#32 Patrick Collison: Earning Your Stripes
01:48:40 Share

Welcome to the Farnham Street Podcast called The Knowledge Project. I'm your host, Shane Parrish, the curator behind the Farnham Street blog, which is an online community focused on mastering the best of what other people have already figured out. The Knowledge Project is where we talk with interesting people to uncover the frameworks you can use to learn more in less time, make better decisions, and live a happier and more meaningful life.

On this episode, I have Patrick Collison, the co-founder of Stripe, which he started with his younger brother, John, in 2011. While Stripe started as a company to make online payments easier, it's morphed into an internet infrastructure company. Patrick is one of the most well-read and thoughtful people I've ever met. After listening to this conversation, you'll realize his success is less about luck and more about thought. I'm pleased to have Patrick Collison on the show.

so

The IKEA Business Network is now open for small businesses and entrepreneurs. Join for free today to get access to interior design services to help you make the most of your workspace, employee well-being benefits to help you and your people grow, and amazing discounts on travel, insurance, and IKEA purchases, deliveries, and more. Take your small business to the next level when you sign up for the IKEA Business Network for free today by searching IKEA Business Network.

You have the unique background of having dropped out of high school and dropped out of university. Can you explain what went through your mind dropping out of high school? Well, I didn't technically speaking drop out, although I sort of practically speaking did. But given my lack of education credentials elsewhere, I do...

I should, for the sake of my parents, insist that I did in fact formally speak and graduate from high school. But I guess what happened is that I'd become very interested in programming and I sort of wanted to spend as much time on it as possible. And Ireland actually has this kind of interesting thing called transition year, this year between the sort of two major exams of kind of high school or at least Ireland's high school equivalent. And in transition year, it's sort of a

It's a formally designated year that's optional where you can go and pursue things that you might not otherwise naturally tend to pursue. And the school tends to be kind of much more permissive of going and spending three months abroad or going and doing some work experience in this area or whatever the case might be. And so in that year, I basically decided to spend as much of it as possible programming. And so I did that. And then I returned to school for kind of the latter half of, again, Ireland's kind of high school system. And...

it felt so much slower and less fun. And so I tried to see if-- well, as part of the programming, I had visited the US for the first time. I had gone to Stanford for the 2005 International Lisp Conference. And it was a fairly small conference.

But it was very eye-opening for me. And I remember walking around Stanford and thinking, man, American colleges seem great. And so back in high school in Ireland, I decided to see if there was some way that I could just go to college in the US the subsequent year. And it was sort of a long story, but I eventually figured out that I could not do it if I did the standard Irish kind of thing.

follow the standard Irish education path, but that I could do it if I did the British sort of terminal examination. And so I kind of resumed my sort of self-education, except instead of programming, I was now studying for these British exams and did that for the subsequent year and ended up starting at MIT the next fall. And how did we get from MIT to where we are today, which is Stripes offices in San Francisco? Well...

It's sort of a long and torturous story, and I'll spare you most of the less interesting details. I guess the overarching thing is while people in the US have sort of grown up in an environment in which college attendance is sort of really prioritized from an early age and sort of, you know, you're optimizing your extracurricular activities from the time you're 14 and you're choosing your kindergarten on the basis of what the sort of downstream college acceptance rates look like and all that kind of stuff.

Of course, growing up in Ireland, that sort of wasn't part of the culture or discourse or environment at all. And so by the time I got to MIT and just to college in general, it didn't feel like that big a deal. It didn't feel like sort of this was the terminal state that I'd sort of spent my entire kind of childhood and adolescence sort of trying to pursue. And so...

as a kind of other things and other ideas and opportunities sort of, you know, across the transom, I think I was maybe more open to them than, uh,

than my peers, not because of, I think, any differences in me, but just because of differences in the culture environment that I'd come from. And so my brother John and I-- John at this point, being a little bit younger-- he was now in this transition here in Ireland. We decided to start a company six months after I got to MIT. And so I'd really just started. And I felt that I had some kind of time despair because I'd started college a year younger than most of my peers.

And that company sort of worked okay, and it's kind of a long story, but it ended up becoming a small acquisition. I went back to MIT because when I started there, I'd sort of

been very interested in math and physics and had kind of been interested in this idea of, you know, potentially becoming or at least attempting to become some kind of academic. And of course, at a place like MIT, that's sort of the default around you. You know, everyone is planning on at least, again, trying to get a PhD or to become a professor or whatever. And so I think, you know, that environment had some effect on me. And so I went back because I felt that I hadn't sort of

really, you know, properly rejected the hypothesis that maybe I should try to become a professor, right? Maybe kind of physics is what I should be, again, at least attempting to spend my career on. And after a year back at MIT, I decided that that was not the case. Progress in physics really felt like it had sort of slowed down pretty substantially compared to the, you know,

1910s, 20s, 30s, the period in which so much of what we were learning about, that sort of broader period of discovery, felt like the period in which we existed in, say, 2010, there really was just not the same rate of progress. And so there was a little bit of that, and then also some amount of appreciation myself that I think I just enjoyed...

programming and software and technology more than I did math and physics, even though to some degree is a little bit maybe painful to realize that. I want to explore a little more about the cultural differences between Ireland and the US and how that impacts you as the CEO of Stripe. I think there's maybe a couple of things in that Ireland is very outward looking, necessarily so, in that

Ireland's improbable rise from poverty over the latter half of the 20th century was very significantly enabled, maybe almost wholly enabled,

by exports, by sort of importing American multinational companies, having them set up factories and bases and, and, you know, hubs of different sorts in Ireland. Uh, one of the world's first special economic zones was created in Shannon, uh, which was very close to, you know,

10, 15 miles from where I was born. Deng Xiaoping visited it and found this quite inspiring and so decided to set up special economic zones in China. And so Shenzhen and the Pearl River Delta, that special economic zone was in some ways directly inspired by what he saw in Western Ireland. And so I think the fact that there's such a very visceral link between betterment and progress and economic development and

This kind of outward looking sense that the possibilities of the rest of the world are sort of much greater than kind of those internally. That's very pervasive in Ireland. And I think that's certainly influenced Stripe in the sense that we really are always trying to emphasize the...

the imperative for and the potential of globalization. While maybe in the mid-90s that was something that was uniformly accepted in at least elite circles, now that's something that perhaps has been questioned somewhat more, but I guess the Irish experience is very much one of seeing it as an almost wholly unalloyed good.

And again, I think that's greatly influenced us here, certainly me. Well, it's interesting too from a cultural standpoint where Ireland has had very high rates of immigration, particularly post the expansion of the EU in 2004. A very large number of Eastern European immigrants moved to Ireland when those countries acceded to the EU. And

That was really not accompanied by any material social strife or conflict or a lot of the sort of challenges that we've seen in sort of other parts of the world. And so again, I think that sort of an appreciation for borders that are more open or more openness to immigrants, more sort of

facilitation of opportunity, things like that. Again, I think that really is the Irish experience. And of course, there's the reverse version where so many Irish people themselves have sort of benefited enormously from being able to go and sort of pursue lives in the UK and Australia and the US and Canada and so on. And that's, again, just really kind of part of the national ethos. And then maybe more softly, I guess, Irish culture places a lot of importance on

Just a kind of warmth and a particular tenor to the sort of interpersonal dynamics and trying to have other people enjoy themselves and be at ease and have a good conversation with them and whatever else. And I think maybe that's something that's influenced us somewhat at Stripe, where we want Stripe to be a warm place. I mean, we...

play music at reception and in the kitchen to just try to put people at ease and to create enough sort of soft noise around them where they feel comfortable having just a good conversation. And maybe that's because of entirely unrelated reasons, or maybe again, in some way we're influenced by the kind of environment we grew up in in Ireland. How would you describe the culture at Stripe? What do you actively try to achieve with that?

I'll answer that with a caveat. And the caveat is that I'm pretty sure the answer I would have given to this would have differed in some material ways two or three years ago, right? And that's in part because I think we're coming to realize things that we just hadn't really appreciated or sort of seen the significance of two or three years ago. And also in part because

Literally, what it is that we need today is just different to what we needed two or three years ago, right? And so I think there's kind of double contingency in the answer where it's a function of just what we've realized at this point, but also sort of what it is that the organization and the company needs given the sort of challenges that we currently face. With that caveat, I think the things that we really prize and try to seek in the people we hire are

a kind of rigor and clarity of thought in that I think so many organizations prize sort of smoothness, smoothness of sort of interactions and trying to reduce, minimize the number of sort of ruffled feathers. And they kind of

at least sort of inadvertently, if not deliberately, prefer cohesion over correctness. And we really try to

identify people who are seeking correctness and who don't mind being wrong and who are willing to at least contemplate things that seem improbable or surprising if true or really divergent to what is sort of the generally accepted status quo. And that's hard to find. And I don't think most of the sort of educational institutions that we all tend to have attended actually do a great job of teaching that.

