why, you know, it's, it's

Steve Hsu: That's the problem.

Philippe Lemoine: and

Steve Hsu: Which is why, like, for myself, like, when I started graduate school, I thought I was, I was originally intending to do string theory. But as I learned more about string theory, I realized this is not going to be testable in my lifetime. So, in a way, it's almost kind of not, it's sort of borderline not physics, because unlike other areas of physics, it's not very falsifiable, at least in my lifetime.

So, I just said, I'm not going to work on string theory. But we have a luxury in physics of being able to select between, you know, things that are readily testable and things which, okay, maybe in 20 years this will be testable, maybe never, right? But in other fields, it could be just never for everything, right?

Philippe Lemoine: yes, that's, I have a pretty, I guess I'll come to this. Let me finish on the plumber thing. So, I think, you know, people, it's much, it's a lot more like learning a trade. Okay. Then people realize, I guess, is the point I'm making. And so, but, but in many fields, and so this actually relates to the, the way the thing you were just talking about is that in many fields, we have no way of knowing that following the script is actually truth conducive.

Precisely because we don't have this response from reality, you know, like you can't check a lot of the stuff, you know, like Even stuff where it looks like, you know Like recently there was those economists who were doing like a victory lap claiming that they had got The, you know, the, because they had written this paper showing that, you know, the economic effects of, on Germany, of the Russian gas cutoff was not going to be, was going to be fairly small. And so recently they were, and I guess, you know, if you compare what they're saying to what some people in the industrial unions and trade unions were saying, then sure, I guess they are right to do a victory lap, but that's because what those people are saying was insane. Now they're saying there's going to be a drop in output of like 10%.

Okay.

Nobody seriously believed that. I mean, I get, you know, I get what they're annoying, you know, because it's true that in the press in Germany and stuff, you had those apocalyptic predictions. So, in a sense, it would be wrong for me to say, even though I do think it's true that nobody seriously believed us, they were featured prominently in the press, and they were part of the public debate.

So, okay, fair enough. But, you know, but nevertheless, you know, beyond that. We don't even know what the effect actually was, you know, they're talking as if we could measure the effects on output of the Russian gas cutoff. And so we could check this measured results against their predictions that they made, you know, before the cutoff, but, but we can't measure it, you know, because it's essentially a counterfactual claim, you know, like, so you, so you just can't, you know, like you can look, you can look at the forecast, the difference between the actual growth and the one that was forecast before the cutoff, but various forecasters like the IMF or whatever, but even that, you know, doesn't measure the, the actual effect, because there could be a number of reasons why the forecast was off beside the Russian could have, I mean, it's not as if like economic forecasts were known for being particularly accurate all the time, you know?

so, so, you know, this is a thing where you don't have this kind of like, you know, it's like you said, it's like you, you're like, you know, testing a superconductor, it's, if it's not going to work, you're going to be able to see it in a fairly straightforward way. I mean, I'm sure this is a simplification that in many cases is not as straightforward, but, surely, you know, it's a matter of degree, but like, it's going to be much more straightforward.

Like in this case, you can't escape that even to measure, to estimate the actual impact, this is going to be model based and it's going to be crucially hinging on the virus assumption you make in a model, but. Why should I trust the model? You know, like, and you go back to square one. So, and it's not just that in many cases we don't have, we don't have any good reason to believe that following the script is actually truth conducive.

It's worse than that because in many cases, we actually, we actually have good reasons to think that it's not truth conducive,

Steve Hsu: Yes,

Philippe Lemoine: but people don't care because that's just, and it's exactly what you were talking about earlier.

Like, and so in the case of, epidemic modeling, that was one of those cases where I was like, look you have this explanation.

And I, you know, there are clear anomalies in the data, like this stuff about, you know, even if we assume that it's like voluntary behavior, they explain the, you know, the, the data we have on this, which again, admittedly is like very imperfect, in many, many cases, you see like, you know, changes in, in, in the transmission rate that don't seem to be associated with any kind of behavioral change in one direction or another. so, but nobody cares, you know, so one of the things I did is, because I was really bothered by this. I was like, okay, this is, this is really weird, you know, like, and, and I wanted to understand it. And so, I figured, okay, I, I think you're right. You know, I think ultimately. we cat, no, because it's just too messy.

And, we also have to learn the data we have is too poor and like, it's just too complicated. and so, you know, if I'd been, you know, one thing I explain all the times that the incentives are fucked, because if you're, if you're a cop, a smart enough person to understand the, those methodological issues, and if you're intellectually honest and a journalist or some government official will ask you, to say, what's going to happen if I do this?

What's going to happen if I do this, if I do that, the honest answer should be, I don't know. But of course, if you do say that. You know, they're, they're not going to, if they're a journalist, they're not going to invite you again, you know, if they're a government official, they're going to find someone who will answer the question, even though he won't know any more than you.

In fact, chances are, if he doesn't realize that he's not in a position to know, he knows less than you, because at least I know that I'm not in a position to know. And apparently, he doesn't even know that. So,

Steve Hsu: Almost, in almost, in almost every human endeavor, the incentive is toward being overconfident, right? Or right? Yes. If, if

Philippe Lemoine: yeah,

Steve Hsu: if you say, I dunno, the journalist is not gonna quote you, you're not gonna become famous guy in your field, et cetera, et cetera, those fields where if you are overconfident someone punches you in the face right away, those are the only fields where the incentives are, is flipped everywhere else.

The incentive is to be somewhat overconfident in what you say.

Philippe Lemoine: So that, so that was, that was the case over there and like in the case, but you know, Even if you're not in a position to say what's going on, I think you should at least, so that's what I try to do, is that acknowledge, A, acknowledge the anomaly or anomalies, and B, try to come up with at least some possible explanation, even if you will never be in a position to verify that those are the actual explanations.

So, one hypothesis I proposed, for instance, was that, to those models that people are using, they assume that human populations are what they call homogenous mixing. Which essentially means that they model human population in which the virus spreads as like an ideal gas in physics. Like a bunch of molecules that randomly bump into each other and that's how they, the virus spreads.

Of course, this is like a completely unrealistic way of modeling the way people interact with each other. And those interactions that give rise to viral transmission. And in reality, you know, people interact, you know, the people's interaction forms a network. And, so it's not random. And so, what I did is like, look, what's going to happen, you know, what I show is that if you, if you model the spread of the virus on a, on a network, then you can see that depending on the topology of the network, the epidemic behavior can be very different.

Steve Hsu: Yes.

Philippe Lemoine: And so, if you have something that's called community structure, we're basically in the overall network. You have some parts of the network that are internally very well connected, but those parts are very poorly connected to each other. What can happen is that the virus is going to seed, you know, it's going to be seed, one of those parts of the network that are, you know, very well, like community is going to be very well internally connected, but poorly connected to each other is going to be seeded by the virus.

The virus is going to spread easily in that part of the network because it's very well internally connected.

Steve Hsu: Yep. Mm-Hmm.

Philippe Lemoine: herd immunity is going to be reached locally. And then it's going to die down, and then what's going to happen is that you have a certain probability, which is fairly low, that it's going to jump to another community, in which case, you know, it will die down.

rise again, or it will just die down because, in that part of the network, you know, herd immunity will be rich. And so, this is one possible explanation for why you would, you had this like cyclical behavior, when in May, including in many cases, because there were cases where behavior like, could be measured as changing, but there were many cases where it wasn't and yet you still had this kind of behavior.

So that would be, that's one possible explanation. And then what I did is that I simulated fake epidemics, assuming that the virus was in fact spreading on this type of network. And, and I showed. And then what I did is, I had fake data based on this fake epidemic of a virus that spread on this type of network.

And then I did, okay, what happens if we use the, the, the, the, the econometric techniques that people are using, in fact, you know, to study the effect of virus government interventions on the spread of the virus on those data. And what I show is that, because implicitly those econometric techniques, so that's an interesting thing where people, those assumptions are almost always left implicit. But if you think hard enough about it, you see that for the methods to recover the causal effect of those interventions, you have to assume that the virus spreads on a homogeneous mixing population. And okay, if the network has a topology such that there is, there isn't like this community structure, it's going to be a good enough approximation.

Even though it's not homogenous, really homogenous mixing, it's not going to bias your estimates of the causal effect of those dimensions too poorly. But when you run those methods on those fake data, that is obtained by running an epidemic on a network that has a structure. What you see is that then you get estimates of the causal effect of those interventions that are really, really, really badly biased.

Steve Hsu: Yeah.

