Uncapped #5 | Sarah Guo from Conviction

[00:00] All right, so thank you for doing this. I realize that you've been spending more time being the interviewer than being interviewed, but I appreciate you doing this, and it's nice to be doing it in your office. Yeah, well, any excuse to hang out with you. The first thing I want to ask you about is... [00:14] What a VC firm sort of is, and the thing I've been thinking about a bunch is when I was doing Lattice, I kind of knew what a company was. Like, you know, you build a product, you sell the product, like there's like a sort of a very tractable thing. And then over the last year doing venture full time for the first time, it struck me that it's a much more sort of flimsy concept. And, you know, there's obviously like a fund that owns shares, but then there's like a brand and there's people and there's these other parts. [00:44] firm, I've realized that thinking about what it actually is, is sort of like a wispy idea. And so I would love just to hear from you as you've thought about, you know, you've been at one, like a very substantial longstanding one. Now you've built your own. How do you think about what a VC firm is? We've taken like two steps toward building them. [01:01] Our own. You've done something now. It's an interesting question. Like, the way we think of it internally is, and it's not like the most flattering view, but venture is a bundle of, like, money. [01:13] I'm thinking about the value props to the customer, right? The customer is the founder. Venture is a bundle of money where everybody's money is approximately green. You can have... [01:23] lower and higher quality LPs. But generally, that's not how founders make their decision. Then you have people and beliefs that you are associated with as a founder. And you have advantage, right? Like what can like can and I wouldn't do the job if I didn't figure like that the last piece could matter at all. But help you can give companies and to try to make things happen

[01:53] It is a weird way to think about venture as a business because for all tech investors love to talk about like... [02:04] you know, like moats and differentiation and sustainable advantage. I think venture, like when you think about those components, money is the purest commodity. The rest of it is pretty squishy. It's a lot of like brand and individuals that feel quite fragile or like point in time. But, you know, some of the great venture brands, they've been around for a long time, and I think it's still possible to build an enduring institution. So yeah, and some of those brands do something is compounding there. And I guess maybe that's an interesting sort of way to [02:34] is like, what is compounding at Sequoia or Greylock or these firms that have been successful for decades? What's the thing that is like hard to knock off balance? Like, what is the thing that's making that? Is it like, is it the brand? Is it the network of founders? Is it something about the way they do business? Like, what is [02:50] holding on through different years and teams and cycles? Yeah, well, I would start by saying, like, I actually, like, really want to try to live like this. And I think your entire, like, personality expresses this. But, like, this is this doesn't feel like a particularly zero sum environment. I think you can try to build a new thing without saying we're, like, completely after somebody else. But in terms of... [03:13] durability, I think you have, you have like, ethos, right, like what matters to the people in a firm that goes from generation to generation and different investors, you have like,

[03:24] Some tribal knowledge, and that is like what has worked in the last generation of companies, good and bad biases, like the history of technology, right? Like here's how Apple or Google or NVIDIA did it or Facebook for that matter. And then you have brand, which is just… [03:41] Obviously, what founders think of the place, how much success you've had, and then you have network. And I think those are the things that are harder to replicate very quickly. [03:50] I think you also, if you are at a platform venture firm, you have... [03:55] a lot of capital available to you, but there's also just a lot of capital in the world. Yeah, I want to come back to what you said about sort of the whole, the zero sum nature of it all, and to what degree, you know, more value gets created versus, you know, venture capitalists are sort of capturing. I've heard some people say that they think that great founders select, like the truly great founders are actually selecting for a different thing than other founders when they're thinking about their VCs. I can't remember, but you know, who all, but one [04:25] And I know Parker at Rippling has said that brand is the most valuable thing that he can get from a VC because it makes it easier to recruit, easier to get the next customer, et cetera. And he's going to do the work anyway. And so at least he can get a brand. Do you think, like, as you've experienced sort of working with companies and founders, do you think that there are different attributes in a venture firm that – [04:46] different founders prioritize? And does that impact at all the way that you want to build when you think about, you know, your founder is the customer, but obviously you're looking for certain founders within that customer set? Yeah, I definitely think we don't have to appeal to everybody. I think if you are going to go to the sort of highest prestige brand, like you want to buy, I don't know, I was going to say like IBM, but I mean this in a flattering way, like you want to buy like the known winner. Apple, let's say. Apple, you buy Sequoia.

