I’m Ezra Klein. This is “The Ezra Klein Show.” This is a great conversation today. But it’s a tricky one to introduce, because the guest I have — I’m not having him on for the thing he’s best known for. So Patrick Collison — by day, co-founder and C.E.O. of the multibillion-dollar payments company, Stripe; by night, by weekend, I think, one of the most important thinkers now in Silicon Valley — certainly, one of the most quietly influential, someone who is forging and traversing an intellectual path that a lot of other people are now following. And it’s this second incarnation and role that I’m really interviewing him in today — the soft power side, I guess, of Patrick Collison. Collison’s work here centers around this question of progress. The argument is that human progress is much more precious and rare and fragile than we realize. We maybe take it for granted. We live in this time when things have been changing, atop decades and decades, even centuries and centuries, even millennia now, when things have kept changing. But for most of human history, that was not true. It was not true. There just was no market rapid advance in human living standards. It’s only in the past 10,000 years, and then practically in the past few hundred — just an eye-blink in the time human beings have been on Earth — that things kept changing, usually for the better. And the question is, why? And Collison’s particular meta question is, given the clear fragility of forward motion here, given how rare it has proven to be — and so how easy it might be to lose — why isn’t the question of the conditions of progress more central? Why isn’t the study of progress in a wide multidisciplinary way a more common and central discipline? Collison has written a few influential essays here, with the economist Tyler Cowen. He called for the inauguration of a discipline — they call it progress studies — and that now has people studying it. There’s people creating journals for it, creating syllabi and podcasts and books around the topic. It’s one of the more singularly successful calls for a research direction I have seen. Separately, in a piece co-authored with the scientist, Michael Nielsen, Collison and Nielsen argued that, though it is hard to measure, it seems like the rate of scientific progress is slowing down, and that’s particularly true if you account for how much more we’re putting into science, in terms of money, of people, of time and technology. Now, these ideas are not original to Collison. The point is not that nobody studied human progress before this or worried about the pace of scientific research. He wouldn’t claim that. It wouldn’t be true. But he is playing a distinctive role in their framing and their popularization, and in creating and funding a community around them. And what I see in my travels here is that it is working. Something is burbling here. But I can’t find many big pieces where Collison really lays out his worldview. There are a couple essays, tweets, interviews, but he’s not been primarily writing this down. What he has been doing is funding it through Fast Grants, which has been successful, but more than that, intellectually influential effort to show you can give out scientific grants quickly and with very little overhead, through the Arc Institute, a big biotech organization he’s creating to push a researcher-first approach to biotech, and through giving a bit of money, and a bit of time, and a bit of prestige, and a bit of networking to a lot of different projects that circle these questions. He’s got this funny quality of being nowhere in particular, but also somehow, almost everywhere, if you’re interested in these questions. So what I wanted to do in this conversation was try to get as close as I could to the Patrick Collison worldview, the underlying theory of the case here that animates his thinking his funding, and the ways in which he’s trying to nudge the culture he’s a part of, or the ways in which he’s trying to actively create a culture he doesn’t yet see. As always, my email — firstname.lastname@example.org. Patrick Collison, welcome to the show.So you’ve made the argument that science — all science — is slowing down, that we’re putting more money and more people into research, and we’re getting less and less out of it. Tell me about that.Well, I want to separate two things. There’s a question as to whether science in its totality is slowing down, in terms of the absolute returns from it. I think that might be true. You can maybe divide up the first half of the 20th century and the second half and so on, and sort of try to compare one with the other. And we had general relativity and quantum mechanics and various other major breakthroughs in the first half. You can ask the question of, well, did we have as many in the second half? But in the second half, we did have the discovery of D.N.A. and molecular biology and lots of other things. So I don’t know that I would claim a total slowdown. The thing that I think is clearer and should be very concerning to us is, as you look at the number of scientists engaged in the pursuit of science, and if you look at the total amount that we’re spending, and as you look at the total output, as coarsely measured by things like papers and number of journals, all of those metrics have grown by, depending on the number, let’s say, between 20 and 100x between 1950 and, say, 2010. And if you look at it on a per-capita basis, or a per-unit-of-work basis, now used to divide all those total outcomes by a factor of 50, and it seems like if you imagine yourself as the median scientist, you’re meaningfully less likely to produce anything like as consequential a breakthrough as you would have, say, in 1920. And so Michael Nielsen and I, in order to try to put slightly more rigor on that question — we went and we surveyed a bunch of scientists across a number of universities in a number of different disciplines, and we presented them with different Nobel Prize-winning breakthroughs. And we tried to compute an approximate ordering of their significance in the eyes of these scientists. And the thing that would kind of have to be true — for the per-capita impact, we remain in constant — is we’d have to be discovering much more important things in the latter half of the 20th century in order to compensate for, to make it worthwhile, for us to be investing this 50-fold greater effort. And we didn’t find that. In physics, in the estimation of physicists, there was a kind of flat-to-declining trend. It’s not super obvious which way it points, but in as much as there’s a trend visible, it’s probably slightly downwards. And in other fields, it was maybe similarly equivocal, perhaps a slight increase, visible in some, but importantly, in no fields that it looked like we’re on this crazy, exponentially improving trajectory, which is what you would have to have for this per-capita phenomenon to not be present. And I think that should be something we’re interested in for multiple reasons. One, because presumably, as a society, we’re interested in just how much more scientific progress and technological progress and so forth, how much more innovation is there going to be over the next 10 years or the next 50 years or the next century. But also, because there’s kind of two possibilities. One possibility is, fundamentally, we’re running out of low-hanging fruit, and it’s just going to be harder to do this stuff. And in as much as we’re setting investment or making investment decisions around to what degree should be pursuing the stuff, I guess it’s important to know what we think the returns should be. Or the other possibility is, somehow, we’re doing it suboptimally. Something changed, and we were pursuing this process of discovery more effectively in the past, and presumably, for inadvertent reasons, something went wrong, and now, we’re just less efficient at it. But either explanation — and it doesn’t necessarily have to be fully binary — but either explanation is important, and either explanation, I think, has prescriptions for what we should do going forward.Let me start with the low-hanging-fruit explanation, which I think is a more popular one. And you have — in the piece you did on this with Michael Nielsen, the sad, but in the very academic way, very funny quote from the physicist Paul Dirac, who says of the 1920s, there was a time when, quote, “Even second-rate physicists could make first-rate discoveries,” which I just kind of love. But the theory there is you can only make a lot of the big discoveries once. You discover quantum mechanics once. You discover the atom once. And most of them have just been made, so what you have now is more complicated, smaller, requires much larger teams of people, much more complicated experiments, with much more infrastructure. So we’re just structurally in a period where it’s going to get harder and harder and harder to make big gains. Do you believe that?I think it’s possible, but even though it’s intuitively compelling on some level, I’m not sure that it’s true. It’s probably true to at least some degree for some particular research direction, right? We go after discovering the various subatomic particles, and initially, without too much difficulty, we discover the electron or whatever. And by the time we’ve discovered the nth quark, it’s now gotten super hard, and even with ever-larger particle accelerators, we’re not necessarily making breakthroughs of the same magnitude. So I think it’s pretty true for a given direction. But obviously, the question is, well, to what degree is progress in any area opening up other directions, right? And so I mean, you mentioned the Dirac quote and, say, physics in the early part of the 20th century. Those discoveries opened up new techniques and investigation methodologies and so on, that then gave rise to molecular biology in the ‘50s, ‘60s and ‘70s. And so there’s kind of a combinatorial benefit, where discoveries over here or discoveries over there might unlock opportunities and major breakthroughs in areas that we could not have foreseen in advance. There are lots of, quote unquote, “low-hanging-fruit discoveries” made in computers and computer science in the ‘70s, ‘80s, and ‘90s. Maybe we’re even still in that regime, right? We’re still making some pretty fundamental breakthroughs. And of course, again, those, quote, “low-hanging discoveries” would not have been possible without a lot of this optimization and discovery in other fields. And so I think it’s probably true for a given research direction, but the relevant question for society is, is it true in aggregate. And there, it’s much less clear to me that it is.I want to read something provocative you said in an interview with the economist Noah Smith. And you said, quote, “Most systems get worse in at least certain ways as they scale. The idea that science could have gotten worse in significant ways sometimes sounds strange to people. Like, we’re doing so much more. How could that be bad? But I think that misses the many examples of sensitivity of scientific processes to institutions and culture. Swiss nationals have won more than 10 times more science Nobels per capita than Italians have. 10 times. And yet, they’re neighbors. And Italy certainly isn’t lacking in scientific tradition — Fermi, Galileo, the oldest university in Europe, et cetera. The ‘how’ of science just really matters.” And this seems, to me, to be where your exploration really goes. So tell me what you think might have gone wrong in the “how” of science.So I think this point about the sensitivity of scientific outcomes to the specifics of the institutions and the cultures is very important and probably underappreciated. At the beginning of the 20th century, not only was the U.S. not a scientific powerhouse, but it barely had a presence in frontier research, whatsoever. To become a credible researcher in the U.S. in 1900, you almost certainly had to go and spend time in, most likely, Germany, and failing that, in France or England — you know, what have you. And by 1900, the U.S. was already a pretty prosperous place, and it had a well-educated society, as societies went. And yet, somehow — and it had universities, right? I mean, Harvard was hundreds of years old by that time. And so it checked many of the ostensible boxes, and yet, the sum total of the U.S.’ research output as of 1900 was still de minimis. When James Conant, who was later president of Harvard for 20 years — when he went to Germany as a chemist, which was his original training, in the 1920s, he recounts how dispirited he was by what he found there and how far ahead of Harvard German research was, as of the early 20th century. And then, for a variety of reasons, all sorts of cultural, institutional funding — various transformations happened. And of course, by the latter half of the 20th century, the U.S. was the unquestioned leader at the frontier of scientific progress. If you look backwards, you see where that locus has been, where the most successful and fertile scientific grounds have been — it has repeatedly moved. As we just said, maybe the 19th century, it was Germany. Before that, in the 18th century, it was plausibly France. They had a couple of these really successful École Polytechnique and Grande École and so on. And so as a kind of first-order empirical matter, we can just notice, huh, this really seems to matter — and then, the example you just gave of the divergence between Switzerland and Italy. And so then, if we kind of accept that, and we try to ask ourselves, well, specifically, what are the mechanisms? You know, what’s actually going on? It’s hard for me to say. It seems like the transmission of research culture by individual researchers matters a great deal. And you see these kinds of pockets of the cultural transmission repeatedly crop up, where Gerty and Carl Cori — you probably haven’t heard of — they ran a little biology lab in Missouri, and no fewer than six of their trainees, of students they trained, went on themselves again to win Nobel Prizes. And if we tell ourselves a standard kind of mechanistic story as to, well, it’s the funding level, it’s how much are we investing in science, or it’s something about whether there’s an institution in the courser sense, that can possibly be amenable to it, it’s very hard to explain these eddies where you see these pockets of excellence really produce these outsized returns. So I think it’s a complicated question. I think all of aggregate culture, funding, institutional characteristics, and so on all contribute to it. But if I had to isolate a single variable, it seems to me that the research culture set by specific people and the tacit knowledge transmitted through direct experience is probably the number-one thing.This, I think, is where I sometimes fall into my own pessimism on this. Because I want to believe, as you do, that we can double the rate of scientific advance, maybe even go further than that. But I think the prediction — if I’m putting this on institutions, on culture, on pockets of transmission and mentorship — I think the prediction I would make is then, even if you believe, say, that America had a great 20th century, but its institutions have become sclerotic, and we’ve slowed down, and everything is piled in lawsuits and review boards now, somewhere else that didn’t have that, that has a different culture, that has different institutions, would be pulling way ahead. So you might think, well, China will be pulling way ahead. And you’ve noted this in some places. We’re getting a lot of peer-reviewed research out of China — huge number of citations out of China. We’re not seeing them dominate the big breakthrough advances of the era. It doesn’t seem like Europe is lapping us. And so if you think this slowdown is somewhat global, then that seems to me to militate against questions of individual institutions, cultures, how different labs work, because there is so much variation that you should have some of these labs that are doing it right, some of these places that haven’t piled on a little bit too much bureaucracy. But I don’t think we really see that.This diagnosis of these phenomena to cultural, institutional, mentorship-related, interpersonal dynamics, and your observation that it’s not obviously the case, that there are other places we can pointed that are doing it so much better — for me, my takeaway is that, well, successful cultures are a pretty narrow path. Homo sapiens emerged 200,000 years ago. And as far as we can tell, for the first 190,000 years of our genesis, we think we were largely biologically equivalent to the people we are today. But as best we can tell, there was some kind of cultural capital that those people lacked for a very extended period of time before human societies in somewhat recognizable modern form started to emerge — agriculture, all the rest. And in a similar vein, we had many billions of lives and centuries elapsed before the Industrial Revolution., and before we started to put together many of the input ingredients or enough of the input ingredients that we can get sustained improvement in standards of living and ongoing economic growth and progress. And so your point about, well, as I look around, I don’t see anything or anywhere that’s obviously better, I agree with that. But again, my takeaway is that that’s what makes the question of how do we improve or how can we do somewhat better so urgent and pressing, where it’s many things have to go right. It’s not easy to be even as good as — or to get to a place where things are as good as they are today. What we have is very precious. And I think the threads and the themes that you’ve been pulling on of late — all of these dynamics underscore their importance.I think that’s a good bridge to progress studies as an idea. And I want to have people hold in their heads that idea that progress is very narrow, that it is a very narrow bridge that we have walked on for a very short period of time. But let’s try to define it. When you say progress here, what are you actually talking about? Is it just shorthand for economic growth or G.D.P.? What is progress?Well, I don’t know that I would claim to put forth some kind of definitive definition. And I think, to some extent, our intuitions around it are probably broadly correct. And so it might not matter to define it super precisely and finely. For, me it is something along the lines of our success in realizing a liberal, pluralistic and prosperous society, and a sense among people that their offspring can and probably will do better than they themselves have, and that more broadly, the future will be better than the past, and that we’re at least making incremental progress towards embodying values and morals that we collectively think we can be proud of. But I don’t think anything that novel in that. I don’t think my conception of progress would differ that materially from some kind of average aggregate over any other group of people in the country.I do think there’s something interesting, though, which is that if you look at eras that I think progress-studies-type people and economic-growth people and historians of economic growth study most closely, actually, some of the periods where people feel a lot of rapid progress don’t fit that at all. You have, say, the Industrial Revolution, where life spans and lifestyle get worse for a lot of the people. I don’t think one will look at that period as unbelievably pluralistic. You have a lot of periods of war when you have very, very, very rapid technological progress, but it happens in context of much more martial societies. So there is an interesting tension, at least in periods — and some of them quite long, actually — where you can have fairly rapid economic progress, but it comes at a cost that I think isn’t always acknowledged, but is an important thing to think about.Yeah. So I don’t think you could point to some of these periods in the past and say that they definitively embody to the extent that we would fully aspire to some of these broader traits and characteristics. But I think the question is more, what are they doing as — you have to judge it relative to the baseline that preceded them. And I don’t know that the 18th century in the U.K. is some ideal as a society. But if you compare it to the 16th century in the U.K., the ideals and ideas of natural rights and religious tolerance and so on — they were somewhat better embodied by the 18th century than they had just a couple of centuries previously. And similarly, in the U.S., say, during either war or the ‘30s or whatever, again, it’s not like that was any kind of perfect society, but assessed relative to the society of 1830, I think it compares relatively favorably. And I think it’s not a coincidence that Adam Smith — his first book, of course, was on ethics and morals and trying to instill better general ideals and behaviors across a society. And maybe after that, he then argued for and laid many of the foundations of what we would recognize as modern economics. So I don’t think it’s perfect. But on average, I think the correlation is positive.So let’s talk about the Industrial Revolution for a little bit here. I think a lot of people locate a takeoff in human living standards — it continues to this day — there. And it’s strange in a way, right? “There” is a very geographically contiguous spot. It’s the U.K. — England, actually, I should say, at that point. And there is a moment in time that probably could have come at another moment in time, depending on how human history plays out in the counterfactual. I know that you have an interest in the theories of why then, why there. How do you work your way through them? What do you think is persuasive for why then, why there?Well, you know, again, I caveat. With all of these topics we’re discussing through this podcast, maybe the first-order banner for all of them should be, I don’t know, these are my best guesses, and I think it’s important that all of us were pretty humble in the claims and the assertions and the beliefs that we hold. Recently, I’ve been reading a bunch of Irish and Scottish writers around then. It’s very interesting, because for both the Irish and the Scots, there was a sort of a pressing and kind of obvious question where England was much more prosperous than they were or we were. And there’s no super obvious explanation for that. There wasn’t an obvious climatic or natural resource endowment that England benefited from that was lacking in Ireland or Scotland. It wasn’t like England was actually a vastly larger polity. The orders of magnitude were comparable. And Bishop Berkeley wrote this book, “The Querist.” He was asking these questions directly, just