Sequoia’s Roelof Botha on Decision Making, AI, and the Next Trillion Dollar Market | Ep. 28

Transcript

When you're in my shoes as part of the third generation to run the partnership, there's this enormous burden that Sequoia has been at the top of its game for a long time. Yeah. And we have these legendary companies that we've participated in. It was something like 30% of the total value of the NASDAQ is comprised of companies we were investors in when they were private businesses. I still don't understand how that happened, but that's crazy. Yeah. And so there is this expectation of, can you keep going? I am super excited to be here with you today. And I was just commenting to a friend that I was going to mess up your name before we started. So I haven't done this before, but could you introduce yourself? My name is Rolof Buita. I'm not even going to try, but I've been really looking forward to this. This was also probably the best pre-chat conversation I've had, where it turns out we both had a detached retina and this miserable surgery. So I feel lucky to have gotten to bond with you over that. It was brutal. I hated it. Misery loves company, and we can certainly bond over that experience. It's like spikes in your eye. Okay, here's where I wanted to start. I was thinking about trying to put myself in your shoes, and you're running what I think is widely considered the best, strongest, most storied venture capital firm. And you've been running it for three years now. And the first place I wanted to start was: Tell me about your mentality on day one, when you became the head, the steward of Sequoia, and how has it evolved over three years, and what has updated for you in your mentality as a leader of the firm? Interesting. So Sequoia has a long history of generational transfer. And so we have a very interesting culture where we, hence the title stewardship, are momentarily privileged to be working in Sequoia, and we have a duty to leave it for the next generation. And so even when I joined in 2003, I had the sense that there were people ahead of me who were willing to invest in me and nurture me and train me in a mentorship fashion. And maybe down the road, I'd be in a leadership position at Sequoia. So I think it says a lot about the culture that we have. And so we don't have a lot of discontinuities in our leadership. So when I joined and Michael Moritz and Doug Leone were running the partnership, Don was still around. Don interviewed me. As the founder, he wasn't overbearing, but he was present, and he provided counsel without having to be in the room. And if there was a disagreement, he respected that. I became the steward, actually, of the US business in 2017. Jim Goetz and I had been running the US venture business since 2010. And in 2017, I took over all of the US business when Doug was our senior steward globally. And then, as you say, in 2022, I became senior steward. But I think the point is really that there's a lot more continuity than you may think from the outside. Yeah. This week alone, I've spoken to both Doug and Jim, both of whom have been leaders here before, about interesting topics that I needed their input on or wanted their input on. Not because they have the right to it, but because I want their help. And that's the kind of spirit we have at Sequoia. When you think about what that steward role means, it struck me when I was speaking with Peter at Benchmark. And one of the things I thought was really cool was there was this sense that it's bigger than you. And when there's this thing that's been going for a long time and it's been passed from leader to leader and you're thinking about the future, obviously that's got some impact. But I'm curious how that shows up day to day, because maybe we can get into this, but take the opposite extreme. I just got started. Obviously, the steward concept makes no sense to me yet. I'm just at the beginning. Well, congratulations, by the way, on your new fundraising. 275 million. Yeah. That's nice. Yeah. That's a start. Another competitor. Absolutely. I'm a big one. 100%. Collaborator. Yeah, exactly. For now. But I think I'd imagine in your shoes that daily there is a presence behind it of, this thing is big, and there's something before me, there's something after me. Does it impact the way you make decisions or how you operate? It comes with an enormous amount of pressure. Yeah. I would think. You know, Don obviously made a wise decision not to call it Valentine Ventures, which was a pretty viable option at the time. And most people would have put the names of the founders on the door back in those days for most professional services firms. And he called it Sequoia for a reason. Sequoia trees can live to be 2,000 years old. He wanted to have a partnership that would outlive him and invest in companies that would endure. So that was an important ingredient. I suspect it surpassed his expectations, just as we are sometimes, often, always surprised to the upside by our winners. But then when you're in my shoes as part of the third generation to run the partnership, there is this enormous burden that Sequoia has been at the top of its game for a long time. Yeah. And we have these legendary companies that we've participated in. Something like 30% of the total value of the NASDAQ is comprised of companies we were investors in when they were private businesses. I still don't understand how that happened, but that's crazy. Yeah. And so there's this expectation of, can you keep going? Yeah. Don't screw it up. And that comes with enormous pressure at some level. And so there's an important dynamic of, how do you leverage this incredible platform that you have? I think we all feel privileged that we can do the best work we could possibly do because we have the benefit of the Sequoia brand. Yeah. We can win investment opportunities, open doors for our founders, help them do things for their companies to help them realize their ambitions that I wouldn't be able to do if I was anywhere else. And that's an incredible privilege. And yet we have to continue to innovate because I think there's a real temptation for leaders in industries to end up resting on their laurels, and you end up with the innovator's dilemma, where they stand still and they quickly become yesterday's winners. Actually, I did this exercise when I joined. If you look at the top venture firms in Silicon Valley in 1990, the majority of them no longer exist. There's a long half-life in the venture business, but there's no guarantee that you will succeed long-term. And so there's this insecurity that we have at Sequoia that really drives us. And it comes out in the people we recruit, the culture we have, the idea that we need to be both performance-focused but also innovative. How do you keep the paranoia or the sense that we're only as good as what we did yesterday? Because what I observe is there's a lot of firms that are good, but not Sequoia, where there is much more of a sense of comfort and maybe resting on laurels is too strong. I think venture is competitive in 2025, and I think most people are working hard. But I would say Sequoia is, if there's a quadrant of how successful and longstanding the firm is and how paranoid, Sequoia somehow seems like it's really got both, which is weird. It's not always the easiest place to be, by the way. It sounds stressful, by the way. It is very. Actually, if you come upstairs to our office where the investors go grab coffee and snacks, on the wall, wall to wall, in each individual's handwriting, we have printed the handwriting that says, "We are only as good as our next investment." And so every single day when I go grab a cup of coffee and I look at that wall, I see my own handwriting, and it's a reminder. We are only as good as our next investment. We cannot rest on our laurels. And so I think some of that is the people you recruit. Some of that is the culture that we hone at Sequoia. We care so deeply if we lose an investment. We are very diligent about looking at the coverage analysis of what our competitors invested in. And did we have a shot at that investment? Did we miss it? Did we not understand it? Is there a category that's emerging that we're slow to identify? We obsess about that. So I think it's a cultural trait. You have to nurture that. And the cost is that it's stressful, right? Yes. Yeah. That's the only way. Yeah. But winning feels good. Yeah, of course. It's interesting where, actually, I'm curious because if you're the Yankees, you expect to win. Is there any dynamic where it's like you become so deeply seated that the pain of losing hurts And so I think some of that is the people you recruit. Some of that is the culture that we hone at Sequoia. We care so deeply if we lose an investment. We are very diligent about looking at the coverage analysis of what our competitors invested in. And did we have a shot at that investment? Did we miss it? Did we not understand it? Is there a category that's emerging that we're slow to identify? We obsess about that. So I think it's a cultural trait. You have to nurture that. And the cost is that it's stressful, right? Yes. Yeah. That's the only way. Yeah. But winning feels good. Yeah, of course. It's interesting where, actually, I'm curious because, if you're the Yankees, you expect to win. Is there any dynamic where you become so deeply seated that the pain of losing hurts more than the joy of winning, or how do you keep culturally this paranoia and, we've got to stay there, and keep it joyful, where it's not just the sadness of losing and a relief when we win, but a joy? I think you'd find that most people who are very driven and competitive will say that there's an asymmetry and that the pain of a loss is far greater than the joy of a victory. I think you'll find that with most people that are driven and successful. And so I don't think you can wash that away or wave that away. I think