Patrick Collison — Why Silicon Valley's most talented should leave

1. 0:00 – 12:58

   Advice for 20-30 year olds

   1. PCPatrick Collison0:00

      And I wonder for people in their 20s if they, like, shouldn't go to San Francisco. The entrepreneurs are held in excessively high regard, in my view, and San Francisco doesn't really encourage the pursuit of really deep technical depth. (whooshing) I guess my general view is most products and most businesses, things can just be done much better, and I think moats are, are typically kind of overrated. The businesses that we serve, which is, in rough terms, 1% of the global economy. I mean, that's, that's about a trillion dollars a year. That then makes us, like, really terrified of outages. We're all trying to impress upon people at Stripe the importance of multi-decadal abstractions. I think people sometimes respond to that thinking that that's implausibly ambitious, but no, I, I think that's actually just what happens when you get this stuff right.

   2. DPDwarkesh Patel0:47

      Okay. Today, I have the pleasure of speaking with Patrick Collison, CEO of Stripe. Patrick, first question. You have an excellent compilation of advice...

   3. PCPatrick Collison0:56

      Mm-hmm.

   4. DPDwarkesh Patel0:56

      ... uh, on your blog for people 10 to 20, and you say there that once you turn 35, you'll write some for people in their 20s. What do you, what advice do you have for us now, uh, the people in our 20s now? (laughs)

   5. PCPatrick Collison1:08

      (laughs)

   6. DPDwarkesh Patel1:08

      When's it coming? (laughs)

   7. PCPatrick Collison1:09

      I haven't really thought about that. Um, though what I've been wondering recently is, um... You know, I, I said for that advice for people in their, uh, teens, they should-

   8. DPDwarkesh Patel1:22

      Yeah.

   9. PCPatrick Collison1:22

      ... go to San Francisco. Um, and I wonder for people in their 20s if they, like, shouldn't go to San Francisco. And I'm being glib, and, you know, I think there's a (laughs) significant set of people who should, in fact, go to San Francisco, but the, the thing that I wonder about is, um, for... There- there is a set of career paths that I think some set of people, um, you know, ought to pursue and would derive most fulfillment from pursuing, uh, and, um, and that are, you know, that are really valuable for the world, uh, if pursued, that require accumulating a lot of expertise, uh, and you know, really, really studying a domain in, uh, in tremendous depth. And I think San Francisco valorizes, and look, this is, this is also San Francisco's great virtue. San Francisco valorizes a kind of striking out on your own, iconoclastically dismissing the sort of received wisdom and, uh, and, um, you know, the, the, the founding archetypes of, uh, and lore of the Steve Jobs and the Bill Gates and, uh, all the rest. And you know, I'm way less successful than those people, but like, to some extent, you know, Stripe, in as much as it fits a pattern, is an instance of that pattern. Um, and look, that's, that's great, and I'm kind of happy that that phenomenon exists in the world, but, but I, I don't think that, um... Just the world needs lots of other things, right? Uh, and I don't think San Francisco particular... I mean, I'm, I'm again using San Francisco as a kind of metonym for a cultural orientation, but, uh, I think that San Francisco doesn't really encourage, yeah, the, um, the, the pursuit of, uh, of really deep technical depth, uh, and you know, we're, we're here, we're recording this in South San Francisco, um, and, um, you know, South San Francisco is most noteworthy in, uh, the, um, in the corporate world for of course being the, uh, the headquarters of, uh, Genentech, uh, and you know, Genentech was, uh, was co-founded by, uh, Bob Swanson and Herb Boyer, um, and you know, they, they, they produced cheap insulin for the first time with recombinant DNA. Um, like, Herb Boyer couldn't have done that, like, at age 23. (laughs) Um, Herb Boyer first had to accumulate all of the knowledge and the skills required to, you know, be able to invent that over the course of a multi-decade career, and then I don't know what age he was when he, uh, finally went, went and invented it, but, uh, he was not in his 20s. Um, and, uh, and, like, I feel San Francisco perhaps doesn't, uh, doesn't culturally encourage one to become Herb Boyer.

   10. DPDwarkesh Patel4:02

       Hmm.

   11. PCPatrick Collison4:03

       Or, or yesterday, at the time of recording this podcast, um, Patrick Hsu, one of the co-founders of Arc, which, uh, maybe we'll speak about, uh, you know, later in the show, this is a, a biomedical research organization we, uh, we started a few years ago, um, he announced, uh, uh, these, um, uh, this, this new phenomenon of bridge editing, uh, which is, um, a new recombinase, uh, where you can, uh, insert DNA, uh, into a, into a genome, and you know, it's pretty early, but it, it might turn out to be, you know, quite consequential. Uh, and you know, in order to do something like that, you have to study for a long time, um, and just acquire a lot of, you know, basic, um, and not-so-basic, uh, technical skills. And so anyway, the, the thing, and I don't quite know how to synthesize it yet, but as I think about advice for, for people in their 20s, look, I- I'm not gonna normatively pretend or, or presume to, um, to know kind of in, like, in which direction one should go in one's life, and obviously there are successful examples of basically every strategy.

   12. DPDwarkesh Patel5:04

       Mm-hmm.

   13. PCPatrick Collison5:05

       Uh, and you know, I'm really glad that you're doing what you're doing. At what age?

   14. DPDwarkesh Patel5:08

       23.

   15. PCPatrick Collison5:08

       23. Um, uh, so that's, uh...

   16. DPDwarkesh Patel5:10

       I mean, the podcast isn't. (laughs) I'm not inventing recombinant DNA here. (laughs)

   17. PCPatrick Collison5:13

       I mean, I mean, well, look, I, look, I think, uh, um, uh, information dissemination, um, is a, is a really valuable thing in the world, um, and causes... Like, the, the guy who, um, who last I heard, um, was, uh, in the, in the lead for Nat's Skoll Prize-

   18. DPDwarkesh Patel5:31

       Yeah.

   19. PCPatrick Collison5:31

       ... uh, uh, uh, Nash told me learned about-... uh, the Skoll Prize, you know, listening to your podcast, right? So I thi- I think kind of cre- increasing the catalytic surface area of, uh, of, you know, certain kinds of information is a, is a valuable thing in the world, so I'm, I'm very glad, you know, you're, you're producing the podcast. Anyway, uh, I, I don't presume to know what people should do with their lives, uh, obviously, but, um, but as I, I wonder in, in as much as I'm trying to give advice, uh, and especially maybe if they're reading my advice and not someone else's advice, maybe if they're kind of thinking about, you know, career paths that look sort of directionally like mine, uh, I think, I think my advice might be, I mean, maybe you should do something like what, what, what I did or I'm trying to do but, but there are other paths as well, and I think a lot of really important invention in the world and s- a lot of the things that I'm most happy are happening, um, actually require a very different trajectory. And I think there are counterver- counterfactual versions of my life where, you know, I, I, you know, pursued that path and, well, who knows how well it would have worked.

   20. DPDwarkesh Patel6:27

       Mm. Um, uh-

   21. PCPatrick Collison6:29

       And by the way, so last point in this. In San Francisco it's just very status-oriented, I feel in this way. I mean, to me status-oriented is, I mean, everything is status-oriented so that- that's kind of tautological, but, um, maybe, maybe really what I'm saying is, I feel San Francisco, um, the entrepreneurs are, um, are held in excessively high regard in my view, and look, I, I guess I like entrepreneurs, um, uh, at and, you know, I, I think, um, I think, uh, look, entrepreneurs as an aggregous, whatever, group in the world, you know, all the, all the companies built on Stripe, you know, I think, I think are, I think are great, but, uh, but there's just a, a strange version of it in San Francisco that I think is, um, um, should not be c- should not be peoples only fixation.