[05:16] Right. And like, you know, number of challengers in the established tier. But I think a lot of people make decision like founders are not a homogenous set. And like, let's just use examples here. I I think a lot of people make decisions very individually. [05:33] Right. If they know and, you know, we also want to back first time founders who've never met anybody in venture capital. And we talk about how you build that trust in those relationships. But Brett Taylor, when he builds his third company, he's just calling people he wants to work with. He does not care where they work. Right. I don't think he like particularly cares about the brand of the firms he is working with. [05:54] firms with nice brands, but he is picking individuals, probably because he doesn't feel like he particularly needs the halo. I think if you asked Parker, because I have like, I'm pretty sure he called Mike, actually, because there's, and maybe that's also a symptom of him. [06:12] also being a third-time founder. I think his specific words here were like, if I were ever to build another company, please God, no. But I'd call Mike. And so I think that at some point, like... [06:23] the things that you feel like you want help with or support on or how much brand matters to you change as a founder. And so we want to attract a particular set and, um, [06:34] I think our brand does stand for something. We're trying to make it stand for something now, but it doesn't have to be like the most established prestige. We actually can't go back in time and do that. Totally. Yeah. OK, I want to shift to brand now because this is it's something that you've done notably. I promised you I wouldn't like ask like a question that was like rooted in flattery. But like you've done a good job branding the firm since you started. And actually, in some ways, I would say, you know.

[06:58] The last sort of really great firm to be built that I can think of that also launched with what I think was a very strong market presence was Andreessen. This is like 2009. And all of a sudden, they were everywhere. They did many things. They dominated PR. They were investing in great companies at what seemed like then high prices. Many of them were actually still really good investments from there. But there were a bunch of things that... [07:23] let them get this brand very quickly and then they scaled rapidly and all of that. And in some ways there was not, you know, another firm that did that. And, um, [07:33] you know, at risk of flattery. You've built a very good brand very quickly and you've kind of been everywhere. And, you know, people see you in a lot of different places, in high value places. They see you on good cap tables. They see you at good events. They see you, you know, they see you just like frequently. To what degree is that just... [07:53] what's been happening by a product of you sort of working hard, having a clear view and whatever? Or to what extent is this all like wrapped around like an intentional goal, building the firm in a particular direction? And this is part of the story. There's a founder that we both know who has like a medium scale SaaS business now. And at one point I lost an investment at my prior firm to Andreessen. It was a series A for this

[08:23] hey, love you guys, but they're paying more. And like, let me like I was and I was like, oh, well, like, it doesn't matter. Like, all of these other things matter. And we sat down in a bar for like three hours. And he explained to me like how he saw the firms. And he basically was like, ah, these things like they look pretty similar, except they look bigger. And like, they're better at PR and they're willing to pay more. I was like, oh, [08:47] That's bummer. I took this like list that this founder gave me and like stuck it on my office door at my old firm for a number of years. I was salty about it or wanted to like try to solve the problem. And so one of the things that – [09:02] I think you just like think about when you start a company and you've worked on early stage companies for a while is like the who said this originally? Like it's such an important thing. Like nobody cares. Nobody cares about Jack or Sarah or their firms or your new startup. And so like you have to make them care. The other thing that like we thought about was just if we think we are good partners to entrepreneurs, like why should they know? We have to like solve an information asymmetry for them. [09:32] network. They say they have a good network. It all kind of looks the same. One of the things that we wanted to do, and all this stuff is like [09:41] I don't know, you can just, like, try stuff and then see what sticks. But one of the things that we thought would be important would just be, like... [09:49] demonstrate who the network is to founders in ways that are like accessible to them without us like