it's a reality. We have tried over the last decade, especially, to do more to celebrate our victories. One of my partners, Jim Getz, used to talk about how we don't do celebration well at Sequoia. I think after the YouTube acquisition, we literally spent about 15 minutes. We huddled around the reception area of our office and we said, great. And then it was back to our desks. Yeah. Yeah. And he was a little bit surprised that we don't take more time to celebrate. So we've tried culturally to embrace celebration more and also to tell the stories. So one of the things I'm really proud of is every time there's a successful outcome, most recently we had the Klarna IPO, the Figma IPO earlier this year, the Wiz acquisition, and we were early investors in all these companies. We would write an email internally and we would celebrate not only the person who was then the board member, but everybody else that had an influence in that success. The person on our talent team that recruited that key executive that made a difference, or the communications team that helped craft the narrative for the IPO story, or our legal team that spent endless hours figuring out some nuance around governance or how the company was structured. Sometimes the people who did diligence or who found the company were not the people who were the board members by the time there was an exit. And so we celebrate everybody's contribution because we also play a team sport here. What for you is the most enjoyable part? Is it about the companies you're working with? Is it something to do with the team? Is it a purpose thing that's greater than the day to day? What drives your satisfaction now? Interesting question. There are a couple of answers to that. Probably the most important one is leaving Sequoia in a phenomenal place. And I'd love nothing more than a decade after I'm long gone at Sequoia being able to look at the team and see them flourish. That is probably the thing I think about most. And if the team flourished, that has a lot of inputs to that. That means that we've maintained a certain culture while we've continued to innovate. And it means that we're serving our founders. It's the development of individuals that I get a tremendous amount of joy in. Whether it's a young person who joined our investing team and a decade later, I see how they're flourishing. This is my favorite part of our own company. And it's individuals. Yeah. Just getting to see your team grow. And somebody joins when they're early in their career and then you just see a completely different person five years later. It's very gratifying. Yes. And the same with founders. The founders I've worked with, I think about MongoDB when we were an early investor in 2010 and what happened to them. I think about Jack when we first invested in Square 15 years ago and the kind of company it's evolved into and how Jack has developed as a phenomenal business leader. I look at some of the more recent investments we've made because I've helped us lead three investments this year in a variety of categories. Not all the investments are publicized yet, but one of them is a young founder, first-time company. He's a solo founder. And we had a fabulous 30-minute conversation yesterday about some of the challenges that he's dealing with. And just having a sounding board of me being able to talk to him to help him navigate a tricky issue, there's so much gratification that comes from that ability to pay it forward. Totally. When you think about Sequoia and its position in an evolving venture landscape, I'd be curious to hear, at a zoomed-out level, how you're reading the playing field today. And obviously we've got a bunch of dynamics happening at once. You've got AI, obviously, as the big tech wave that's going at the moment. There's others too, but that's the dominant one. You've got a bunch of firms that are really scaling capital. Some are new, some are older. You've got firms that are institutional and have been around for a long time. You've got new up-and-coming founder-led firms. I'm curious, when you think about that whole landscape, are you able to give a summary lay of the land of how you maybe see things, or what's your framework for the venture world right now? So we have the benefit of seeing many cycles over 50 years plus at Sequoia. Obviously the scale of what technology impacts today is different from what it was a decade ago, two decades ago. Technology just affects far more of the world than I think we could have imagined. So the scale is different, but history doesn't repeat, it rhymes. And I see echoes of 1999. I see echoes of what happened in 2008. I see some echoes of what happened in 2021. Happening now. Happening now. Yeah. The claims that are gravity defying, this is different, and I go back to the principles of investing. You think about the truisms of Benjamin Graham's teachings. I think those remain, and things got a little unhinged in 99 and it was different this time, and it wasn't. And then it was different in 2008 and then it wasn't. And so I just worry about that a little bit, that AI will have a tremendous impact and some of the other innovations that you mentioned, whether it's in robotics, whether it's happening with stablecoins and how that might change financial services. There's a lot of innovation happening. I'm really excited about the future we're going to build. We inevitably overestimate these in the short run and underestimate them in the long run. I'm reminded of the fact that it seemed pretty obvious in 2000, 2001 that e-commerce was going to be dominant. And yet here we are 25 years later and e-commerce is not even 20% of US purchases. It's crazy. It's incredible. It's taken a long time. Think about how long it's taken us to do unbundling of cable. Didn't it seem obvious that media should go this way? And so I think human behavior changes a little more slowly than technology is presenting us with. So I think it'll take a little bit more time. I don't think venture is an asset class. Why not? It doesn't support the numbers. So there was a lot of analysis back in the 1970s and 80s with the capital asset pricing model, and people figured out that there's this asset class that supposedly has uncorrelated returns, and a bunch of asset managers deem that they need to invest a certain percentage of their endowment or foundation or pension fund into this thing called venture capital. If you look at the data, there are basically 20 companies per year on average over the last 20, 30 years that have ended up being worth, in realized exits, a billion dollars or more, just 20 companies. Despite a lot more money plowing into venture capital, we haven't seen a material change in the number of companies that are outcomes that are that large. And I think part of that is that there's a lot more talent than really interesting ideas or interesting companies to be built. And I think we're spreading a lot of that talent thin right now, similar to what happened in 1999, by the way. Yeah. So when you look at the data, the amount of money going into venture capital right now in America is in the order of $250 billion a year. uncorrelated returns and a bunch of asset managers deem that they need to invest a certain percentage of their endowment or foundation or pension fund into this thing called venture capital. If you look at the data, there are 20 companies per year on average over the last 20, 30 years that have ended up being worth, in realized exits, a billion dollars or more, just 20 companies. Despite a lot more money plowing into venture capital, we haven't seen a material change in the number of companies that are outcomes that are that large. And I think part of that is that there's a lot more talent than really interesting ideas or interesting companies to be built. And I think we're spreading a lot of that talent thin right now, similar to what happened in 1999, by the way. Yeah. So when you look at the data, the amount of money going into venture capital right now in America is in the order of $250 billion a year. And the numbers, these are all estimates. Let's just say it's $250 billion a year. Need a lot of exits. Well, let's do some very simple arithmetic for a second. $250 billion going in every single year. If you assume that the firms generate 12% IRRs net, net of fees and carry, which isn't that great, by the way. Over the last three or four years, the NASDAQ has compounded at 16, 17%. Let's just say 12%, not spectacular. You're basically just average performance. You'd need a 3.7x, roughly. On 10 years. On a seven-year exit horizon. So I'm being a little bit aggressive. Maybe it won't even be that good. So 3.7x on $250 billion, that approximates to a trillion dollars a year. Coming out. Coming out. By the way, that means, and that's what the investors own. So let's say that the investors own two-thirds of the company to make the arithmetic simple. That's 1.5 trillion annually in company exit value. Yes. Just think about that for a second. Where's that coming from? Well, Figma's worth what? It's a lot, but it's... 40-ish billion. That's... Let's say it's worth 0.03 trillion. So if you start thinking in trillions and Figma gets you 0.03 trillion... You need 30, 40, 50 Figmas every year to make that arithmetic work. I don't see that many companies of that scale every year. So the only thing that breaks is the return assumption doesn't hold. Yeah. And so venture is a return-free risk. Not a risk-free return. That's terrible. Yeah. You are better off investing in the index or holding T-bills, honestly. And so I don't think venture is an asset class. Asset classes scale if you add more money. You can build more real estate. There's a lot of equities. There are trillions and trillions of bonds to be purchased. Venture capital doesn't scale with more money. I agree with you. But just to take collective, just to take, what would the collective argument