   22. DPDwarkesh Patel7:14

       Yeah, I mean, well, what I like about this and what I like about you is just like you have this sort of, uh, a sense of like contrarianism of like the things people are expecting to hear from you in a given moment, you just like really wanna just tell them the opposite. (laughs) Of, I don't even know, like when, when, when, uh, uh, I feel like, yeah, when, when, yeah, it was a little bit more popular you were like, "Here's the problems, here's why progressivity is important," and (laughs) when it was down in its depths, like, "Hey guys, pay attention." Um, uh, but okay on this particular piece of advice-

   23. PCPatrick Collison7:36

       Well, Mik- Michael Milson says that, you know, every field in science, uh, has, uh, you know, way too many adherents or way too few-

   24. DPDwarkesh Patel7:42

       Yeah, yeah.

   25. PCPatrick Collison7:43

       ... uh, but like, you know, the market is almost never in, into the, the, the right equilibrium, and I, I think something like that might be, I mean, I, I think the reflexive contr- (laughs) in a contrarian way I will say that I think reflexive-

   26. DPDwarkesh Patel7:53

       (laughs)

   27. PCPatrick Collison7:53

       ... reflexive contrarianism for the sake of it, um, is, is also tired and, uh, you know, if you, um, if you, uh, if you're just contrarian to the prevailing-

   28. DPDwarkesh Patel8:02

       Sure.

   29. PCPatrick Collison8:02

       ... mood then, you know, you're, you're, uh, you're just following the prevailing mood but, you know, with a, with a sign bit inversion or something.

   30. DPDwarkesh Patel8:07

       Yeah, yeah.
2. 12:58 – 23:07

   Progress studies

   1. DPDwarkesh Patel12:58

      Uh, before we get back to Stripe and Arch Institute and everything, I wanna, uh, just, uh, touch on the Parker Study stuff for a second.

   2. PCPatrick Collison13:02

      Hmm.

   3. DPDwarkesh Patel13:03

      Um, so there, there's a view that says, listen, if we improve the NIH 10% or whatever percent, are we, are we really making a dent in the fact that ideas are getting harder to find over time? And how much of, um, like how much of a difference do institutions make anyways?

   4. PCPatrick Collison13:16

      Right.

   5. DPDwarkesh Patel13:16

      Can it, just like if it's just about a number of researchers and how many people in your society you can put into research-

   6. PCPatrick Collison13:21

      Yeah.

   7. DPDwarkesh Patel13:21

      ... it's like, it's not like Singapore can have a much more effective scientific institution that lets it compete with America in science or something like that. Do you... What, what, what's wrong with that intuition?

   8. PCPatrick Collison13:29

      Noah Smith and others have talked about, um, I can't remember the term he used. Something like, uh, moneyism. Um, uh, he had a, a funny, a funny phrase. Um, but, but sort of this idea that we, um, that we assume there is some, um, kind of constant elasticity between investment in some particular outcome, like building a semiconductor factory in Arizona or a new bridge or whatever, and the outcome of, you know, the factory or the bridge. Uh, and one, the conversion rate between, you know, those, um, uh, th- th- those inputs and the output is, uh, you know, is, is not a cosmological constant. Uh, like maybe any of these things could be done for half or a tenth or, you know, whatever of the cost. Um, but two, uh, there's, you know, there are even deeper questions as to like, you know, (laughs) is it possible at all or, uh, you know, what else would have to change for it to be possible and what are the other constraints? Like by kind of just talking about these things in funding and dollar terms, you're kind of making the implicit assumption that the only relevant constraint is, is the financial one, where in practice maybe it's permits or it's, you know, labor shortages or, you know, it's, it's, uh, it's other things. Anyway, in the context of, uh, of the NIH and science and, and R&D, I, I'm really skeptical of, um, of this kind of this, this same approach being brought to bear where, you know, we can, uh, where we can, um, we can just talk about the amount that we're spending on R&D and think that that's implicitly a useful measure of the, uh, of the output. Uh, and, you know, and, and to a fairly close approximation there were around one percent as many, uh, practicing professional scientists, uh, in the US pre-World War II as post-World War II. Um, and, um, uh, or say even 1950, uh, and, you know, the other kind of epi-phenomena, you know, in papers or patents and so forth, you know, it tends to, uh, you know, yeah, follow pretty similar ratios. Um, and, you know, we, we got a lot of, like, pretty good stuff, uh, in the, uh, in the first half of the century. Um, and, uh, you know, despite increasing the amount that we spend by, you know, between two and maybe, maybe ... I mean, it's slightly more than two orders of magnitude, um, not quite three. Uh, it's, uh, it's just, it's not clear to me that there is, like, a direct linear relationship. And so, you know, when, when analyzing the NIH or how we should pursue any of this stuff, I'm inclined to try to get way more, I guess, concrete and tactile.

   9. DPDwarkesh Patel16:05

      Mm-hmm.

   10. PCPatrick Collison16:05

       And try to think, okay, like, what would success here look like at ... Well, what is happening today at the micro scale?

   11. DPDwarkesh Patel16:12

       Mm-hmm.

   12. PCPatrick Collison16:13

       And, and what are the, like, what are the actual problems? And then what could success look like at the micro scale, uh, and then what might it look like to scale that up? And just to give one kind of pointed example of that, we ran a survey of the Fast Grants grant recipients, um, after Fast Grants, asking about their normal work.

   13. DPDwarkesh Patel16:32

       Mm-hmm.

   14. PCPatrick Collison16:32

       And not about anything to do with Fast Grants itself. And we asked them, um, if they had flexible funding, um, that is to say if they could spend their research dollars, uh, their current research dollars however they wanted, so not if they, like, had more research dollars, just if they could spend, uh, um ... If they could direct their current dollars however they wanted, how much their associated research program would change. Um, and we gave them three options: not much, a little, and a lot. 79% said a lot. Uh, so four out of five said that their research agenda would change a lot, um, if, uh, if this constraint was removed. And so, like, should the NIH funding level be, you know, X or 1.1X or 1.2X or whatever, that seems to me like a bad way to analyze this question as compared to, you know, for example, perhaps how bound and constrained should an NIH grantee be, uh, in choosing their research agenda? You know, may- may- maybe, I mean, if their...... judgment was way better, uh, than, uh, than that of the committees. Not saying it is, but maybe it is. Who knows? And maybe there's a 5X improvement (laughs) , you know, to be generated just by making that one, one switch. Uh, so, um, so yeah, I'm just, I'm very skeptical of these, um, of these financially oriented, uh, frameworks.

   15. DPDwarkesh Patel17:41

       Hmm. I mean, uh, m- m- uh, uh, uh, m- maybe the financial is not the right word for it, but just like trying to map inputs to outputs is the framing which you're using to compare the pre-

   16. PCPatrick Collison17:49

       Ah.

   17. DPDwarkesh Patel17:49

       ... World War II-

   18. PCPatrick Collison17:50

       Yes.

   19. DPDwarkesh Patel17:50

       ... um, uh, uh, inputs to, um, what's happening now. And if it was particular to the scientific institutions, you'd expect, for example, that things that are disconnected from, like, the NIH-specific structures-

   20. PCPatrick Collison18:01

       Yes.

   21. DPDwarkesh Patel18:01

       ... I mean, obviously there, you, you've talked a lot about the R&D is getting harder to find paper.

   22. PCPatrick Collison18:05

       Right.

   23. DPDwarkesh Patel18:05

       And, you know, you know, they, like, uh, sector through sector, it's not like f- NIH is running Moore's Law progress, right? But even there you see you need exponentially more researchers-

   24. PCPatrick Collison18:13

       Yes.

   25. DPDwarkesh Patel18:13

       ... to keep up the same level of progress. Um, so, eh, i- it does seem important to have these, um, level effects, uh, that are one time in the case of something like COVID where like, yeah, we need that level effect right now. Um, but when, uh, if we've, if we're framing it in terms of like, hundreds of years from now we're gonna be, uh, this is gonna be a thing that increases growth rates, um, which is a, sort of framing that is also, uh, supplied when talking about these progress or these things. Does that make sense in that context wh- when like all these sectors are seeing these slowdowns which seem consistent with just like, yeah, this is how, uh, the, uh, economy and science progresses over time?