Philippe Lemoine: And so, and so, you know, again, but we don't know what the topology of the network is. So, you know, if you're intellectually honest, you have to say, okay, I mean, you can use those techniques if you want, but You know, you don't, the validity of the conclusions you get rests on the assumptions that the network on which the virus is actually spreading has a certain topology and we don't know that it has it, you know, and, and so, you know, that's, and that's just, and of course, you know, in a sense, it's a negative result because, and so it's the kind of thing that, as you were saying to earlier, like nobody at least wants to hear, because it's kind of like you're ruining the party and saying, yeah, we can't, You can't know for sure, but

Steve Hsu: it takes a very mature and on average high G sub field where people appreciate negative results. So, in math, you could prove a, like a no-go theorem. Like it's impossible to prove this or like, as you're familiar with from logic, but, or in physics, oftentimes that you, if you, if you show that, you know, there's no way to do better than this particular bound in, in, You Are characterizing something through this experimental technique.

People actually appreciate it because they understand, like, it's actually valuable for reasoning. But, in most fields, look, just to be honest, in biology, not a single person understands this stuff, basically. So, you won't get credit for negative results in, in, in most, almost all fields, actually. Oh, there's only a few

Philippe Lemoine: Yeah, so my, my, my, I have like this taxonomy of scientists.

If you look at, based on my experience of love with epidemiologists, so you have, I have like this three, I, I, I distinguish like three classes of epidemiologists. And this is based on like. what I've observed, like the kind of paper they write, the kind of emails I've received because I did get like a lot of emails from various people, you know, will tell me things if I emailed that they wouldn't say publicly often and also the response to the stuff I write, you know, I wrote about this, exactly like how they were, how receptive or non-receptive they were to those kinds of like negative results they were, and so I think you have one part of those people are just dumb.

Like they don't even understand. They don't understand, you know, the. The limitation of those methods, because they're literally not intellectually equipped to understand it, understand it. And I mean, I've had, I've had like surreal conversations. I've had, there is this woman. So, to be fair, she's not an epidemic model.

She's an epidemiologist, which, you know, as you know, it's like this huge field where people do completely different things and she doesn't. I don't think she worked, I hope anyway, that she doesn't work on anything that's math intensive, but you know. I had this conversation where she could not understand like high school level math.

I know it sounds like an, like I'm shitting you or something. I'm exaggerating, but no,

Steve Hsu: no, I'm not, I'm not, I'm not surprised. I've had these experiences

Philippe Lemoine: assure you it's true. And so, okay. So, you have those people, but, but this woman was invited constantly on, on, on, you know, by television channels in France, et cetera. And she would opine about this stuff. And I confidently say stuff that she was in no position to say, it's crazy.

But like, so, I mean, I didn't have a very high.

Steve Hsu: She's the winner. She's on TV. People,

Philippe Lemoine: She's, yeah, I mean, she's literally the winner. She actually got a gold medal, from our institution, from our institution for, for our work. I kid you not, you know, it's like, it's, she works at the most prestigious French biomedical research institution, the INSELM, and she

Steve Hsu: Honestly, Philippe, this is why some people are praying for the arrival of AI because, because the AI might not be this dumb. The AI may be able to read all these papers during the epidemic and then advise the president. Actually, we know very little, like we cannot actually extrapolate confidently.

Philippe Lemoine: Use a, use a very simple heuristic and don't trust anyone who has like those. But so. So, you know, you have, so you have one class of people, they're, they're just dumb, to be honest. I'm sorry, but there is no other way to put it. They, they limit, they literally don't understand the limitations of their method.

They just apply them mechanically like a bunch of automatons and they don't understand the conditions on which, under which those, will be reliable and the conditions under which they won't be. Then you have a second class, which are people who are smart. So, the problem here is not that they're stupid, but they're not, in part because the incentives are bad and in part because of their character, I guess, they're not very inclined towards those like methodological questions.

So, they will write papers, they will understand the limitations of their method at a very high level of abstraction. You know, they will write the limitations section of their paper. They will say all the right things about it, et cetera, but they won't really internalize how deep the problem is, how, how seriously this undermines the reliability of their conclusions.

You know, they will just go through the motions of writing this little section, you know, stuff. And so, but they will never win, in part because they're, it's not in their character, but also because they're not incentivized to do this, you know, like you don't need to do this to get stuff published.

And in fact, if you do, you're going to waste time, you know, and get less stuff published, and you're going to get scooped. With your fake result by another person is going to publish the fake result before you so, you know, because of that, you know, they just won't really think very deeply about those issues and because they won't think very deeply about those issues, the results will be that they won't have a very deep understanding of how serious the problem is.

And then you have a third class of people who are both smart and inclined to think deeply about those issues and to care about them. Despite the bad incentives and those people will understand exactly how fake all this stuff is unreliable is, but look, you know, they have only two, two, possibilities, you know, two possible choices here at the exit of the field, which is what you were talking about.

I would say that those people tend to select out and I think you're right. You know, often this is what happens, but I think in other cases, they won't leave because look, I mean, now they're in the field. You know, their whole life is organized around this. So they will just like go through the motion, but at least unlike the people in the second class, and especially people in the first, they will understand, you know, but it's kind of like, it must be bad for them psychologically, I guess, like, because they understand the, what they do is like, just, Kind of fake in that sense, but like, but they, and they will, you know, so I got those emails from people who would like send me emails, say, yeah, you know, I'm in the field and like, I mean, yeah, I know this is fake, but what are you going to do?

You know, like, and they're right. You know, what are you going to do? It's just like, there's no, nobody's going to be interested. You're not going to get your career and are not going to get advanced by publishing this sort of like negative result, because this is not, it's not part, it doesn't follow any of the scripts.

And, you know, if it doesn't follow one of the scripts, you, you don't get the stuff published.

Steve Hsu: Yeah, I think your categorization is 100 percent accurate and there are people who are both smart and epistemically careful, but they tend to be a minority in most fields because incentives are against them succeeding. And for them, just think how frustrating it is to spend 30 years in a field where you realize the field is completely noise dominated like this.

And, and you're just like. Gradually eking out some tiny results that you can actually believe in out of all this crazy activity that goes like this, right? That's sort of a sad story. I think in a lot of fields whether it's psychology or climate science or Epidemiology that that breakup into three categories is completely accurate.

Philippe Lemoine: So, yeah, so, you know, that's, that's kind of what I did a lot of, I mean, so I was already, those are ideas that at one level or another already had before the pandemic, but the pandemic really radicalized me in that sense. Like, cause, and I guess, you know, one reason is that I, that in the case of the pandemic, those results also had some direct, in many cases, some direct, like, Impact on public policy.

Most of the time, this stuff, you know, it's not that it has no impact, although often it doesn't have any, but even when it does have an impact, it's much more indirect. You know, it's like, very downstream of the actual research, and it's more like the accumulation of similarities. Results that, changes the culture of people, you know, in public policy and like affects the actual policies are implemented and, but then you get all the messy stuff of like democratic or non-democratic as the case may be, decision making that, you know, will also like Effect policy often much more than this stuff, et cetera.

So, but in this case, you know, like you could see people would like, like I would get, you know, in France, we got some pretty strong, like measures, you know, like, like, for a month, you know, for like six months, more than six months, I was you know, there was a curfew in place, which, you know, it's a curfew.

It's not some anodyne thing. It's like, it's the type of things we do during wars, you know, where after 6 PM, I would be considered a criminal and be fine if I left my home. And, And you know, everything was closed, you know, I couldn't do like several times I forgot to buy my groceries in time and then I was like, I didn't see the time, you know, and I was like, fuck, it's six, you know, I have nothing to eat because it's just like, so you can order, you could order food, you know, because the, at least after the first lockdown, you know, the, the delivery service still work, but it's just like, it was a crazy thing.

And then I had people who would go on TV, you know, scientists. We do like press conferences with the government saying, oh, our model shows that the curfew has saved, you know, X number of lives. And I was like, are you, are you serious? Like your model is total crap. You know, like anyone who knows, understands, really likes the most basic statistics, you know, can tell that this doesn't show anything.

But you're using this, look, I'm not even saying, look, if you want to defend the curfew on like prudential grounds or whatever, I think it was crazy at this point, but, okay, I, you know, at least, you know, we can agree to disagree on this, you know, but like, that's not what they were saying. They were pretending that those completely garbage methods, like demos scientifically demonstrated, whatever that meant that, you know, this was effective and was saving dozens and thousands of lives. And this is just like, this is either like total incompetence or, or dishonesty, intellectual dishonesty. In that case, I actually think it was intellectual dishonesty because the woman I have in mind, it's a different epidemiology and this one, does epidemic modeling.

And I've read some old papers she wrote like. 15 years ago, that, that tells me that she understands those, those, complications, you know, so it's just, but yeah, of course, but then she would get, she'd get to be on TV to give a press conference, you know, to get the sanction of the government, that you get money for a lab, et cetera, et cetera, you know, just, you're just playing the game, but the game is crap.

And so, so I guess he had, I already had like most of those ideas about, you know, the problems with social science before, but this truly radicalized me like in. In a way, you know, that's, previous like, issues, you know, I've not, so.

And so, to continue with what I'm up to, now I'm working more.

So, you know, COVID, as I said in, during our pre-recording, like discussions, I've kind of like put this behind me because I spent too much time talking about it. You know, I'm still, as I was telling you too, I'm still really interested in those like meta science issue issues, you know, this is because this is not like covid specific.