be that the herd is all betting is happening? I guess there could be two things. One argument would be these companies are going to be bigger than ever. This was roughly what Marc Andreessen said on the podcast, was something like, there's going to be multi-trillion-dollar companies in a way that there haven't been in the past. And you look at OpenAI and SpaceX and Anthropic and Anduril and so on. And there's a lot more coming out than there ever has been before. That's argument one. I guess the other argument is everybody thinks they're a better-than-average driver. Everybody thinks they're a better-than-average allocator as an LP and a GP. And so, yes, the overall asset class isn't doing well. But I, as a particular LP, know how to pick the good ones. And so the top best style is going to be great. Do you think that there is wisdom in the number of dollars going in and where the overall industry is? Does it make sense? Or is it collectively smart or collectively stupid? Well, firstly, I agree with Marc that the scale of the outcomes are completely different today and will be even bigger in the future than they are now. So I'm not trying to be a Luddite saying that we should go back to the way the business was. Yeah. When I was at PayPal, there were roughly 300 million people on the planet that had access to the internet, and most of them were on dial-up. And today we have, what, four billion people with high-speed mobile devices connected to the internet. We have all this data. So the world is completely different, and the scale of the outcomes is much, much bigger. So I completely agree with that. It's just that I don't think there are enough of them. Right. No, I hear the nuance you're saying is there might be one-and-a-half-trillion-dollar companies, but is there one every year? That's tough. Doesn't seem like it. That's the thing that is really challenging in this equation. I just don't think there are enough to make the math work for it to be an asset class. Is it an industry? Absolutely. Does it account for a lot of innovation in America to keep our economy competitive? Absolutely. Does it drive an enormous amount of job creation? Yes. All of those things are true. I just don't think it's an asset class. It's that particular word that I have an issue with. And the amount of money, I think, is in excess of what the industry can bear to generate good returns. So that is my quibble. People thought that we'd gotten rid of gravity in 2021 and that we're going to have so many more spectacular outcomes. And it just did not prove to be true. Not enough to merit the amount of money going in. So I think there's a little bit of prospect theory. If you think about Kahneman and Tversky's original paper, I think there's a little bit of people are risk-seeking in this domain. We'll take a chance. Maybe I'll be lucky. Maybe I'll get the double zero on the roulette table, and now's my chance. I think that drives more of this behavior. Let's say you were on my board at Alt Capital. What would you be pushing for me if you were trying to give me advice or trying to push on the thing that a new manager getting started needs to do to be successful? If you want to build something that lasts, what would you be harping on? The network you build, the tributaries you develop for getting access to interesting emerging investment opportunities. This is not a business that you do sitting at a desk. You need to get out there. You need to be able to meet people and understand where interesting new founders are thinking about building businesses. And part of that is being smart, where they want to have a conversation with you. So you need to study up. If it's not your field necessarily, can you study up in the categories that they're innovating on so that you can have interesting conversations that are memorable to them? Because it is a competitive business, and then you need to be congenial. People want to do business with people. Yeah. When you think about the best investments that you've made over the last, I don't know, let's say five years or something, do you feel like more of them are deals that were controversial at the time and you picked something that was hard to pick as a firm? Or do you think it was deals that were clearly good and competitive and you won them through relationships and effort and all the rest? A bit of both. Yeah. Yeah. If you had to bucket them, do you feel like it's a good mix? It is a good mix. I mean, Zoom with a Z. Yeah. Or Z as I grew up saying. It was one of those investments that was not controversial. We all thought it was spectacular. The company was already generating cash. Eric had built a truly differentiated product in video conferencing and belied all the naysayers that that was a tough category. And the challenge was winning the investment. And we were able to win that investment. And that was a case where we were all above the line, so to speak, in our internal voting system. And then there are other ones that are maybe a little more controversial, where it's a little more fuzzy. We had one actually last year. It's a company called Aspora. They provide non-resident Indian remittance services, and they're built on stablecoin infrastructure. So they're really reinventing the whole process and not dealing with the interbank system that's expensive and cumbersome and slow and non-transparent. And honestly, that was a case where I didn't quite get it. I was the one person who listened to the presentation, and I really liked the founder path. I thought he was super dynamic. And I was a little bit worried that there were a lot of risks in the business. And the challenge was winning the investment. And we were able to win that investment. And that was a case where we were all above the line, so to speak, in our internal voting system. And then there are other ones that are maybe a little more controversial, where it's a little more fuzzy. We had one last year. It's a company called Aspora. They provide non-resident Indian remittance services, and they're built on stablecoin infrastructure. So they're really reinventing the whole process and not dealing with the interbank system that's expensive and cumbersome and slow and non-transparent. And honestly, that was a case where I didn't quite get it. I was the one person who listened to the presentation, and I really liked the founder path. I thought he was super dynamic. And I was a little worried that there were a lot of risks in the business. And then I looked at the assessment from the rest of the team, and I figured out I'm the one who doesn't quite get it today. Maybe I didn't sleep well. Maybe I got out of the wrong side of the bed this morning. And so we looked at the vote distribution, and I said, despite my inclination, we should absolutely make this investment. We did. And when we did the first portfolio review, I was relieved that we did make the investment because the company is absolutely flourishing, and I'm glad I didn't block it. Which is actually probably super important as a firm leader, to be like, you've got things about your vantage point that are going to help. But trusting the team in those cases, obviously, that's a good reminder. Yeah, and I think, thank you for saying that. The one thing we've realized, when there are controversial decisions, you sometimes have one or two people who don't quite get it. And we make consensus investment decisions at Sequoia. Actually, when I joined Sequoia, it blew my mind that that was the case. I thought that committees were exactly where bad ideas happen. And then I understood that Sequoia has this approach to teamwork, which means that every time we make an investment, it's our investment. It's not your investment, it's our investment. And it means that six months down the road, when you need help hiring somebody or you have a strategic question, I'm helping you. I'm not going to brush it off and say, it's your problem. You made that investment over my objection. So I understood the power of us making investments as a team. But that means that sometimes you have a controversial investment where one person is a negative and everybody else sees it. And Airbnb was a little bit like this. When the three founders came in, it was Airbed and Breakfast at the time. It was just the three founders. It was a nascent company. And there were a few people in the partnership who struggled with this idea that strangers are going to stay in each other's rooms. Is that going to happen? Those people were willing to go with the conviction of the sponsors and gave the full support after we had the conversation and didn't block it. So I think we've tried to harness that kind of decision process. You still have a forthright conversation, but we empower those who have conviction. Do you need to ultimately get a unanimous yes from something like, as part of the deal there, that there's a sponsor and they have to convince other people? Or can you get through somebody who's like, I still think we shouldn't do this? What we've done with our voting distribution as we discuss investment opportunities is we want to have a full-throated, full-contact conversation, as we call it. And it has to be about the merits of the investment, by the way. It's nothing personal. As soon as you walk out of the room, it's as though nothing had happened. Yeah. Because it's very important. It's so hard to have that. It's so powerful when you have it. It requires a lot of, maybe you could do what people don't like each other, but I feel like you have to like each other and you have to trust each other. The key is to trust each other, by the way. And one of the things we do when we have our offsites, we do these check-ins with each other. Yeah. And almost every time we do a check-in, somebody cries. At least one person cries. And people talk about what's happening with family health issues. They talk about challenges they're having