   26. PCPatrick Collison18:44

       I don't know is the short answer. I think it's really puzzling. Um, uh, I think the, um, yeah, like the, uh, the constancy of, uh, of, uh, US GDP growth, um, is I think just like one of the weirdest things.

   27. DPDwarkesh Patel18:56

       Yeah.

   28. PCPatrick Collison18:56

       And I've also got explanation for it. Um, but, but, but also I, I don't think that it's, or sorry, a- an obvious thing to do would be to shrug and say, "Okay, well just it's, it's, it's over determined or something and that's just like how countries work."

   29. DPDwarkesh Patel19:06

       (laughs)

   30. PCPatrick Collison19:07

       Uh, but you can look at other countries where, you know, it's obviously manifestly not the case. And so, you know, what is it that's weird and special about the US? The, the thing that I wonder about in a lot of these cases is, um, you, you could get many of the observed system phenomenon characteristics, uh, if we, like, (laughs) if we weren't actually adding productive capacity. Um, like that's, that's a simple way to explain a lot of it. And that if you're just adding, um, exponentially more unproductive capacity-
3. 23:07 – 35:13

   Arc Institute

   1. PCPatrick Collison23:07

      question.

   2. DPDwarkesh Patel23:07

      Okay. I, I think that's a great, uh, that's a great point to talk about Ark Institute. Um, so yeah, I mean, th- the, I think you just kind of answered this question basically, but, um, uh, it's not exactly like biology research is, uh, uh, is, is something that society has neglected. So what's the theory of change here? Is it just a story similar to Stripe in that if you get the right people, even though bi- you know, there's like tens of billions of dollars of biology funding, getting the right people and the right culture and right dedication i- is what it takes?

   3. PCPatrick Collison23:32

      ... even though there are lots of scientists and, um, and lots of, lots of universities, um, there's a lot of homogeneity, uh, today, um, in, in how science and in particular how biomedical science is pursued, uh, where in kind of basic, you know, uh, uh, research, you know, in, in an academic context, you know, before there's any commercialization, uh, or prospect of it in sight. And, um, and I don't know that the model is necessarily a bad one. Certainly, you know, we're not particularly claiming that it's a bad one, but sort of the, the construct of universities, labs, uh, PI, a principal investigator running the lab, that person, you know, applies for grants primarily to the NIH, maybe supplemented by, you know, other, um, other sources, uh, and, uh, and grants reviewed by, uh, committees with, uh, with pretty, um, uh, study sections as they call them with kind of pretty rigid scoring criteria and so on. Like, that's the structure. Um, and, um, it just seems suboptimal to me. I mean, uh, homogeneity is bad in basically any e- e- ecosystem, especially ecosystems where you're, um, uh, you know, where you're, where you're producing or excuse me, where you're seeking tail outcomes. Uh, and we thought that for a variety of reasons, uh, like, well, from first principles that other models should be possible. And that like we had, we had specific ideas as to how, you know, one particular model, uh, might be a, might be a good idea and, you know, complementary to the status quo in very short terms. Um, uh, what's different about Arc is, one, scientists are funded to pursue... uh, scientists are funded themselves to pursue whatever they want. So it's, it's curiosity during research, uh, whereas NIH grants are, are given for projects. Um, second, we build a lot of in-house infrastructure so that scientists can draw upon other platforms and other capabilities that they don't have to go and, you know, build and maintain themselves. Whereas again, in the kind of standard university academic context, uh, scientists, uh, you know, would virtually always have to, have to do that in-house. And because of the natural skill constraints on any given lab, you know, that, that, that effectively circumscribes the ambition of a, of a, of a possible research program. Um, and then thirdly, we try to provide career paths for people to, uh, to remain in science, even if they don't wanna become principal investigators, where the, uh, the university structure kind of commingles the training purpose of, uh, of, I don't know, the, the, of academia with the, uh, with the, um, with the execution, uh, where the people who are doing the work are typically the, uh, um, the grad students and the post-docs, uh, who are both themselves at least nominally, uh, on the career path of themselves eventually becoming principal investigators. And, um, you know, there are lots of people who for all sorts of different very valid reasons like love science and love the pursuit of research, but don't wanna be a manager running a lab, choosing their own research programs, and dealing with all of the overhead and typically grant applications, uh, that are concomitant with that. Um, and so with Arc, we have a real emphasis on hiring, uh, scientists who have finished their post-docs, finished grad school and just like that, that's what they wanna do in their lives, and there again, isn't really a, a, a career path for them today. Um, and one of the things that's actually really exciting about the discovery that we, that we mentioned that came out yesterday, this new bridge editing technology, is that work was led by one of these senior scientists, (laughs) uh, who, um, who'd finished his post-doc, uh, and, you know, it's, it's not clear to me that he wanted to, uh, to go on to become a PI, um, but he loves science and he's an amazing researcher, clearly. Um, and, uh, and so he's able to go and to, you know, have that, have that career, um, at Arc. And, you know, it, um... and in addition, uh, the, you know, the, the, the prospect of these, uh, mobile elements being, uh, you know, usable in this way for this like genomic insertion or whatever, that's a pretty speculative, you know, out there thing. And, you know, had he applied to the NIH to go and pursue that, I mean, you know, he didn't, so I don't know what the outcome would have been. Uh, but, uh, Jennifer Doudna's, uh, work was, if I recall correctly, funded by DARPA, uh, because, you know, her CRISPR, (laughs) um, uh, NIH applications were, were rejected. And of course, Katalin Karikó's NIH applications, you know, for, for mRNA, uh, vaccine work were, were famously rejected. Uh, so, um, so it at least seems very plausible that, that it wouldn't have, uh, it wouldn't have worked out. Um, and, uh, and so look, it's, all these things are random and, you know, I can't make any definitive claims about, you know, the what would have counterfactually happened. But it seems plausible to me that this thing announced yesterday wouldn't have happened or would've been less likely to happen, uh, in, um, yeah, in, in a different environment.

   4. DPDwarkesh Patel28:12

      Mm-hmm. Whe- when we think forward, um, 10 years or 20 years, uh, the, the, the specific line where you, uh, of research where you understand the effects of, uh, the genetic architecture on different traits and also you can edit, in- invert, uh, insert, whatever, um, uh, the DNA arbitrarily. Uh, y- you know, you saw sickle cell anemia, you've done the obvious things. What does that lead to? What are you excited about?

   5. PCPatrick Collison28:36

      Well, the thing that I think is really interesting about it is, um, is using it as a new kind of telescope, by which I mean y- you know, when people hear about CRISPR, um, there's, um, y- there's an obvious excitement and legitimate excitement around, you know, using this to cure things, you know, directly in the body, using it as a kind of therapeutic. Uh, but you can also use CRISPR to, um, to try to figure out what's going on in cells and in cell cultures, uh, in a, in a kind of structured way. And so, excuse me, the, you know, the body is interesting in that it has this, you know, this, this switchboard, um, of, uh, like, like, you know, the, the DJs I guess at the, um, um, you know, with th- those fancy mixing sets, uh, uh, you know, of, of 20,000 genes. And with CRISPR, you can systematically go and sort of perturb each gene one by one, like mashing all the keys, uh, i- in sequence, um, and try to figure out well, you know, what the effects of, of perturbing this versus that are. Uh, and, you know, if you do that in a cell culture where, you know, you can-... subject of cells to some stressor or some treatment or, you know, whatever. You can kind of see differentially how different cell, how different perturbations affect different cell outcomes. Um, or you can just kind of use it for synthetic data generation more broadly, where, you know, you could p- perform all these perturbations and then sequence and kind of see what's happening in the cells and so forth. And, you know, single-cell sequencing has come a long way. Anyway, point is, there's a lot you can do with all this gene editing stuff for, um, uh, uh, for discovery and for, and for data generation in the broadest sense. And, uh, and I, you know, that's really compelling because, you know, for, for a lot of diseases, um, you know, that they're complex, um, in, uh, in the sort of fields jargon, meaning, I mean, yes, they're complex in the colloquial sense, but they're specifically complex in that they're not, um, they're not infectious, um, uh, like they're not just like some pathogen getting into you, um, and they're not monogenic, like, you know, Huntington's where it's like one specific mutation. Instead it's like some combination of environmental factors, but like maybe some genetic factors as well and, you know, it's, it's, it's somewhere in between. Um, and by figuring out ... And, uh, you know, that includes most autoimmune diseases, um, most cancers, um, uh, to some extent cardiovascular disease and neurodegenerative disease, like the big ones we haven't yet solved. And, um, and, you know, so then coming back to these functional genomics technologies, what's interesting I think is trying to figure out the, the how it is that the genetic, uh, component of those diseases happens and works and so on. And even if that's only a small contributor, it can potentially shine light on just like what the general pathway is, and so the question would be ... And look, this is, this is speculative. None of this has actually happened. But by figuring out the genetic interactions between, uh, genes and, say, Alzheimer's, can y- can you figure out the, you know, how the, how Alzheimer's arises, which we don't understand today. And then once we understand how Alzheimer's arises, maybe you can use kind of conventional technologies, uh, and targeting to, you know, figure out how to, um, how to inhibit that or to, or to sort of, um, you know, um, modulate those pathways. Uh, and, uh, and so yeah. The, that, that's what we're re- really excited about from a, from a functional genomics standpoint, and there's kind of an AI angle as well that, you know, we, we could talk about if you want.