[09:56] man-to-man combat, women-to-women combat, proving it. And so we did a bunch of what you would consider to be, if you were in a company, it was like a SaaS founder of Flatus, it would be like partner marketing. Yep. Right? That's the way to look at it. And we just were like, if it was a business, it would be a funnel as part of marketing, [10:13] Like, what can we do that is efficient with people who have the distribution that we want that like us? Like, OK, let's go show up with NVIDIA and Snowflake and OpenAI and whoever it is that, like, is part of the community we want to be part of. I mean, there's also the reality that you doing, you know, to the Parker point, there's reality that you doing the brand work actually also helps the companies in some ways. Because, you know, you doing branding can help your, at least at the early stage companies, you know. [10:38] there's value to it. So you talked about, so there's like partner marketing, which is sort of like express the network. You also touched on something before, which was like stand for something and then like... [10:49] you know, share that. Is that like an essential part of it? Is that like the center of it is like believing something in the world and like that's what the marketing is about? Like, you know, I know like this was like a Steve Jobs thing was like, you know, we stand for something at Apple and like that's what Apple's brand is, is we believe in like, you know, we believe in, you know, X, Y, Z things. And I think like that that could be the analogy here. But do you think of it, you know, is it network? Is it stand for something? Is there a third pillar? Like what are those components [11:16] in your mind? Be willing to take risk on having an opinion is an important one for us. I think a number of founders choose to work with us because they believe that like the firm is AI native, in terms of understanding of these sets of things that the companies are trying to do. And then some is access to network, some is understanding of the research itself, some is just like company specific challenges. I also think that maybe entrepreneurs, like if you if you like want to go one level deeper, like I don't know what the world looks like.

[11:45] 20 years from now. I really hope that like conviction as an institution exists. But if you think about the like, name we chose as well, it is like have an opinion. Maybe we're in the like, post AI abundance age and like venture is irrelevant, or like we're on to the next big technology thing. And I hope the people at the firm and maybe it's me like are on to the next big thing too. But part of the core ethos for me about venture is like, I want to understand the new thing. And then like, we take risk on it. Like we published our LP letters recently, [12:15] are going to be wrong and look very stupid. But for a generation of entrepreneurs right now, where everybody is in like a very dynamic environment, and I think people are unwilling to like make any claim about understanding or about prediction. I think that can be like, very grounding of like, well, at least these people have an opinion and I can decide if I have that worldview or not. It's actually funny that you said that because about conviction, before you said that I had been thinking of the brand as AI native. And now that you say it, [12:45] can immediately re-understand it as have strong conviction and what we believe right now is that AI is the most important thing. But if in 10 years that's solved and now it's about space travel, like conviction could be about space travel. And that actually, that resonates. That makes sense to me. Do you think that in venture, do you think like brands typically like need to follow returns or do you think like, do returns follow brands to some extent? Is there a strong correlation

[13:15] should be success. [13:17] and success for a venture firm is like the quality of the companies. I actually don't even think that most founders have that much visibility into like how early you are or like how much of the companies you own. And like, you know, sometimes I think like that would be useful to founders to have more transparency around where like, hey, that person that like says they're part of this journey bought secondary at the Series F, right? And that's like probably different than working on the company [13:47] It also brings up a funny point, though, of like from a founder's perspective, what is a VC being successful? What does that mean in the way they care about? Like, should they care more that you partnered with important companies or that you deployed a lot of capital or should they care more that you got more? [14:01] crazy returns in companies they never heard of and didn't play a role. You know, so it's not so obvious. And I do think that there are firms that – [14:10] that can kind of impact the direction of capital flows and firms that are working with the most important companies that maybe don't have as good of returns as firms that are, you know, not doing those things. And so I do also, you know, as we're talking about it, wonder from a founder's perspective, which type of success is actually important. You know, like I joked recently that somebody was talking about like, oh, why should a founder want to work with that company? You know, they're. [14:31] They're huge and they don't have big returns. And I was like, if I'm thinking from a founder lens, I'm like low returns for them means that that was less dilution for me. So that sounds great. You know, but it's not so one to one. Yeah, I think it's a really good point of like, I actually don't know that founders should care about the quality of our multiple. I think that is a...