with one of their portfolio companies, whatever it is. We have an incredibly transparent conversation with each other. And we've done all these other things to build trust. Because that's what you need. You need that well of trust if you're going to have these kinds of investment conversations. Yeah. So the idea is to lay it all out. And if I don't quite see it the way you do, I'm going to give you my full opinion. And then at the end of the conversation, if I feel that you've really listened to my objections, you've weighed them, you've digested them, maybe you've come back a week later and you've actually answered some of the diligence questions. And you say, listen, Roloff, I hear your objections, but I see it this way. I then have to decide, am I going to block it? Or am I going to say, I'm glad you heard it. I don't quite see it the same way, but I'm going to ride with your conviction. And that's part of the nuance of a subtle investment conversation that we have to get right. Yeah. And if you don't have that, the alternative is like, I think this is a terrible idea and I'm just going to graciously bite my tongue and then complain about it to somebody else in the firm later. Which I think is a common state of dysfunction in venture. Yeah. We have this phrase, front stabbing. If you're saying something bad, say it to them. Yeah. It's really hard. It requires a certain culture. It's hard. And that's why I'm exhausted at the end of Mondays, by the way. Yeah. At the end of a Monday partner conversation, I think our whole team feels this way because we're thinking about making really consequential decisions. By the way, it's part of why we keep our Mondays very light. Mondays, it's like the Olympic finals. This is where we make investment decisions. This is what we're getting paid to do. How do you avoid people being afraid to give you the front stab? Because I'm sure you want it. You're right. Because we want the triumph of ideas, not the triumph of seniority. It's really important. So to give you one sense, we had an offsite in 2019 where we were debating certain strategic decisions for Sequoia. And we asked each team member to write a pre-mortem and a pre-parade for Sequoia in 2030. So this was in 2019. So it was roughly a 10-year horizon. And this pre-mortem, pre-parade, we put in our investment memos as well. And then we anonymized each person's write-up because we wanted to be untainted by the name or the seniority of the individual. We only wanted to focus on the merits of the ideas and the criticisms and what might go right, what might go wrong. When we do an initial vote on an investment on a Monday, we do it anonymously. How many people are in that Monday meeting? How big a group can you have looking at things together? So we separate our early team and our growth team's conversations. And then we get everybody's vote, even if they're not yet a managing member, because we want everybody's input. So it's roughly a dozen people. For each fund, we only have six technical decision-makers because we think you want to have a small number of people actually feel the accountability and the heat for the final investment decision. Six is good. But we have a dozen people who would vote because somebody who might have joined a year ago might have the right insight. It might be the person who just graduated, who has a fabulous connection or a fabulous insight on a new emerging technology. We want to harness their insights. But we look at an anonymous vote distribution. And if you look at that and you go, we have three people that are below the line. That's interesting. And how do we tease that out in the conversation to really have a proper conversation? When you think about natural limits to scaling a venture firm effectively, obviously, we just talked about the money-in, money-out thing. It would strike me, and obviously, we're very small, but it would strike me. How many people do you have in your firm? There's three others besides three. You plus three. Yeah. And are they all on the investing team, or is some of that support? I mean, it's so early that we're all working together, and they're all earlier in their career than me. But it's a team that, like what you're describing, I trust a lot and who will tell me when they think what I'm saying is stupid or when I don't. I will very much defer to their judgment. I don't have any particular reason to think that my judgment is any better. When you think about natural limits to scaling a venture firm effectively, obviously, we just talked about the money in, money out thing. It would strike me. And obviously, we're very small. But it would strike me. How many people do you have in your firm? There's three others besides me. You plus three. Yeah. And are they all on the investing team, or is some of that support? It's so early that we're all working together, and they're all earlier in their career than me. But it's a team that, what you're describing, I trust a lot and who will tell me when they think what I'm saying is stupid. I will very much defer to their judgment. I don't have any particular reason to think that my judgment is any better. And so, there are advantages and disadvantages to all this stuff. But I think of it very much as a team where I know about myself that I'm too optimistic, and I believe in people too much. And I think any idea can work, and I am dreaming about the future all the time. I just know that about myself. You and I are alike. I need to be balanced. Naive optimism. Yeah. I need to be balanced with people who remind me that most companies aren't big. And that's just how it works. That's my starting point in the world, is I go around liking everybody. So you just have to be careful if you know that's your state. Interesting. But by the way, it's useful sometimes to appoint a devil's advocate. So now and again, when we have a situation where everybody is above the line, I might privately message somebody in the Zoom chat and just say, listen, do you mind being the devil's advocate, just so we can actually have a proper conversation and reassess? Yeah. So you might want to think about that as a technique. It certainly helped us a bit. But I would think that past a certain size group, it becomes, I felt this with my exec team at Lattice, where we would go through these evolutions where the exec team grows and grows, and then at some point the exec meeting becomes useless. And then we would split into a broader exec and a senior team meeting, or whatever. And that evolution happened many times. Yeah. I would think there's some limit to the number of people you can have working on a team. You talked about going deep enough with people that people can share things that make them cry. You just can't do that with a big group. Yes. I guess that's a limiter to what a venture firm can do. I completely agree, or at a subunit decision-making level. So we have an early team and a growth team at Sequoia. So the early team, we have about a dozen people in the room. The growth team, we have about a dozen people in the room. Right. No more than that, I think, to make effective investment decisions. Because you need that trust. You need that sense of camaraderie. You also need enough time with each other. Of all the GPs at Sequoia now, I think everybody's been here for at least five years. So we've had many years of working together and understanding each other's nuances and traits and quirks. You can have a heated discussion and still be friends afterward. Exactly. And I think there's a lot of behavioral science research on group size for effective decision-making. And a dozen is about the upper bound of what you can do. There's this phrase that the camel is the horse that was designed by the committee. We don't have a committee making decisions. We have a small group of people making decisions. And I think, by the way, that's one of the conclusions of that 2019 offsite we had, was for us to be a leading firm in 2030, we thought we had to keep our investment team small. It was so important for us. In total, we have about 25 investors at Sequoia. That's it. And what we've done is we've invested in technology to give us more superpowers so we don't have to grow the size of the team. To the point about what's coming out, there are not that many companies that are going to drive all of the returns. And so you shouldn't need that many people to find your way into those over time, I would think, particularly when you can invest in many stages. The caveat here would be at early stage. It seems extremely hard to catch everything. One of the things that I'm jealous of that you have is multi-stage. You can miss things at seed and the A, and you can track it, and then you can go lead the B and the C if you want to. And that does seem like a powerful thing. Obviously, the earlier, the better. And we chatted for a little bit before about how you always need to be in the business of early. I strongly believe that. But it does seem like you have the advantage where, on some level, I imagine if I'm in your seat, I would be like, no matter when, in the end, I need to be in every important company at some point, roughly speaking. Is that kind of how you think about it? At the right price. We're not in the business of buying posters. We want to be business partners to founders that make a difference and build great long-term companies. And in almost every single case, we are on the boards of these companies by the invitation of the founder, who wants us at their side as they navigate tricky issues in their business. Alfred is still on the board of DoorDash years after it's gone public, helping them think through their international expansion strategy and becoming the last mile of commerce in general. I'm still on the board of Mongo and Square, which were investments from more than a decade ago, because the founders still want me on the boards. So I think we