And so we look for that kind of combination of sort of openness and rigor. I don't exactly know what the right word is, but kind of determination and competitiveness and I guess willfulness in that sense.

Just doing anything of significance is hard. I mean, anyone who's tried to do anything that they themselves consider significant knows that very viscerally, right? And I mean, especially for a startup, like the default outcome is you're relatively...

near-term non-existence like the the default outcome is that you do not survive and to to survive over the medium or you know, even with even more difficulty over the long term and That is that's like an unnatural act right? And so you need to find people who not just are willing to sort of push against the sort of the expected trajectory

of non-existence, but people who actually enjoy that, who want that, right? Because if they're merely willing to do it, but they don't actually enjoy it, then the work is probably going to be less fulfilling for them over the medium term. And I really don't think that is for everyone. I don't think that's a bad thing, right? In that the cliche, of course, is that startups are extraordinarily hard, and they just are. And you want somebody who finds that

who's at a stage in their life where that's the kind of challenge that they want, where the fact that the particular area in which they're going to be working is sort of undefined or significantly underbuilt out or significantly broken or whatever the case might be, that that's what they're looking for, right? And then we try to find people who just have a kind of, again, to return to this word,

interpersonal warmth and a desire to make others around them better and just a degree of caring for others and a desire to be nice is a kind of anodyne word, but to be nice to them, to make them better off, right? We really try to find people who we just actively enjoy spending time with, right? You spend such a large fraction of your life inside

the walls under the roof of, you know, whatever organization institution you're working at. And so given that, I really think it's worth prioritizing this. And I think, I mean, I of course don't know for sure, but I think we go to sort of some greater lengths to find these people than other organizations tend to do. And there's other things as well. I mean, you know,

It almost goes without saying, but we really care a great deal about ethics and integrity in people. But I think so too do a lot of other organizations. I think the three that really stand out to me are this kind of rigor and clarity of thought, this sort of hunger, appetite, willfulness, determination, and this, again, warmth and desire to make people happy

around them better off. Those are three that really stand out to me. Take me back to the early days of Stripe and the struggles you were having and maybe walk me through some of the things that you've learned since then or some of the mistakes that you had made. Sure.

I mean, the kind of background context here is that by almost every sort of under almost every kind of ostensibly sane analysis, Stripe looked like a bad idea, right? This was a crowded market. There were tons of existing incumbents that were significant regulatory and just kind of partnership institutional barriers to entry. We had no experience in the domain. We were very young.

We weren't even US citizens in an ecosystem that, again, just because of the regulatory dynamics, that sort of adds further complication. We had no sort of obvious mechanism for gaining sort of significant distribution. And we were not a sort of naturally viral product or one that would have sort of organic adoption the way maybe a social network or a consumer product might have. And so for all those reasons, I think a lot of people sort of very reasonably thought that

Either Stripe was a bad idea or us pursuing Stripe was a bad idea. And they certainly didn't hesitate to tell us that. And to be clear, I think they were doing something reasonable by telling us that. I mean, they were giving us their sort of honest and, again, reasonably justified assessment. And so it all started in the background context of that. I think the thing that primarily gave us the confidence to actually attempt it was that

It just seemed so strange that something with Stripe's character didn't exist, in that we really looked for Stripe before we started it. It felt that it must be the case that there is some service, some company somewhere offering infrastructure and APIs and payments and economic tools that are straightforward to use for a developer, right? I mean, this is one of the sort of top

top needs that any business operating on the internet has, arguably by definition, sort of a business on the internet must have access to these tools. There are

tens of millions of developers operating on the internet. And so just given the magnitude of that market and the sort of obviousness of the business model, it really felt like this had to exist. And so we'd kind of forlornly Google for it, you know, with different sort of permutations of keywords. And then sort of after a couple months became, you know, somewhat resigned to the fact that no, it did not in fact, you know, improbably exist. And non-existence was so kind of strange to us that actually

kind of discouraged us where if there was sort of such an obvious idea and such a surprising, you know, absence of a kind of solution, maybe there's some kind of latent force that we're not seeing that actually makes sort of solving it impossible, right? And that, you know, for example, we were also interested at the same time in why kind of consumer banks were so bad in that just, you know, they weren't really keeping abreast of technology and the fees were really high and they were getting fined by the CFPB and, you know, et cetera, et cetera, et cetera. And, you know,

And as we looked into it, it became apparent that actually there was a good reason as to why the problem had not been solved, where A, the banks are subject to sort of such onerous regulation where it's very difficult for them to do anything themselves, right? And so, for example, the difference between a checking account and a savings account, which might seem sort of quite unfriendly from a consumer standpoint, that's actually kind of essentially mandated by law. And so it's kind of not on some level the bank's fault. And the second reason is the Office of the Comptroller of the Currency, which is the entity that...

issues federal banking charters, had basically stopped issuing new banking charters post-financial crisis. And so if you came along and you're like, well, I'm going to go solve all these problems in consumer banking, you're essentially blocked from doing so by the kind of regulatory apparatus.

And so we kind of wondered in this kind of similar vein, is there some force like that? Not necessarily regulatory, but just like there's some constraint that kind of we aren't observing or weren't. And after a couple months of investigation, we decided that no, there didn't appear to be at least. I mean, of course, you can never kind of definitively reject it, but we really couldn't find one.

And so we decided to build a prototype. And the prototype was built on top of and with existing payment systems. And so it didn't do anything overly ambitious. It was just enough to get a sense for what-- it was almost like a concept rendering of what a solution could look like rather than a solution itself. But it was sufficient to get just a couple of our friends started using it.

And I think the particular thing we realized that caused us to really go take it a little bit more seriously and concretely to drop out of college was the realization that the sort of problem that we perceived in kind of developers like us, people building some little side project or with this kind of very nascent startup or something like that, that the problems we perceived for developers

that segment of the market, were actually the problems that larger companies had as well. That kind of what we thought initially might be a little lake of opportunity was sort of more akin to an ocean. And when we talked to companies doing hundreds of millions or billions in revenue or companies in other countries and so on, and we just asked them to kind of recount their problems and what they wished existed and everything else, they basically give us the same roster of features.

When we thought about it and looked at the macro figures, we saw that at the time, 2% of all consumer spending in the world happened on the internet. Even though we were 20 years into the web's evolution, and even though we'd all engaged in lots of e-commerce and so on, when you looked at it on a macro basis, it was apparent that we were still barely off the starting blocks.

And so I think the combination of those things where we kind of decided that there didn't appear to be some sort of dark energy preventing a solution and that the set of problems we could see actually seemed sort of very pervasive rather than just sort of a microcosm. And then thirdly, that actually this whole market and environment was still actually at a sort of surprisingly nascent stage when you looked at sort of the full picture.

Then we decided to drop out. You guys went from two employees, you and your brother as co-founders, to 800, 900 now? About 1,000 now. 1,000 employees. What have you learned from scaling the business? I think on some level, scaling a business is both relatively straightforward and extremely hard. I mean, it's relatively straightforward in the sense that it's usually not that difficult to see what the problems are.

And to the extent that you don't see what the problems are, it's usually because there's some kind of subjective blindness rather than it being actually difficult to see the problem, right? And so it's more sort of a question of what are you oblivious to because of your own biases rather than what is particularly difficult to observe and kind of what are your corrective mechanisms to sort of account for that. So it's, I think, straightforward in that sense. And I guess straightforward in the sense that

Usually solving the problems is not outlandishly difficult. I mean, it's not easy, but you need to hire someone in this role. You need to figure out how to raise this capital. You need to build this system, whatever the case might be. I mean, none of those are easy things, but they're also not sort of scientific breakthroughs. There are other companies that have done it. There are generally playbooks that exist, but

And while sort of your particular strategy might need some sort of correction, refinement, and you might hit some walls along the way, it's rarely unprecedented. And I think it's extremely difficult in the sense that you don't get to really choose the clock cycle and the time horizons necessarily.