[14:51] proxy for like something founders should care about, which is taste. I do think some of the very best people, if you ask them, like when they say they're choosing brand, they care about association with quality and with success. And so if you do a small number of companies, and the hit rate is high quality, and the people is like a set of people that you would be excited to learn from, I think that is attractive, right? And there is often a debate about whether or not [15:21] I believe you learn more from success. Me too. And so I think people want the tribal knowledge going back to like, what is a venture firm really of like, what is working at like the handful of companies that are really working in this era. And so I think we want that to be, you know, part of the value prop and then part of the brand. But to your point of like, should founders choose the same thing? And are the incentives fully aligned? Like, I think not, like, there's a [15:51] Thank you. [15:52] my my partner mike of like i think it is very reasonable if you think about if a bunch of vcs can argue about whether or not they're good operators but once they've been investors like how might they raise money they might raise money from their friends they might raise money from the like highest taste people and people who are most helpful they might also raise money [16:15] the most money they can at the least dilution possible with the least control. It really depends on what you are looking for and what your view is of what you're looking for from your investors. And so I also think it's very stage specific. I don't think that founders should be optimizing for our returns. I think that's just a proxy for other things they might care about. As you think about the current market landscape –

[16:38] And you'd been investing, you know, professionally through the teens. And then you got to, you know, live through sort of the Zerp escalation and like sort of late stage SaaS before AI and all this other stuff. You know, for me on the founder side, what I remember in 2020 and 2021 was all these crossover firms that I just thought were hedge funds, like came and were, you know, doing all these crazy market distortions. And as founders, everyone was like, cool. But like that was what was going on. [17:06] Then everybody kind of went away in 2020, 23. And now back in 2025, [17:12] We have what looks to me like a similar but different dynamic where the crossovers don't seem to be back in a particularly important way, at least. But you have the sort of bulge bracket institutional VCs that have gotten as big as those firms were, and they're now – [17:27] you know, three or five or nine billion or whatever. So you've got that going on. [17:33] The other thing that I think is interesting that you sort of were the – [17:36] maybe the first of the most recent set. But there's like a number of now successful GPs that have left multi-stages, started their own firms. So now there's going to be, you know, there's always new firms being started, but there's now a set that is sort of like ostensibly going to be slightly more institutional over time, probably smaller. But so you have these two sort of ends of the market going on at once. And you have, you know, AI mania, which, you know, as we know from, you know, Internet in 1999 could be could be undervalued still, could be overvalued. [18:06] Who knows? What do you make of the shape of this market right now? What does it mean for where you think people can successfully play? What do you think it means for...

[18:15] you and where you want to be playing, but like what's your read on that sort of market map? [18:19] Starting a venture firm has like made me slightly more cynical about the industry. I'm like a I'm the opposite of a cynic in many ways. Like I'm a deep optimist about technology. I like working with really earnest people. I try to be an earnest person. But it is clear to me that raising money like the capital markets are really deep and raising money is an awful lot easier than making money in venture from a returns perspective. [18:49] You could be a like a best in class in any vintage investor delivering, let's say, like respectable eight to 10 times multiple returns on a small fund and make not that much more money than somebody who's. [19:04] charging traditional venture fees on a fund 10 plus times your size, right? And so I think there's a very... [19:12] rational thing happening in the maturation of certain asset platforms where it is very easy to unconsciously be attracted by growth because like, you know, you've built a company, growth feels good. Growth is not a natural state for venture. It changes your ability to invest, right? But people are growing their firms. It's also like a natural organizational thing, right? If you hire young people and then you hire 10, then you hire 50, like they got to go somewhere. And that also [19:42] So I think that like a very natural thing has been happening where these things that were little boutiques are acting like businesses. The businesses naturally grow. There's capital available to them. The feedback loop for whether or not the returns are supported are also distorted by the fact that lots of people see the backwards looking multiple of the best returns.