COVID was just like, for the reasons I just outlined, a particularly striking example of the kind of damage that this stuff can do. But those are issues that, you know, will stay with us even now that

Steve Hsu: Wait, wait, wait till you get into climate modeling or some other things

Philippe Lemoine: So, I never, I never did. I never did that, but, uh,

Steve Hsu: Don't do it. Don't look don't look too carefully Angus

Philippe Lemoine: So yeah, so, I, you know, those issues are still interesting to me. And also, because they're like, you know, they're just related to like, they're really interested, they're connected to really interesting, epistemological issues. So, I think the next thing I'm going to do, I'm get back into macroeconomics.

I'm really more interested in it, yeah. So, you know, recently there was this famous Nobel prize economist called Deaton. I forgot his first name. Angus Deaton. Yeah. He published this paper that was very, like he criticized, you know, the profession of economics and he got a lot of it from people and you know, I think some of his arguments are just not convincing. But I think he makes a point that's just true, you know And nobody's like saying wants to admit that it's true or you know Like he gets a lot of shit, but I haven't seen a single person say by the way this is actually a good point that you know, I mean even though some people have made this point in the past It's not a novel point.

The point he makes is that now, we've moved recently. There's been this thing that the economists call the credibility revolution and which they are very proud of, and it's basically, it's, it, what this describes is the shift to like quasi experimental methods. estimate the causal effects of various things in a way that's like where you have more credible identification than if like the type of structural models that they used to in the heyday of like the Kohl's commission, that type of stuff.

and it's true, you know, that you have like more credible identification, although I think often even those methods have their own problems that are kind of like underestimated, but okay, it's certainly More credible than like the previous methods in general, but the point he says the point he makes, which I think has not been acknowledged enough is that those methods focus the intention of the profession on local effects at the expense of looking at more what he calls like slow acting mechanism.

Steve Hsu: Yes.

Philippe Lemoine: That can be really important, but that will, like, act with a lag and over, like, a very long period of time. And those methods are not going to tell you anything about this. And, you know, there are some arguments that, that some people make, like, Angrist, who coined the term, like, the credibility revolution, I think, made this argument that, yeah, you can use a method, you can still get generalized based on those methods by multiplying the studies that use those methods in different settings.

But I don't think that you can really generalize a lot from this. and you need a theory that the problem is that you need theory to be able to support the kind of policy relevant counterfactuals that people are really interested in because ultimately that's what we care about. You know, it comes that we want to be able to, to know what would happen if we did this, or if we did that, you know, and, and you need a theory.

But the problem is that. To empirically validate this theory, when you try, you will run into the same kind of identification problem that motivated the shift to those quasi-experimental methods in the first place. So, you get trapped basically because you have on the one hand, if you, if you want to have like the, the kind of like more general theory that, that can support the policy relevant counterfactuals, you, you can't get credible identification estimate causal effects. And on the other hand, if you, you know, in response to this fact, you shift to like those quasi-experimental methods. Then, you will more credibly identify, you know, causation, you know, and for like, and be able to estimate local causal effects. But this will leave out a lot of questions, which honestly are really more relevant or more interesting to most people than the ones that those can address.

And so you get stuck and, you know, and, and, and, And I thought it was a pity that, I mean, sure, there were some bad arguments in the short piece, you know, in question, but, I think, you know, look, when someone makes like a good and important point, I think it's better to focus on the good and important points, you know, rather than like pick at the, The bad examples he, you know, or the bad arguments he also made in the thing and like, but, but also we did that.

So that's one thing I want to explore more in the future, but at the moment, what I'm doing is something completely different. I've been writing a lot about the origins of the Russo Ukrainian war. So, when I was, you know, studying political science, you know, that's cool. I guess, you know, I specialized, you know, in international relations.

So, I've been interested in that stuff for a while. And so, and I've been particularly interested in post-Cold War, Western Russian relations for a while. And, and so I think, you know, this is another, it's funny, you know, you have, it's very different, of course, from like, Epidemiology or economics or sociology, it's more like history and political science, but the non-quantitative type.

but I also think that, you know, this is like, there are two, there is this narrative that I think is very simplistic where, and then, so I think the. You know, basically the standard narrative is that it's what I call eternal Russian essence theory, where, you know, the Russians have always been bad and like to invade the neighbors and it's in their DNA.

and so, you know, that's, that's just like their natural expansionism. And that's what explains everything. And I think this is wrong. You know, it's not to say that There aren't strong imperialist undercurrents in Russia because there definitely are. but you know, it's just like real history is more complicated than that.

And basically, I think that the West in general and the U. S. in particular have had very stupid policies toward Russia in the post-Cold War era. And I think we wasted an opportunity. I don't think so. I don't think, you know, it's plausible that everything would have gone perfectly well between Russia and the West, and we would have been like best friends forever and that sort of thing.

That's not what I'm saying, but I thought that if we had been smarter about it and if they'd been smarter about it, but you know, it's a two-way street and we only focus on what they did wrong and never look at what we did wrong. we could have reached a modus vivendi that would have avoided this Total catastrophe that is going, because it's terrible.

Honestly, this thing is like, obviously it's terrible. The worst affected of course, is Ukraine was going to get completely wrecked. I mean, I think at this point it is already certain that no matter what happens next, they will not recover demographically. For the rest of the century, at least, but it's a complete disaster for them.

It's a disaster for Russia, which is becoming a weirder and weirder place. It's such a tragedy. I mean, both of those places have huge pools of human capital, potentially at least. That's, you know, not just them, but the rest of the world could benefit from it, and it's just going to waste because of this stupid shit.

And, and it's, it's terrible for the rest of the world too. I mean, for Europe, we're going to suffer economically a lot from this. I mean, I know there's this magical thinking that for some reason, now that we don't get the Russian gas, it's going to be even better for us. I don't know exactly what, you know, it's funny.

You have those people who correctly, I think, point out that, say tariffs. will like really damaged, damaged the economic well-being on average. So, then you have, I understand you're like distributional issues that are too often like neglected, even though they're important, but you know, on the whole, they're going to damage the welfare of, of people in like, and, or at least it doesn't even matter if it's true.

I mean, I do think it's true, but what matters is that those people are the ones who constantly make this point about tariff things. And then like, you know, even though the cause is different, the mechanism is going to be very similar or the effect is going to be similar to the Russian cutoff.

We got to, we had a relatively cheap, you know, energy source that no, we're going to have to pay for. And it's going to last for several years because we're going to have to rely on LNG and, you know, it's a different production, more costly production function, basically. And so, it's going to be, it's like a, it's like a permanent supply shock.

It's gonna hurt you a lot. I mean, it already has hurt us a lot, although as it was like, I don't know. I think that was before we recorded like the, or maybe not, you know, like, the, it's hard to estimate exactly how coffee is going to be, but there is no question. That's going to be very costly, but we're supposed to be basically because people are happy that we're cutting them off.

Russia, you know, we're cutting the, the, the, our relations with Russia politically, they want to ignore the economy because we're pretending it doesn't exist. But it's just like, even by their own assumptions, you know, this is going to be large. And even, even for the U. S., because look, there is this common argument that the U.

S. is the only one that benefits from this, and China, maybe. China, it may actually be true, that they benefit from this. Although even that is debatable, in my opinion. But, the U. S., I mean, sure, they get to sell LNG to Europe, I mean, that's, and they get to sell weapons, but you know, in the total, like us GDP, this is very small.

Whereas the effect is going to have, I mean, Russia at the end of the day, still the only country on earth that can literally wipe the U S off the face of the map. There is no, not even China, although. Unfortunately, soon it will be the case, I think, but because they're going to build up their nuclear arsenal, but, but it's still right now, it's still the only country that can do that.

And there was this fairly elaborate system of arms control, especially nuclear arms control that had been put in place at the end of the Cold War or starting with the detente. And then, you know. Even more so under Gorbachev, and, and early Yeltsin that that like started to unravel, in fact, before that, and largely because of the U. S. So that's another thing, you know, we, that's part of this whole thing that we don't, that we don't consider, you know, in how we got to this point where, you know, this, this whole apparatus, this arms control system, the, the ones who started to dismantle it was not the Russians. If only because, you know, they arguably benefited most from it because they were poor.

And so, it's harder for them to maintain a large nuclear arsenal than it is for the U. S. And the U. S., in fact, in part for that reason, started to dismantle it, after the end of the Cold War. And now this war is going to completely destroy this. I don't think there was going to be much left of it.

Yeah.

Steve Hsu: an, there's an effective, sort of expected utility calculation you can do where you say. Well, the amount of trust now between the U. S. and Russia is at an all-time high, well, is very low, certainly lower than if we hadn't had this conflict. And as you say, we're dismantling all these safeguards against nuclear

Philippe Lemoine: Yeah. Yeah. I mean, so

Steve Hsu: And so, I just, just to finish, if you, if you just say, like, it increased by a few percent the possibility of the U. S. being wiped off the face of the Earth, well, of course, we will wipe Russia out too, but still, very terrible outcome. Hundreds of millions of people were killed. that totally outweighs anything that could have happened pro or net, you know, positive or negative in Ukraine.