It's a category of sort of flash games, desktop tower defense games, where you're sort of building little towers that shoot missiles, and you have all these little critters sort of scampering across the board, trying to sort of break into your fortress or whatever the case might be. And a startup feels a little bit like that, where...

you fundamentally don't control the rate of problem appearance. You just control the other variable of the ratio of which you're building defensive or mitigatory or mechanisms to deal with those problems. And sometimes the rate of the problem creation can outstrip

the rate at which you can solve them, even though in principle any one of them is relatively manageable, right? And so I think that really adds a lot of difficulty. I think just on a, even if on this kind of very abstract level, dealing with the problems is tractable, the character of having problems materialize at sort of, at every level of

of the organization or at every kind of level of abstraction or, you know, at every kind of magnitude and, and so on. That's just a kind of unnatural thing that I think is just on a, on a psychological, emotional level, difficult to deal with. And so while you might recognize sort of on some contemplative stoic level that this is how it goes, uh,

you know, that's not necessarily how it feels in the moment, right? And it kind of feels like that way every day. And some days you almost have to smile at the sort of unreasonableness of the swathe of problems and challenges that have, you know,

materialized on your desk or in your inbox and that you know it it sort of in the same way that you see the constellations in the stars you know the sort of constellation the problems looks so implausible and so unreasonable that like someone must secretly be screwing with you right um and so there's that kind of emotional sort of self-management and then of course there's the challenge of dealing with uncertainty where you know it's i mean it's kind of uh i guess um well you

You're operating in sort of the weird zone where you're often making decisions that have sort of significant long-term impact or that are at least difficult to reverse or to course correct in the face of great uncertainty, right? And the uncertainty is often unnecessary in the sense that

you could in principle go and significantly reduce the uncertainty. You could go and study the question more. You could go and obtain more information. You could go and run an experiment. You know, it's not like cosmic uncertainty where there's just, it's true sort of nighty and unknowability. And I think when it is like true, deep, unmitigatable uncertainty, then I think it's not too hard to say, well, we're just going to choose something and, you know,

make the best decision we can. I think it's a more frustrating kind of uncertainty where it's actually not necessary, but the thing that's sort of limited is essentially the cost of obtaining further information, reducing that uncertainty. And so you're left in this sort of dissatisfying situation where I have to make a highly consequential decision. There's a lot of uncertainty. We could have less uncertainty. We could take steps to mitigate that, but we just don't have time to.

and making a lot of decisions in that zone is somewhat dissatisfying, right? And I think kind of correctly so, and that one is correctly reacting to the fact that it could be otherwise, right? And then lastly, maybe...

you're playing this sort of multi-armed bandit problem where you're sort of constantly trying to balance exploration and exploitation or sort of, you know, just optimization of that which already exists instead of doing it better and better with trying to figure out what are the things that, you know, we aren't doing or that we don't know or we haven't even considered or, you know,

if we were doing what would make this other part of the organization vastly more effective and so on, it's very hard to know what the optimal rate of exploring those things is while also basically operating outside the system and operating inside the system, or optimizing outside the system and optimizing inside the system.

And it's very hard to know what the right kind of rate of doing those things is. And so, again, I think a lot of the challenge of scaling the organization is sort of finding at each kind of moment the right way to balance those things. But without ever having kind of sat down before to try to sort of, you know, in any way to kind of distill it into any unified theory, I think that a lot of the experience of scaling an organization is kind of specific versions of

or specific applications of sort of those dynamics and just figuring out how you yourself or how the organization or how your peers and colleagues sort of deal with that and what the kind of structural mechanisms for doing so is or are. And then maybe very lastly, I mean, those are all kind of the structural ones. I think there's just also a personal version where you certainly don't start out being well adapted to or...

at least in my case, particularly skilled in organizational management and leadership. And I mean, depending on the rate of growth of the company, you sort of need to acquire those skills on, again, a timeline that's largely out of your control. And, you know, depending on the rate of growth of the organization, that might be a pretty difficult thing. And so, you know, certainly in my case, I think I've just had to accept my sort of

managerial inadequacy relative to what either is required in the moment or sort of will in the near term impending future be required and just figure out strategies to try to acquire those skills and abilities as rapidly as possible. When I go back to the explore, exploit kind of comment that you made, which we can probably just relate to focus.

How do you think about focusing on one thing and being exceptional at that or doing a variety of things and trying to be exceptional at all of them? You mean in the organization or personally? In the organization and maybe personally if that's different. I don't know of a better answer other than using course heuristics and then being willing to revisit or make an exception if something

if something seems particularly promising.

Roughly speaking, we invest most of our effort, we don't have a precise number on it, but let's just say 70 or 80 percent in optimizing that which we already have, that which we already know is producing returns, that which there's a sort of relatively clear line of sight from sort of the input, the work, the optimization, whatever, to sort of the output improvement. And then, you know, some fraction of the work and the sort of distribution of output.

of bets skew, but some fraction of the work, let's call it 20%, into things that are more speculative, right? And I think that's kind of necessarily the case because I think it's necessarily the case that, call it again, 70% or 80%,

is devoted towards optimization of that which already exists. If we did not do that, then this kind of default non-existence we just discussed would be guaranteed, right? It's very easy to sort of fly the company into the side of a hill. And so I think really the question is just, do you spend 20% of your time on things that are more speculative or do you spend 0%

And then maybe secondly, to what degree do you allow those answers to be different at different levels of the company and sort of in different places? And how much is it sort of a uniform answer and how much heterogeneity do you permit or do you design for? And I think as we've grown, we've tried to shift into a model where it is somewhat less uniform and in certain teams,

less optimization of what already exists is going to be required. It's going to require more exploration. And in other parts of the company, it could be tilted in the reverse direction. And I think that kind of recursive decomposition, I think, is really required to avoid the diseconomies of scale that otherwise set in as you grow. How do you decide which speculative projects you take on?

Are they based on disrupting your business? Or these are things that I want to do or I want Stripe to do? I don't know that there's a better answer beyond given all of the axes of constraints and returns, which ones seem like a good idea. I mean, I think it's kind of like investing. When you ask, you know, what are the criteria for investing in a company? It's, well, when you kind of

normalized down from the sort of really high dimensional space of market and founders and idea and all these things, you normalize all that down into kind of what do you think the return profile looks like while you invest when the return profile looks good enough, right? I think kind of similarly when you decide which ideas to pursue, of course, on each axis there are many things you prefer or you don't want or whatever.

And for example, something that requires less effort rather than more or entails less downside risk rather than more, whatever, you know, those are all good things. But I think kind of where it all nets out is, well, when you take account of all those factors, which things just seem like a good bet, right? And so just, you know, to give a concrete example, Atlas, the service we launched for helping entrepreneurs

new founders in corporate companies, and in particular, sort of, without the geographic restrictions that tended to exist before, so it's essentially open to founders anywhere in the world. There was no kind of one reason as to why

that was a good bet. You can't just measure that on any one axis, right? But when you look at it overall and you see that, well, if it doesn't work, it's hard to see how it could cause that much downside for Stripe. It's not going to require an enormous fixed cost investment in order to learn as to at least whether it's initially working. If it did work, it seems like it could produce quite significant returns.

The kind of things we'll have to do for it are actually things that are probably valuable for us in other parts of the business and so on. So we'll learn interesting new capabilities and skills in the course of doing it, etc., etc., etc. I think the reason there aren't more good bets made in the world is because making good bets is difficult. And again, I think you can have in different areas...

Difficult in terms of recognizing them or difficult in terms of acting and executing on them? Or what do you mean by difficult? I think both. Well, I think most organizations are sort of institutionally resistant to bets in that because most people are necessarily optimizing things that already exist, they're

And again, that's correct, you're not making a mistake. Things that are not optimized along the way, especially things that are not being fixed and optimized and patched up and corrected as they burgeon, those are going to break. And so the optimization is critically important. I don't mean to sound remotely dismissive towards it. But BAT's a very different character.

And this is sort of a continuum of betfulness and riskiness. Well, startups are bets, right? Exactly, right? And large institutions and incumbent organizations sort of dislike them, right? Structurally speaking. And find them difficult to understand and difficult to interact with. And so I think there's a whole host of reasons there in that, you know, people in startups are sort of less worried about the risk of failure, whereas people in sort of existing systems must understand

quite a bit about the risk of failure. You know, newer things tend to operate on sort of faster sort of clock cycles. And so, you know,

Dijkstra talked about the idea of the Buxton Index, the time horizon upon which an organization makes its decisions. Maybe a university makes its decisions with a decades-long time horizon, whereas maybe a company makes decisions on a quarterly time horizon, and maybe an individual makes decisions on a weekly or monthly time horizon, whatever.

The observation was that organizations with very different Buxton indices find it difficult to work together. If an organization with a really long time horizon is working with one that's rapidly updating and rethinking, there's just a fundamental impedance mismatch.

And so I think that, to your question as to sort of why it's hard and why there aren't more good ones in the world, I think there are lots of different kinds of impedance mismatch like that. It's not just the time horizon thing, but I think there's just like a fundamental deep intrinsic difference between sort of existing incumbent systems and the actions and the mindset required to optimize them and this sort of the exploration of figuring out that which is totally orthogonal, different, and new.