[20:06] funds of the vintage right before ZERP and were under invested in venture and are still trying to invest in venture. And I like the outcome of all that to me is there's obviously secular increasing opportunity in technology over time. And that's the thing we index as venture investors. And so I think there should be some growth in the venture industry. I think this is like [20:31] askew from that, and it will not end well in terms of returns for folks, but it is also just maturation of the asset class. And so I don't know that it goes away. If you like take private equity as an asset class prior or those platform businesses, it's [20:47] it is very hard to make money in private equity at a top tier, and people are not moving money out of the asset class. Does it skew the ability for small firms to make money? Because these big firms, though, are going to play at every stage. And in the same way that, you know, the crossovers got into late stage growth. [21:03] and sort of we're sort of starting to do non-economic things, do the venture firms that are huge, [21:09] Or is it rational for them to do non-economic things at seed because, you know, it fills a pipeline and they know that they're just going to play, you know, a different game on the winners and take sort of like a different IRR on that? And does that impact – [21:23] the whole ecosystem in a big way? I think it does. And so I think the, like, if you ask me what is a... [21:32] What could I be wrong about in the whole like strategy or the ecosystem? I think there is a point of view of like you can't make money in venture anymore in terms of like an outsized multiple in early because it's just a sourcing funnel for the late stage firms and the expectation on multiple like the cost of that capital is lower. Right. And so like and you can just spend resource on it subsidized by the growth firm to, you know, be more competitive early. Yeah. Right. In terms of like marketing or people or whatever.

[22:02] We have eight people today. We had six people until last week. I guess like I would... [22:09] Posit until proven wrong. Like, I think in the very early end of the ecosystem, like we're not trying to win everyone. We only have $200 million of capital. We need to have a couple really important companies to have a hugely winning fund in every cycle. You don't need that many people and that much money to compete. [22:26] because people are looking for different things. And the experience of working with you or working with a like a very specific set of people that have an opinion and have some certain understanding is quite different than working with a very large business. And like people choose different things. It's not a homogenous market. We talked about how like there's a certain type of round that is [22:47] obvious. [22:48] Sometimes obvious is Brett Taylor at the seed, but maybe more often it is... [22:53] cursor or something like that, like a company that is really working, you know, and there's things in the middle, there's things past that. But as you think about [23:01] you know, these types of companies ranging from completely not obvious to extremely obvious. And you think about where money is most likely to be made. [23:10] and sort of line that up against sort of the dynamic we've talked about with, you know, [23:15] Sequoia is able to sort of play a dominant strategy of like they can go win the obvious things at a very high rate. You know, small firms can do different things. Where do you think that that leaves if you had to take your best guess of where in today's market like the most alpha exists? Like is the winning strategy right now to become able to win things? Because to this recent point, there's the non-obvious feeders, but there's more money to be made at doing the first obvious round.

[23:45] Actually, the $10 billion companies look much better than the $1 billion companies because we now live in this [23:51] world where things are gonna get way better way bigger than we thought like what is your current read on where the pockets of value are at the moment i think it's a super interesting question but i don't spend that much energy on this question in particular because my orientation is actually just like my taste is my taste i'm an early stage person i actually think my taste is like [24:10] pretty mainstream in many ways, right? Let me contrast this. I think we will have opinions on companies that other people will not have. And like, we might choose to be earlier or make like, [24:23] It's very hard for me actually to tell what the mainstream opinion is, right? Because if you remember like the GPT rapper era, we're like nobody's going to make any money in the application layer. And now it's like you're only going to make money in the application layer. It was crazy how that flipped very quickly. Yes, exactly. And so one thing is like I don't think a lot about – I think great investors do this, but not me. I don't think a lot about like where is the risk reward and like – a lot does. Like where's the risk reward and like the mid-stage, the late-stage, the early-stage? Because I'm like, well, I'm only going to – It doesn't matter. I've got my game. It doesn't matter what – [24:53] I am pretty sure being great at early is going to be valuable no matter what. And I'm not like globally optimizing. Right. And then I think your point of view on product and people just is what it is. I do think that. [25:05] If you think that, like... [25:09] you let's say you just want to be able to win whatever opportunities you want to work on then you have to like build your ability to win right so brand does matter to us i think there are like seemingly very effective alternative strategies right you can own a particular community he'll hate that i even mentioned his name out loud but i think zach frankel is very good at this right he works with a really small number of people he has good taste he does well if you look at um