want that association. You're right that it's convenient on some level that you might catch them later, but that's also risky because it means that there's a risk of complacency. Yes. It's like, oh, I don't have to make the seed investment. If it works out, I'll catch them in 12 months, or I'll catch them for the B. And so we obsess about making the right decision at every single stage. Just never let yourself off the hook. You can't. And by the way, part of the joy is being able to participate in the company-building journey from the get-go. Helping those founders from that initial idea before a line of code was written. It's a totally different relationship. And navigating all the tricky issues. There's so much, by the way, just from a fun point of view. Those are the formative, that's the formative time. It's so much fun to be part of the company from those early stages. And so we love that, and we love doubling down on companies when... I do think when you join a Series B or C and you've got special people, like Pat Grady or Ravi Gupta or Andrew, there are people who can still become the partner of record, even at the B or the C. Oh, absolutely. Even with a great A investor. So it's not so late that all the glue is dry or something like that. Oh, absolutely. There's a tremendous amount. That's part of why we remain on boards of companies sometimes after they go public. There's still a lot to do from then. There's a long runway. And that's why we created the Sequoia Capital Fund in 2022, because we believe so much of the upside can be in these companies long after they go public. Yeah. Just to give you a sense, ServiceNow, HubSpot, MongoDB, all in recent memory, Palo Alto Networks, these are all companies that are 10x returns after the IPO. So we believe there's a lot of money to be made for LPs after the IPO, but there's also a lot of fun company-building to be done. So you're absolutely right that there's a lot of company-building all along the way, but it's so much fun to be part of it from inception. And so we can't rest on our laurels. We just can't take comfort in the fact that we might catch them later. Figma is an example of a company that Andrew helped us lead the investment in from our growth fund. We led the Series C. And boy, the early team, we were kicking ourselves. At the time that we first met Dylan, this idea of the browser being sufficiently performant to realize Figma's vision, it wasn't quite yet obvious. So we believe there's a lot of money to be made for LPs after the IPO, but there's also a lot of fun company building to be done. So you're absolutely right that there's a lot of company building all along the way, but it's so much fun to be part of it from inception. And so we can't rest on our laurels. We just can't take comfort in the fact that we might catch them later. Figma is an example of a company that Andrew helped us lead the investment from our growth fund. We led the Series C. And boy, the early team were kicking ourselves. At the time that we first met Dylan, this idea of the browser being sufficiently performant to realize Figma's vision wasn't quite yet obvious. And we didn't get it right. And boy, did we beat ourselves up for not getting it right. Yeah. We're delighted we got it later. Yeah. But we keep trying to learn from that lesson. It's also when you do it early and then you lead, I think with WhatsApp, you led a bunch of rounds. And you end up with these unbelievable results. You can only have those dynamics happen when you really just keep doing all those early rounds, I guess, too. So there's that. Yeah. We've often doubled down, whether it was at ServiceNow, whether it was at WhatsApp, whether it was MongoDB, Unity, Airbnb, DoorDash. A lot of these examples are where we've kept on investing in great companies down the road. I find that more impressive almost than a single good investment decision. Even one seed investment, obviously, it's still impressive when you get a good one, or it's sad when you miss it when you do it wrong. But there are so many of these decisions, a first decision, a first-round decision, where you're on the cusp and you barely pass or you barely say yes. To me, when firms can do the round over and over and keep making a correct decision in a market-contrarian way, or they seem to be overpaying, or they own 25%, but they lead a whole other round when they don't really have to. They're already so exposed. There's something more impressive about that to me. Yeah. It seems harder to do, especially I would imagine once you're already a few rounds in. There's something about that that really impresses me. Yeah, we don't always get it right, but I agree with you that it's impressive when it can happen. And part of it is the psychological anchoring we're all subject to. Yes. We just invested in this company's seed. Now it's 10 times more expensive five months later. Really? I mean, why? So we doubled down on a company called Listen Labs recently, Alfred and Florian building this company to bring AI to market research. The company's on fire. Love this investment. We made a seed investment roughly two years ago, and then we doubled down on the Series A. But we own a lot. We're in the seed. And why don't we let somebody else lead the A? And it's a very interesting conversation to have. And then you plunge in and you make the Series A investment and you're delighted. Now the next question will be the B, obviously. I would just think it takes a lot of, it takes a very clinical mindset to be untethered to that stuff. Yeah. Clinical, unemotional, without being a sociopath. Yeah. That's a tough balance. Maybe a little sociopath sounds better for some people. I don't know. One of the things I've really thought about a lot is behavioral economics and psychology and how it applies to investing. And we're all subject to it. I'm as guilty as the next person of falling prey to heuristics. The key is when you can identify and name them. And when you educate your team about it and we can all hold each other accountable to it, you can get over these biases more effectively. Yeah. And we actually have started to put this in some of our investment memos where we'll actually write down the biases the author thinks they're guilty of as they're recommending a particular investment. Because the more you can name it, the more you can discuss it and look at it clinically, to the point that you're making. I've heard of some firms where the partner who did the initial investment is not allowed to make the call on a double down because it's too hard. They believe it's too hard to dissociate yourself. And that you can either have the version where you just become too obsessed and you go native and you're just like, this is back to the end of the earth. Or the other direction where you know all the warts and nooks and crannies, and so you just don't see it as cleanly as somebody who's coming in with fresh eyes or something, which I thought was kind of an interesting way to go around it. Yeah, we definitely get fresh eyes on it. We still want to harness the insight of the person who is working closely with the company and complement that with somebody who brings a slightly third-party perspective. A lot of focus right now in AI software companies goes to the product, which is obviously very new and interesting. And AI is powering that. I'm curious to talk to you about the other side of it, which is the cost structure. I know you've spoken about this in other places, but I find it very interesting and kind of under-discussed. Before getting into how it applies in AI times, which I think is new, interesting dynamics, how do you think about the component of costs for a startup building a company, its product, its go-to-market, everything? It's a very unsexy thing to talk about because certainly what most journalists want to write about is the snazzy new product innovation and the features that are so cool and whiz-bang. And then I tell people cost is the secret of Silicon Valley. And they look at me very puzzled, and I unpack it for them slowly, which is I think relentless cost reduction is actually a far bigger ingredient to Silicon Valley success than most people realize. And some of that makes technology available to many more people. And it democratizes access, if you want to think about it that way, whether it's the little Square reader that turned every mobile phone into a credit card terminal, whether it was the way that SpaceX reduced the cost of travel to space by an order of magnitude, or Google's innovation in its data centers. And we just look at industry after industry. Cost reduction has enabled technology's ubiquity. And so I think people need to obsess about that. There are two pieces of it. One is the gross margin that goes into the cost of your particular product. And the other one is the fixed cost of running your business. I think today, partly because of cloud infrastructure, mobile technology, and AI, the basic cost of running a business is lower than it's ever been. I think it's a continuum, by the way. So 25 years ago, when I was the CFO at PayPal, we would cut checks to Oracle for databases and Sun for servers. Five years later, when YouTube came around, we were using MySQL, Memcache, a bunch of really good open-source software, and starting to use the beginning of cloud infrastructure. We didn't actually have to stand up a Colo facility. We could use commodity servers. Google obviously took that to another extreme. When YouTube was acquired, it had about 50 people in the company. I think when WhatsApp was acquired by Facebook, they had about 30-odd employees. When Instagram got acquired, another company we invested in early on, Kevin and Mike, that company had about 20 people, roughly speaking, at the time it got acquired. And I think there's this potential that you're going to end up with a company that has literally single digits of employees