How do you keep the mentality? I mean, when Stripe started, the cost of failure was really low. Now you have a thousand employees, they all have families, you have a business, you have people who've invested a lot of money in the business. How do you maintain that ability to place massive bets? It's really a question of...

How do we make sure that we can place bets that don't have excessive downside or sort of fatal downside, right? Or a cumulatively fatal downside across maybe a whole portfolio of bets. And I think that actually, I think the impediments to...

placing good. Well, again, I'll caveat all this by saying it's not like Stripe has a long track record of sort of making really good, you know, investment bet decisions. You know, we are, I am, we are far from being, uh, the, the apples or the Berkshires or whoever, you know, a multi-decade sort of track record of making good bets. We're back here in a decade. We'll, we'll reevaluate. Uh, if we are here in, in, uh, three decades, which, you know, as established, uh, would, would, uh,

will not be the default outcome, and we have a great portfolio of successful such decisions, then perhaps we can opine, you know, with, I don't know,

It's a modicum of confidence. But it feels to me, and we'll see if this is right or not, it feels to me that actually the reasons that organizations don't tend to make more of these or make more good ones is it's more kind of sociological, more institutional, and less that it's fundamentally too costly. Because in most cases, the downside cost is not that large.

And either in terms of like just direct financial cost or in terms of the sort of the broader damage to the organization, whatever form that might take, it's much more the mindset of improving that which already exists is just quite different to the mindset of screw the old system. Let's do something that's fundamentally new from scratch. And so I think the challenge is in significant part, how do you reconcile these two mindsets?

how do you have the, I mean, Stuart Brand talked about pace layering in buildings and sort of different parts of the building have to change at different rates and how do you design for that? And I think the kind of analogous question for an organization is how do you do organizational pace layering? How do you have parts of the organization that can try to do something fundamentally different to and

and hopefully superior to that which already exists? And how do you have people who are trying to, who basically disagree with people trying to do something new, who think that, no, the way we're currently doing it is in fact the right way. We're just going to do it better and better. And because these people fundamentally, structurally disagree with each other and must have significant conviction in their respective approaches, otherwise they do great work. How do you have those people at the end of the day have dinner together and fundamentally feel like they're on the same team? How do you do that? Come back in 30 years. Yeah.

I think I recall one of the interviews that I was watching as prep for this, where you talked about one of the first five or six people worked at Bridgewater. No, one guy in particular did. And over time, we've hired more people who have. But I would not say we were particularly Bridgewater influenced. Did you come at this sort of...

notion of thoughtful disagreement before that influence and how if so how did you well yeah it's hard to know exactly where to attribute it and it's probably kind of over determined um and maybe they're just kind of some sort of underlying personality traits that we each had sort of um

come to in different parts of our lives in sort of somewhat coincidental ways. I mean, for a start, to your earlier question, Irish people are always disagreeing and always arguing. And so, again, there's a cultural dimension to it. It's not something that people tend to shy away from. Because they don't see it as an attack on them. Exactly, right, right. I think that... I think there was just a common shared personality trait

A lot of the people who helped establish the culture of Stripe, where they enjoyed sort of disagreement and trying to find the boundaries of an argument and the places where it's not the case and what the exceptions might be. And just trying to kind of get a feel for the topology of that space and kind of stumbling in the dark, try to construct a map of what

where different intuitions and heuristics apply and where they don't and so on. And I think one kind of deep mindset difference in people is often those who enjoy finding the limitations of arguments and beliefs and those who don't. And Tyler Cohen talks about, I think it's his second law,

that there are no knockdown arguments. There are no arguments that are just uniformly completely true. There are always the limits to it. There's always the other side. And I think that's kind of very deeply true, but I think there's kind of just a...

a question of sort of affect and again, personality as to do you enjoy finding those limits and the exceptions and thinking about, well, maybe this is less true than I think or where is this less true than I think or is that just like a stressful process? And I think that sort of getting that kind of rigor and clarity of thought requires...

sort of a joy of discovery, like, ah, this thing I believe, this rule that I thought existed, like, it's actually not good in this place. And having that be an enjoyable discovery rather than sort of something stressful and threatening. And I think globalization is a good example there where, you know, as we discussed, I think that globalization...

is, on net, overall for the world, a fantastic thing and something that support is rising for on a global basis and has propelled more people out of poverty than

almost any other force ever. And yet there are people like Danny Roderick and others who are sort of prodding at the edges of that and showing, well, but not in this place or not in this way or Autour and these other folks at MIT, like maybe it has this sort of underappreciated downside. And I think that's great. I think those are important questions and really interesting work. And I think the kind of

Again, the underlying sort of sentiment is sort of interest in where the heuristics and the intuitions and the rules and the arguments are wrong. I want to come back to some of that a little bit later. I think one of the questions that people want to hear from you is what would you say is the biggest difference between the Patrick making decisions today and the Patrick making decisions maybe five years ago in terms of how you actually make those decisions?

I think there are four big differences. The first is, and I'll just place more value on decision speed, in that if you can make twice as many decisions at half the precision, that's actually often better. And then given the fact that sort of the rate of improvement of decision making with additional time almost necessarily tends to kind of

flatten out, I think that most people, certainly the Patrick of five years ago and potentially even the Patrick of today included, should be operating earlier in that curve. Make more decisions with less confidence but in significantly less time. And just recognize that in most cases you can course correct and treat fast decisions as a kind of asset and capability in their own right.

And it's quite striking to me how some of the organizations that I hold in the highest regard tend to do this. The second thing is not treating all decisions kind of uniformly. I think the most obvious kind of axes to break them down on are degree of reversibility and magnitude. And things with low reversibility and, you know,

great impact and magnitude, those ones you do want to, you know, really deliberate over and try to get right. But I think it's very easy, sort of absent care, to have maybe this mechanism you put in place for those decisions to seep into decision making for the other categories. And really in the other three quadrants, you can afford to be sort of

much more flexible and much more fluid. And again, really just to prioritize speed, because obviously if it's very reversible, then, you know, by definition, you can always correct it later. And if it's, you know, of low import, then who cares, right? And so that's kind of the second one and just being kind of cognizant of that. And before making the decision, trying to categorize, well, what kind of decision is it? The third thing is

I now try to fairly deliberately just make fewer decisions in that why am I making the decision? And for some kinds of decisions, there are some...

good reasons for that. I mean, there are some decisions the CEO ought to make and is kind of fundamentally on the hook for, but there are some decisions where if I'm making it or if I have to make it, that probably suggests that something else organizationally or institutionally has broken. And I think the need for a decision from anyone, not just from me, is often like only a sort of an epi phenomenon. And there's really some other underlying issue that's causing you to have to make in the first place. And so thinking about that.

and concretely doing more to push others to make decisions and pushing them back to people who ought to be the domain experts. And then fourth,

When I realized that I would make a decision differently to how someone else is making it, not even really discussing the decision itself, but trying to dig into what is the difference in our models such that you want to make decision A and I want to make decision B. And one thing we're currently spending a bunch of time on here at Stripe is having different parts of the organization write down

what they're optimizing for essentially, like what their mission is, what the long-term key metrics are for kind of their part of the organization, who their customers are, either internally or externally. And so things of this kind of persistent, ongoing, underlying nature such that hopefully once there's agreement on those longer-term things, then maybe a difference on...

on any particular decision might just be, well, we differ sort of on what the most instrumentally effective way to achieve this outcome is, but we're both really unified on what the desired end state is. And there, I think, disagreement over sort of instrumental efficacy, you know, well,

that's really that problematic a disagreement because, well, if you're right, then we'll soon learn that. If you're wrong, reality will probably sort of make that pretty clear in short order. I think the more troubling ones and the ones that tend to cause more kind of persistent friction in an organization are where sort of there is latent disagreement in what you're actually optimizing for, but that's kind of never explicitly surfaced and uncovered. And so now, I guess, again, in decision-making,

I place kind of more importance on making sure that we have the right sort of foundational agreement such that the kinds of disagreement that then tend to arise are of the sort of essentially more superficial sort and their agreement is actually less important. Part of culture is...

learning from the decisions the organization makes? What do you do at Stripe to make sure that people are learning? And what do you do personally to make sure that you're learning from the decisions that you've made, both positive and perhaps ones that you, in retrospect, would have wished you could make differently?