[25:37] uh what matrix has done the quality of some of ilia's companies i think what he would say is like hey we're not here to compete in the like general brand game and put a lot of energy toward that it's not really me but but he's an incredible company but he's an incredible companies and like them when times when they weren't obvious absolutely there is definitely um alpha in choosing to be controversial on purpose and like i look at some of his investments and i'm like [26:07] any sense to me when I met it. Shame on me, right? I was just like, this is a weird company, like cool guy, weird company. I think like your taste is what your taste is. And we are probably more willing to take technical bets and then like go further down the thread of like, hey, we have a thesis around this particular area, or we'll take a bigger slug at something than I think one of the benefits of being an early firm versus a multi-stage firm is a specific [26:37] firms out there is we want to invest in the most important companies of tomorrow as early as it's clear that they're the emerging market leader. But that means you could do it like as soon as the company's in traction or like right before the IPO. Right. And because we do not have this luxury of like waiting until the market has panned out and all the co-gen companies have like fought it out and one is waiting at the end. And then like maybe you're paying a 20 billion dollar entry price for it. We're like, [27:06] We're just going to choose, right? Like whenever we think that company like enterprise value can be built, we'll just take the competitive landscape as it is.

[27:16] On this point about, you know, you're now running your own thing, you can sort of, you have the luxury of now investing the way you want to invest. Figuring it out. Figuring it out as you go. But you have, you know, you get to adapt as you see what you like and what you think. I want to sort of reflect on your learnings because we're investing very differently at Greylock, but they're objectively very successful. And you worked closely with Ashim, who, you know, as you know, is like really different style of investing where, you know, I saw that, you know, people talking about this recently. [27:46] that he's like never lost money. And this is like also not, you know, somebody who's just making safe investments. He obviously, he did great stuff with that, but it's a certain type of investing. To me, in some ways... [27:55] What you're doing now is, you know, maybe opposite is too dramatic, but it's like quite different where, you know, if that was very linear, very clear, you know, he could probably name, you know, several companies that ought to exist that don't yet and how they might be built. You know, you're operating in sort of like the most shifting of. [28:11] you know, sort of sands underneath, you know, the space. And so I'm curious what you... [28:16] you know, what your reflections are going from that experience to now investing this way. And maybe more interestingly, is there anything that despite how different they are, is there anything from that that you take to this that somebody who hadn't had that Greylock experience wouldn't have? So much. I mean, like I learned to be an investor at Greylock. It is one of the great firms. I think Ashim is one of the greats. And I think having his hit rate on both [28:42] Uh, [28:43] Capital, both, all three of capital preservation, doubles and home runs is like, you know, that's that's fucking amazing. Right. What to learn. So I think that there are.

[28:55] patterns of investing that are like we were talking before and you're like, Oh, it's nonlinear. I appreciate like, um, [29:03] Linear thinking is clear thinking, right? You can follow somebody's pattern of logic. And one thing that is like characteristic of traditional firms that I think is a very valuable thing is your understanding of how an existing market works. [29:33] also amazing at this, right? And then you or a Neil Bustery when he was a full time investor, right? And it would be like objectively true that for any point in time, like the dominant markets are incumbent existing markets, right? And you just look at what the transitions are that are macro technology. And then you make the bets on the right types of people, given that transition in an existing market. [29:58] Three checkboxes, right? I don't know, like you have amazing intuition for investing, but like maybe you didn't practice this specific thing, seeing the pattern of it. And like, let's take an example. If you look at the biggest categories of security spend, you might say, you know, category one of security spend, given the shift to the cloud, how does that change?

[30:28] How does that change? And then who are the right people? A lot of those people work at the incumbents. I'm not saying that's like all of traditional venture investing, but I think there are patterns that are really powerful where the existing network and the understanding of how those businesses work. Like, I'm really glad I got to learn a little bit of it. [30:45] and try to replicate some of it. I do not think that what I'm doing now is the opposite of that. But again, like venture is like such a, you know, companies are an expression of the like personalities and taste of founders to a great deal. Right. And I think venture investing is also a very personal thing. Even when you're responding to customer needs, it's still like, you know, Lattice, it was a lot about you. I'm curious about like the things that change most. [31:15] that there was actually an opportunity for a new firm was that the like AI technology shift would actually mean that the the most interesting markets were not necessarily the markets that already existed right like let's say enterprise infrastructure and like systems of record in the high end and I love businesses that look like that by the way but some of the categories of [31:45] markets that are not traditionally software markets. Harvey in law, Sierra in like actual support work. Heijan is doing the work that, you know, SMB video agencies did before or just creating new capability. And so I think if you were very traditionally market focused, you might not, you might not have the time to see them in the same way. Are you trying to bring some of the like equivalent