like, what’s going on? What’s wrong with Ireland? You know, why can’t we do this? And then, you have the Act of Union in 1707, uniting Scotland and England — and sort of similarly, of all these Scottish thinkers being like, all right, we’re now literally the same country. Why are we so much more impoverished? And then, if you shift to England, there’s Joel Mokyr and — you’ve read his work — and more recently, people like Anton Howes. And in a similar vein, they go back to — I mean, the word, improvement, came from Francis Bacon, or it was kind of popularized as a concept by Francis Bacon. But that’s noteworthy, right? Like, that was not a pervasive broad concept in the 15th century. I mean, literally, the word, improvement, in this broader societal context, came from word, “translated,” at the beginning of the 17th century. And the ultimate conclusion that these historians and scholars and analysts of the Industrial Revolution come to — and I think it’s a correct one — is somehow, whether it’s through Bacon or Newton or various of the tinkerers who produced some of the earliest technological breakthroughs, that somehow, this improving mind-set became pervasive. You had societies explicitly — like the Hartlib Circle or the Lunar Society, or the Select Society, and the club, and so on — all these societies explicitly devoted to figuring out ways to advance the state of affairs that prevailed. And these societies were comprised of many of the leading people and thinkers and so on of the day. And it seems maybe a bit satisfyingly squishy to attribute it to something so hard to pin down. But as you run through all the possible other explanations, it’s differences in IP law. It’s difference in the Malthusian conditions. It’s difference in the prevalence of coal, you know, et cetera, et cetera. Through various cross-sectional analyses, you can exclude most of these in looking at all of Ireland, Scotland, and England. It really does seem to me that differences in the mind-set and in the culture are where you have to net out. And that’s not to say maybe that it’s fully sufficient. There might be other preconditions that are important. And then, maybe as a last thing to say, it is striking to me that many of these kind of original 18th-century economic writers and thinkers — and again, the kind of people we look to as the founders of much of the discipline — that they themselves were kind of centrally preoccupied with this. And yeah, they were in favor of free trade and specialization and human labor and lots of these concepts that we’re now very familiar with, but they really thought that general mind-set played a big role, too.So let’s talk about Joel Mokyr ideas for a minute. So Mokyr is an economic historian. People should read his book, “The Culture of Growth,” which is really fascinating. He argues, as you’re saying, that in this period, this mind-set that we can increase the store of usable knowledge, and then use it to alter nature, to better the human condition, takes hold. That’s a new mind-set. It’s different than religious ideas of the past. It’s different than cultural ideas of the present. And that, plus a bunch of other things, particularly the republic of letters, the way people are writing letters back and forth, kind of combine into a culture that is able to grow. But one of the things that I really take from his work, that sits in my head, is he believes it’s all very contingent. He really believes it might have not happened. But the other is that I think it opens up this question that as a tech person, I’m curious to hear your thoughts on, which is, he really believes — Mokyr really believes — that there is a communications infrastructure that arises at that time, that has a kind of culture of generosity and argument and honesty in it, and is built on writing letters slowly to one another, and then copying those letters over to other people. And that culture is really good for intellectual advancement. I think one of the promises of the internet and the age we live in is, it’s all faster. We can write to people immediately. Things we write can go viral and be seen by 5 million people all of a sudden. And that was going to speed up economic growth really, really rapidly. And I would say, you don’t see that. So I’m curious how you think about communication cultures here and what you think for all the advantages of ours we might not have.I mean, I think it’s hard to say in aggregate. I feel it’s pretty likely that the effects are very heterogeneous across different populations. And you’ve made the case that you think Twitter is bad for journalism and for journalists. And I guess you live this yourself with your now mostly inactive Twitter account, I guess, apart from announcements. And I think in the case of the internet, that it’s almost certainly a tremendously large gain that billions of people now have access to educational materials. And some of the otherwise hard-to-communicate tacit knowledge — that things like YouTube videos now made legible and available. And I think it’s true that there are various gravity equations that we see across different disciplines. I mean, in economies themselves, in trade, where you rapidly decline in propensities to trade as countries get further from each other — but you have versions of this in academic disciplines as well, where geographic distance correlates inversely with likelihood of the exchange of ideas and so on. And I think it’s clearly the case that the sort of reaction surface area has increased substantially by the internet there and represents a kind of efficiency gain for people looking to exchange in ideas. Many of the companies that Stripe works with are remote companies, and they might employ people across myriad countries, and that’s a kind of communication and efficiency gain that would certainly not otherwise be achievable. I think it’s worth recognizing that the aggregate amount of G.D.P. that we are creating or gaining every year is so much larger now than — I mean, the percentage might be the same. But the total amount of stuff happening, or the increasing amount of stuff happening, is so much larger now than it was 100 or 200 or 300 years ago. And so for all of those reasons, I think we should give superior communication technologies and faster communication technologies a significant amount of credit, even though the ways in which those are manifests might be hard to measure and somewhat prosaic. Take my mom, for example. My mom works with a hospital in Minnesota. Our youngest brother has a physical disability. And in the course of that, she trained herself in treatment for cerebral palsy, this condition, and she wrote a book about it, and she did a master’s in this. And now, she’s trying to improve treatment for this condition throughout Ireland, in the U.S. and other countries as well. She’s a retired Irish mother who spends some of her year living in the U.S. near her sons, spends the rest of her year living in Ireland, working at a hospital in Minnesota, who just got a proposal to have her book translated into German a couple of days ago. And that’s a relatively prosaic story, but literally, millions of these stories exist in kind of aggregate form around the world. To circle back to the initial thrust of your question, though, I think it’s at least possible that the internet is bad for civic discourse. I’m not saying it is, but it’s certainly in the realm of plausibility — and that perhaps both things are true, where there’s some kind of iceberg where there are these enormous welfare gains that are not that legible, not that visible, lie beneath the surface, and then certain of the most visible manifestations, like what we see on cable news or what we see written in the papers — perhaps that is worse, and perhaps, slightly more structural judiciousness would be desirable there.I want to try to flip that and suggest that — because I’m going to push some counter ideas on why we maybe don’t see as much progress as we wish we did. But one is that I think possibly, very large welfare losses lie beneath the surface. And beneath the surface of stories like the one you just told about your mother, I think we all have stories of ways or people for whom the internet has unlocked a possibility. I mean, my whole career is built on the internet. I was an early blogger. I got rejected from my student newspaper. And if there was no blogging, like, god knows what would have happened to me. [LAUGHS] I mean, nothing too terrible, probably, but I wouldn’t have the career I have today. And at the same time, I think that the group of people who, by luck or by temperament, proved very, very good at using the internet, to some degree, distracts from the many, many, many people for whom the internet is fundamentally a distraction machine, or for whom the internet is creating, because of what we built on it. You know, shorter attention spans — how many people would have had an idea, sitting in a room by themselves, or taking a walk, that they never have now, because they never have to have a moment where they’re thinking alone? And so one thing that I think we’re all loathe to do is we’ll talk a lot about how it’s weird that we have so much more knowledge, but productivity isn’t increasing faster. It’s weird that we have so much more rapid communication between researchers, but science isn’t advancing faster. And then, the idea that maybe there are things happening to us that makes us less able to use that increasing stock of knowledge well, or makes us less able to collaborate in a useful way, I think, gets dismissed rather quickly. But I don’t think it’s totally implausible. Now, I don’t want to say, like, the greatest technology we ever had was letter-writing. Obviously, the greatest technology we ever had was blogging in the early aughts when I became a blogger. And whatever happened in your 20s is, like, as good as it was ever going to get. But I do wonder about these questions. And I think something Mokyr is right to put a lot of attention on is communicative cultures. Communication is how we collaborate. And if communication is in any way getting worse, it’s going to have pretty big macro effects. The other thing is if you believe these cultures matter, weirdly, as big as we’re getting, the internet allows a certain disciplines culture to stretch boundaries and borders in time in a way that it would have been harder. I suspect that labs were more different 50 years ago than they are today. The countries and the disciplines of researchers and the cultures of researchers in countries or cities are more different from each other 50 years ago than today, which is great if we have the best of all cultures today, but it’s not that great if you actually think variation is really important.Let’s wrap up there. So first, I agree, as a basic matter, that there are welfare losses occurring across society that we should be worried about, and probably everybody listening to this is familiar with the Stephen Pinker case for optimism, and rather than focusing in the headlines, you zoom out, look at these long-term time series. And once one does that, things seem a lot more encouraging, whether you look at it by income or life expectancy or infant mortality or choose your metric. Something that’s been striking to me of late is if you change the x-axis on those time series, and look at many of those phenomena and trends over a much shorter window, the valence changes substantially, and life expectancy in the U.S. is now, in fact, declining. According to C.D.C. data, 54 percent of teenage girls now report persistent feelings of sadness and hopelessness. And you could say, well, teenagers were never stereotyped as the most cheerful lot, but we do have some degree of longitudinal data here, and that number is up from being in the 20s as recently as 2009. 