I'm inclined to say, I don't know if I actually believe this, but I'm inclined to say in response to that question that decision making in organizations is slightly overrated in that organizations are not like investment entities or funds or managers in that

Well, with investing, it's fundamentally very binary. There is a moment at which you either buy or don't or sell or don't or whatever. And maybe it's somewhat more continuous in the case of, say, public market investing and so on. But given sort of constraints on just decision-making time, I think you have to treat it as a bit more binary. You assess the stock and you make a buy or a sell.

or not decision. Whereas in organizations, everything is much more fluid and continuous. It's much more about, I think, designing the feedback mechanisms. More biological. Yeah, exactly. And, you know, there's the famous sort of...

water model of the economy. You know, with the sort of circulating fluids and you can vary the interest rate or the inflation rate or whatever, but just kind of try to get a sense for the overall kind of biological apparatus. And I think an organization is much more like that. And so I think the things to optimize are the incentive structures and the

and the definitions of the goals and the feedback mechanisms from the outcomes to the inputs and the work and the operations themselves and all of those things.

And less the binary decisions. And I don't want to kind of completely dismiss, obviously, the importance of decision making in that there are times where you decide, well, are we going to launch this product or not? Are we going to start this project or not? Are we going to replace this system or not? And so on. So there are, of course, real decisions. But I think it tends to be much more, well, I guess maybe it doesn't feel like the right unit of analysis to me. I think the right unit of analysis is decision.

is that of the cell? And the question is, well, in an organization, what are the cells and what are the organs and how do they interact? What are the feedback mechanisms between them? Let's geek out a little bit on the feedback mechanisms here. What sort of feedback mechanisms do you try to make sure are in place? What point in the process do you try to acknowledge what they are? I really think that, and this is not to evade the question, but, um,

I really think it's too early to answer that in the sense that, I mean, I can kind of tell you what I think today and the sort of changes we've made over the last year and things like that. But like Stripe has been a thousand person organization for, or has been a more than 500 person organization for just over a year. Right. We're, we're beginners at this and, you know,

Three years ago, Stripe was under 100 people. And I think either to opine as if or to even more problematically believe that we can have it figured out would be real hubris. And so in what we've been talking about, I think that's maybe some of where our and my thinking comes from. But I don't know what the right answers are yet.

And we spend a lot of our time sort of scrutinizing other organizations, trying to find out and kind of reverse engineer what works for them and why. And I think that part of what's interesting about the tech industry is that it's a kind of pure knowledge work.

that we're still, I think, quite early in sort of figuring out in terms of how to optimally coordinate it and collaborate on it. In that you can sort of draw a lineage of HP and Intel and Microsoft and Google and Facebook and so on, WhatsApp. And there are all these sort of suggestive

examples that I think at least again suggest that we may not have it all figured out I mean the fact that WhatsApp was such a minuscule team and Instagram too of course despite operating at such a scale or the fact that the way of a new paradigm yeah yeah and the way kind of Facebook operates is very different to the way you know HP operated outside of Stripe which company cultures do you admire the most not business models but culture and why

Well, I admire cultures that are strong, first off. Cultures that when you ask somebody who's in the culture, can you describe it, that they can expound on its merits for more than half an hour. And in almost every case, describe at some length all the things they don't like about it, right? Because if it's strong, it's improbable that every aspect of it is something that the person really agrees with or feels an affinity for. And so whether it's

the New Yorker or the military, a shared characteristic of those cultures is that they're strong, right? So I think that's the first order thing. And I don't think that describes most organizational cultures. I think most organizational cultures are some kind of milquetoast averaging, right? So that's number one. The second is cultures of perfection. And so both The Economist and Apple are

have extraordinarily high standards for themselves. And really, in both cases, the work has a kind of primacy. And so who designed the latest iPhone or who wrote that article, in both cases, that's anonymous because there's such a belief that the work speaks for itself. I have a lot of admiration for that. And then cultures that have longevity and really sustained success.

And so I think that one of our major investors is Sequoia Capital. And Sequoia has been the top firm or in the top three firms. Obviously, it's a subjective ranking, but call it unquestionably a top three firm for essentially its entire existence.

And there was no other VC firm that has been a top three firm for, you know, call it four decades. And so I think the obvious question is, well, why is that? What's different about Sequoia? There have been tons of VC firms. And lots of different firms have had at any moment in time a strong claim to being a top three firm. But what are the underlying institutional characteristics that enable that to be sustained? And, of course, this applies to some of the other firms.

organizations we mentioned, like say The Economist or The New Yorker, or even this is one that I've been trying to read more about of late, Koch Industries, in that Charles is of course, or Charles and David are most famous for their political activities. But if you just look at the company, that has kind of compounded from 20 million in annual revenue to now, according to public estimates, 100 billion.

over, you know, call it five decades. And there aren't that many organizations that have compounded like that for that long without there being kind of

one driver of success there's no one thing that enabled their rise they didn't like stumble upon some resource that they kind of cornered there was no kind of iphone for them etc it's clearly something kind of deeper and more sort of institutional uh and and the fact that that's been kind of sustained for so long i think is interesting in its own right as in what is it that sequoia capital coke industries and the new yorker share uh and i haven't quite unpacked the answer to that yet

Can you give us an example of what you've learned from studying Koch Industries? It's very striking to me how Warren and Charlie at Berkshire and how the folks at Koch Industries are so into a kind of epistemology and structuring of doubt and accounting for biases and mechanisms for clarity of thinking.

Like to a very striking degree. I mean, obviously, if you read the public writings or you go to Omaha and you listen to what, you know, Warren and especially Charlie talk about, you know, it's sort of half investing and half applied epistemology, half philosophy, right? And that's been the case as well to a striking degree with Koch, right?

And I don't know them well enough by any means to sort of opine in any deep sense, right? Like I've never been to one of their factories. I've never looked at one of their financial statements. And so I'm not qualified to assess in any kind of comprehensive way. But just in terms of what it seems that the leadership prioritizes, it's strikingly consistent across sectors.

two of the most successful multi-decade institutions in the US. There's something to be said, going back to your point earlier, about learning from companies that have consistently demonstrated over a period of time without these huge kind of like one-off hits that have caused most of that track record. Right. You're a huge reader. Where did this love of books get started?

Well, we had crappy internet when I was growing up because our house was so remote. There was so much noise on the phone line that we didn't have internet for years. And then we got it was treacle slow and so on. And, you know, I was fortunate. My parents were very willing to pursue all these harebrained schemes. And so we eventually got an ISDN line, which was ferociously expensive. But, God, that was sort of the...

the fiber of its day, at least as far as I was concerned. 7.6k a second was majestic. I barely keep up with the speed. And then we eventually got a satellite internet connection, which was really a game changer. But it effectively meant that for the first, I don't know, 14-ish, 15 years of my life, there was no internet. And we lived in a very rural part of Ireland,

I was quite distant from even my friends at school. And so all there really was for us to do was play in the garden, which we did a lot of, and to read.

And, you know, it's funny. I often wonder about this in the context of, you know, if I had kids or when I have kids, what's the optimal upbringing for them? And, of course, you think, well, you kind of want them to grow up in a stimulating environment and have all these, I don't know, experiences and extracurriculars and everything else. But certainly that was not my upbringing. My upbringing was...

a kind of... Get out of the house, go play. And, I mean, there's plenty of stimulation around. You know, our parents had lots of books, and so, you know, we could just kind of burrow our way sort of sequentially through the shelves. But, you know, it was pretty unfettered, and I think our parents had a kind of... They followed our interests and supported them, but they didn't choose them. It felt like they...

They pushed from behind rather than pulling in front. And so, yeah, I think that's where the reading thing came from. And I think that, well, I don't know, I run quite a bit. And I don't even run because I enjoy it that much. I mean, I enjoy it, but...

In the immediate moment, it's not like it's euphoric or anything close to it. I mean, it's pretty painful. And the Greg LeMond quote about how it's very dispiriting when you think about it. And it is very deeply true that how it never gets easier. You just go faster. That's true of running. Like, if I stay running for the rest of my life, it will never get easier. I will just go faster.

But it feels like something I ought to do. I vastly rather having run than not having run. And so I sort of continue to do it. And with reading, basically, I don't feel like I'm weird. I feel like everyone else is weird in that there's just like so much stuff to know. And I guess I just feel stressed out by reading.

Like it feels important or it's obviously important and I don't know it. And so shit, like I better get to work. But it's not when I'm reading, I'm not in this like especially blissful place. I mean, I enjoy it perfectly fine. But it's more like I think there are extremely important things that I really should know and I don't. And that feels problematic. How do you filter what you read? Yeah.