[32:10] market linearity to being like, you know, you're spending all this time understanding AI right now. And you obviously, you know, have spent a lot of time just knowing regular business markets. Are you trying to intersect those and think like, you know, I'm still going to be founder driven, but this year I want to find a company doing, you know, something in XYZ vertical, because like, my triangulation is that this should be happening soon. Do you do it that way? Or are you like, following founders and networks? We do both. There are a lot of people out there with really [32:40] I think it's like better use of your energy as a venture capitalist in general to listen to founders ideas. Right. And just like be educated by them and be able to make the right decision when you see it. That being said, like you spend a lot of time. [32:53] Talking to customers. [32:55] out in the world, like let's say you've ever been an idea person. We made one investment in the portfolio we haven't announced yet where I like carry around my little like set of PowerPoint slides trying to convince founders to start this company for five years. No luck. Well, we finally found the company. Oh, you got it. Yeah. Well, I mean, we didn't convince them to start the company. They'd already started the company. But I was just like so thrilled to meet them because I was like, no, no, no. Stop your presentation. Let me show you like we believe.

[33:25] my partners now, um, [33:26] Arnpunov and Mike are working on, like, a space where they want something to exist. And so I think we'll actively go hunt for people that we think would be great to start the company and just be like, you should start this company. A couple times that I've done that and put, like, a little beacon out in the world, I have found that, like, it does – [33:44] help attract the people who are some it somehow makes the people who are working on that more likely to come find you and you're a much more prepared mind when you do it so i do always think i should be spending more time doing this yeah than i have been yeah i kind of want to shift into like ai and not necessarily like prognosticating about the future but like you do spend your time here yeah and so like i want to try a little bit or at least to sort of like hear your thoughts on some recent commentary um there was like a good quip recently that i was like that's a good reframe from satia talking about like you know [34:12] this AGI stuff is great and we should think of it that way, but like I'm looking for like 7% to 10% growth. And I think it was like a soundbite that sort of – [34:21] It did something where I think it sort of changed the conversation from like inputs to outputs, basically. I think it in many ways gets at the overall question of just like, is this market big enough for... [34:33] you know, the tens of billions or hundreds of billions that's going in from the hyperscalers and the hundred billion that's going to go in from VC and all this other stuff. When you think about [34:41] where we are and how you'll see the signs that like the outputs are following the inputs [34:45] is [34:46] Are you thinking of this in a, I just want to see, you know, more companies that are at, you know, 50 or 100 million that are durable? Are you thinking about something on the intelligence layer? Like, what is your frame of mind when you see like that kind of thing from Satya?

[35:00] I think it's a super rationalist decision from Satya. I think it is a statement that he does not believe that owning... [35:09] a particular lab or research effort that gets to a reinforcing fast takeoff is going to lead to a lot of economic value capture or that the probability of that is not worth spending many billions of dollars. Right. But I think the view of like, hey, we're just not seeing enough revenue adoption to make. [35:29] $50 billion of capital outlay. [35:32] into model training every year worth it for us. [35:35] is completely reasonable, right? He's responsible for public shareholders. I have like a much easier problem, which is, I am quite sure that there's going to be enough economic value created to return a best in class venture multiple on $[redacted address] like we try to stay rational about it, but like, I think if you wanted to back up the truck and just like only do CapEx heavy foundation model companies, you just raise a lot more money than we had. Yeah, right. Yeah. [36:05] are like actually going back to like, what do you take from traditional venture? I'm like, I really like to see revenue. One of my companies talks about EBITDA. Do you remember EBITDA? It's a very, very exciting concept. There are companies that are creating user value very rapidly. [36:19] And they're not spending a lot of money to do it. I think you can be more or less expensive as a business to grow. Like, we'll do both. But it's really nice to see that some companies are being capital efficient about it. One of the things that I think is interesting is that we're back in sort of like [redacted address] where like 100 times ARR rounds are like not particularly uncommon on –