1/2 the population now is either prediabetic or diabetic — again, according to the C.D.C. Basically, point is, when we look at more recent windows, I think there are plenty of aggregate, emergent, complicated outcomes and phenomena that should give us concern. On the degree to which we should attribute the diagnosis to the internet or to our kind of communication media more broadly, it’s less clear to me in that — not saying it’s not true, but presumably, the life expectancy one is not — or at least if it is, the mechanism has to be very complicated. There are a bunch of other health-related ones. So take, for example, say, the incidence of diabetes or pre-diabetes. And you could say, OK, fine, all those things might be true, but they’re totally different. I guess the question I wonder about is, well, we know that lots of basic biological outcomes are correlated with mental states and so on. And so to what degree is there some more nuanced and complicated relationship there? But I think it’s a fair question, and I wonder a lot about it myself.Let me ask you about how you think, over the long period here, about the relationship between technology and equity or egalitarianism. And something specific is in my mind. I flicked earlier at the way the Industrial Revolution, for an extended period of time, seems to have reduced a lot of people’s living standards. And it wasn’t till later you had changes in redistribution in labor unions and labor protections that the amount of material prosperity that was generating created more broad-based prosperity, particularly at a very high level. I don’t know that you can sustain that kind of thing today. We have much more a small-d democratic culture. If things aren’t working for people, it’s much easier for them to organize and be heard. I think there’s a much more direct and complicated relationship now between whether or not people feel benefited by technology, and whether or not they are going to accept the conditions and the risks of rapid technological advance. But I’m curious, from your vantage point, how you see that both kind of historically and currently.I agree with that. I worry a lot about the basic stability of a society that does not successfully generate and make sufficiently broadly accessible the benefits of economic growth. The world simply has too little prosperity. And if it is not the case that people in the U.S. or people in any country — if they either feel like things aren’t progressing, or if they feel like maybe somewhere distant from them, things are progressing but they personally will never be able to benefit from it, I think we put ourselves in a very dangerous and likely unstable equilibrium. And if you go back to — well, you don’t have to go back very far in history to see, obviously, plenty of instances where this kind of instability brought the whole house of cards down. And so as a consequence of that, I worry a lot about, how do we simply make sure that — or one of the small things we each individually can do to try to make sure that society is generating enough economic gain and enough broadly experienced welfare gain that the whole compact can be maintained? And again, I don’t think there’s a ready neat kind of singular answer to that. Maybe Stripe as part of our small little contribution in one little fissure. But I think the central question you’re getting at is super important.And one of the questions I wonder about there — we’ve talked about the way progress has been very geographically lumpy, let’s call it, right? There’s a lot that happens in very small places, and it ends up affecting the whole world. Obviously, then, the gains of progress sometimes have that quality, too. And I do think of one of the politically destabilizing effects of the past, let’s call it, 30 or 40 years of digital progress, is being the concentrations of wealth. We just used to have a lot more spread. Even putting the questions of rising inequality aside, just where rich people were was different. And so where they were giving a lot of money to the local hospital was more spread out, say, across the country or in other countries across the land. But here, even as the internet is supposed to democratize distance, and in many ways, has — I mean, telework is not a fake phenomenon. It has really concentrated the wealth of that to, literally, where we’re sitting, but to New York. There’s a lot of money now in Austin. And then, on top of that, you often have barriers of entry, in terms of how many homes can be bought. So it’s not even like people can move to the place where all the economic opportunity is happening. And I do think that creates some of the skepticism you see of technology. I don’t think a lot of people’s — I think people are really excited about a lot of the goods they’ve gotten from it. But in this kind of macro political sense, as you’re saying, in a period of a lot of change, a lot of folks with real backing in the data don’t feel life has gotten better at the macro level. Life expectancy, happiness, political stability — it’s not like you can look around and say, well, I got this computer in my pocket, and everything else is going great, too. It’s like, I got this computer in my pocket, and what it keeps telling me is that everything is going to hell. Now, maybe it’s telling me that a little bit too much, but there is validity to the narrative.Yeah. So again, vehement in agreement on the sort of central importance of making sure that improvements in the standard of living are actually broadly realized across the society. That, too, I think, could serve as a manifesto for some of these Progress Studies ideas. On the internet in particular, or on technology and the technology sector and so forth, I think it’s complicated and difficult to try to sort of fully collapse or linearize it or something, where on the one hand, you have some of these concentration dynamics you identify. At the same time, of course, it is also a tremendous and incredible dispersal agent in making some of those possibilities and opportunities be more broadly available. I think it’s dangerous to take an excessively U.S.-centric perspective here. If you interact with or look at survey data, or otherwise try to assess what’s the sentiment of people in Poland, what’s the sentiment of people in India, or what’s the sentiment of people in Indonesia, they view the internet extremely positively. And I think correctly so, where their opportunities for advancement would be substantially curtailed in the absence of much of what the internet makes possible. I think in places like the U.S., or actually, even at home in Ireland, some of this story is complicated by lots of other things, but including — and I think, substantially — dynamics around housing policy, where there’s an extensive literature showing that, for example, a very substantial moderating effect on the increase in wealth inequality that would otherwise have happened, or income inequality, was geographic reallocation and people responding rationally to, OK, things in New York are going great, or things in California are going great. And if you look at the rate of increase of the Californian population, say, through the 1960s, that was a tremendously potent mechanism for us redistributing some of the economic gains that were being realized at the time. And of course, now, we have this crazy position, where California is losing population at the same time where the market caps of these companies and the profits of these companies are increasing very rapidly. But as one assesses that dynamic and tries to ask the question of, well, why aren’t these gains being better or more broadly distributed, it’s certainly not clear to me that the answer even lies in the realm of technology qua technology. And I think it’s certainly more broadly, again, some of these considerations like geographic allocation.Let me ask one more question on the geographic dimension, and then I’ll move on to it. And it’s on my mind, in part because when I try to think about progress, when I try to think about what inventions and innovations are coming really quickly, I actually see a bunch here. And I’ll use A.I. as an example. I’ve met people who are trying to automate a bunch of legal contracts. It makes a ton of sense. People pay a lot all over the country — to some degree, all over the world — to get fairly basic legal contracts drawn up — wills and real estate documents and merger agreements and all kinds of — from the small to the large. If you imagine that getting really effectively automated, though — you think about Saint Louis, Missouri, where some of the people who are important pillars of the community work in law firms there, and what they do is contracts. They do estate planning and all the things that people have to do in contracts. And if it actually does get concentrated to really, really great contracting firms in the Bay Area or in New York, on the one hand, the democratizing potential will really be realized. Those contracts will get cheaper. There’s also a theory in crypto of smart contracts. And on the other hand, you really will have a lot of that — the gains of that, economically, going to smaller areas and aggregated across a bunch of different domains. So graphic design, in all kinds of areas of the country — midlevel graphic designers get paid to make logos for local businesses. It’s pretty clear they’re going to be able to do that really, really easily on things like DALL-E pretty fast. So you can imagine a lot of that area getting wiped out. And you kind of run through a couple of these. And before you get to really unbelievable and sci-fi-like dimensions of artificial intelligence, you just have a thing that is going to democratize a lot of capabilities in a way that’s going to put the money for those capabilities both a little bit back into the pockets of the people who need them, and then a lot into the people who run the best A.I. rigs and is going to have a really weird geographically destabilizing effect. And that paradox of the internet both democratizing geography, and then concentrating wealth and capital in very small areas is, to me, a central challenge. Because if you get that wrong, if it goes too much in the concentration area, I think we’re going to lose a lot of the political stability we need here. But you’re more on top of these technological advances than I am. Do you think the trends there are going to play out differently than I’m worried they will?First, yeah, it’s not — I don’t think it’s foreordained whether or not these are going to be centralized technologies. I don’t know. And on the one hand, there’s, I think, an obvious feature we can contemplate, where there are only three A.I. models, and they are rooted in the hegemons, the citadels of Silicon Valley technology, and we all are digital serfs who are subsistence-farming on their