and is worth a billion dollars. Definitely. I think that's around the corner. And it's this incredible availability of infrastructure for scaling. And it's a magic time to be a founder. So that's on the fixed cost. But you still need to think about your marginal cost. And I think the founders that really succeed, in my mind, are obsessed with understanding how they drive down the cost to serve their customers. And they end up with very high gross margins. Why is it so important to have high gross margins? On some level, you could be like, oh, well, you want to make more money drop to the bottom line. But I think it's deeper than that. So I'm curious, why does this rise to the level of a secret of Silicon Valley? I think there's a misunderstanding where people often think that price is a competitive advantage. What's the secret to your company's success? Well, I'm going to price lower. That's not an advantage. Cost is an advantage. If you have a fundamental cost advantage over a competitor, you might price the same and just end up with high gross margins. Maybe you price a little lower. But you still need to think about your marginal cost. And I think the founders that really succeed, in my mind, are obsessed with understanding how they drive down the cost to serve their customers. And they end up with very high gross margins. Why is it so important to have high gross margins? On some level, you could be like, oh, well, you want to make more money drop to bottom line. But I think it's deeper than that. So I'm curious, why does this rise to the level of a secret of Silicon Valley? I think there's a misunderstanding where people often think that price is a competitive advantage. What's the secret to your company's success? Well, I'm going to price lower. That's not an advantage. Cost is an advantage. If you have a fundamental cost advantage over a competitor, you might price the same and just end up with high gross margins. Maybe you price a little lower. Maybe you have the same gross margin percentage as your competitor, but you can price below them and gain market share because your costs are fundamentally lower. And so I think the reason I obsess about cost is it gives you the degrees of freedom to choose how to play the game. And when you think about being powerful in business, when you dictate the rules, that's how you can end up succeeding. Yeah. If you're talking to a founder about this and they say, I don't have lower cost right now, but I will. And in the interim, I can raise a ton of money because of X, Y, Z. Does that argument land for you? Or do you say that is a hope and a dream and we don't think about it that way? The question there is, what is the logic train, and is there evidence to support the hypothesis? So when we were an early investor in DoorDash, we led the series A. Tony showed us the unit economics that he had at a city level. And there was a point in time when the company needed to raise expansion capital where the burn rate was increasing. And so if you looked at the company financially superficially, you might have concluded that it isn't building a viable business. But if you peel the onion and you went down to a basic unit level and you actually understood the profitability in a particular town, town by town, you saw that the business was working. And the reason it was burning more money is they were expanding and building new markets, but they perfected the playbook. And I think that's the kind of diligence that you need to understand. So that's why we doubled down on DoorDash, to the earlier conversation. And we're delighted we did because Tony understood the unit economics of his business. And he has a detailed understanding of his business, which is unrivaled. I also think, in the long term, this margin question, to me, tells you what you can afford to spend building your company. One day you want to make money as a company. So let's just say you want to be profitable. You've got 100% of your revenue that you can spend on everything. If you've got 80% of that left for R&D, go to market, everything else like that, that affords a certain envelope. Versus if you've got 20% left, it's like, okay, how many companies have a really good R&D org that's 6% of revenue? Not many. Most of the great ones, for whatever reason, the R&D orgs seem to be a lot higher than that. And maybe that'll change over time. But I also think of it as this offense, where it lets you spend a bunch more on project marketing. Profits or power? Profits or power? That's what you want to focus on. That gives you the interesting thing about empires, by the way, relative to just regions or countries or nations. Empires have flexible borders and relentless ambition. And I think about the great companies having those two characteristics. Flexible borders means they keep on innovating and pushing the boundaries on interesting new innovations. They expand into categories that maybe were unanticipated, novel, unexpected. And they're relentless. Yeah. And that ambition is part of why you create these great companies. But you need profits to fuel that. I want to just connect really quickly to the cost structure in the AI world. Because obviously margins, like if SaaS companies before AI were often 80%, a lot are lower now. They're not zero, some are zero, but a lot are 60 or 40, whatever. But there's these costs in there. And then everybody's kind of collectively assuming that the price of intelligence will go down at a certain rate. And so you can kind of capture market at a certain speed. And these markets are blue ocean right now, but they won't be for long. And so when you're triangulating and reasoning through these dynamics right now, how do you look at this when you see a company that's a SaaS company with a 40% margin because they're using a lot of compute? So I think this is a really good question. It's a contemporary question because a lot of people are looking at the current margin structure of these, especially AI application companies, and wondering if it's a sustainable or a good business given the margin structure. And there's so many parallels to think about. In general, in business, there's this idea of an experience cost curve, that as you increase production in a particular industry, production conceptually, you end up with a very predictable curve in how costs decrease. And there's an economist, a British economist, back, I think, in the 19th century, who originally coined this and actually modeled it very accurately. If you go back to what happened in the photovoltaic industry, solar today is less expensive than we thought it would be 15 years ago. And solar is actually producing more electricity today than we had predicted 15 years ago. You don't read about this often. People often think that solar was overhyped 15 years ago. It was actually underhyped. We fail to understand how it compounds. And a lot of people have written about humans not intuiting compound interest. Anthropologically, understanding compounding didn't benefit us when we were hunter-gatherers or whatever the case is. And so we don't quite anticipate that. And so I remember in the early days of cloud infrastructure, people were dismissive. I remember at MongoDB, when we went from being an on-prem software company to building Atlas, our cloud database as a service, initially our gross margins were basically zero. And people were wondering, well, are you going to build a real business being a cloud database as a service because the cost is so high? But we had confidence that you would drive those costs down. There was such a clear curve that you could walk down. And today the company has fabulous gross margins. The same is going to happen in AI. I think if a company has product market fit today, the cost of tokens is going to keep on coming down very aggressively. I think the algorithms are going to improve. The scale is going to increase. You have open source models that are going to compete against some of the proprietary models. And I would not invest or not believe that the company isn't going to be able to improve its gross margins. I would bet on that all day long. Do you think that we, in some sense, need the rate of progress of frontier models to slow? Because the faster that they're progressing, the more people are willing to pay for the frontier, and then the margin improvements never come. Or do you think they'll come even if the models keep getting better and better for the next 20 years? Oh, I think they'll keep coming because the application space is very... And I spoke to one of our founders, Max, who runs a company called Fair. And he described to me how they're using an ensemble of different models for different use cases in the company. And they use some cheap ones in some cases and expensive ones in some. Exactly. In certain cases, the application is relatively low value, and paying for the latest, greatest frontier model doesn't make economic sense. But they can sacrifice the quality and speed with an open source model. And the cost benefit works in that particular application. And in other applications, it doesn't. They need to use the frontier model. So it'll basically be in a world where you'll use the frontier for some percentage, but not 100%. Probably. Some very cheap stuff for some large percent, probably. Depending on the cost benefit. And I love the fact that if the frontier models keep pushing, you'll get a continuum. You'll have so many choices for picking the right model for the use case that you have. To go back to something you said earlier that I thought was interesting, that empires are relentless ambition and open borders. Flexible borders. Flexible borders. Not open borders. Not open borders. Let's cut that. I'm just kidding. We can keep that. There's companies that are like that, too, which presents an interesting problem, which is conflicts that the founder feels that the VC doesn't feel. A