   6. DPDwarkesh Patel31:56

      Mm. Um, how, how do you think about the, uh, the dual use possibilities, uh, o- of biotech? I, I, I am sympathetic with the idea that, like, if you think of prior technology, just like Google Search or even, like, just the computer itself, right? Uh, there, there's ... You could, like, forecast in advance, like, "Oh, this has all this dual use stuff." Um, uh, but for some reason, just, like, history has been kind to us and maybe we should just ... The, the, the, I mean Chestrin's meta-fence here is like, keep doing science. But, um, uh, y- y- l- y- but with biotech there's d- I mean, we don't have to go into specifics here, but there's like specific things you can think of with the specific technology where you could imagine some nefarious things. H- h- how do you think about, uh, just, uh, why not just like, uh, focus, let's say, on a- ameliorating the risks, uh, first or something like that?

   7. PCPatrick Collison32:36

      Well, um, I don't think that, uh, I don't think the binding constraint on harmful use of, um, biotechnology or bioweapons today is pure biological capabilities. Like if some set of incredibly capable, uh, intelligent people wanted to, uh, wanted to, you know, cause tremendous harm with, uh, with ... well, presumably with pathogens, but with something biological, um, you know, they, they wouldn't necessarily need to invent anything new. They would just need to apply currently known techniques in, in kind of a malevolently directed fashion. Um, I think there are some concerns and some risks there with respect to, um, with respect to things that don't invent new technologies, um, but do, um, make them more accessible. Uh, and so, I mean, I think the sort of, I think the question of is, um, you know, how ... What would the effect on the world be if there was a sufficiently sophisticated LLM, uh, that, uh, you know, it could, it could help anybody, you know, synthesize and disperse smallpox? Um, like I don't, I don't know that the laws of physics, uh, prohibit such an LLM existing. I, I presume they don't. Um, and would the world be fine if such an LLM was, you know, widely distributed? Like, maybe but, you know, maybe not, right? So, so I think, I think there is that kind of threat factor. But my point is, I don't think, um, knowledge at the frontier of biology is the, is the relevant margin here, and if we take seriously what, um, what ... Um, you know, that, that this is ... I mean, we don't need crazy AI risks to motivate this. You know, where (laughs) the world is perfectly capable of, uh, of originating, you know, really severe pandemics and pathogens itself, plus all the other diseases that are, that are not pathogenic. So, you know, we, we got other problems. But, you know, whether we care about the, the possible, you know, kind of dual use harms you just mentioned or we just care about the things that already exist, to ameliorate both of those, we do need enhancement of our capabilities, um, like we ... Um, there are, you know, there are a lot of biological problems we don't today know how to solve.

   8. DPDwarkesh Patel34:40

      Mm-hmm.

   9. PCPatrick Collison34:40

      Uh, and so I think in that respect, if one were to do what you're proposing and try to advance the, um, the, the, the defensive side of this, I don't know that what one would do would necessarily be that different, um, uh, because there are just fundamental capabilities that we would presumably need to have, um, that we, that we don't today have.

   10. DPDwarkesh Patel35:01

       Mm-hmm.

   11. PCPatrick Collison35:01

       Uh, and by trying to solve current human diseases, I think you're probably also pursuing something pretty close to the best steps to solve the potential diseases that, you know, malicious actors

4. 35:13 – 44:32

   AI & Fast Grants

   1. PCPatrick Collison35:13

      could cause in the future.

   2. DPDwarkesh Patel35:14

      Yeah. That makes sense. Um, uh, I mean, uh, z- zooming out from bio risk in particular, just like how are you thinking about AI these days?

   3. PCPatrick Collison35:22

      Well, I, I, you know, I th- I think, I think everyone has to be sort of high perplexity, um, in the sense that, I mean, um, the, the verdict that one might have given at the beginning, you know, we're recording this here pretty close to the beginning of 2024. The verdict one might have given at the beginning of, you know, '23, '22, '21, you know, back, say, the last eight years do-... those would all, I think, have looked pretty different. Um, I mean, maybe Gwern, uh, might've, uh, might've scored the best, um, uh, from 2019 or something, uh, onwards. Uh, but, uh, but broadly speaking it's been pretty, pretty difficult, I think, to forecast. Um, and so I think, I think the basic position to our first order has to be one of, you know, some degree of humility. Um, I think as your blog post identifies, the, the, the big question right now is, you know, to what degree, um, to what degree scaling laws hold. Uh, and I guess if they hold then, uh, you know, what exactly is it that we're, um... Well, asymptoting is maybe a presumptuous word. Maybe it's not an asymptot, but, but, like, what, what is it that we're approaching? You know, it's not... We don't necessarily know the shape of that thing, whatever it is. Um, and, yeah, I, I think, I think it's a lot of... Um, yeah, I think, I think how one should feel, um, needs to be, uh, uh, or ought to be very sensitive to, like, the exact parameters of, you know, those curves. And I just don't think anyone, anyone knows what, uh, what the, like, the, the true value of those parameters actually are. So, you know, it's, it's, it's clearly going to be, um, important. Uh, is already important today, and it has a pretty central bearing on, you know, both, uh, both Stripe and, and Arc. Um, and we'll see.

   4. DPDwarkesh Patel37:04

      Yeah. I, I, I wonder if the meta lesson here, and I totally agree with that sort of general sentiment, but I wonder if the meta lesson that we got from COVID, for example, and with things like Fast Grants was, um, um, you, you obviously can't predict these things in advance, but the most important thing, even in addition to these, like, specific sort of countermeasures trying to come up in advance is, like, when the thing is happening-

   5. PCPatrick Collison37:24

      Yeah.

   6. DPDwarkesh Patel37:24

      ... having, uh, competent i- uh, individuals who can synthesize and organize information.

   7. PCPatrick Collison37:28

      Yes.

   8. DPDwarkesh Patel37:28

      And also having these, like, new initiatives and institutions-

   9. PCPatrick Collison37:32

      Yeah.

   10. DPDwarkesh Patel37:32

       ... to, um, uh, get the right thing done.

   11. PCPatrick Collison37:35

       Yes, yes. The, the adaptability premium-

   12. DPDwarkesh Patel37:36

       Yeah.

   13. PCPatrick Collison37:36

       ... is probably gonna go way up over the next decade.

   14. DPDwarkesh Patel37:38

       Yeah. And with that in mind, and I know you have already a couple of day jobs. (laughs)

   15. PCPatrick Collison37:42

       (laughs)

   16. DPDwarkesh Patel37:42

       Um, but, yeah. Yeah, I feel like something like Fast Grants, like, when the time comes down to it, like, I, I don't know, you're like, it should be like, uh... You know, you'd be, like, the, uh, one of the top people you could think of in terms of having expertise and respect in a wide range of domains and competency as a leader.