gains. I think there’s also a very plausible story where these technologies prove substantially less defensible than we might have expected, and where, instead, they have this enormously decentralizing effect. Because otherwise, economies of scale that only large firms could benefit from can now be realized and pursued, even by massively smaller firms. And if we look at the recent history of A.I. — I don’t think any clear story there, but it does feel to me that it has been more biased towards the second story than the first. There are now multiple companies with large language models. There are a number of very successful open-source A.I. efforts. Actually, there was a really cool example from Replit, which is a service — it’s a programming I.D. in the browser, used by kids learning to code, but also increasingly used by people who are pursuing serious programming. And they recently released a GitHub copilot-like technology, where it will kind of autocomplete your code in the editor, and where you can do some pretty cool things. Like, you can highlight a block of code and ask it to be explained, and it’ll turn code into natural language, into English, and say, hey, here’s what this code is doing. Anyway, they wrote a blog post about how they built this, and they describe how it was built by one guy over the course of a couple of weeks. And maybe that’s only the case in the early days of this AI technology. I mean, in early computer games, the first games were built by a single heroic person, and now, it’s these gigantic studios and enormous CapEx budgets. And so I think the fact that this is the case today doesn’t mean that it will remain the case through time. But my takeaway is that at least not foreordained that AI or any of these other technologies will be centralizing forces. And then, the other thing to observe is that when we talk about these being centralizing, I think there’s a question as to, do we look at it in relative or absolute terms? And my contention would be that, both from a moral standpoint, but maybe more importantly from kind of a political-economy standpoint, what will matter is whether, on an absolute basis, people feel like they are realizing opportunities, their lives are improving, that things are getting better, that their kids will be in a better situation and so forth. And exactly how much value is realized by the companies themselves doesn’t actually matter that much, compared to that former question. And whether A.W.S. or whether any of these organizations has super high or super low profit margins, I don’t know is nearly as important as what is the actual effect on these communities and individuals across the society. And so in as much as one means — by centralizing, one means a large share of the profits, I think it is probably a more useful framing to look at it instead in terms of absolutes, and in particular, the absolute surplus generated by the users. [MUSIC PLAYING]What have you come to believe about the relationship between progress and war?I am somewhat skeptical that war is as conducive to breakthroughs as we might intuitively conclude, or as is sometimes claimed. You’re probably familiar with Alexander Field’s work on the ‘30s here. And his basic claim is, the productivity gains we often attribute to the Second World War in the U.S. — like, those foundations actually were laid in the ‘30s, and then the first half of the ‘40s were a period of decreasing productivity as we massively, inefficiently reallocated our economic resources for the purposes of winning the war, which was probably a good thing to do, but inefficient in narrow economic terms. And he has a new book coming out, I think, next month, that sort of extends this argument into the ‘50s. And as one takes stock of the scientific breakthroughs — and so Stripe Press recently republished Vannevar Bush’s memoir, where he takes stock of this. And obviously, you have, say, the Manhattan Project, and that’s a big deal, certainly. But it doesn’t feel to me that had the Manhattan Project not occurred, that peaceful development of nuclear technology would have been massively stymied. Maybe it would have taken another 10 years, but it was already happening to some meaningful extent. And then, as you take stock of all the other breakthroughs that took place in the U.S. during the Second World War, there were some meaningful stuff like blood plasma and blood transfusions. There was some significant breakthroughs there. Some of the first antimalarial medications, radar, the proximity fuse, which I’m not sure is all that useful outside of military applications. But by the time you get down to invention 6 on the list, I don’t know that as you compare that list to, again, some counterfactual of what would otherwise have ensued, that it looks radically better as you take stock of the Cold War and the enormous fraction of our economic resources and human capital that were devoted towards us, that the gains necessarily look that impressive. And so again, it’s super hard to judge. You don’t have proper controls and so on. But I’m skeptical.Let me take the other side. So there’s a question of, during war, how much did we invent during World War II. And that’s a question of how much the threat of war or the competition with an adversary ends up charging up innovation and convinces us to put resources, both in terms of people and in terms of money, and maybe in terms of institutions, into projects we wouldn’t otherwise have done. I think there’s an argument, at least, that we went to the moon because of the Soviet Union. It would not have done that for some time. Probably would have eventually done it, but also, who knows? As Derek Thompson, who I’m working on a lot of these ideas with, likes to point out, the Apollo Project was unpopular. It was not something that commanded wide popular support. Even now, if you look at the CHIPS Act that passed, it passed, with all that spending on semiconductor research and other kinds of next-generation technologies, under the framework of, let’s compete more effectively with China. If you look at all the things Darpa has done or been part of, the fact that “defense” is the first word in the Darpa acronym, I think, is meaningful. There’s something about what threat persuades societies to do, and persuades them to do technologically or what risks it allows otherwise-more-cautious governments to take, or what failures they could justify that allows them to have big successes. Something there doesn’t seem to small to me. And maybe it’s my political side, where I so often see scientific funding justified in Congress in terms of countries we’re competing with or are adversaries with. And I see what the defense industry can do that other institutions cannot, because they don’t get a lot of political blowback. But I don’t know. I worry a little bit about how much we seem to need the threat of another to accelerate things. And in a small way, maybe, we see what the pandemic — where we were willing to move much, much quicker on things like mRNA technology than I think we would have outside of it.Yeah. So I think it’s certainly true that the crisis can cause the discontinuous shifts that have large effects, which in your example, say, are probably super beneficial. I don’t know. If you take Darpa as an example, it started as Arpa, as a more open-ended research institution and set of programs, and then with the Vietnam War, had the D pretended to it. And we decided, in the face of threat, to make it more applied, to take more seriously its translational and kind of, quote unquote, “competition-oriented mandate.” And I think that was bad for Darpa. And the internet, which arose under Arpa — it’s hard to think of innovations of similar magnitudes that then occurred in then-Darpa’s subsequent, say, two decades. If you take, say, U.S. science in general, the war — the Second World War — to some extent, the first, but much more so the second — precipitated an enormous centralization of U.S. science in its aftermath. Because we really marshaled together all of the — or a significant fraction of the scientific capacity of the U.S. in service of the war effort. For, example the 50 percent overhead, the fraction of government grants that goes to universities — that was chosen in the early days of the coordination of the war effort, and has now become a kind of a pillar of academic and research funding in the U.S. And in the aftermath of the war, we sort have this question of OK, we’ve kind of pulled everything together. Now, what do we do? And that became, in various ways, the N.I.H. and the N.S.F. and so on. I mean, the N.I.H. predated it, but the growth of the N.I.H. really occurred after the war. And the federal government, shortly thereafter, for the first time, became the majority funder of US science. And all that centralization — and I mean, you pointed out the benefits of variety and of experimentation and of heterogeneity, and having some degree of institutional and structural diversity and so on, I totally agree with all of that. And I think all of that was very meaningfully curtailed by, again, the aftershocks of some of the threats that we faced during the war. And if you think about the things that we’re maybe happiest about having happened — the founding of the major new U.S. research universities in the latter parts of the 19th century or the revolution in health care and kind of medical practice that first happened at Johns Hopkins, and then kind of codified in the Flexner Report, or the great industrial research labs of Bell and Park and so on — or excuse me — Xerox — they didn’t obviously come from a place of fear or a threat. They came from a place of hope and optimism and opportunity. And maybe there are some inventions that you’re more likely to get to from some of these external pressures. But yeah, if you gave me a dial, and I can kind of turn up or down the threat or fear index of society, it’s not super obvious to me that one would want to turn it up if what one cared about was the aggregate rate of progress.That’s a good bridge, I think, to the question of institutions. And I take one of the main concerns of yours, of progress studies, as being around institutional slowdown. You have this idea that we don’t meta-maintain institutions very well. They start in one place, and then over time, they crust over, and we don’t really know what to do with that. Give me a little bit of your thinking there.I think institutions, the cultures they instill and act as kind of coordination points and training sites for — those of enormous consequence — I think much of the success of the U.S. and of various other Western countries has, in substantial part, been attributable to successful institutions. Most people would accept, I think, that there is, to some extent, consistent trends that tend to happen with institutions through time. Just maybe most basically, the problem that gives rise to an institution in the first place is probably a pretty real and significant problem. No one would have taken the time to found the institution if it wasn’t. And there can be some degree of drift there, where we don’t necessarily decommission the institution once the problem has subsided or abated. And then, you tend to attract a certain kind of person in the early days of an institution — people who are slightly less status