There's millions of books. There's one of you. Right. Well, I discard a lot of books. I like the insight that there's a set of great books that are really worth reading, right? And there's a subset of those books that...

that are really enjoyable to read. Maybe it's like 10 or 20% of them say. And the subset, the intersection of really worth reading and really enjoyable to read is actually still more books than you can read in a lifetime. And so I sort of decided, well, I will read all of the...

the books that are really worth reading and really enjoyable to read. And then when I run out of those, then I'll go back to the books that are merely worth reading, right? And so, you know, fairly quickly, you can decide if this is an enjoyable book to read or not, and not discard it. I think reading is like a, you know, should be treated as a kind of more active process, sort of, you should skim, you should skip, you should backtrack, you should

discard and potentially return. You are not subject to the book. You're not a passive consumer. The book is there for you. You bought it. It's yours. And jump back and forward. Tear it in half if you want. Annotate it wildly. Use it. I wholeheartedly agree. And yeah, I maybe start half the books I get and I probably finish half

a third of the books I start. And that works out to

you know, finishing one to two books a week. But if I finish it, that, you know, it's, I guess, well, it's probably been recommended by somebody in the first place. And then it looked interesting enough upon some very superficial skimming to start. And then, you know, if I finish it, it meant that it was quite interesting. So it's actually like a lot of selection that kind of happens along the way. And then I think just the other thing worth pointing out is, you know, there's the line from Basho about the Japanese poet that you shouldn't follow...

the people you most admire, but you should follow what they admired. And I try to do that. I try to figure out for the people who seem to be doing really great work or have really interesting ideas or just who I admire in whatever regard, to figure out how do they get to who and what they are, what influenced them, what's upstream. And often it's quite obscure. But I try to kind of

When do you typically read? Always. I mean, in the morning, in the evening, while walking. While walking is a good one, actually. Like, your peripheral vision is such that you can actually quite functionally read a book while walking. And there's other people that strive to do this and do it much more and faster than I do. But you spend a lot of your time walking. And so being able to do that I found to be quite...

Quite valuable. Often while eating. So you're sitting at home on your couch. It's after dinner and you pick up a book for the first time. Walk me through how you process that book, what you look at. Yeah, normally I'll jump sort of midway through it and just start reading and see like, would I like to have ended up here? And almost certainly like a bunch of the terms I won't recognize or the antecedent ideas I won't be familiar with or whatever. But like, do I? Yeah.

do I want to be here or have gotten here? And if after a couple of pages, it seems like the answer is yes, then I might sort of backtrack to the start and start kind of pursuing it a bit more seriously. I mean, John has this insight that

and it's kind of related to the previous point that, um, at every moment you should be reading the best book, you know, of in the world. Um, I mean, I don't, I don't mean kind of the absolute best for everyone, but sort of the best book for you. But like, as soon as you discover something that, um, that seems more interesting or more important or whatever, you, you should absolutely discard your current book, uh, sort of in favor of that. Uh,

Because any other algorithm necessarily results in you reading kind of quote unquote worst stuff of our time. Suboptimal. Yeah, exactly. And so I'll be reading the book on the couch and then maybe after 50 pages I'll...

I don't know, be in my room and I'll stumble across something else and I might just switch rails. The other thing that I think is actually quite valuable is just leaving books out. And so when somebody recommends a book, I'll very often pick up a copy, ideally a used hardback copy,

Because the hardback books, they're more durable. And now with Amazon, used hardbacks are really cheap. And I'll leave it out. And so this book's in the kitchen, this book's in my bedroom, and this book's on my bed, and just strewn everywhere. And surprisingly commonly, either someone else will recommend the book or some aspect of the book, whatever, and it's still salient, it's still around you.

And you're like, oh, yeah, I really should check out that thing. Or something else triggers its relevance. You read an article. You just start appreciating a point or a question or something, right? And so part of the reason that I still really value physical books is because you, I mean, for now at least, we still exist in physical space. And it creates a kind of idea space for you.

that makes kind of productive collisions more likely to happen. What types of things do you typically mark up in a book? And how do you, what does that look like? So I tend to just make notes in the margin. I tend to underline stuff, but in the margin. And I underline it

I'm missing the term, I annotate it, mark it, highlight it in the margin, because then you can flip through the book, just like quickly see the parts you marked. And then the other thing is on the last pages, like kind of in the inside cover at the end, I tend to very quickly note page numbers for particularly interesting points or things that jumped out or whatever, so that I can easily go back to a book and I have a list of the 30 things that I found most interesting.

Um, so you keep the book, a book that you completely read that you like. Yep. How often do you come back to that book? Um, if I want to make a particular point or be reminded of a particular aspect or, you know, whatever, um, maybe I will, but, but generally speaking, I don't. And I think, you know, part of the value of making the annotations is of course to, um,

imprint them more firmly in your mind so that you sort of don't need to come back as much in some sense. If it's really good, I don't often do this, but if it's really good, I might write a review for friends and just share an email or a Google doc or something, or just share snippets with friends. And that's valuable both because, again, sort of the act of summary or summarization sort of aids the kind of synthesis.

and better recollection. But also, of course, it triggers out pointers and further suggestions from those friends. And so if you want to identify candidates in adjacent-- or if you want to perform the clustering and figure out what sort of adjacent candidates might be interesting for further exploration, writing a review is a good place to start.

What sort of books have you written reviews on for Friends this year? One that I really enjoyed was A Culture of Growth by John Mokyr. Sorry, Joel Mokyr. Apologies. It's basically a book about why did the Enlightenment and the Industrial Revolution, really the Industrial Revolution, start when it did and where it did.

And he basically makes the case. I mean, obviously tons of different arguments have been made for this. And because it only happened once, it's sort of we can never know definitively. And, you know, was it the abundance of

Was it the abundance of coal in the UK? Was it something like the intellectual property system and patents? Was it the high cost of labor in the UK that sort of created more sort of... that made productivity-enhancing improvements more valuable? Was it something about trade? You know, and so on and so forth. And Mokir basically makes the argument that it was...

that it was primarily intellectual and more than sort of quote unquote economic. And secondly, that it was sort of specifically a kind of synthesis of the importance placed in kind of scientific knowledge where we kind of realized that scientific progress knowledge about the world is, exists and can be important and that progress is possible and that we're not just kind of imperfect scientists

I don't know, imitators or receivers of the knowledge of the ancients. And so kind of a belief in scientific progress coupled with a belief in sort of the practical importance of engineering and of the more prosaic aspects of industry.

and of practical pursuit. And Moquire offers the example of Bacon, who both inspired the Royal Society, who was one of his followers who created it, but also intended to catalog the practical knowledge of all of the craftspeople in the UK and the implicit

knowledge that they had. And it's kind of this interesting combination of the sort of really high-minded and the very practical, right? And so kind of, I mean, Mochir kind of teases through all these arguments and the kind of Republic of Letters and the sort of nascent, you know, rise of science on the continent and so forth, but all in service of this question of why the Industrial Revolution then and there, and, you know, talks about versions of it in China and so forth. And anyway, so Mochir

I mean, I think it's a very important question and Mokir's kind of discussion of it is, I thought, particularly interesting. And so, yeah, I summarized it for my friends. That's awesome. Which book or books would you say have most influenced you? So I asked this question on Twitter back a couple of weeks ago and some of the responses I got were really interesting. And a lot of people responded, like many more than I expected to.

And I didn't actually, embarrassingly, I feel guilty about this, I didn't post a response myself. And I thought about it, and it's actually just a very hard question to answer. Like, I actually worry that it may not have been a good question. Because, like, it's hard to know, did the book influence you? Or did you have an inkling or a leaning, and then you read something that really resonated, but sort of, it's actually not...

Like the book is just the artifact upon which you project the sort of the characteristic that had already arisen or the belief that had already arisen. And the book is not actually causal in and of itself. Right. Now, maybe it's still interesting to talk about the book as a kind of symbol for the belief. But, yeah, there's that kind of question. And then also, yeah.

what I've often found is I think the books that perhaps did in fact influence me the most in a causal sense are often not necessarily that good, right? And that maybe I'll read a book that sort of triggers a realization or some idea or something and that will kind of jolt me in some direction. And then I'll go read better things about that question. And so it probably would have been better if I just started with the better stuff. But in some kind of

truthful descriptive sense yeah it was like the worst one that actually influenced me right uh and so um like you know maybe maybe a better version of the question is uh like which books do you wish you'd read sooner or something right well let's answer that question um

I actually, I don't think I can even answer that one now that I think about it. You can't emphasize this question. This is your question. Yeah, yeah, yeah. No, I hoist by my own petard. Like, it's also just sort of clusters of books in that, you know, I think about programming, for example. Like, it would be hard for me to answer this question and not

cite any programming books. I mean, it's been kind of so influential in at least my mindset and my life. But I can't really point to any single programming book. I can name 10 that I think in aggregate work together, like Paradigms of AI Programming by Norvig, and Structure and Interpretation of Computer Programs, and KNRC, and Books about Operating Systems, the Tannenbaum book, et cetera. And in aggregate, those hugely shaped me. But I don't think I could single out just one.

Even two books on PHP, which were written by a guy who now works at Stripe. I mean, one of those books is the book that taught me to program. And so in answering this question, I could hardly not cite those, right? But it's kind of really the cluster. And you have a similar cluster about science, or about economics, or sociology, or whatever. And so we all have to just get back with a better version of the question.