Right? So, if you take that perspective, everything is just stupid. Right? We, we, we, we increased the existential risk, from nuclear weapons over a very tiny little border region. Right? That we could have

Philippe Lemoine: yeah. And it's worse and it's worse than that because it's not just that the bilateral arms control apparatus between the US and Russia is being destroyed by this. Like I said, it had already started to be dismantled before that. but, no, it's the, the, we're basically finishing the job of this conflict.

but the problem is that, look, this bilateral system, we could have hoped, if it had stayed in place, and even if it had been, you know, we can dream, reinforced, then I think it would have been a good basis. To be turned into a military system that would have included China to prevent, you know, because, but now it means that the prospect of this gets even more remote.

Steve Hsu: Yes.

Philippe Lemoine: You know, the U S really benefited because it's selling like a few billion dollars’ worth of LNG and weapons to Europe, thanks to this war. I mean, I find that ridiculous. So, you, I'm willing to listen to an argument that, you know, this couldn't be avoided, although I think it's false. but the idea that, you know, this is a good thing for the U.

S. ultimately, it's not something I find very serious. I also think that in the end, it's, you know, we're not going to win this. I mean, nobody's going to win this, but I don't think that, I think that, look, the, the argument that the proponents of like restraints before the war, I still think that ultimately it will be proven right.

Even though it will have proven much more complicated than anticipated in the sense that the argument was basically, look, the Russians care more about Ukraine than we do. So, if we go to a clash, we will get tired of it, and we'll give up before they do. Ukraine will get wrecked. And we won't, you know, and, and, yeah, I mean, sure we'll impose a high cost on Russia, but again, it also imposes a high cost on us.

So, what's, what's the point. So, you know, it was dumped. We should have, I think we should have done more to try to prevent a conflict in the first bit, because once we get into a conflict, it's just not plausible that we're going to win this in any meaningful sense. In fact, in any meaningful sense, I don't think anyone is going to win this.

I think it's going to be a disaster for Russia too. And for what? Yeah. I mean, you know, they're gonna get like, say they end up having like the four Obelisks that they officially annexed in totality, you know, they are able to conquer the rest of them. I don't even think it's obvious they'll get that, but it's possible. I mean, most of the most productive people on those territories will have left by then. Everything will be destroyed, so it will cost them a shitload of resources. To rebuild that stuff and rebuild it for nobody because there won't be anybody left except like low productivity people will like, you know, and then, you know, they will get the sanctions because we're not going to remove them because now it's become a moral issue.

You know, people don't think of this in anything like utilitarian terms. They don't see the sanctions, for instance, as like a tool, something that can be lifted to get Ukraine a better term for which is, I think, how we should think about them. But I don't at all think that this is how most people think about them, and I don't think this is how they're going to get used to it.

I think if you try to live them though, you're going to have like a whole concert of people telling you, oh, you're rewarding aggression, blah, blah, blah. You know, it won't even matter that this would get potentially, this could get a better term for Ukraine. It doesn't matter because this ceases in moral terms.

It's about punishing Russia. It did something wrong; we need to punish it. It doesn't matter that, you know, I could hear someone say, okay, actually, this would not really help Ukraine get better terms. I think it's wrong, but, you know, we can certainly discuss to what extent it would help. I think it depends on a lot of other things.

so, it's a complicated question, but it doesn't matter. I don't think people think of those sanctions in the way they think of them, like in, you know, very, I say in purely moral terms, but like, I don't think this is really moral. I mean, moral in a very narrow and short-sighted sense, you know, so I think this thing is going to be a disaster in every, in every possible way.

And. Again, my view is that it was not unavoidable. It was not, it was something that could have been avoided if we had been smarter after the Cold War. And so, this project that I'm working on at the moment is like, it's basically a whole history of post-Cold War, Russia West relations. And, my interpretation of how we got to that point.

To the point where this war was even possible and even happening. And I go back to the end of the Cold War. So, I go back to the, the end of the, the negotiations on things like the unification of Germany, the dissolution of the Soviet Union, all the way to, the eve of the war basically.

Steve Hsu: Can I ask you some quick, just quick questions where I'll throw out a hypothesis and then you tell me high conviction, medium conviction, high conviction, it's not true. Something like that. Just tell me whether you think it's true or not and how much, how confident you are. So, Maidan was to some extent the US effort, in order to, weaponize Ukraine against Russia.

Philippe Lemoine: don't think, I don't think that's true for the most part. What I think is true is that I think Biden happened for pretty much purely domestic reasons in Ukraine. But then what I think, I think what I think happened is that once it started, then the U. S. I mean, some people, you know, the U. S. no state is like states are not like unitary agents.

Like you have various factions and stuff. And I think some people within this, the U. S. national security apparatus. So that is an opportunity. To weaken Russia, and, and, you know, but this is complicated. This is mixed with their ideas about promoting democracy. And I don't think in their mind, they really make the difference very often, but.

It doesn't mean that conceptually there's no two different things. And, you know, there has been like going back to like. you know, the grand chessboard type of arguments, you know, people understand both in Russia and the U. S. The importance of Ukraine. And so, this was certainly in the back of the mind of some people.

And so, I don't think that I don't think my then was caused by anything the U. S. Did except like in very indirect ways. You know, it contributed in very indirect ways. You know, it found some NGOs and stuff, you know, in Ukraine that helps the opposition organize and get this thing started. But like, I don't think, I think people in general overestimate how easy it is to influence the internal developments of large countries.

And so, I'm not saying this stuff has no effect whatsoever. I think it was more, I think the orange, I think the U. S. was more important in the orange revolution than Maidan. But because in that case, I think the recount and everything, this was more, those NGOs were more important. In the case of Maidan, I think, no, I think the stuff was organic.

And then after it already happened, then the U S you know, hatched on this and made things worse. but that's a different claim. You see, that's a different claim from saying that they, I don't think they like, okay, it's really the thing is what I

Steve Hsu: Okay. I don't want to push you down this rabbit hole, but just in case you've done some research on this, I would like to hear your thoughts on it. The snipers that were operating in Maidan,

Philippe Lemoine: So, so I haven't read the most recent stuff. and you know, especially I know there's been like the, the trial, like, you know, as concluded, so you have like more evidence, knowing stuff. So, I haven't tried the more recent stuff. I'm still where we were like a few years ago. Because I've been too busy working on doing the rest of the research for like older stuff, but I'm gonna get to this, by the time I get to the part where I discuss Maidan.

but based on where we were a few years ago, I do think that actually it's, I'm not sure you know what happened, but I think there is a non-negligible probability that it was actually in part, at least in part, that at least some of the shooting was by, you know, far right groups were opposed to, Maidan. Opposed to the government and not the government itself. you know, so it's, but like I said, I haven't tried like the most recent stuff, but I do think, you know, there is like, there is a lot like going back a few years ago, so without even taking into account the most recent stuff, there's just a lot of stuff that's troubling in that, you know, people like want to ignore, but look, you know, you know what, it's funny, one of the.

One of the reasons I think that there is a serious probability that something like, is at least partly true, this, this theory, is that what's their name, um,

Steve Hsu: Bellingcat. No, of

Philippe Lemoine: Yeah, yeah, Bellingcat. Is that Bellingcat never did the debunking of the thing?

Steve Hsu: course. A

Philippe Lemoine: And look, I mean, I think that to me, that that's just like, I mean, I'm not saying that's even the main thing, you know, like, but I'm just saying I find that interesting in there. Yeah. so, I think, yeah, I think, you know, look, I mean, Yanukovych was really a bad guy and was, but here too, you know, the standard narrative is too simplistic.

You know, people say Yanukovych was pro Russia and they had this idea that it was fully aligned with Russia, but you know, that's just false. Yanukovych was pro Yanukovych, first of all. And he tried, he actually, in fact, solved the problem, you know, I have this view that if Yanukovych had signed the association agreement, at the end of 2013, none of that would have happened.

The Russians would have been really pissed. They would have taken sanction and everything, but none of that would have happened. The problem is that Yanukovych tried to play both sides. Because it was in his interest, at least in his, if you're short sighted enough as he was, it looked like it was in his interest.

And then got pressured a lot by Russia, and the EU was also really stupid in that whole business, you know. They tried to play the geopolitical game like a bunch of bureaucrats, and so of course they lost. And, But, you know, so he was, he was violating the constitution, you know, like he did all sorts of illegal stuff.

He was gradually turning Ukraine into this weird authoritarian state. I'm not denying any of that, but it's still the case that, you know, people have this idea of Maidan, forget about the sniper thing. And I don't think it really matters to what I'm about to say here, which is the most important part of the whole story, is that, look, people present Maidan as this kind of like revolution where, you know, the whole Ukrainian people, you know, overthrew this like, authoritarian leader, et cetera, you know, and then like for democracy and blah, blah, blah.