and reputation and procedure-oriented, because a new institution almost never has that. And then, through time, the sort of collective or the mission-oriented incentives of the institution can kind of drift somewhat from the individual incentives that particular people are subject to. I think all this stuff exists. And the thing that I observe, or that I just find myself thinking about is, we’ve had eras of institution formation in the U.S. It has not been kind of a constant rate through time. And the New Deal maybe, and say, the 30 years afterwards, and the Great Society — we bookend it with those start and endpoints. That was a period of tremendously active institution construction and formation in the U.S., Darpa being — or Arpa originally being a good example, and indeed, NASA. And I guess I find myself wondering, one, if we didn’t have any of these institutions — and I’m not saying we should get rid of them. But if we didn’t have them, what institutions would we found today, first, and how high in the list would NASA be, for example? And then, secondly, in as much as we accept that some of these institutional dynamics exist, like the fact that sclerosis as an emergent property arises, what do we do about that? And lots of people have told us it’s pretty — doesn’t need a lot of teasing apart to see it as one compares NASA and SpaceX and the respective budgets, and the respective achievements, and so forth, I think it’s hard to not at least wonder about their respective efficiencies. And say, if society could only have SpaceX or NASA, which one would we choose, and what should we conclude from that, and to what extent do those phenomena generalize elsewhere? I don’t have answers to these questions. But I find that in the political discourse — not that anybody is celebrating that, but in the discourse, it’s very easy to get, I think, very wrapped up in questions of optimal funding levels, and should this number be 10 percent or 50 percent or higher or whatever, whereas to me, a lot of our satisfaction with the outcomes seems to hinge on deeper questions about the nature of the institution. I don’t know that the problem or benefit, or anything good or bad about NASA is attributable to the budget, per se. It seems more, kind of, resonant in some of these deeper cultural questions. And then, in the recent pandemic, or in the — I don’t know. I was going to say, ongoing pandemic. But I guess as of two days ago, with the President’s verdict, it is now over.Exactly. But anyway, I think that was maybe a vivid demonstration of many of these dynamics, where I don’t know this any of the story about the institutional response to the pandemic should be primarily one of funding. I think it’s much more about the dispositions and the attitudes and the cultural biases of entities like the N.I.H. and the F.D.A. and the C.D.C.I find the NASA SpaceX example an interesting and provocative one. Because on the one hand, I think what you’re saying is completely true. And where a lot of the NASA programs and projects have gone in recent decades, is just — it’s sad. It’s just a sad story. And on the other hand, the idea that you — the thought experiment of choosing between NASA and SpaceX — the thing that it immediately asks is, well, you can’t. Because without NASA, there is no SpaceX. And one way the private sector handles a lot of these questions — I mean, I’m always struck by how much of the way biotech research works is that big pharmaceutical companies acquire small biotech firms that have made a breakthrough or have come up with a very promising candidate. This is kind of an accepted thing that the big companies — they do a fair amount of research, but a major, major innovation transmission there is small groups do more, quicker, and they’re just going to buy them. And the NASA SpaceX example has a little bit of that dynamic to it, although with a different mechanism of financing. And I don’t know. I wonder if there aren’t deeper lessons there.Yeah, I don’t mean here in the NASA example — like, I don’t think reducing it to a simple binary of this-or-that is correct. It’s more, what should we make of the differences in these two organizations? And given those observations or beliefs, what do we then think an efficient outcome might look like? And do we think that where we are today — this prevailing status quo — is optimal? Or are there other things we can do better? And maybe an important thing to say within all of this is, to the extent that these are all kind of inevitably determined outcomes, maybe it doesn’t really matter if we think things would be better or worse. We’re going to end up in the same place, regardless. But I think for all of these, it’s super contingent. And various aspects of both funding decisions and, kind of, the precepts and methodologies of the N.I.H., how we design I.P. law, how we regulate and require and run clinical trials — there are tons of individual contingent decisions that we kind of have collectively made that give rise to the biotech and to the pharma ecosystem. And certainly, in the case of space, you know, like, it doesn’t have to be this way other. Universes, no pun intended, are possible. I think perhaps the thing that people underappreciated with science in the U.S. is, it has been very different in the not-too-distant past. Peer review is a relatively recent invention. Modern journals are a relatively recent invention. As I mentioned, the federal government being the primary funder of basic research is a relatively recent invention. And molecular biology was, in significant part, a thesis by Warren Weaver at the Rockefeller Foundation. There’s a thing here, and we should aggressively pursue it. But importantly, it was not — it required an institution, an organization, that was not part of the standard apparatus, for want of a better term. And the Broad Institute, over the last 25 years, has been enormously successful in the field of genomics and functional genomics and CRISPR, et cetera. And the Broad Institute is itself a kind of structural innovation, breaking somewhat from the more traditional prevailing university model. And so I think the fact that so many of our successes are associated with some degree of structural and institutional change should be somewhat thought-provoking for us.You’ve been trying to work in the space of institution-building here, too. I want to talk about Fast Grants and about Arc a little bit. So let’s begin with Fast Grants. What is it, and what has it taught you?Well, it’s mostly “what was it.” In the early days of the pandemic — well, I should preface all of this by saying — well, I’ll reaffirm my preface that I don’t know, to every question. But more importantly here, I will say, my now-wife is herself a scientist. We’ve known each other since we were teenagers. We spend a lot of time talking about science in various forms.You met — am I allowed to say this? You met at a science competition.That is true. We met at a science competition, 100 teenagers, and —And yes. I was the runner-up, and she was the winner. I had created a programming language and a new dialect of lisp, and she had created a new treatment for urinary tract infections. And so I really don’t envy the judges for having to figure out what framework one should use to make all these comparisons and lots of other people.[CHUCKLES] I was gonna say, but no, we can all agree this the correct outcomes ensued. Anyway, so we were living together in March of 2020, holed up. And a number of her friends and colleagues were unsurprisingly with, I guess, a large fraction of all biology scientists, were trying to urgently repurpose their work to figure out, well, could they do something that would be somehow benefit to accelerating the end of the pandemic? And by early April, so a couple of weeks into lockdown, when it was becoming apparent and striking to us, which was it is difficult for these people to get funding for their work. And that might sound a bit, kind of, surprising, because you think, well, don’t they have some degree of money already? And couldn’t they just go and just spend that? But there are, obviously, significant rules around and restrictions around that which one can do with one’s grant money. This is money provided by the government for a purpose. And so it’s not like you can go and readily spend it on something totally unrelated. And the money is administered by the university, and so you have to go through their proper procurement processes. Point is, lots of restrictions on scientists’ pecuniary ability to suddenly repurpose the research agendas. And we kind of thought, well — we assume maybe in the early weeks, that presumably various bodies — I don’t know who — some kind of amorphous other, some combination of C.D.C., F.D.A., N.I.H., philanthropies — whatever. Somebody will come along and just give these scientists the obvious money that society clearly should, so they can go, and they can pursue these programs. Didn’t seem to be happening. I mean, to be fair, I don’t want to give us too much credit. Various people were doing things right off the bat in various different places, but we just personally knew of lots of specific examples of really good scientists who were unable to make progress of their work to the extent that they would like. So we tried to set up what we thought would be a pretty small initiative, and called Fast Grants. The basic idea would be, you send us some kind of proposal. And initially, within 48 hours, you would get a funding decision and either receive money or not. We started out with a pretty small amount of money. The initial donors — we were among them, but there were a number — contributed, best I recall, about $10 million. Launched the website early April 2020. Quickly inundated with, I think, four and a half thousand applications, which, given our promised 48-hour turnaround, was somewhat challenging. I should say this was myself. This was Silvana, my wife, and this was Tyler Cohen. So we had an immediate question as to, how do we actually run a philanthropic endeavor? And how do we stand it up in very short order? And he, through Mercatus and through Emergent Ventures, had some experience of very efficient and somewhat-scaled grant-giving. And so the three of us worked together to put it together over the course of a week or so. We proceeded over the course of, roughly speaking, the next year, slightly more, to make about 200 grants, eventually dispersing almost — or slightly over, actually — $50 million in total, to universities around the world, though primarily in the U.S. And you ask, kind of, what did we learn? A big surprise was how slowly other parts of the establishment mobilized. And various of the projects we funded or the labs we funded and so on — they’ve gone on to now do — none of them were directly implicated in the vaccine research project that ended up yielding so much fruit. So again, I don’t want to give Fast Grants too much credit. Eventually, the thing that really mattered, we had nothing to do with. But versus the projects, things like Saliva Direct, which was in the summer an early discovery that saliva tests work basically as well as the nasopharyngeal swabs we were all being subject to, or various discoveries around possible therapeutics, some of which are — still continue to go through clinical trials, and may still