company thinks that a space that they're not in today is one that they may or may not be in in the future. They need to use the frontier model. So it'll be in a world where you'll use the frontier for some percentage, but not 100%. Probably. Some very cheap stuff for some large percent, probably. Depending on the cost benefit. And I love the fact that if the frontier models keep pushing, you'll get a continuum. You'll have so many choices for picking the right model for the use case that you have. To go back to something you said earlier that I thought was interesting, that empires are relentless ambition and open borders. Flexible borders. Flexible borders. Not open borders. Not open borders. Let's cut that. I'm just kidding. We can keep that. There's companies that are like that too, which presents an interesting problem, which is conflicts that the founder feels that the VC doesn't feel. A company thinks that a space that they're not in today is one that they may or may not be in in the future. And they'd really prefer that you don't stamp the nice green Sequoia logo on that company because it might be on the roadmap in a few years. Please don't do it. What do you say? How do you handle this situation? Because I think going back to limiting factors for venture to scale, conflicts present another one, which is when you're deep enough with the founder. It feels treasonous to invest in a competitor. But those borders are fuzzy when companies are empires. And I think we at Sequoia have a distinct challenge when it comes to this because we're your business partner. I think sometimes if you're an early seed investor, seed investors can maybe invest in several companies in a category. Yeah. So can the super growth investors. So can the super growth investors because they often aren't taking board seats and not involved with helping with the strategic guidance of a company. And so I think it's a firm like ours. It's in the middle where we really are your business partner of choice. But you've chosen me and you are my thought partner to help me figure out how to achieve world dominance. Yeah, exactly. I don't want to share you with anybody else. So we feel that pain probably more than most. In some situations, we've invested in companies that subsequently became competitive. So we invested very early in both Stripe and Square. And so John and Patrick at Stripe at some point realized that they had ambitions to move into spaces that Square was competing in. And Jack over time realized that he wanted to compete in some areas that Stripe is competitive in. I think candidly, the companies have more in common and can work together very well. And I've introduced them to each other. And I think there's more partnership opportunity now. But there was a time where the two viewed each other as competitive. And so when John and Patrick would give the partnership an update, I would actually not join the meeting. And I was not allowed to read the investment memo. And in our internal systems, I actually couldn't access any of the information about Stripe because we wanted to make sure that we've preserved the sanctity of their information, distinct from anybody who was working with who they deemed to be a competitor. So sometimes it happens subsequent to our investment because we were early investors in both. They converge. So this happens to us frequently. I think the real challenge is at entry point. Yeah, because the convergence is no harm. I mean, yeah, it happens subsequently. And what do you do when it's a very mature company that has an ambition on a space? Yes. And those companies often have 17 ambitions going at once that look adjacent to them when they get to a certain size. Yeah. So it comes down to a conversation and having, is this really in your bullseye or is this an adjacency and an option value? Because, by the way, often if we don't invest, somebody else is probably going to invest in this company. And so you're going to deal with them as a competitive issue. And so we want to have a conversation with our closest business partners to really understand, is this really one of your top five issues? Yeah. Or is this a peripheral issue that you can live with? Yeah. And it's a conversation because we're in a relationship business with our founders. So sometimes we do walk away from investments for competitive reasons. And I guess it's probably a judgment call about how deep is the partnership? How much do you agree or disagree with the founder? How upset are they? And you just have to make these ongoing business judgment calls. Yes. But that's what we get paid to do, is render judgment. One of many. Thanks. I want to talk about a couple of the areas around AI. I want to start with robotics. It seems to me that it really ought to work. A couple of the podcast guests I've had have walked out articulations where I left being like, this is the next trillion-dollar market for sure. I'm curious where you think about a market like that where it seemed like it could be an expensive, ubiquitous, completely new type of product. How much time do you spend on something like that? Vinod and I agree on many things. We've worked on a few companies together. And on this one, I agree with him too. I'm very optimistic about the future of robotics. Some of that is enabled by AI. We made an early investment in Deepak and Abhinav, who are the founders of Skilled. They're building a foundation model for robotics out of CMU. And it's incredible to see what they're able to do with commodity hardware and infusing it with the knowledge and the systems that they have. You get these off-the-shelf robots that are imbued with the ability to just walk stairs, climb stairs, navigate new environments they've never seen, open doors, do dishes, do house cleaning work, all the things you'd expect from a future of robotics. So whether it's that, Roman, who's the founder of Robco in Germany, he's building automation for small, medium-sized enterprises. Often they're struggling to hire people. And so robotics is a fantastic solution for them. And I have two immediate recent examples in our portfolio where we partnered early with these founders. And I'm seeing the results. Cobot is another example. Brad Porter, who built most of the robotics at Amazon. We backed that company. They're seeing traction with their business too. Is it like self-driving cars where there's going to be some large number of years between, oh my goodness, the tech works, to, oh wow, we're riding Waymos around? Is it that kind of ramp? No, I think it's a little bit different. The challenge with autonomous vehicles is the risk of fatal crashes, which is part of why we don't see Waymos on freeways just yet. The challenge in robotics is a little different. Most of robotics until recently, obviously robotics is a massive, massive industry, but they're typically very large machines that are relatively dangerous, and they're cordoned off from humans. I don't know if you've visited a Tesla factory. Actually, when I bought my first Tesla, it was such a fun thing to go see the Tesla factory and you see all these robots. You're like, oh, this is so cool. But because of the risk of harm to humans, they had to be kept separate. I think the beauty of this age is when you get computer vision in these robots and they can interact with us in our normal environment, that unlocks so much more possibility. And if you can get them into a form factor that is not likely to cause human harm, I just think it unlocks possibilities. Look, at some level, we already have robots. I mean, you have Roombas running around cleaning floors and things like that. At a small scale, we already have some of the early incliners. I think we're going to stair-step our way into this. I mean, if it gets to something generalizable in the home, it should be truly enormous, I would think. You're right about that. But both Skilled and Robco, well, all three of them, Skilled, Robco, Cobot, all of them have revenue today actually serving customers in a variety of different use cases, whether it's hospital systems, whether it's airport security, whether it's small manufacturing firms that need robotic assistance to help them scale. All these companies have revenue. They're actually up and running today. And so I don't think this is a lab experiment where you see the demos, but who knows if it's real. It works. It just has to get everywhere. It works. And it's slowly progressing into more and more areas where the cost benefit of it makes sense. And a lot of it comes down to the cost of human labor, by the way. Yeah. And part of the reason Japan has roboticized so much more than other countries is because the cost of labor got very high. And so it made economic sense for them to do so. But both Skilled and Robco, well, all three of them, Skilled, Robco, Cobot, all of them have revenue today, actually serving customers in a variety of different use cases, whether it's hospital systems, whether it's airport security, whether it's small manufacturing firms that need robotic assistance to help them scale. All these companies have revenue. They're actually up and running today. And so I don't think this is a lab experiment where you see the demos, but who knows if it's real. It works. It just has to get everywhere. It works. And it's slowly progressing into more and more areas where the cost benefit of it makes sense. And a lot of it comes down to the cost of human labor, by the way. Yeah. And part of the reason Japan has roboticized so much more than other countries is because the cost of labor got very high. And so it made economic sense for them to do so. That should be the case here too, I suppose. And that's part of what's changing, is the cost of labor in America is changing, and that necessitates that. So if you think about the way the minimum wages have gone for the restaurant industry in California, for example, it's driving