But if you look at polls, At the time, you know, and if you look at the sequence, the political sequence that led to his overthrow, what happened really is that there was one side, it was a popular movement, but it was like one side, one half of the country, that was opposed to another half of the country.

But the half of the country that was opposed to Yanukovych was concentrated in Kiev. So, it was disproportionately represented in Kiev. And so, they were able to overthrow him. They didn't overthrow him, you know, in anything like a regular way. I mean, this, like, this was a violent overthrow of a democratically elected government.

I don't say that, you know, I don't say democratically elected in the sense, I know the guy was himself violating the constitution and the law all the time. I agree that he was turning the country into, like this authoritarian, weirdo, authoritarian state. I don't say that as if, you know, Ukraine, Before Maidan was this kind of perfect democracy.

And I'm just saying none of this was regular. Both sides, you know, violated the constitution and it didn't, you know, I mean, people say, yeah, it was, he was impeached, you know, by a regular vote of the parliament. I mean, that's not true.

They didn't have the requisite number of votes. If you look, you know, you have pictures of like Western governments where you see members of parliament.

Of the Rada of the party of region Yanukovych, the party was beaten up by the crowd on their way to the votes and people like, and they're like, oh, no, there was a vote of like 300 something to zero. And I'm like, I mean, the very fact that it was zero on the other side should tell you something here is like, so, you know, my point is like, look, nobody's looking good in that story.

And, however they want to, they present this as this kind of fairytale of, you know, like heroic demographic, democratic revolution, but no, it was much messier than this. And, you know, and it was like one half of the country that won basically using irregular means against another half. wasn't asked to do this, you know, and, and then I think, you know, well, I guess maybe if you want to keep the game going, maybe I should let you ask the

Steve Hsu: Yeah. I have more questions. So

Philippe Lemoine: Yeah.

Steve Hsu: could the conflict be avoided at relatively low cost by some simple security guarantees given to Putin and say 2021 or 2022?

Philippe Lemoine: It depends on what you mean by could, I think it's hard to say at this point. so, I think there is a serious possibility that if we had, if we, you know, I'm not saying that. It would have been necessary to, like, fully agree to the draft agreement that they published in December 2021. but I think that the main thing in that draft agreement, in my opinion, the thing that they really cared about, is, like, some kind of, like, treaty guaranteeing that Ukraine would not join NATO.

And it's hard to say for sure, of course, because we're not in those people's head and the, the ways of the Kremlins are impenetrable. and so, but it's very hard to know for sure, but I do, I think there is a serious probability, possibility that, you know, it would have prevented the war. So, in that sense of, in that, in that sense of could, yes, maybe, I mean, I'm not, again, I'm not saying we could have agreed, we should, we would have had to agree to the whole package.

I think if we had agreed to this, then we could have negotiated on the rest. it's, it's at least there is a serious possibility in my opinion that this could have been done, but that's, that's one sense of could, but then, you know, so in a way we could have theoretically, we could have done that and maybe it would have prevented the war.

I think there is a much higher chance than most people think, in my opinion, that it would have prevented the war. But the problem is that this was not, this was not, really like in practice, this was not going to happen because this idea that, you know, the open-door policy of NATO by then had become completely solidified into a dogma.

And there was just no way the West would ever agree to this. I'm not saying, I don't think it was for any good reason. I think it's stupid. I think this whole dogma is completely stupid because basically everybody knew in the West that as long as Russia would get so opposed, so upset at the idea of Ukraine joining NATO, it wasn't going to happen.

We told them privately, and I think we were telling them the truth, the Russians, I mean, but because of this dogma, we have to pretend in public. That it was a real possibility and we have not to sign any kind of treaty that would have foreclosed that possibility. And so, we, we basically like, we basically forgo a possibility.

You don't even have to say that it would have stopped the war for certain. Maybe it wouldn't have, but it was at least possible that it would have. And we had to. Not even try because of the stupid dogma that in private people will tell you, yeah, I mean, of course,

Steve Hsu: But this, this dogma, you know, it's funny as an American, I would say even among our political class, even within our foreign policy elite. How many Americans really knew or cared about this, this open-door policy of NATO? Most Americans actually would not, if the president had for some reason just decided, yeah, we're going to, we're going to publicly guarantee that they're not going to join NATO because we think Putin really might do something crazy.

So, we're going to do that. I don't think I'm American. I don't think there's any particular interest group other than the people around Victoria Newland or something in the US who really care. So, so is it, is it the Europeans that were enforcing this document? Who was

Philippe Lemoine: I actually, I actually think that at least France and Germany would have been fine with this,

Steve Hsu: Yeah. So, who was enforcing this document? I

Philippe Lemoine: the, it's, it's the blob. I think, I think you're right. When you say, when you talk about Americans, you're right. If you talk about regular people, regular people in general, just don't care about, don't care about foreign policy in general.

Steve Hsu: generals and professors of IR and senators, I think

Philippe Lemoine: I mean, I, I think the, I think in the blob, you know, there's kind of like a confluence of like think tankers, national security officials, academics working on that stuff. I think there is a huge Very strong consensus in favor of that dogma that that dogma by that point had really solidified. And so, I don't think it was just like a handful of people, around Victor or new loan, et cetera.

Although of course those people were like the core of this stuff. So, I think that I agree, you know, if Biden had done it, I'd say like, fuck it. I'm going to ignore all those people and do it. you know, I think in the end it wouldn't have nothing dramatic would have

Steve Hsu: Yeah, he would, he would have to suffer a few editorials against him in Foreign Affairs or something, right? Or

Philippe Lemoine: yeah.

But the,

Steve Hsu: Journal, but that's

Philippe Lemoine: but the thing. Yeah, but the thing is that this is not how foreign policy decision making works, you know, like the president is not, you know, of course in theory is the guy who is the ultimate decision maker, but the way it works is like it goes through all those, you know, yes, that's the way it works, you know, it's funny, you know, I want to write something on my sub stack about this, like about people have this wrong idea about what a democracy is.

Because they have like the classical model of a democracy where policy is determined by the voters and the voters vote so as to maximize the satisfaction of their ideological preferences. This is completely wrong. You know, it's not how people vote. People don't vote to max. People are not ideologues.

Like most, the vast majority of people don't even have a coherence. And stable ideology, really, I think much better model is that people identify with certain social groups and they think, okay, how does a person who is a member of that group is supposed to vote again, that they think, and, you know, so that's why you have things like, you know, in France, my favorite example of that phenomenon is like in France, you will talk to a grandmother and you will talk to her about immigration.

And you'll feel like you're talking to Himmler, you know, it's like, she will say the most racist shit about immigration that you can imagine. And then you're like, oh, so you're going to vote for Le Pen. And then she'll be genuinely horrified that you could make such a suggestion. Of course I won't vote for Le Pen, I'm not a fascist, I'm not a racist, blah, blah, blah.

That, she doesn't, she identifies as a moderate, you know, like a, and, you know, a person like that doesn't vote for someone who's labeled an extremist. So, you know, if, but if she voted to maximize the satisfaction of her preferences, she would vote for Le Pen. But that's not how people vote. And similarly, if you look at the decision making in a way, think it's wrong to think that of democracies as uniquely responsive to, public opinion, arguably, I think what, what differs more between, you know, between democracies and authoritarian, autocratic regimes are more how public opinion is shaped rather than how responsive public policies to public opinion.

And. And if you look at, and in fact, arguably, democracies are perhaps, less responsive or at least governments in democracies are able to be less responsive on some topics to public opinion than in autocracy, because the magic of the election, it's a, it's a wonderful, the election, the competitive elections, it's this incredible technology to provide legitimacy, it's the, it's, it's like the new version of like the divine right, but it's much better.

And, and so you can have a French president.

Steve Hsu: It confers, it confers legitimacy magically like this,

Philippe Lemoine: Yeah, yeah, yeah, it's really magical and, and so that's why a French president can run the country with like ratings at 25%. If you're, if your ratings are 25%, you're probably dead. Yeah, you're out in the best-case scenario in the worst-case scenario, you're dead, you know, so.

Steve Hsu: the election.

Philippe Lemoine: Yeah, yeah, exactly.

But so, so, so you know, and in particular in foreign policy, you know, like in any regime, basically you have ruled you have people who rule and you have people who rule, and democracy is no different despite the, you know, for the people by the people. No, you know, the best way to think about elections and democracy is, I think the, the purely procedural, conception that Schumpeter defended, where it's like a purely, you should think of democracy as a mechanism to ensure the peaceful transition of power, the peaceful competition for political power between different elite groups. And elections, like, fulfill that role. And, in particular, those elite groups are the ones who really determine policy. And the reason why the, the, the, the mass of voters can't, for the most part, is that they're too low in formation. So, they can't, you know, so it's not that they don't, their opinions don't matter because, you know, the big issues like they do create some constraints, you know, if you raise taxes too much, you know, okay, this they can understand.