   17. PCPatrick Collison37:56

       Oh.

   18. DPDwarkesh Patel37:56

       Um, I, I don't know. Just, like, keep it in the back of your mind or (laughs) maybe in the middle of your mind given how far we are into the transition.

   19. PCPatrick Collison38:02

       Well, uh, Fast Grants was, um, was, uh, was three beloved, uh, uh, squirrels, uh, in a, in a trench coat. Um, or I guess, well, I was one of the squirrels, so, um, I don't know. Self-love it. But, um, uh, uh, it was, um, also Tyler Cowen, uh, who's uh, who's an amazing person, a great friend, uh, and then my wife, um, uh, who's also, uh, one of, uh, one of our ex-co-founders. And so, you know, Fast Grants was not this, like, giant, you know, impressive edifice-

   20. DPDwarkesh Patel38:26

       (laughs)

   21. PCPatrick Collison38:26

       ... that would, uh, would qualify me for, for anything at all.

   22. DPDwarkesh Patel38:29

       But it doesn't have to be giant, right? To have that kind of big impact.

   23. PCPatrick Collison38:32

       Um, I g- yeah. I, I guess as a, as an objective matter, that's true. I mean, look, uh, I... John and I try to be very self-aware of the, um, limits of our expertise, uh, which are, which are, you know, very... (laughs) which are very proximate to us. Um, and, uh, I'm sure if, if, um, you know, something like that was necessary, they'd be... I mean, look at Operation Warp Speed. Um, they chose a super effective domain expert, uh-

   24. DPDwarkesh Patel38:56

       Mm-hmm.

   25. PCPatrick Collison38:56

       ... Moncef Slaoui to run that, and it was just, like, monstrously successful.

   26. DPDwarkesh Patel39:00

       Yeah.

   27. PCPatrick Collison39:00

       Uh, uh, like, truly remarkable. And, um, I don't know who the Moncef Slaoui of... I guess it would depend whatever the problem in question is, but I think my recommendation would be like figure out who Moncef is, and, like, go hire Moncef. Um, and, uh, I, I, and I think it is extremely unlikely... I think anybody who deemed me the Moncef of that thing, um, you know, uh, uh, is, is probably mistaken.

   28. DPDwarkesh Patel39:21

       I, I think you're being too humble. But, um, just staying on Fast Grants. Uh, um, you, you, now we have the retrospective of what, how effective the Fast Grants recipients were, uh, compared to the other grants that were given out by, let's say, the NIH or NSF. Um, to your knowledge, what has been the reaction of these institutions to, uh, the discrepancy between, uh, the speed and effectiveness of Fast Grants? Have they, like, analyzed their protocols and, like, what happened during COVID? Is there any sort of, uh, um, retrospective there on their part?

   29. PCPatrick Collison39:49

       Not to my knowledge, but, but I, I don't want that to sound like an indictment. Like, maybe they've done a lot of reflection and, uh, and, you know, I just don't know about it. Like, I, I don't think I would know about it even if it had happened. Um, so, um, I don't know. Um, uh, I... I mean, look, most... Um, well, I don't know anything about the response at CDC or FDA or NIH or NSF or any of the relevant organizations or their international equivalents. And so none of what I'm saying, uh, should be taken as, like, specifically... Not, not only not critical of them, but, but not even a comment on them. I just, like, don't know what they did. Um, but, but in general, organizations are, are, you know, not awesome at, uh, at self-reflection. And I, I, I think... I, I, I assume as a default prior that some of the dynamics we discussed at the beginning of this are rooted there, where none of the people who started those organizations are there today. Um, and so, you know, what exactly are the incentives of those leaders? And, um, you know, uh... You know, I haven't... Yeah. Um, uh, it's, it's, it's not clear to me who would have the incentive, uh, to, uh, to, you know, really take stock in a fully objective and self-critical way, to, to figure out, you know, what, what was done well and what was done poorly.

   30. DPDwarkesh Patel41:16

       Mm. I promise not to be too myopic about AI, but one more question. Um, uh, long term we can't forecast, maybe even medium term we can't, but near term it looks like, you know, we might have things that look like AI agents, um, and they might need to trade.
5. 44:32 – 56:30

   Stripe history

   1. PCPatrick Collison44:32

      that plays a role.

   2. DPDwarkesh Patel44:32

      How long before Stripe was founded do you think a product like Stripe could have been invented?

   3. PCPatrick Collison44:37

      Hm, that's a good question. Um, well, in ... m- you know, depending on what exactly you define Stripe as being, I think, um, conceivably decades earlier, uh, in that, I mean, on some level PayPal is a kind of Stripe, um, and, you know, there were many payments companies before PayPal, and you know, you could go all the way back to, you know, cash registers or something, right? So, so it depends on, on, on, yeah, the- these definitional questions. Um, I mean, the particular secular tailwinds that we benefited from around the rise of app stores and, you know, the on-demand economy and maybe the startup boom post, you know, YC and, uh, after the financial crisis, you know, th- those, those particular tailwinds were, yeah, idiosyncratic and specific to Stripe and I guess, you know, the, the GFC was, uh, '08, '09 and, uh, Stripe was founded in 2010 and so, you know, (laughs) inasmuch as you define those as being, uh, core, then n- not that much earlier. But, uh, but mo- mostly my story with Stripe is one of, is one of market inefficiency and, uh, I do wonder, yeah, why, why much of it didn't happen sooner.

   4. DPDwarkesh Patel45:44

      Yeah. I, I, I always find it really interesting when there's these cases where, um, it wasn't even the case that like, "Oh, well, it could have been started sooner but there was nobody in the market." There were, like, many people in the market and they weren't just like random people.

   5. PCPatrick Collison45:54

      Right.

   6. DPDwarkesh Patel45:54

      There were technology companies headquartered in San Francisco-

   7. PCPatrick Collison45:56

      Yes.

   8. DPDwarkesh Patel45:56

      ... who were in the market. Uh, d- do you have some explanation for why it didn't occur to them?

   9. PCPatrick Collison46:02

      I'm hesitant to generalize too much because well, I only have maybe, uh, you know, n equals one experience and so I think it's, it's dangerous to over-extrapolate from that. Um, m- maybe n equals two now with Arc as a, I mean, a very different kind of organization, but, but an organization nonetheless.

   10. DPDwarkesh Patel46:17

       Or if you include all the features of Stripe, n equals like 10, 20, something. (laughs)

   11. PCPatrick Collison46:20

       Yeah, right, okay, so, so, so yes, depending on your definition maybe, maybe there's some kind of samples out there. I guess my general view is most products and most businesses, um ... like things can just be done much better, um, and I think, I think moats are, are typically kind of overrated, uh, and I mean, payments are a great example of a domain where on a logical basis you would say that there are, you know, so many sources of defensibility, where there's, there's you know, the network effects of the account holders and there's the, uh, data network effects slash economy of scale, you know, for fraud and so forth and there are regulatory moats and barriers, you know, and, and, and. You know, and yet, (laughs) uh, you know, not, not only does Stripe exist but there are lots of other ... I mean, there, there's a whole fintech ecosystem today, right? Um, so yeah, I, I think it gets down to, you know, kind of deep questions of, you know, what are the, uh, what's the binding constraint on just the number of effective organizations that exist in the world and, you know, for any given sector why is it that number of companies, you know, rather than twice that number of companies, uh, and so on. I think it's about motivation and ideas and peoples', you know, willingness and determination to organize talent and so forth, but, but these kinds of more sociocultural explanations rather than ... I mean, Hamilton Helmer is probably the leading scholar of some of the, um, the sources of defensibility for businesses. Um, uh, he has this, uh, you know, kind of niche but very well-known in the niche book, uh, uh, uh, called, uh, Seven Powers, uh, and, uh, it, it kind of attempts to disaggregate all the various sources of, uh, of market power in this respect, um, and, you know, I think that is, well, true and important, uh, insofar as it goes, um, but...... none, nonetheless, it's kind of, yeah, strange to me that nobody had done Stripe before Stripe.