turn out to matter to a significant extent. And that 500 people are still dying in the U.S. per day from Covid, and — despite the existence of the vaccines and so on. So anyway, various discoveries ensued that I think will prove to be important. And the second thing we learned, which is not really related to Covid or the pandemic, but has certainly been significant for us, is — it just got us thinking more deeply and broadly about the questions of, how do scientists choose what to do? And what are the constraints they’re subject to as a practical and applied matter? And this gets back to all this discussion about both culture and institutions. And towards the end of Fast grants, we ran a survey of the grant recipients. And these are essentially all people who don’t normally — certainly don’t normally work on Covid. Covid didn’t exist. But they don’t even normally work on viruses, for the most part. These are basically kind of broadly drawn as a cross section across biology. And we just asked them, as a general matter in your regular research, if you could spend your grant money however you want, how much would you change your research agenda? So not an increase in the funding level, which tends to be what we discuss in as much as we’re discussing science policy across society. But much more specifically and narrowly, if you had complete autonomy in how you spend whatever grant money you’re getting, how much of your research agenda would change? And our intuition was that maybe a third of people would like to be doing something meaningfully different to what they actually are. But of these scientists, and these are really good scientists, four out of five told us that they would change their research agendas, quote, “a lot.” We gave them three options. Not much, or not at all, a little, and then a lot. Four out of five chose the maximum option on our survey. So I just find this incredibly thought-provoking. Basically, we seem to be in a situation where most of our top scientists aren’t doing what they think would be best for them to do. And we could say, no, our various committees and governing bodies and decision-making apparatus and so on, they know better. And I’m not saying it would be completely unreasonable for one to maintain that. But that would seem to be a very central question about the construction of our scientific apparatus. And I think that should give us some pause.There are a couple things there. One is that it is a consistent observation I have learning about new areas that there is a way we’re taught the thing works, or people think the thing works, and there’s this huge middle layer. Right? So in politics, which I know very well, and legislation, you have the “Schoolhouse Rock” version of how a bill becomes a law. And then it’s, like, a filibuster is how a bill becomes a law or does not become a law. We were talking about drug innovation earlier. I think the folk way people think it works is we make a discovery about a drug, and then, like, we make a drug out of it after some tests. But you talk to people who work on pharmaceuticals and just clinical trials. And in science — I think if you had asked me as a high schooler, had some science classes, I’d have told you something about the scientific method. And then you talk to a scientist, and it’s grants. Like, grants are how science works. Grants are the middle layer between — you are a scientist, and you can do some science. And grants are how the N.S.F. and the N.I.H. work. They’re how a lot of the universities work. There’s fund-raising. I mean, there are different ways that it happens. To make the question of “Are we doing science well?” a little bit more precise, I think one version of that question is, “Are we doing grants well?” And I think that question is more tractable. People don’t feel as defensive about it. But I’ve talked to a lot of scientists in the course of my work. I’ve covered health care for my entire career. And I don’t know any who think we’re doing grants well. I don’t know any who will not complain to you for hours. And they may be wrong. And I do want to note — because they also just have somewhat different incentives. I mean, I was noting earlier, and I think it’s very real. The government, particularly when it gives out grants, needs to worry about the reputational cost of the grant. If the grant goes wrong, if not enough of the grants pay out into useful research. If Rand Paul can stand up in Senate and make what you did sounds silly, these things really end up mattering. And so you get a process that is optimizing for a lot of different things. But the question of whether or not we do grants well ends up being really, really, really important in every country that does major capital science that I know of, and is just not the main question for a bunch of different reasons we ask.Yeah. Another question we asked in our survey was how much time they spend on the grants. I think to some extent, this is perhaps — at least, of those who’ve spent some amount of time interacting with scientists, kind of more broadly known than perhaps the finding with respect to how they do — or the degree to which they can choose what they work on. But we found that — or they reported to us that they spend on the order of 40 percent of their time on grant administration. And even if one were to maintain that the decision-making apparatus around what scientists do is somehow efficient, I think it is a very tenuous position to also try to argue that 40 percent of the best scientist’s time is optimally allocated towards grant applications, authorship and administration. And we’re not talking about an inconsequential 40 percent here. I mean, this is 40 percent of the time of this super-elite 10,000, 100,000, whatever it is, some relatively finite number of people. And we’ve chosen to take and to redeploy almost half of their time in service of technocratic, bureaucratic undertaking. And getting back again to this point about people perhaps falsely assuming that things have been more inter-temporally consistent than they have, that percentage has increased very substantially over the last couple of decades as the overall edifice of science has grown, and as the kind of acceptance rates and the various thresholds for various grants has become more exacting.How we allocate people’s time is really important. But also, just how we allocate talent is really important. And it brings me to something you said that I wanted to ask you about. This was in response to a question about whether big tech companies are hogging all the talent in society. And you said, quote, “I don’t think that the ambitious upstarts who go into high speed rail in America, anyway, are going to have a great time or have much success in convincing their friends to follow them. And I suspect that for various reasons, too many domains look somewhat like high speed rail.” And so you go on to say that there’s a view that the internet is a frontier of last resort, and that you don’t think that’s totally wrong. So tell me about that. Tell me about the idea of the internet as a frontier of last resort. But behind that, this idea that other frontiers where talented people might want to go and make their mark on society have closed.You’re familiar with and you’ve probably written about the Stephen Teles idea of kludgeocracy. And I kind of like the term “kludgeocracy,” because rather than making some of the inhibitions that people might encounter in pursuing something like high speed rail, rather than casting those as being deliberate, the valence is more that it’s this kind of emergent, inadvertent and kind of complicated phenomena that nobody perhaps particularly wants or chose. And I think the case of California’s high speed rail is quite striking, where — you’ve written about this and kind of similar projects and the New York subway expansion and so on. And congestion pricing and so on. But it’s striking where it’s not actually obviously a question of first order political will. Like, we’re willing to fund the high speed rail in California. We’re clearly willing to invest in building the subway expansion in New York. But somehow, somewhere between that first order decision and desire and our actual ability to kind of instantiate it, something really goes wrong. And you contrast that with stories of — in the case of, say, California, Henry Kaiser and these various other early part of the 20th century operators in the physical realm. And for a variety of reasons, but mostly prosaic state and county-level complications and things that would extend the time horizon of one’s project, it has simply become meaningfully less-appealing for those people to undertake these initiatives. I mean, Foster City, not too far from where we are now, that’s named after the eponymous Mr. Foster. He was a developer. He decided, well, with reclaimed wetlands, I’m going to build a city. California is growing quickly. The Bay Area is a — kind of propitious and will be a long-term successful area. And I think this place simply needs more housing. And he, with that kind of founder energy, was able to give birth and rise to the city that now bears his name. I haven’t met anybody pitching me on a similar city on the shores of the Bay in the last couple of years. But I would imagine that were one to adopt that ambition today and to propose that maybe the San Jose Marsh wetlands should themselves be an expansion of San Jose, I don’t think one would get very far. And in fact, even for much more sort of limited things, like additional runways or runway expansions at S.F.O., even they have now been stymied for decades at this point. That ability to translate that into something enunciated has dissipated and deteriorated. And then I think the kind of individual version is, and if I want to be that heroic solar farm entrepreneur or railway magnate, that my practical ability to do so has been meaningfully curtailed.Yeah. I mean, that’s what I’m getting at here a little bit, which is talent really matters for a society. Where the most talented people go really matters for society. And a lot of those people want to go somewhere where they can have a really big effect. I think there’s been a huge rush to digital land because you can build on digital land. You can build quickly. I mean, it’s interesting to some of the dynamics we’re talking about, the temporal dynamics we’re talking about, that you see this dynamic even within the tech world. There was a while where it was really exciting to go join Facebook, go join Google, go join one of the big companies. Because you could do so much. And that’s still, to some degree, true. But they got really big. And so crypto got — whatever you think of crypto, one thing that is exciting about it to people is the idea that it’s open land. That you can go in there and have a really big effect on it. Build something new just with a couple of friends that might change the whole direction of the field. And on some level, it’s always going to be harder for, say, putting high speed rail through the middle of California. Right? I mean, just building things in the world is just going to be tougher. But on the other hand, if you make building things in the world too hard, if you make grants too difficult — if you — I know a lot of doctors who their advice to young people is don’t become a doctor. And