But most of the issues, you know, they don't have, they're too low information to be like an effective check on, on, to, to ensure this responsiveness of policy. So, in effect, what that means is a policy is determined by this competition between different elite groups. And, and in foreign policy in particular, because this is something that everywhere, not just in the U S people care, particularly not a lot about, then those elite groups are even more free.

To determine policy, in ways that ignore public opinion. And so, in the case of the US, I think that, you know, what people call the blob, which I think is a great concept. I think that this dogma, you know, the open-door policy for NATO, this dogma as like, you know, has had a lot of take up and basically is very consensual in the blob.

And so, since those people ultimately are the ones who determine foreign policy, I think that in that sense. It wasn't possible. And in fact, I will tell you something before the, I think it was on December 17, 2021, that the Russians published this draft agreement before that, I thought it was, I thought it was impossible.

They wouldn't invade, not impossible in the grand scheme of things. I wrote something in 2014. It was like making them. Argument for restraint I was talking about, you know, that if we continue to push, eventually there will be a war and there will be nothing we can do because they care about Ukraine more than we do.

So, it's not as if I ever thought it was impossible in principle that Russia would invade Ukraine. But before December 17, I thought there was no chance it would happen now. You know, And December 17 is the one time where I started to change my mind. I still like it. Until the last moment, I thought it was more likely than not.

They wouldn't do it to be honest, but December 17 really is the thing that changed me, that made me take it seriously. The possibility is serious for the first time. And the reason is that surely the Russians have to know about this dogma. They have to know that Biden is not going to do this. So, if they publish.

And, and the fact that they do it publicly, because, you know, when you do this stuff privately, you give yourself more room for like maneuver, you know, like, you know, negotiating some stuff, you know, compromising and stuff. But the fact that they do it, they publish, they make this stuff public. I was like, damn, I mean, it, I still think it's crazy.

It's still thought it's crazy that they wouldn't invade, but like, it really looks like what they would do if they wanted to have some pretext, because they know very well that Biden is never going to accept that. Again, I'm not saying it's for good reasons. They had to know. I may be wrong about this. Maybe they didn't know it.

I've heard a theory which I've heard from someone who heard that from a high level, fairly high level Russian official. And his theory is that after Biden left, withdrew from Afghanistan, the Russians interpreted this as some new era where they would be ready to like, reject some very old dogma, like the open-door thing.

And so Maybe they really did think that he could do this. Honestly, I find that hard to believe though. I think like I, so at least at the time, you know, that's what I thought. I remember thinking it's still so crazy for them to invade it like that, I don't believe they're going to do it. But now I don't know what to make of this thing because it was like another anomaly.

Like we're talking about anomalies, you know, and it was like, it was a big anomaly for me because I was like, I mean, why, why would they do this? Why would they publish this stuff? Why would they make it public and then risk losing face? If, as was certain to happen, Biden refused to, to agree to this, then they would lose faith.

And then what would they do? I mean, you know, it seemed to me that I still didn't think they would do it, but I was like, that's, that started to make me take the hypothesis a lot more seriously. But

Steve Hsu: Bye. I think. Sometimes when you watch Putin, I mean, obviously he's a professional, but sometimes I think some of his emotional or psychological feelings leak out. And I do think he felt for many years, the US just was not taking them seriously and ultimately just felt he had to do something very drastic in order to basically show the Americans that he was serious.

So, um,

Philippe Lemoine: so, Putin, Putin has this theory, which I think is basically correct. for, if you look at the first, say, eight years of his presidency, where we see at first, the first year it was Clinton, but then it was Bush. And people have forgotten, but Putin and Bush were really good friends. I mean, they were genuinely friendly.

Like it wasn't just fake, just for the job, et cetera. They really liked each other. at a personal level. And so, and Bush, he really wanted to get some, reach some modus vivendi with, like, Russia. And Putin believed that, and I think he was right to believe it, because it was true. But what happened is that, and this was particularly clear in the thing that really made him like to convince Putin of that, which I think is the anti-missile controversy where Russia really bent over backward to try to find a way that would satisfy, that would both allow the U.

S. to have this anti-missile system, defense system without, you know, but like including Russia and the stuff so that it would, assuage, you know, Russia's like security concern over, over this thing, which frankly were not unreasonable. Yeah. They were, as always, exaggerated, you know, they inflate the threat, but everybody does that.

This is the result of, like, the anarchy of the international system, you know, like, it makes people paranoid. And the Russians, perhaps, are particularly paranoid because of their history, etc., but, like, fundamentally, they're not that different, you know, and, like, in the U. S., would not react kindly to that kind of stuff on doorstep, either. You can argue that it's irrational and stuff, and I agree, to some extent, it is. But it doesn't matter that everybody, at least every great power is like this to some, to a large extent. So anyway and, and, and , Bush was really receptive to the proposals that the Russians put forward and some other people in the administration were receptive to it.

But then what happened is that. Between the times, you know, the, between the time between like Bush initial reaction, fairly positive reaction to the initial proposal and the time when they made a counter proposal, it had gone through the blob basically, and they just destroyed the stuff. And, and he told, and Putin actually told, so this is like, you can read that in the Philip Short's biography of Putin, which is really good.

And, Putin told Bush, Look, George, you and I, we want to reach an agreement. We, we want Russia and the U. S. to, to make some kind of deal and everything, but the truth is that the, the people who work for us, they don't, you know, and he was saying on both sides, you know, he was saying that about the Russians too, and he told Bush, you wouldn't believe the crazy proposals directed at the U. S. that people put on my desk every day. And he says, and for you, I know it's the same for you. And that's what happened. And he was right, you know, like it was. So, he's convinced, basically his conviction is that, you know, the U. S. president doesn't make foreign policy. And he is right, you know, like, it's the blob, and the blob doesn't like Russia.

And, and, and I mean fundamentally this is right. I think the conclusions he drew from this, we’re crazy and will harm Russian in, in, in the long run, let alone Ukraine and everybody else, but, but he was right about this in any way, that's his conviction. And, but I, and, you know, given that this was his conviction, I find it pretty hard to believe that he thought that this December 17 like draft agreement is something that Biden could ever agree to, because this is a conviction that he reached by the end of Bush's second term. And so, I see no reason to think that he changed his mind in between this. In fact, I see every reason to think that this view was reinforced by what happened subsequently, like Syria in particular, Libya in particular, and also Syria, but in Syria, and they stopped it.

But, so, yeah, um. I, you know, I, yes, I think I do think that something like this is a true, but I also think there were some, you know, what I'm basically my interpretation of like the proximate causes of the war is that in 2014 after my then, he annexed Crimea without thinking a lot. out of a combination of fear and opportunity, opportunism, fear that they may lose the, the Sebastopol base, Naval base

Steve Hsu: Mm hmm.

Philippe Lemoine: new government and opportunism because he thought, you know, he had after his reelection, his popularity was somewhat down after the protest that was in 2011, et cetera, you know, the election was rigged and everything.

he thought that you know, the, the, he knew that this would be popular among a lot of people in Russia. And so that was like the bonus. And so, it was a combination of fear and opportunism in that sense, you know, but he didn't think carefully about the long-term consequences of this and the long term, and then, you know, he lets weird people tear up shit in, in Donbas.

Steve Hsu: Mm

Philippe Lemoine: And then realize, you know, he figured, I think I'm conjecturing here. I do think this is the best, the most consistent story based on the available evidence, which admittedly is limited. So, I'm not saying I can be sure of this, but my interpretation is that he figured, okay, I'm going to let those people like, Gherkin, you know, Strelkov stir up shit in Donbas and who knows, maybe the same thing will happen.

That happened or something similar happened in Crimea, and I can get this as a bonus too. But then fairly quickly realized that this, it wasn't the same thing because the population was very different in, in Donbas in Crimea, where in Crimea, you know, the support was much stronger support for Russia, at least for annexation by Russia than in Donbas, where you know, the, it was something, if you look at the polls.

The last polls that were done before it was, it became impossible to do, to make, to do polls over there. You had like, at, at, at best, one third of the population that was in favor of annexing Russia, but not, not the rest. So, you know, that's why it didn't work out. And then he got scared, like in April, he realized that, you know, this wasn't working and that if it continued, He would endanger his relationship with the West and particularly the Europeans to whom he was selling gas.

And so, he tried to stop it. But those people, you know, it's very interesting to read the memoirs by those separatists over there. They were really upset by what he, when he tried to stop their referendum for the, you know, in the early May of 2014. They saw that as a betrayal, which in a way it was, you know, except that Putin is not a Russian nationalist.

It wasn't really one of theirs, despite what they might have thought. And so, they went ahead anyway, and then, you know, Putin clearly wanted to stop this because he wanted to preserve his relationship with the West. He was like, okay, I got Crimea, I'm going to just keep that and that will be it. And I will, I will keep my relations with the relationship of the West and more or less intact.