   12. DPDwarkesh Patel48:06

       Mm-hmm. Wh- when you think about the, uh, the fact that moats are overrated and just, like, doing the thing is underrated, does that, uh, like what is Stripe's moat in that context? Does that make you, uh, you know, think differently about Stripe's moat?

   13. PCPatrick Collison48:17

       I, um, yes, one, um, and I guess I, I guess I do think that, um, that w- w- one can have, um, organizational and cultural moats, um, and maybe this contradicts what I was just saying or maybe it's consistent with it in the sense that it's a kind of cultural explanation. Um, and, uh, I think that in as much as we have a moat, it's because we have a very good understanding of our domain and a set of people who actually care about solving the problems, um, and who are, uh, continually paranoid at the prospect that we might be forgetting something important and sort of trying to figure out what the important thing that could supplant Stripe's approaches is and making sure that we build those first, uh, and so forth. I, I think, I think, um, organizations that are... I mean, there's, um, you're familiar with Conquest Laws and there's Conquest's Third Law, I guess, which is that one should model organizations as if they are run by a cabal of their enemies.

   14. DPDwarkesh Patel49:20

       Right, right.

   15. PCPatrick Collison49:21

       Um, and, you know, obviously it's or presumably it's tongue in cheek, uh, but, you know, it's interesting to try to think about, like, well, kind of what is the kernel of truth in that and why would it be there? And I think what's going on is that I think most organizations when they start out are actually trying to achieve their stated goals. Um, like, somebody started the organization for a reason and, and probably it was for the stated reason. Um, but then over time, you know, that person and that, you know, set of people who initially populate the organization depart and some set of new people come to, you know, take their place, and there's, there's, you know, multiple versions of that. There's, there's generational turnover on a, on a continuous basis. Um, but say for the fifth generation, like, why are they there and to what degree do kind of their particular specific local incentives align with the nominal originally stated, uh, goals of the organization? And I, I think there can be a lot of misalignment there, right, where they're following a, a local path, um, and, um, and conceivably even the leader of the organization, not even through any, like, fault of their own per se necessarily, just that they're a human and they have their own incentives and again the original kind of constitutional incentives of the, of the organization might be quite different. Um, and so yeah, I, I think this, I think this phenomenon is, is kind of a fact of life and, uh, and I think these kinds of explanations for me, um, are, um, are much more explanatory in trying to figure out sort of why some of these things either happen or don't. Uh, and to your question, like, in as much as Stripe has a moat, what is it? Uh, I think it's that, um, I mean others can judge to what degree it's actually, you know, manifested and rooted in practice. I think it is, but, you know, I'm a biased observer, but, uh, I think it would be that people at Stripe really care about solving the problems that we say we are trying to solve.

   16. DPDwarkesh Patel51:06

       Mm-hmm. Yeah. Uh, the, the point about the misalignment over generations or over time is interesting. Um, and a- actually, do you have examples of institutions which have for decades or hundreds of years managed to keep their original not only mission statement, but the organizational, uh, competence? Because if you think of tech companies, they're, even the oldest tech companies have not been arou- around that long, right? And they're some of the biggest tech companies in the world and the median age of the corporation is, like, famously low. Uh, what, what is a good example here?

   17. PCPatrick Collison51:35

       I think some of the explanations around the, uh, the effects of shareholder capitalism and sort of the idea that shareholder capita- capitalism as a mechanism does in fact have, um, have, you know, uh, has some consequence with respect to the, uh, incentives or- of organizations and their long term fates. Um, I think those theories have some credibility and I think it is very plausible that, uh, shareholder capitalism even attenuates, uh, the, the, um, the durations of these organizations. I'm not saying that's definitively true, but I, I find it, you know, credible the, the idea that it is. Um, it's not clear to me that that's necessarily bad even if it is true, right? Uh, in that are we on the side of the humans or the kind of aggregate innovation in the world or on the side of, like, the corporations, you know, qua legal entities? Um, and, you know, yeah, it's not, it's not clear to me the answer should be the third. Um, at the same time, uh, uh, or maybe in fact consistent with that, you know, if you look at say Europe or, you know, some other places like, you know, in, in Denmark, um, there's, uh, for, for reasons, uh, related to the, the tax code there, a lot of organizations, um, are either controlled by or, you know, very substantially held by, um, by, uh, uh, nonprofit foundations. And so Novo Nordisk, for example, you know, the, the GLP company, uh, but, uh, but Maersk, the shipping company, um, and I believe also Lego, a, a lot of these corporations are, are, um, are controlled by, um, and again usually have a lot of their stock held by foundations. That has the secondary effect, um, in many cases where they actually do embed in a legally binding constitution their mission. Um, and so, you know, I'm not an expert on Novo Nordisk, but, um, I was, I was, um, happened to get a book about it over Thanksgiving, uh, and, um, my, um, uh, and actually there's also a book on the Danish industrial foundations. Um, uh, but it's enshrined in their constitution that they have to make insulin, you know, broadly available, uh, really cheaply or at least cheaply in Scandinavian countries and then I think they're allowed to charge market prices elsewhere. Um, and, uh, and I think that, uh, a- and then the, the, you know, the rest of their, of their profits, they have to, uh, they're, they're again legally obligated to reinvest in R&D. Um, you know, is that somehow causal in the fact that they then invented, you know, one of the most remarkable, uh, pharmacological discoveries of the last 20 years, uh, in these, uh, in these, you know, GLP-1, uh, agonists? I mean, you know, plausibly, um, and, uh, and so I think, um, yeah, I, I think, I think these questions around, you know, the, the, you know...... why it is that the median age of organizations, you know, and corporations is what it is, are, are definitely interesting, and, um, I, I suspect it's a somewhat contingent aspect of how we've chosen to organize large corporations, you know, in the US today.

   18. DPDwarkesh Patel54:18

       The, the thing you were mentioning about, uh, this firm seems very similar to the export-led growth in Asian, uh, uh, subject.

   19. PCPatrick Collison54:23

       Totally, 100%. Yeah, yeah. Got it.

   20. DPDwarkesh Patel54:24

       Right? Like the, uh, the FTRFs, the, this one company you're tasked with making the cars, so you better make the cars good. Um, y- w- you ha- you have no competition but you gotta invent the best car in the world.

   21. PCPatrick Collison54:31

       Yes, yes, yes. And I think, I mean, you know, we are all fans, uh, of, you know, Smith and Ricardo and, you know, all these characters, um, and, um, and, you know, even they, I think, are, are sort of less dogmatically a- attached to free trade, uh, than perhaps, you know, people today interpret them as being.

   22. DPDwarkesh Patel54:49

       Ah.

   23. PCPatrick Collison54:49

       But, but I think p- people like, you know, Friedrich List, uh, and, you know, those other not quite contemporaries but, you know, quasi-contemporaries, um, uh, you know, are, are maybe on a relative basis underrated and I do think...

   24. DPDwarkesh Patel54:59

       Ah.

   25. PCPatrick Collison54:59

       I mean, in as much as you believe the kind of sociological, cultural skill, whatever, um, even vague alignment, um, uh, not in the AI sense, but in the-

   26. DPDwarkesh Patel55:07

       (laughs)

   27. PCPatrick Collison55:07

       ... just in the, in the-

   28. DPDwarkesh Patel55:08

       Sure.

   29. PCPatrick Collison55:08

       ... more kind of interpersonal sense, in as much as you think these are, are important and explanatory, then, um, then yeah, I think, uh, you know, you, you end up thinking about some of the things, the things you just raised.

   30. DPDwarkesh Patel55:18

       Tha- that's really interesting to hear you say that, because if you think about Stripe's mission, right? It's to, uh, uh, facilitate global trade-
6. 56:30 – 1:02:25

   Stripe Climate

   1. PCPatrick Collison56:30

      yeah.