Switching gears a little bit, what's the smallest habit that you have that makes the biggest difference? I reach out to people whose work I admire and tell them that. And often it leads to a dialogue. And in some cases, I've gotten to know them pretty well. And so I'm fortunate that Tyler Cohen, who I mentioned, is a friend. But I was never introduced to him. I just...

I actually randomly emailed him years ago. I actually invited him to a Bitcoin meetup that I held in 2011. I did not, however, buy any Bitcoin. But I invited him to that meetup and he replied and apologized that he couldn't make it. But we sort of ended up in kind of a dialogue after that. And when you reach out to these people, yeah, half the time they don't respond. But, you know, half the time they do.

it's asymmetric. It doesn't really cost you much when they don't, and it can be incredibly rewarding when they do. And so, yeah, if I did not do that, I would have missed out on a huge amount. How would you answer a question about what your personal values are? Probably by evading it. I'm now about to... Do you think perhaps this disproved that answer? By actually answering it, but...

I guess I just think it's so, it feels like too important a question. It's kind of the book question. It feels like too important a question to answer simplistically and too complicated a question to answer briefly and thereby perhaps unsuited to something extemporaneous. And I'm sure whatever answer I gave, you know, when I'm thinking about it in an hour's time, I'll kick myself and realize I'd left out, you know, this critically important dimension to it. And I think I've

So I can cite some things I value, but the sense of giving a complete answer is very oppressive.

I mean, this is, of course, the value of Twitter, where because of the constraint, there isn't the same... Because the system chooses when to cut you off rather than you choosing when to stop, that's quite liberating. And so maybe if you allowed me 20 seconds to speak about values, I could do that. But I could blame the constraint on anything I omitted. We have 10 hours of recording left.

Okay. What would you say is the most common mistake that you see people make over and over again that you wish you could correct and you have 140 characters? Maybe not having the right peer group or not having the right...

Mentor isn't quite the right term because mentor implies something kind of quite active, but not striving to be more like the quote unquote right people or not just being kind of in either case deliberate enough about that. Of course, who the right peer group is for you is, I mean, that's an entirely kind of personal and subjective question. But whoever it is, is going to be massively formative and influential in determining the

where it is that you end up. I mean, Drew Houston has a quote about how you end up the average of your five closest friends. I think there's a very deep truth to that, right? But if you accept that, then of course, who your five closest friends are, I mean, choosing that, and we do, though we may not think of it this way, we do choose those people. Like, you're choosing who you are. And of course, that's a kind of sort of bi-directional process where

who you want to be is determined by who you're around, which determines who you want to be around and so on. But... Five best people that will accept you as... Exactly, right, right. But I think, like, certainly my mental model when I was 18 is that my five closest friends are, you know, people I ran into who kind of like me and I like them and there's a kind of, you know, we're cordial and close and all those things, but that it's kind of fundamentally mediated by sort of happenstance,

And I think people should kind of invest more in it than they do. And related, once you've found those people, you should really invest in it because you

If you accept they can shape you and you think they're the right people to shape you, well, then embrace that shaping, right? And then kind of on the mentor point or on the latter one, I think almost all of us, at least subconsciously, have a set of people we hold in really high regard or would like to be more like in at least some ways and so on. I see people, in my opinion, they've kind of...

They haven't either found the right people or just like the right relationships and so on. And if they had someone who was steering them more or in better ways, could just be much better off. I want to talk a little bit about the future of e-commerce and maybe Silicon Valley culture. And I know we've got to end soon, but talk to me about how...

You foresee them changing from not only the customer perspective, but from the merchant perspective over the next, you know. Well, I think there's two levels to this maybe in that there's, there's all this just like the basic mechanical stuff about payments where you,

we kind of forget just how much friction still exists and how many business models and transactions and businesses and everything sort of are impeded for fundamentally kind of stupid reasons, right? In that because microtransactions are impossible, both because the fixed costs are too high and because just like the friction's too high, then things that one would pay for with microtransactions just don't exist, right? It's not that they pursue a different monetization model. In some cases they might, but as a general matter, a significant number of them just won't exist, right?

Or because maybe it's hard to purchase things that are really expensive in a way where the kind of risk of fraud is sufficiently low, then, you know, don't pay your rent online, say, right? And so, and then I think maybe the most important dimension to that is the sort of geographic kind of thing.

balkanization and sort of inefficiency that ensues where it's extraordinarily difficult for somebody in Brazil to buy from somebody in Germany or somebody in Germany from somebody in India, et cetera, et cetera. And so you get this kind of unnatural sort of subclusters existing, not because of sort of deep necessary limitations, but because of something much more arbitrary and contingent.

And, you know, economists talk about sort of the gravity equation and the fact that the sort of proclivity of any two countries to trade falls off with the square of their distance. And, you know, there's all these like big questions about like, well, is that about something kind of fundamental in culture or about, I don't know, just surprising returns to proximity or, you know, what have you. Assuredly, there's, you know, some of that stuff. But I think...

talking about the challenges and kind of complexities and hidden costs of pain methods, that doesn't feel like a very deep thing. It doesn't feel like something that is kind of significant enough on some level to have such kind of far reaching and deep consequences. But I think a lot of these sort of ostensibly, um, quote unquote cosmic phenomena, uh,

are actually consequences of these very prosaic and straightforward limitations. And so I really think that solving this aspect of commerce on the internet, like literally just making it easy for any two parties, a business and a consumer, in arbitrarily chosen countries, making it easy for those two entities to transact, will have consequences.

enormous consequence for the world. And that sounds like such a sort of straightforward idea that it almost sounds cliched. And the fact that it sounds cliched should not blind us to the fact that it is still extraordinarily far from being the case today, right? We have had commerce on the internet for decades at this point, but it's still...

90 plus percent of Brazilian credit cards do not work online outside of Brazil. Brazil is not some backwater. It's not some inconsequential country, right? Obviously, one of the top economies in the whole world. And Brazilian consumers basically cannot purchase outside of Brazil. And so it's difficult to overstate the magnitude of the sort of limitations and inefficiencies that prevail today.

So that's kind of the kind of payments level. And then kind of on top of that, I think there's, or beneath it, depending on how you look at it, there's maybe just like a deeper question of what determines how many firms there are in the world? And what determines the character of those firms? Are they doing something innovative and novel? Are they doing something professional?

prosaic that has existed for a long time. What determines who starts and why and the probability of survival? What determines the growth trajectory and the expansion rates into other markets and other products and so on? And I think part of the Stripe hypothesis is that things like that that seem very sort of one would think are very difficult to move are actually movable and that

And really macro measures, like the number of people who start a company or who start a technology company, or again, the success rate of those companies. And just to give some kind of...

maybe intuition pumps here. When we survey companies started with Atlas, 60% of them tell us they would not exist if not for Atlas. And, you know, they could be wrong. Like maybe some of them actually secretly would. But maybe some of them are actually overstating their own resourcefulness or overstating, maybe they're underestimating the challenges they would have faced. And so I think that number could either be too high or it could be too low, right? But let's be conservative and say that it's actually only 40%.

If Atlas is causing 40% of those founders to start companies where they otherwise would not have, and if the subsequent success rates look similar, that's a huge deal, especially if Atlas itself gets big. Over time, that can have real economic significance. Or if we can make it the case that businesses sell to twice as many global markets as they would otherwise sell to. Again, integrated over an entire portfolio, that's a really big deal.

Nick Bloom at Stanford did this really interesting work has done a whole bunch of interesting work about management practices do management practices matter you know is good management merely correlated or in fact causal in terms of leading to the advent of better outcomes and

They did an RCT, a proper trial in India, where they taught better management practices to a cohort of firms and did not to a sort of control group and saw double digit percentages in revenue over a multi-year period. I don't recall exactly. I think it was 13% over three years or something like that, right?

That's an incredible low-hanging fruit. All they did is teach better management practices, 13% more revenue, 13% more value provided by the company as assessed by their customers just from better management practices. And so when we think about Stripe and what to do in the future and the possibilities that exist and so on, it's much more, I think, about how do we...

this overall system to move some of these kind of macro outcome measures like number of technology firms started, survival rate of these companies, expansion rate of these companies, magnitude of the value provided to the end users, consumers, customers, and so on. And kind of mediated by payments as this kind of foundational layer because it's something every business necessarily has and because it's

gives us good sort of understanding of the dynamics within the business and so on. But it's on some kind of fundamental level and not about the payment, even though we think that kind of per the first point, the impact of just solving the payments will itself be enormous. Do you think reducing friction across the board is a good thing? Or do you think friction in certain parts of it actually serves the system? Well, it serves it for who? Yeah.