And, but then, you know, the Ukrainian military eventually reacted, and they started to defeat the separatists. And then he got caught because it was like, you know, it was hard for him politically, although he could have done it. And I think he should have done it to let them be defeated by the Ukrainians, but he couldn't bring himself to do it.

So, we intervene, you know, on the down low and. And then, you know, try to feed it. And then basically what happened after this is that the problem is that by taking out most of Donbas and Crimea, it took out the most pro-Russian regions of Ukraine, which completely changed the balance of political forces in Ukraine and made it inevitable that Ukraine, it left its own device, devices would go west.

Which is precisely what he didn't want to happen. But he hadn't thought about that when he annexed Crimea. Because again, he did it like it was a spur of the moment kind of thing. Of course, they had military plans. That much is clear because of the way it happened. But this, everybody has like military plans for pretty much any possibility.

That's not really surprising. So, and then for years, he tried to find a formula to square the circle. Meaning, preventing Ukraine from going west. Without having to go to war to do this, to like full blown war. Because again, the Russian army had already intervened in practice, but without, he never, at the time, he didn't admit it except for Crimea.

and so, but it didn't work because look, the Minsk agreements, the truth is that because people say Ukraine like violated them, it's true they violated them, but of course they violated them because the only reason they signed them like at gunpoint basically, that's the only reason they signed them.

So, and I, I think, you know, to, to finish like my interpretation of the proximate causes of the war is, I think that when Zelensky was elected, Zelensky was elected on a platform that, in addition to fighting corruption, was also to make peace with Russia. Although he didn't know, he didn't tell how he was going to do this.

And I don't think there was a realistic way for him to do this given the constraints he faced. and especially the nationalist pressure in Ukraine at the time. And so, but, but I think Putin, you did the, did hope that he would work, that he would work out something with Zelensky and that Zelensky was going to help him square that circle. And so, you know, he gave him a deal on gas early after his election, if you can tell that he was really hopeful that this would change the relationship and would help him square that circle. But then, it didn't happen, you know, like the Zelensky popularity crashed, especially with the pandemic.

of course it didn't, the fight against corruption didn't work out like it never does. and, and so what he did is that he did the easiest way for him to get back up in the polls. Was to go after the, you know, the low hanging fruit was the anti-Russian voters. And so, in the early 2021, it did this like an anti-Russia U turn.

and then I think this is the moment where sometime after this, I'm not sure exactly when. Again, this is, this is a lot of conjecture here. I'm not, I'm not sure. In the Kremlin's head, you know, Putin's head or anybody else, but I think that sometimes after this, Putin basically realized that this was, this was not going to work, you know, like if it's not, if it's not working with Zelensky, it's not going to work with anyone else.

So, I won't be able to square the circle. So no, I have to choose either. I let what's left of Ukraine go West. Or I go to war, and he decides to go to war. I think it was stupid and criminal, but that's, you know, that's, I think that's the kind of calculation he had in his mind at the time. And so, you know, the, the stuff of the, the relations with the West and NATO and all that stuff, I think it mattered more before.

It might have been 2014 because of what happened before. It explains how you could get to the point where you would have the kind of theory we would do this. but after this, I think he got trapped into a trap of his own making and he ended up with this dilemma that I just described and he decided to go for the branch of the dilemma that was war, instead of like letting Ukraine go.

And now. everybody is going to get screwed for first of all Ukraine, but you know, Russia to us to like,

Steve Hsu: What is your interpretation of the negotiations in 2023 involving Naftali Bennett?

Philippe Lemoine: You mean 2022, right?

Steve Hsu: I'm sorry, March of 2022. Sorry.

Philippe Lemoine: Yeah, it's, it's hard to tell, you know, I'm really annoyed by the discourse on this because people on both sides, I mean, of course, in the West, especially the people who think that, you know, it could never have worked, et cetera, because that's by far the prevailing view. talk like, again, with like a great deal of certainty about this, but the problem is that because journalists have been strategically incurious on the topic, it's remarkable, I mean, how little investigation of this topic there has been, given how potentially important it is.

And I think, you know, my interpretation, that's why I say they've been strategically incurious. It's the same thing with the Nord Stream stuff, is that people don't look too much into it. Less they find out something they don't want to find out. And that was a recent, the most interesting piece about this was published only a few days ago by the wall street journal.

So, two years after those events, which is kind of crazy again, when you think about it and they described the deal, the problem is that the, so, you know, according to all the accounts, including the most recent and more, the most thorough we have, which is the one by the wall street journal. Part of the deal that was floated by the Russians, and that was at least in part agreed to by the Ukrainians, included some kind of security guarantee by the West.

Steve Hsu: Mm

Philippe Lemoine: And, and, you know, but because of wall street journal didn't publish the actual, non-paper, you know, like we, we don't have all the details about what, what this means, you know, because it could be something like as empty as the, something so empty that he didn't actually commit Western countries to anything like definite, in which case, you know, they might have agreed to it because there was no cost for them to do so, but suppose it's something as strong or similarly strong to like article five in NATO.

Yeah. I mean, the thing is that I don't think that the I don't think that Western countries would, nor do I think they should have agreed to this. So, if, so you know, there is a hypothesis that the Russians put into the deal a very strong security guarantee, knowing that the West would not agree, and so that they could blame the West for the failure of negotiations.

Steve Hsu: Mm hmm. Ha.

Philippe Lemoine: possibility is that not the guarantees were essentially meaningless as often happens for like this type of stuff. And the reason why Ukraine, like, stopped the negotiations is either because, you know, they were deterred by being dissuaded from continuing by the West, which is like the theory of like Johnson's visits, or, because they thought they could win.

You know, or like it came from, you know, it came from the West and Ukraine, but we don't have information that's publicly available. I don't think we are in a position to. Determine which of those hypotheses is correct. And it kind of drives me crazy that, that, that nobody's looking into.

It's like, look, this is not this, what I'm talking about. Like this is something that in principle could be, if not settled with certainty, at least we could, we could get a much better idea if we look at precisely what those clauses were, you know, people have the document, like the wall street journal have the documents.

So, it's just crazy to me that we don't, but I think that people are way, what I will say is that I think people are way too quick to assume that the Russians were not serious about negotiations. I think there's a fair chance that Putin thought by then that he didn't because people assume he had those maximalist aims, but I think the evidence that he has is very poor.

and, and I think it's, there is a fair chance that he thought he had made his point. And that he could have stopped with a guarantee that, you know, Ukraine would not pursue NATO. I mean, I think it was stupid because look, he would have gotten a piece of paper. No, nothing could have stopped the Ukrainians from like violating yet later, you know,

Steve Hsu: Mm hmm.

Philippe Lemoine: unless, you know, it depends on the detail of, for instance, the limitations on armaments and stuff like that.

That was also included in the deal, but depending on the detail, you know, they could very much have validated that stuff later. So it was, it would have been a guarantee in a limited sense, but anyway, but it's, it's, Look, it's certainly a, you can make this argument. It's a reason why I thought he wouldn't invade too.

I mean, because I thought he had no way of enforcing whatever it would force people to sign. So, so, you know, you can make this argument. This is the same argument for saying he wouldn't have invaded, but he did invade. So that's not an argument to say that he couldn't have been serious about this, this kind of deal.

and so I think, you know, there is a fair chance that he was serious and, you know, recently there was this, article in the New York Times about how, in the fall of 2022, after the successful counter offensive by the Ukrainians that took back a lot of territory, U. S. intelligence got wind of, you know, serious discussions apparently in the Russian government about using a nuclear weapon.

Against Ukraine, if, if, you know, they wouldn't like to stop them from, for instance, taking Crimea. I mean, I think it was unlikely to happen, but, apparently, you know, the U. S. intelligence caught some stuff saying that there was some serious discussion about this. So, they got really worried, as you can imagine.

And so, they sent Burns, the director of the CIA, to talk to Naryshkin, which is like vis a vis in Russia. And apparently, so the Russians agreed to the meeting, and when Naryshkin arrived, apparently was convinced that Burns was here to negotiate an armistice.

Steve Hsu: ha, ha. Yeah.

Philippe Lemoine: And it took a while, according to Burns, it took a while before he could convince him that no, actually, the reason he was here had nothing to do with this is because he wanted to send us, too, to let the Russians know that if they used a nuclear weapon in Ukraine, there would be consequences, but he had to convince him that he had no intention of, like, negotiating a ceasefire at this point.

So that's another.

Steve Hsu: This is a good example of one side really not having a good understanding of the, you know, strategic intent of the other side, right? So, I think that would be more likely to be the case because, I mean, we don't even, I don't even know how our block is thinking about something.