   2. DPDwarkesh Patel56:30

      All right, that's so interesting. Um, is, uh, I'm trying to... I think th- there's an interesting thread here in how it relates to Stripe Climate in that you're, uh, I don't know, like subsidizing these learning curves that these, uh, East Asian countries did for their own c- internal companies. Um, I mean, you, I mean, you haven't picked out like a specific company that's gonna necessarily be the Kia of (laughs) -

   3. PCPatrick Collison56:48

      Yeah.

   4. DPDwarkesh Patel56:48

      ... um, uh, carbon sequestration, but yeah, how do you think about this?

   5. PCPatrick Collison56:51

      Well, may- maybe a way to unify s- um, the, the, the two points, and I'll speak about Stripe Climate in a second, um, is that I think, uh, I guess it's, um, it's Say's law, um, about, you know, demand-

   6. DPDwarkesh Patel57:02

      Right.

   7. PCPatrick Collison57:02

      ... creating supply and in as much as Stripe aggregates more and more global demand, uh, I guess part of the, I don't know, it seems too self-aggrandizing to call it, you know, the, the theory of Stripe, but you know, some, some vague hunch in Stripe, uh, is that, uh, that, that, that aggregation of demand can have important, uh, expansionary effects with respect to the ensuing supply. And yeah, Stripe Climate is, is some version of this, you know, hypothesis applied on a much scal- smaller scale than Stripe itself, but, but, but still, you know, real and, and well, we'll see, maybe important.

   8. DPDwarkesh Patel57:30

      Right.

   9. PCPatrick Collison57:30

      Uh, and the, the, the basic idea just for, for folks who aren't familiar, which I assume is most of your audience...

   10. DPDwarkesh Patel57:34

       (laughs)

   11. PCPatrick Collison57:35

       Um, uh, so we, we, we observed in 2018, I guess, that, uh, that everyone seems to agree that carbon removal will be very important, um, and, you know, even if we decarbonize the economy on the kind of time scale that optimistic people, um, you know, on, on the most optimistic, uh, time frames, there'll still be, you know, an accumulated stock of carbon that, you know, is a problem. Um, it sounded pretty weird, uh, there were, there were virtually no, uh, uh, carbon removal companies in the world, uh, in 2018, maybe there were, you know, two or three or something. Uh, no companies had ever purchased from a carbon removal company. These were, these were really sort of science projects. And so we thought, "Well, you know, (laughs) somebody's got to start, uh, and e- it might be valuable to, you know, not only transfer some dollars but to kind of confer some credibility on this sector." Not that Stripe is the world's most credible company, but you know, it's better than nothing. Uh, and so we started contracting with some of these carbon removal companies. Um, that went pretty well, and, and, and they seemed kind of appreciative of us and so we, we thought somewhat more about this and we, uh, then in 2021, uh, formed Frontier, which is an AMC, an advanced market commitment. So inspired by, um, the, the first ten, yet AMC, which was a, a pre-commitment to purchase vaccines for, uh, uh, developing world countries for diseases that, I mean, uh, well, e- either were kind of market failures where pharma companies hadn't purchased the vaccines or, or were just like the profits weren't sufficient to, to pay for the program. So we said do this for carbon removal. We raised a billion dollars, uh, Stripe was, um, uh, the, you know, the, the, the first investor. We're not actually investing, we're just buying, so the, the first company to commit. But then we're joined by Shopify and Alphabet and Meta and JP Morgan and a, a bunch of other companies. Um, and now there's, now there's like a fairly active sector of carbon removal companies. Uh, there are, I think S- uh, Frontier has contracted with 40, betw- between 40 and 50 companies, um, the overwhelming majority of which didn't exist when we started, started out with this. Actually we, we ran an anonymous survey, uh, back, um, the end of last year and, you know, we asked them to what degree wa- was, was the existence of Frontier, you know, uh, somewhat causal in, you know, their starting the company in the first place. Again, this was in an anonymous survey and I think it was 70, 74% of the companies said that Frontier, you know, played a causal role, uh, in their starting the company. So yeah, I think, um, I think this, these inducement effects, uh, can be, can be pretty significant.

   12. DPDwarkesh Patel59:57

       Yeah, tha- tha- that's huge. Well, what are other ideas you've come across where an AMC would be a, effective instrument of moving forward the tech?

   13. PCPatrick Collison1:00:03

       Hmm, that's a good question but we've, we've actually been having some of that discussion internally, um, just it's not that we plan on doing it ourselves necessarily, but, um, just wondering are there-... are there, are there people we should share our technology with? Not that it's even technology per se, uh, but, but share our experience with or something and, and try to help along. Um, I mean, there's still, I think there's still a lot of stuff in, in, in the biomedical field. Um, and I mean, patents are, um, you know, p- patents are, are pretty useful insofar as they go, but you know, they, they, there's a lot of, um, there's a lot of innovation that seems like it would be, you know, socially beneficial that p- patents don't provide a way to cover the cost of. And so you know, th- there was some excitement, um, a few years ago about, uh, mannose, uh, which is a, it's a sugar. Um, and, um, there's one paper that, uh, or maybe a few papers, uh, I, I can't remember, that, that suggests that maybe, maybe tumors will selectively take up mannose rather than glucose, uh, but they can't actually metabolize it properly and so th- they just die. Uh, and so you know, maybe this could be an effective, um, you know, uh, onco treatment of some sort. Um, but mannose is, is like, it's a, it's a, it's a generic sugar. It's been, you know, understood for, I, I guess more than a century, uh, and you couldn't patent it, importantly. Um, and so it's not clear who has the incentive to even, you know, fund the work to test, you know, whether or not this would actually work in practice. And you know, (laughs) this is not an endorsement of, uh, of mannose, but just there, there are things of this shape, uh, where there's something where you can clearly see, wow, that, that might be very beneficial, but it's not totally clear how the kind of economic structure of the market can, can make it possible. So I think there are still a lot of those, uh, across the biomedical landscape. I mean, look, there are still a lot of vaccines, uh, that, you know, could in principle exist that, that don't. I mean, Lyme disease. Uh, you know, there's no, um... There's, there's one vaccine that, uh, that was withdrawn from the market over some safety concerns that I, I think were misplaced, but you know, it's still not a vaccine. Um, uh-

   14. DPDwarkesh Patel1:02:10

       It's not even that well understood, right? Like people have chronic Lyme disease.

   15. PCPatrick Collison1:02:13

       Right. Yeah.

   16. DPDwarkesh Patel1:02:13

       We don't know like if it's legit or not. (laughs)

   17. PCPatrick Collison1:02:14

       Exactly. Yes, yes, yes. Um, and, but it's, it's a good question. I mean, look, may- maybe some of your listeners will know, uh, yeah, will have ideas for, for fields for, you know, we, we sorely need an AMC.

7. 1:02:25 – 1:12:37

   Beauty & APIs

   1. PCPatrick Collison1:02:25

   2. DPDwarkesh Patel1:02:25

      Mm-hmm. Yeah. Um, uh, I want to go back to Stripe for a second. So, uh, you're famously, um, uh, you know, appreciative of craft and beauty, but also you appreciate the power of scale and growth. Is there a type of-

   3. PCPatrick Collison1:02:38

      And speed.

   4. DPDwarkesh Patel1:02:39

      Oh, interesting. Okay, yeah. Th- yeah. Yep. Um, but is there a type of craft, uh, that is just not amenable to speed, growth, scale? If you think of like a Japanese chef, he's like learning to cook rice for, you know, a decade and then he can move on to the sushi or something. Uh, is that just not competitive in the modern world?

   5. PCPatrick Collison1:02:55

      Craft, um, scale, and, and speed, I don't know if they're strictly necessarily in tension in every case, but they're definitely frequently in tension. So just yes, I think is, is kind of one short answer to, to that. At the same time, um, a lot of the most successful companies, um, are those that I think are distinguished by the extent to which they, they exhibit, um, appreciation for and like skill in realizing craft, um, and, and beauty. Uh, and so LVMH is one of the largest companies in the world and like that's literally their business.