That's a good question. I mean...

the incremental gain of the constituency is substantially outweighed by the social utility loss of the rest of society. Right. Uh, and so, you know, bad teachers, um, do great in the U S. Um, but almost certainly that's kind of a, a net bad trade for society, but the bad teachers care more about, uh, sort of, uh, their ongoing employment, uh, than the rest of society cares evidently. Um, uh,

uh, about correcting that. And, you know, the same thing applies to fishing policy where, uh, perspective makes all the difference. Well, or, but, but, but, uh, you know, people driving fishing stocks to extinction care, care more about their ongoing, you know, right to do so than the rest of society cares about, um, about sustainable ecosystems. Uh, I mean, I think that's just, that's the character of political economy. Um, and so, yeah, absolutely. I think, um,

I mean, to return to our earlier example, it's not even clear that the-- well, one could look at the fact that essentially no new banking charters are being issued in the US as a bug. Or of course, depending on your perspective, it's a wonderful feature. It's great for the regulators, and it's great for the banks. Providence of consumer banks are higher than they've ever been. MARK BLYTH: Until they all get wiped out in the next crisis.

And then because they're even more systemically important than they were in the past, to the extent there was a systemic argument for building them out in 08, there will presumably be an even stronger argument in the future. It's almost like we were talking about this earlier, but bets, when you get big, you have more loss aversion. And so your goal is not necessarily to get better from your customer's perspective. It could be to prevent competition, prevent new entrants. That might be a more...

Without a moral judgment on it, it might actually be a more effective business strategy than innovating for your... No question. And I think that... I think we're very dissonant on this point as a society. On the one hand, we decry lack of innovation. On the other hand, in our society,

collective action, we do so much to ensure that it doesn't occur, right? And so, you know, on the one hand, we decry the state of the sort of medical system

industrial complex and the 18.5% of our GDP that is spent on healthcare costs and the plateau or even decline in life expectancy and the declining rate of drug discovery and so on. And yet, on the other hand, we sort of...

through regulatory structures, make it harder and harder to engage in drug discovery. I mean, you can't even start a hospital unless you get a certificate of need. But if you observe that, well, hey, medical care in San Francisco doesn't seem so great and it seems extraordinarily expensive, you know,

even though it seems like a very thankless undertaking, I'm going to try to do better. Well, first you'd better get approval for that. You can't just enter the market. And so I think that kind of, and I'm not making kind of a normative judgment. I mean, I have my personal preferences, but I'm not casting normative judgment as to kind of what we ought to do as a society. The thing that I feel strongly is that

we were inconsistent, um, uh, in our state of desires. There's like a perpetual sort of seesaw, if you will, where success sows the seeds of its own destructions. Would you, how would you make an argument right now that San Francisco or Silicon Valley is doing that? Oh, I mean the obvious one. Well, the two obvious ones, I guess, um, are in culture and in housing. Um,

and cost in general. I mean, on the ladder, well, on cost, on the ladder, everything is getting more expensive. And nobody seems to quite understand exactly what's going on, right? And that is this, I mean, if you take healthcare again, for example, I mean, the case has been made that this is not, in fact, a bad thing, that what would you expect an enlightened society that has solved all of its other material needs to spend its money on, but healthcare, it's kind of, it's the last thing, it's the last frontier.

And perhaps we are actually getting sort of commensurate improvements if you sort of disaggregate appropriately and analyze the right way. Or perhaps not, right? How much of this is some kind of Baumol cost disease where some things are getting more efficient and that higher productivity and higher wages are sort of causing cost increases elsewhere to pay for opportunity costs and all the rest.

But I think sort of specifically in Silicon Valley and specifically on cost of living and housing, Silicon Valley is the sort of

greatest concentration of wealth creation that I think has ever existed in the US on a per square mile basis, potentially that has existed ever in the world, right? Facebook, Google, Apple, Intel, you know, they're all based in a fairly small number of square miles, right? And if you sort of

If you were to look at Seattle and the Bay Area together and look at that aggregate urban zone separated as they are by a two and a half hour flight, then of course you can layer in Amazon and Microsoft as well. Obviously what you see is that their rise in success was enabled in part by cheap mobility and cheap expansion. Again, through just political economy and collective decision making,

that no longer exists. Cheap mobility no longer exists and cheap expansion. And you can see it now in the sort of latest generation of upstarts, you know, be it Twitter or Uber or Airbnb or Lyft or whatever, who are, you know, facing these really significant kind of

structural headwinds. And so much of the wealth that's being created, this improbable fountain of wealth creation is accruing to the sort of lottery winners of the existing landowners rather than to the people who are actually doing the work. And because of that accrual, the sort of the barrier to entry for newcomers

is getting progressively higher, and you see it in declining rates of mobility. And furthermore, the other people in the city not in the tech industry who might otherwise benefit from it are, of course, getting priced out. And, you know, this is not necessary. I mean, you can look at places like, you know, obviously, Tokyo has, over the last couple of decades, been an improbable, well,

not especially improbable, but has been such an enormous economic success story. And, you know, you had the boom and the bust and the supposed stagnation of Japan in the kind of early 90s on. But sort of broadly speaking, it has done really well. But because of vastly fewer limitations on housing supply, have had just very stable housing costs, have not had the same displacement, right? And so the kind of the issues we face and we see here in San Francisco where it's getting ever, you know,

40% rise since we got to San Francisco in 2010. That's not necessary. It's not natural. And it's a function of our sort of collective decisions rather than kind of some secular and unavoidable economic force. And I guess I find it sort of dispiriting because it's a negative sum in the sense that it's not just that these gains go to

go to these sort of existing landowners, but actually there'll be fewer future gains. Like I think you should be mad about this, you know, if you don't live in Silicon Valley and you don't have the slightest interest in doing so because it's much less likely the next cool technology that you'd like to take advantage of will exist. It's sort of, it's a suffocation of future potential and of future gains.

And there aren't many places, well, if you believe in increasing returns to scale, that sort of, you know, this is kind of Paul Romer's work and others, that because of the sort of the collision of ideas and people in cities makes them more productive than if they were elsewhere. If you believe that to be the case, and there's like pretty good empirical data that it is, then it's,

You can't just move elsewhere. You can't just move to Nevada or wherever in the South. You actually will be less productive in those zones. And so, again, I think it's a real loss in terms of spillover gains to the rest of society in service of not building six-story buildings in San Francisco.

What do you think your role as a large employer and thoughtful citizen of San Francisco is in this? Well, I don't make any secret of the injustice and, well, the moral injustice in terms of the displacement that's occurring and the sort of economic wrongheadedness of the prevailing policies. And, you know,

I'm a landowner in San Francisco, John and I own a house together, and I hope its value declines. I think it's impossible to answer what the price of land should be. But I think it is very clear that on a marginal basis, the social returns of cheaper land in the most productive region of the country

would vastly outweigh the reduction wealth to existing landowners. But going back to the banks, everybody has a system that they want to protect. Totally, right, right. I mean, of course, you can try to estimate the magnitude here. And so over at Berkeley, this guy Enrico Moretti has estimated that 50% of US GDP growth between 1964 and I think 2010...

was left on the table, as it were, by sort of inefficient land use and land allocation. And obviously 50% is a high number and quite speculative, and it's very difficult to measure the counterfactual. But even just the idea that one can, with a straight face, hypothesize that it could be anything remotely in that vicinity, I think gives you a sense for how high the stakes here are, right? And yes, we can decide that, you know,

We place such an enormous premium on the aesthetic appearance of the San Francisco of today Recognizing that it is of approximately a third of the density of even just Greenwich Village in New York, right? We're not and you know that the the sort of the other extreme is not Hong Kong You you can triple San Francisco and get to Greenwich We can decide that that's our preference and but sort of sober estimates are measuring the cost of that in our

in double-digit percentage points of aggregate national GDP. And of course, when you look at our revealed preferences in terms of where we like to take vacations to or where we dream of spending a summer someday and things like that, it's to European cities, which tend to be of very significantly higher density. Paris, London, much, much higher density than San Francisco. And so, again...

I'm hesitant to cast normative judgment, but I personally feel strongly. I think that's a great place to leave this. This has been a phenomenal conversation. Thank you so much for coming on the show. Where can people find more about you? Well, if they want to start a business, they should head to Stripe.com. But if they want to subject themselves to more of the particular detritus that I post, they can head to my Twitter account, which is just Patrick C.,

Thank you so much. Thank you.

Lastly, if you enjoyed this or any other episode of The Knowledge Project, please consider subscribing and leaving a review. Every review helps us make the show better, expand our reach, and share our message with more people. And it only takes a minute. Thank you for listening and being part of the Furnham Street community.