it always breaks my heart a little bit. We need really great people to be doctors. And their point is not, don’t go heal sick people. Their point is, being a doctor is too hard now. The amount of time you spend dealing with insurance agencies and malpractice insurance and boards, and this and that, it’s just too much administration. When industries become very complicated to operate in, you want to select for people who are good at operating complicated industries, which may be different than the people who are good at moving really fast and changing things dramatically. But two, you kind of subtly bias where different kinds of people in your society go. I think in China, if you want to change a lot, you still probably go into infrastructure construction, among other things. Right? The idea that you might be a genius rail mind, in China, that’s great. There’s probably a lot of rail you can make. That’s not true here. And the point is not to make too much of the rail example, but to make a lot of the idea that talent flows towards where it can have an effect and people can live the kinds of heroic lives they want to lead. And if we have subtly pushed a lot of people into maybe not the right — not the socially optimal directions, that over time will have a pretty big effect on a society.I think a constant is that some number of ambitious young people will want to do something, as you say, heroic. And yeah, I think maybe two things have changed. For one, for whatever reason, our predisposition to putting those people in positions of authority has diminished. And then secondly, even if placed, their ability to actually execute, again for various reasons, has been attenuated. I’ve been reading about the university founders and presidents and those associated with some of the great US research institutions. And one thing that is striking is how many of them were so young when placed in those positions of authority. You know, Daniel Coit Gilman at Johns Hopkins, or William Rainey Harper at the University of Chicago. I think he was 32 when he was appointed president of the University of Chicago. Even in the recent past. So my dad was in the first year of the University of Limerick in Ireland. Or at the time, it was called N.I.H.E. It kind of acquired university status later in its life. And the Irish guy who founded it and was really the dynamo behind it, I think he was 29 when he was put in charge of that project. And I think it was in 1970 or ‘71 that he was charged with this mission. But as recently as 1970 in Ireland, we were willing to put a 29-year-old — I mean, that’s a person meaningfully younger than me in charge of the project of overseeing the creation of a major new research institution. And I don’t know that I have compelling or confident observations to offer in terms of the etiology underlying these changes. But I think the changes themselves are important, or at least we should assume they’re important if we come from a place of humility, where this is what has worked in the past. Enabling these ambitious young people who are willing to contemplate spending multiple decades in pursuit of some ambitious and idiosyncratic vision. And maybe we’re more enlightened now. Maybe we figured out how to get all the same innovation and all the same breakthroughs without unleashing that force. But I guess my starting point, at least, would be, well, we should — before getting super confident in that or before really being deliberate about it, I think we should give some kind of credit and credence to the prescription and the methodology that’s worked heretofore.And before books, let me end on this. We’ve talked a lot about scientific slowdown, about technological slowdown. But let’s say in the next 15-year time frame, what are the three technological or scientific possibilities you’re most excited by? If in 20 — I guess it’d be 2037, we’re having a conversation about how dumb this conversation was because it was right on the cusp of so much incredible stuff happening, what do you think is likely to be on that list?I don’t know that I’ve super non-consensus answers. I think that there are fundamental a priori reasons to believe that the rate of progress in biology could increase substantially over the years, and to your question, kind of decades to come. So if in 2037 we are enormously impressed and struck by the discontinuity there, that would not shock me. Clearly, over the past couple of years, there’s been acceleration in progress in A.I. And kind of far for me to try to point estimate for kind of where that is in 2037. But I would be surprised if that is not somewhere on that list. And then I think there’s something about education in the broadest sense that feels to me like a very significant, and hopefully very positive change happening in the world right now. Maybe best embodied by YouTube. But also by Twitter and by blogs and Substacks and even Zoom and kind of the growing ease of being in some kind of cultural proximity to people one aspires to emulating, or following in the footsteps of, or otherwise kind of being more like. And to the extent that one believes my story about the significance of sociology, and culture, and mentorship, and the kind of delicate transmission of tacit knowledge, it has until very recently only been possible for that to happen to a meaningful extent through physical co-location. And the fact that we’ve now thrown open those doors to such an extent feels to me like a really compelling and plausibly transformative change. And if it were the case in 2037 that we have multiplied by 20 the number of people who can — who have the initial mental models and understanding to become successful entrepreneurs, or successful scientists, or successful writers, or successful in whatever one might choose one’s domain to be, again, I think that would not be shocking. And I think it’s a pretty hopeful fact about the world.And then always our final question. What are the three books you’d recommend to the audience?Well, I’m right now reading “Revolution and Empire,” which is a book about Edmund Burke. And it is just fabulous. I very highly recommend it. Edmund Burke, Ireland’s foremost political philosopher. And I’m embarrassed to say that I have known less about him than I feel like I ought to have. So I recommend that very highly. And the autobiography by Warren Weaver, who I mentioned, at Rockefeller. I can’t remember if it’s called “Scene of Change” or “Scene of the Action.” But it’s Warren Weaver’s autobiography. That’s not a great book in the sense that you don’t read it — you don’t find it to be a vivid, compelling page-turner. And your mind is not blown on every page. But I find myself thinking back to it quite a lot and having various parts of it sort of ricochet to my mind. And then it all depends on what people are interested in and all the rest. And you should read the things you like. But I have on my desk at home right now “A Widening Sphere,” which is a history of M.I.T. And I was re-reading it recently. And my —Who doesn’t re-read the histories of M.I.T.?[LAUGHS] Well, William Barton Rogers, the founder, was the son of an Irishman, and started M.I.T. substantially with his brother. And I find it very inspiring, I guess back to what we were saying earlier, how motivated he was and they were by a kind of broad-based desire for societal betterment. Like, M.I.T. didn’t inadvertently end up being a significant contribution to American prosperity and ingenuity and welfare. He was really immersed in that milieu. He paid a lot of attention to some of the cultural dynamics we were describing in England, and the Darwins. And the early writing on M.I.T., if you go and just read the first two pages of the founding manifesto, it wasn’t utopian in some kind of implausibly lofty sense. But it was somebody who knew they weren’t founding a run of the mill nth technical college. And I feel like it’s easy to get cynical always. It’s easy to assume that the things that really worked out worked out through happenstance, as opposed to optimism and ambition. But yeah, I find the history of MIT to be a kind of inspiring reminder that sometimes these implausible, lofty, ambitious, long-term initiatives can work out much better than one would hope. [MUSIC PLAYING]“The Ezra Klein Show” is produced by Annie Galvin and Rogé Karma. Fact-checking by Michelle Harris, Mary Marge Locker and Kate Sinclair. Original music by Isaac Jones. Mixing by Sonia Herrero, Isaac Jones and Carole Sabouraud. Audience strategy by Shannon Busta. Special thanks to Kristin Lin and Kristina Samulewski. [MUSIC PLAYING]
Why do some countries produce far more science Nobel laureates than others? Why did Silicon Valley happen in California rather than Japan or Boston? Why did the Industrial Revolution happen when it did and where it did?
These are just some of the questions that have inspired the formation of a new intellectual movement called “progress studies.” The basic idea is this: For hundreds of thousands of years, human history played out without any rapid, marked advance in material living standards. And then, suddenly, in just the past few hundred years, everything changed: Humanity achieved a truly mind-boggling amount of progress in the evolutionary blink of an eye. In the early 21st century, we are all living in the world that progress bequeathed. And yet we understand shockingly little about what drives that progress in the first place.
That’s important because, at least according to some metrics, progress seems to be slowing down. We spend far more on scientific research but that research results in fewer breakthrough discoveries. Key economic indicators such as productivity growth have slowed. Many have argued that the technologies we’ve invented in recent decades, while highly impressive, aren’t as transformative as the technologies from the last century. All of which means that the questions animating progress studies aren’t mere academic exercises; they are central to understanding how we can bring about a better future for all.
Patrick Collison is the co-founder and chief executive of the multibillion-dollar payments company Stripe. But for years now, Collison has also been developing and advocating a worldview that has become the intellectual backbone of this new discipline. In 2019, Collison, alongside the economist Tyler Cowen, called for “a new science of progress.” And since then, an intellectual ecosystem has sprung up around it, full of its own magazines and thinkers and syllabuses and podcasts. And Collison himself is putting its theories into practice through organizations (like Fast Grants and Arc Institute) that he’s founded and funded.
This conversation is an attempt to better understand Collison’s worldview, and more broadly the worldview of progress studies. The ideas that animate progress studies are worth taking seriously on their own terms. But they are also important because they are becoming increasingly influential among a wealthy elite with the power and resources to shape all of our futures.
“The Ezra Klein Show” is produced by Annie Galvin and Rogé Karma. Fact-checking by Michelle Harris, Mary Marge Locker and Kate Sinclair. Original music by Isaac Jones. Mixing by Sonia Herrero, Isaac Jones and Carole Sabouraud. Audience strategy by Shannon Busta. Special thanks to Kristin Lin and Kristina Samulewski.