   6. DPDwarkesh Patel1:03:32

      Yeah.

   7. PCPatrick Collison1:03:32

      Um, I mean, I think Tesla is pretty good at this. Uh, I mean, they're good at many things, but you know, including this. Um, obviously there's Apple. Um, I mean, TSMC is a kind of, you know, uh, uh, it's not the, the Japanese, uh, sushi chef you, you, uh, you mentioned, but it is the, it's the TSMC chip sushi chef-

   8. DPDwarkesh Patel1:03:51

      Yeah, yeah.

   9. PCPatrick Collison1:03:51

      ... uh, uh, in, uh, in, in, in Taiwan. Uh, and, and so much again, tacit knowledge and, you know, difficult to, to, to transfer skills. Um, so I, I think it might be the case that, um, that craft, uh, and the pursuit of it is, is as important as, as it's ever been. And, and certainly as, as Stripe has gotten larger, I think we ourselves have come to greater conviction in this where... I, I think part of what's interesting about, um, these aesthetic qualities is they're generally speaking unquantifiable. I don't know if they're intrinsically unquantifiable, like maybe you could like train a model to do so, uh, or, or something. But today, they're, they're broadly speaking unquantifiable, and yet they are actually, um, you know, they, they influence people in significant ways. People very demonstrably care about aesthetics. Um, and they care about, you know, if they're a company, they care about the aesthetic characteristics of the products that, you know, they produce just like on an intuitive p- uh, level, people know that that's true. But you know, yeah, it's, it's difficult to manage that a- a- at an organizational level where there isn't a PnL associated with it, and if you're screwing it up, you, you don't see a neat time series, uh, decline. But over the 14 years of Stripe, you know, we have, I guess through a kind of, not exactly trial and error, but just by studying cases where things worked well at Stripe and cases where things worked less well and what customers responded well to and so on, it really seems clear to us that even in a domain like ours where we are selling primarily to businesses, um, that this is something that's, that's truly important. Um, and also that in as much as, you know, getting back to what we were discussing previously, you want, um, in, in as much as the, the sociology, um, and the kind of cultural explanations of defensibility are real, the best people, you know, consider themselves craftspeople in their domain. And they really, above almost all else, want to work with the best other people. And so I think it's, it may almost be true that even if from a customer-facing standpoint, uh, craft was not valued by the market, you actually might still want to build an organization that indexes very heavily on this because you, you just want the best people for...... other reasons.

   10. DPDwarkesh Patel1:06:18

       Mm-hmm.

   11. PCPatrick Collison1:06:18

       And now as it happens, I think customers do in fact value it, and I think the evidence is, is broadly consistent with that.

   12. DPDwarkesh Patel1:06:23

       Mm.

   13. PCPatrick Collison1:06:23

       Um, but, but yeah, I think, um, I think it's very hard to assemble groups of the best people if you don't take the practice of the work super seriously.

   14. DPDwarkesh Patel1:06:32

       Mm-hmm. Uh, what kind of beauty, or craft, or simplicity is more important, interface or implementation? There's famously that essay that Unix is successful because the implementation is simple and not the interface.

   15. PCPatrick Collison1:06:44

       Yeah. The, I guess the interface is kind of simple, um, but it, there's a lot of, like, asterisks, and caveats, and, you know, edge cases that, that, yeah, I guess Unix doesn't handle for you. Um, uh...

   16. DPDwarkesh Patel1:06:55

       But Stripe does, right?

   17. PCPatrick Collison1:06:56

       I mean, look, p- p- presumably it depends what you're, what you're building, right? Uh, for TikTok, it's probably more important that their interface is simple, and even if their implementation's a mess, you know, that's, that's probably okay. Uh, nothing it is, I have no idea. Um, uh, whereas for Stripe, uh, yeah, people are, people are unsellable purchasing our architecture-

   18. DPDwarkesh Patel1:07:15

       Mm.

   19. PCPatrick Collison1:07:15

       ... or, you know, purchasing their ability to do certain things, and some set of things rather than some different set of things, you know, be- because of what our architecture makes easy and what, and, uh, and makes possible. Now, I don't think, I mean, if by interface you mean the, the visual, you know, gooey interface, um, then, you know, maybe we can draw some separation there, but I guess we, we don't really draw that distinction. Like, we think of the interface to Stripe as being the architecture.

   20. DPDwarkesh Patel1:07:37

       Right. Huh.

   21. PCPatrick Collison1:07:38

       Uh, we're, we're selling, um, you know, (laughs) no one else seems to agree with me, but I often think of Stripe in the co- um, as similar to Mathematica, (laughs) uh, where we're selling kind of a, a, um, a self-contained universe to model whatever it is, uh, um, is of interest to you, and that, that, you know, you care about. And we're providing some primitives and some, um, some yes, kind of interfaces, and tools, and so forth to enable your modeling, um, but fundamentally, we're helping you do something in your own terms. Uh, and, you know, in that sense, I don't think kind of the architecture and the, and the interface are, are necessarily that separable.

   22. DPDwarkesh Patel1:08:18

       That, that, that's really interesting analogy. Um, although, I mean, if you think of Mathematica, the, the, like, the entry that that's giving you to the, is, you know, just like the platonic objects of math.

   23. PCPatrick Collison1:08:28

       Right.

   24. DPDwarkesh Patel1:08:28

       Whereas for you guys, it's like the entry is that to, like, Visa error codes, right? (laughs)

   25. PCPatrick Collison1:08:32

       Yeah. Right, right, right, right. Yeah.

   26. DPDwarkesh Patel1:08:33

       Like, the, the end object is not the, uh, the, you know, the platonic, uh... (laughs)

   27. PCPatrick Collison1:08:36

       Tha- tha- that's, that's true, though in, in both cases, yeah, I think y- yes. Uh, so, so the analogy falls down in, in a few respects. Um, but, um, but look, I mean, the idea of a transaction is pretty fundamental, um, and is, you know, roughly as old as, you know, the, the quadratic equation, uh, or something.

   28. DPDwarkesh Patel1:08:53

       (laughs)

   29. PCPatrick Collison1:08:54

       I, I guess the transaction's older. And, uh, and Mathematica (coughs) especially today, excuse me, today, now supports all kinds of, I mean to a very impressive extent, supports all kinds of like crazy arcane stuff. Like, if you go through the more obscure packages of Mathematica, you can definitely find things that are, uh, I think much less broadly, uh, employed and, and understood even than, than Visa error codes or something.

   30. DPDwarkesh Patel1:09:15

       (laughs)
8. 1:12:37 – 1:29:02

   Financial innards

   1. PCPatrick Collison1:12:37

      to do.

   2. DPDwarkesh Patel1:12:37

      Is Visa and Master an example of this? And one might even say that one of the downsides of being able to use an implementation for many decades in the future is, even if it's, it's self-sustainable and you have this ecosystem and equilibrium set around it, uh, if you, um, you know, can't modify it, uh, just because of people's locus and incentives, you get stuck in this, uh, uh, equilibrium that's worse than it could be otherwise.

   3. PCPatrick Collison1:12:58

      I see. I see. Um, I think the card networks generally, Visa and Mastercard, um, are, are, are, are a pretty good equilibrium. Um, where it- it's easy to judge today, um, with, you know, the world as it exists in 2024, but I think you have to look at the world as it was when they started out, uh, and the particular problems that they're solving.

   4. DPDwarkesh Patel1:13:19

      Yeah.

   5. PCPatrick Collison1:13:19

      And I think when you compare the, I don't know, financial landscape in, in the US or in the Western world to those in other places, it's certainly not clear to me that the US has, um, has gotten, you know, a, a, a bad hand, so to speak, or is, is somehow stuck in any meaningful way. Um, so, you know, the current works to a couple of things. They, uh, they... originally, they, um, you know, they were, they were designed to, uh, to replace store credit-

Episode duration: 1:55:31