many more restaurant chains in California to try to figure out solutions for automation because otherwise the business models don't work anymore. Actually, one of the favorite books I read this summer is How Economics Explains the History of the World. And just because you invent something doesn't mean it'll be applied. It needs an economic driver. And I think that's precisely what's happening now with robotics too. We're back to costs. Exactly. Other than robotics, what are some of the other big societal-level things, whether it's education or government or healthcare? What are other big, big areas that you think, outside of business applications, AI is going to matter a lot in the next few years? Health care is one I'm enamored with, in particular genetics. We sequenced the human genome 25 years ago. It's actually the 25th celebration of that. That was an enormous research initiative to get to that point. By the way, the cost of sequencing the genome follows one of these phenomenal cost curves. If you've ever seen these plots, it's actually faster than Moore's law in how the cost of sequencing has dropped dramatically. And that's opened up so many more possibilities for us, whether it's prenatal screening, which is basically ubiquitous now in America, whether it's helping do oncology screening, organ transplant rejection screening, newborn analysis. One of my friends at Stanford in the genetics department has got this technology to be able to do a whole genome sequence he developed originally for the NICU unit at Stanford, where they can sequence a full human genome in hours because they have these babies where you need to understand their genetics very quickly to make clinical decisions that are life-saving. And so I think about all these possibilities around the field of genetics. They're incredibly powerful. Yeah. And that's a dividend we keep on benefiting from as that cost curve rides down. So I think genetics is super interesting. And then I think about AI, by the way, in the healthcare industry. We only have about a million physicians in America. It's stunning. It's not that many. And we have a population of roughly 350 million. We only have a million physicians. What can we do to make them more productive? Partly by giving them tools. So we partnered early with Daniel at Open Evidence. That company is now being used by something like 40% of American physicians. Which is great because, by the way, if you speak to physicians who have been doing it for a while or you speak to a new one, it seems clear that a newly minted doctor today is exhausted and overworked. And there's too many patients, and they're rushing through things, and they're documenting, and the patient experience changes. I don't know why it seems somehow different. I don't know if the ratios changed or something else changed, but it seems like if you talk to older doctors, something's different. Well, I think the burden, the ratio of patient to physician, has changed because I don't think we're educating as many physicians as we need to support the level of population we have. So the burden per physician has increased. So we need tools to help physicians make better decisions, but it also needs to alleviate their administrative burden. Yeah. So the U.S. healthcare industry is something, I think it's 16, 70% of U.S. GDP is spent on healthcare. Yeah. Roughly one in five dollars in healthcare is for the administration of healthcare. Yeah. So if you think about that, it's almost a trillion dollars a year. I mean, they're typing in Epic a lot. So we have a company called Freed that's helping with automating the workflow for a physician. So it's not helping with the clinical decision-making the way that Open Evidence is, but Freed is helping automate the workflows. How do you take a dictation of the patient visit and automatically understand how to write the record? What are the follow-up appointments that need to be scheduled? What are the referrals you need? What are the prescriptions to follow up? And if you can save a physician an hour or two a day of admin work, that is pretty dramatic in their day-to-day lives. Yeah. Are there any other areas in particular outside of robotics, healthcare? Are there any other societally important things that have you particularly animated? This may sound cold, but I think money. What's cold about that? Yeah. Sometimes people feel that money is dirty or whatever. But I think, at the end of the day, commerce and capitalism has been the— The financial system. The financial system is the biggest boon to human welfare. And when we improve the efficiency of the commerce system, I think it has massive societal benefits. And one of those, in my mind, is stable coins and the way that we are using stable coins to rewire our financial infrastructure. A lot of our systems were built decades ago with what was the best technology at the time, but it's dated. And it's slow. And so we have this ability to use stable coins to reinvent that. And Stripe is one of the companies at the forefront of this with some of what they've developed. They also bought a company called Bridge that we were an early investor in. And Bridge is being used for international transfers. I mentioned this company, Aspora, that's helping at a consumer level with international transfers. And so I think there's a really interesting possibility for us to completely change our financial system and make it far more efficient, ultimately for businesses and for consumers. As a final topic, what I wanted to ask you about was your team and how you think about building the team, the composition, who you want to attract, what values and activities you want to promote or dissuade. My first guest on Uncapped was Sean McGuire, who I love. I think he's awesome. And he's also a polarizing figure. He says a lot of stuff online that I'm like, go Sean. And a lot of people are frustrated. And on first glance, it could seem like Sequoia is this extremely professional, not overly bombastic place. But Sean says what he feels online and is an important member of the firm and is a great investor. And so I guess my first question is, is that a tolerated edge of the center of Sequoia's culture? Or is that, in some ways, the heart of it? I think it's the heart of it. And it actually goes back all the way to Don and his own streak of irreverence. Don told me a story when I joined where he was going for an interview at IBM, and he was sitting in the waiting room for this interview, and he noticed how everybody was wearing the same navy blazer. And he just thought, wow, this is a place of conformity. And he realized then that he needed to move to the West Coast and break out on his own. And I think that streak of irreverence actually runs very deep in Sequoia. It's part of why Don backed the founders he did. When he backed Steve Jobs, who at the time, I believe, wasn't wearing shoes and maybe had just come back from a trip to India and was not bathing regularly, maybe didn't smell that great, Don saw through that. And so we've always had this view at Sequoia that we're backing the underdogs, the unknown, the defiant. Our founder prototype is somebody who's a little unusual. These people want to change the world, the audacity to change the world and the will to do it. And we need to reflect that on our own team. And so we have always had a band of quirky people. My partner, Doug, always said, we want to recruit people to Sequoia who want to be pirates, not people who want to join the Navy. And to me, Sean is like that. We want to have a band of pirates at Sequoia. And that's how we think about our team composition. Yeah. It's really hard to build something outlier as a founder. Don saw through that. And so we've always had this view at Sequoia that we're backing the underdogs, the unknown, the defiant. Our founder prototype is somebody who's a little unusual. These people want to change the world, the audacity to change the world and the will to do it. And we need to reflect that on our own team. And so we have always had a band of quirky people. My partner, Doug, always said, "We want to recruit people to Sequoia who want to be pirates, not people who want to join the Navy." And to me, Sean is like that. We want to have a band of pirates at Sequoia. And that's how we think about our team composition. Yeah. It's really hard to build something outlier as a founder. I think, if you yourself don't have some outlier or out-of-distribution traits. Absolutely. And by the way, an outlier is not one standard deviation from the mean. Alfred, my partner, talks about this. An outlier is not one, not two, not three, probably four standard deviations from the mean. And these people are crazy enough to start companies, and they actually want to change the world. Yeah. The other thing I'd say about Sean, and as it applies to the rest of us, and this is my partner, Pat Grady, who I think you've spoken to, he had a great way of saying this, which is we look for people who are hyper-competitive, but who have a heart of gold. Because you can sometimes get people who are hyper-competitive, who are individual contributors, don't want to play team sport, and don't necessarily look out for the others. And we need people who have a heart of gold at Sequoia. People who bleed Sequoia green, people who go do whatever is necessary to help a founder succeed, but also look after the well-being of their own teammates. And anybody who's actually had an opportunity to speak to Sean would know that too. Somebody recently used a term to describe a similar type of person as a killer teddy bear. And it really stuck with me that it's somebody who is ultimately very high integrity, high caring, but they're playing to win and they're going to do what it takes. I love that. Yeah. All right. This was extremely fun, and it means a lot that you took time for this. So I want to say thanks and really enjoyed it.