At a given moment, how would you expect the Russians to know exactly right? So

Philippe Lemoine: Yeah. Yeah. Yeah. I think, I think, I mean, you know, that's I mean, look, if I had to bet, I think that I would bet that, I wouldn't bet a lot, but I would bet that, that they are in fact serious about stopping it because, it's become more costly to them than, than is worse, you know? And they, I think like they realize at this point that the cost of like really forcing a decision to impose their terms, whatever those are, and I don't think anybody knows.

I actually suspect that if they don't know the term there, their actual goals would be too costly. And so, they would be open, like, again, they think, they think like, it's, it's, a lot of it is a status issue. Like they think if they can get, if they can force the other side to agree to something like no NATO expansion, explicitly in a treaty, it would, it would like to satisfy their quest for status, which I think is driving a lot of that stuff, even though there are also some security components.

But, and so I, I think, you know, if I had to bet, I would say that he actually, they are serious about it. I certainly think that we should. Look, you know, try, but I also think we won't because, you know, people, if you try people like, oh, you want to reward aggression and blah, blah, blah, you know, I mean, oh, is this going to help you, but and you know, the, the, you would have to, you would have to twist the arms of the Ukrainians.

People are also not willing to do this because at this point, the Ukrainians don't want to. Negotiate yet, it will take more death before they do. And, and look, I mean, there's also a fair possibility that the Russians don't really want to either, or that it will also take a lot more death before they're serious about it.

Cause I mean, you know, of course everybody's willing to negotiate if they can get everything they want without having to compromise on anything. So, when I say, when I'm talking about the hypothesis that the Russians are serious about negotiation, I'm talking about the hypothesis that they would be willing to make serious compromises too.

To make a deal and it's really hard to tell, you know, like, but anyway, I think no matter what the truth is, it's not going to happen because we're, we're not going to, because we're very dumb, you know, I mean, the Russians are dumb, but we are too, which is, I guess we're just less directly affected. So, you can argue they're even dumber than us.

But that's hardly a consolation. If you ask me.

Steve Hsu: So, you know, we're over two hours and Maybe I should not take up too much more of your time.

But how about some long, longer-term predictions? So plus, one year plus two years and anything that you feel high conviction about that you think is going to happen Either in Ukraine or in Europe in geopolitics

Philippe Lemoine: I mean, I, you know, I don't know if I like it. I'm very calm. I wouldn't say I'm very confident in this, but I think I'm reasonably confident that the war is going to last several more years. I think probably, probably at least an extra five years. But I also do think that the intensity will decrease because I think on both sides A, they've lost most of their best men, in the first, like, two years, and, and I think this matters a lot.

And, you know, if you look at World War II, I was reading Blood and Ruin by Richard Auvery. It's like this big history of World War II, like his latest book, which is really good. It's like this huge synthesis from World War II. And he was making this point, which I think is interesting, that other things being equal, the fighting ability of units during World War II, on all sides, decreased over the course of the war, because they lost their best men.

And the new recruits were just not as good, but this was more, other things were not equal because this was more than made up for by the fact that there was like absolutely gigantic war production. And so, they basically substituted labor for capital. Like if we're to use those economic terms, like in the context of warfare.

And so, this more than made up for the loss of other things being equal quality of the troops themselves. And I think about this. In this war, we do have the loss of the best men. We also have the loss of equipment. Like a lot of it, especially on the Russian side. but also, on the Ukrainian side , the war production is modest.

It's nowhere near the kind of absolutely gargantuan levels that we saw during World War II. So, I think it's not going to make up for the loss of troop quality. So that's why I'm, I think that even though I think the war would like to continue for a very long time, my expectations are that the intensity of the war is going to gradually decrease over time.

and I think the most likely outcome is that Eventually it will be something like what happened during the Korean war, where they will just freeze it. It won't be like Ukraine won't formally acknowledge the loss of territory, but they will just agree to freeze it and that will be it.

But I think it will take longer than in the Korean war, because in the Korean war, the U S was fighting directly, which meant that during the ceasefire negotiations, they were the ones pulling the shot on their side. Like the South Koreans, they were basically They had to accept whatever the US was saying, because they were the ones doing, the US was doing the bulk of the fighting with the UN troops.

And even so, it took two years. Remember, it took two years before they could get an agreement.

Steve Hsu: Yep

Philippe Lemoine: So that's why I think that this time, we're gonna have to first reach the point where enough people have died on both sides that they're like, okay, you know, both sides agree that, There's no point in continuing, but what I'm, the point I'm making here is that even after they reach that point, I think it will probably continue for several more years because then they have to find an agreement.

And this won't be easy, especially because again, since the US is not directly involved in this case, they have less leverage on Ukraine. And so, it will be harder. And especially given the environment, the media and intellectual environment in the West about this, the idea that it's okay to twist the arms of the Ukrainians is like really a nonstarter.

and so, it will be hard, but then who knows, you know, maybe if Trump is elected, it will change, but I don't think it will fundamentally change things. My, my, my, my, like, to me, the most likely scenario is this, you know, it would like to go on, although the intensity will gradually decrease, but gone for several years, and, you know, they will reach a point where they both agree that it's time to stop.

But then, if you're not good, the negotiation will drag for years until they get some kind of ceasefire agreement. You know, there is another scenario, which is worse for Ukraine. But I wouldn't rule out either, which is that the Russians, you know, because like the scenario I just outlined is based premise on the assumption that the Russians will not drastically increase the number of resources you allocate to the war, which I think is the most likely thing because it's difficult for Putin to, to do so too.

But there is a possibility, you know, there's a scenario where they do, like to force a decision and attrition works and Brex, you know, like there's just a collapse in the Ukrainian army and then they're forced to accept much worse terms. I don't think even in that scenario, you know, that you Russia, I don't think they can occupy the whole of Ukraine.

It would be too costly. And I think, I like to think that by now they understand how crazy it would be, but who knows, You know, so that's the, that's another, and then, you know, this would be bad for Europe because, you know, there was this dumb argument in favor of supporting Ukraine that, yeah, we're going to, we're going to like, destroy the Russian military on the cheap.

That was kind of like the, I mean, first of all, already based on what we spent already, it's not cheap because people only look, even if you, people only look at the direct like military financial economic assistance, even that at this point is already pretty sizable, but that's not the worst thing, you know, in Europe.

It costs us probably close to a trillion dollars, you know, in like, government stimulus, you know, to, to, deal with the, the, the, the rise of the price of energy. and like, and again, this is a permanent shock, you know, this is going to be. It's going to be like this economic effect is going to be affecting us for a while.

so, you know, there is like in the scenario that where, there is a Ukrainian collapse eventually, which is especially likely, although again, I don't think it's by far the most likely scenario, but it's certainly possible, but it's more likely if the U. S. can get the ad, you know, processed to Ukraine, then we'll end up, instead of having like destroyed the Russian military on the cheap, we'll have like a, a much larger, much better, better equipped, et cetera, et cetera, Russian army closer to a borders.

I'm talking about the Europeans and we'll have to waste a lot of money increasing, you know, defense spending. Look, I mean, it's just like a waste of money, not in the sense that. You know, of course, this also goes to people's salaries and like, you know, it does contribute to the economy, but like this is my resource.

Those are resources that would be much more productively allocated to other uses and, but we'll have to do this because look, even someone like me who is very critical of this like hysteria, but like, oh, we need to increase defense spending right now. If we end up with like, with like a Russia that's armed to the teeth that has been radicalized because we forced it to, to fight like a very bloody war for years and it's right on our doorstep.

Look, I mean, I still don't think it's very likely they would attack NATO, but like, if you're in government, you have to, you have to deal with even an unlikely scenario. And we can't just in that, in that, in that case, you know, we can't just ignore the possibility that it will. So, we'll have to spend a lot more on defense.

And it's just such a waste of resources, you know, just such a complete waste, the whole thing is a complete waste. So yeah, that's my long term, my long-term predictions for this. That's basically, I mean, there is also a scenario where attrition breaks Russia, which I think is also not totally impossible, although I think it's much less likely at this point.

Steve Hsu: Great. Well, you know, if the war drags on, the one good point will be that you can turn your articles into a book.

Philippe Lemoine: Yeah, that's, I'm kind of thinking about it.

Steve Hsu: I mean, if you, if you do the work for these long articles, you may as well just, write a book on,

Philippe Lemoine: Yeah. I mean, you know, the first part is already a small book, really. So, when I put all of them together, it's going to be a big book. Yeah.

Steve Hsu: Great. It's been very enlightening talking to you, Philippe. And I thank you

Philippe Lemoine: talking to you.

Steve Hsu: and, when, when are, when are you next in the us

Philippe Lemoine: Actually, I'm in the U S right now. I arrived a few days ago.

Steve Hsu: Oh, you're kidding. Sorry. Are you in Ithaca?

Philippe Lemoine: no, I'm in DC. I'm in DC until the beginning of April, basically.

Steve Hsu: I see. Alright, well I wish you the best and, I, I'll let you know when this episode airs.

Philippe Lemoine: Great. Thank you.

More episodes

Chapters

Creators & Guests

What is Manifold?