[36:39] things that are growing fast and have a reasonable base and whatever. I remember in 2021 sort of the [36:46] kind of the general wisdom was, hey, we all thought there were going to be X number of $10 billion companies, but there's actually going to be like 10 X and there's going to be a lot more $50 billion companies than we thought and so on. And, you know, we all just had like the wrong idea of the size. And it now looks like that was like mostly wrong. And I think in some ways now we're back to thinking that in a way that, you know, [37:09] at the moment at least, feels more right. That could change in the future. But are we still investing, even at the early stages, in ways that require there to be a lot more... [37:20] $10 billion companies than they're like ever have been and like are we collectively betting on that? I [37:24] Absolutely. And I think it will be true. Like, let's talk about the mechanism of action when like people thought this in the SaaS world. Right. I think some of it ended up being and like speak up because you actually did this. But, you know, as a as an analyst. Right. Some of it ended up being true. More people than ever before, like in businesses got to consume software for their business because it became cheaper to use. [37:50] because you didn't have to like buy package software and operate it because it filled many more functional gaps. [37:57] Right. And because you had Internet distribution. Right. You got cheaper to actually sell to people. That's how I think about like, oh, why should we have more SaaS? The thing that I think the. [38:08] market got wrong or one of the things that the market got wrong was there was not infinite ability to like

[38:15] cross sell more software into the organization in any segment of business, right? SMBs, mid-market, enterprise businesses. Enough became enough at some point. Enough became enough. And like, they didn't want to buy any more darn SaaS. And there was just a limit to budget and a limit to consumption. And like, that seems rational, right? Like, in contrast, well, that seems rational in [38:45] Like – [38:46] I have... [38:47] poo-pooed on... [38:49] uh generally working in health care tech for a long time not because of lack of interest but i'm just like okay i was one of like three people at graylock that 10 years ago got staffed to like go look at digital health and we looked and it was like not um it was not a good look right you want to sell to providers it was slow there wasn't much budget it just didn't change the business that much uh i now sit on the board of the health [39:12] care provider tech company and it's growing really fast and there are other companies that are growing really fast. I think that people just didn't build stuff that was worth buying enough for a long time. [39:22] like, you know, it's sort of mimetic now to say, like, AI eats services. But I think if you just, like, think about the value provided, it is not necessarily like service industries. It's just technology does way more. And your willingness to pay as a consumer or as a business for these things that you like, again, like,

[39:47] could have been part of your job being done for review. Could it could be a legal service could be like the skill to be creative could be education. Like those were different budgets. Yeah. What's interesting on that point is, [39:59] that makes sense to me now why it's happening, but I'm not sure how it plays out over time, is at the moment, agents, for lack of a different term, are... [40:08] replacing [40:09] labor. [40:11] instead of software. [40:12] And so the price that they're comping to is labor. And so instead of SaaS being, you know, [40:18] $100 a year, $200 a year, [40:21] for a person's seat. An agent can be like, oh, it's 40K a year, which is way cheaper than your software engineers, or way cheaper than your accountants. You know, you're gonna save 70K a year by just giving us 40 per agency. And so there's this, [40:36] very strong pricing potential that you can have. [40:39] And that seems, at least at the moment, like it could be rooted in sort of the existing paradigm. But in the future, does that pricing stay, you know, as competition goes? And so that's one thing that I, you know, and I say this as somebody who's like investing in companies building agents. But I'm like wondering how that ought to play out because it's like – [40:59] will this eventually price relative to like cost and like an appropriate margin or like it happened to used to be done by people. So we're just going to kind of peg to that forever. I think generally there'll be like absolutely very strong pricing pressure. And I really think about like, what is the [41:14] basis of competition some number of years from now. If you have a unique offering,

[41:21] where the customer really wants what you are offering. And I think, like, people don't really know what, like, the modes with agents will be. But, like, let's assume, you know, there is uniqueness in some set of offerings. Like... [41:33] If you're Figma, pricing pressure is different than if you are one of many different companies. People just want specifically Figma. There's nothing quite like it, right? And your team is in it, and you're going to use it. So I think there's some set of companies that will end up with that dynamic, and they will be able to retain some sort of value-based pricing. And then I think the majority of companies will probably gravitate toward cost of compute or energy or intelligence plus. [41:58] But I have no idea how long that takes. Right now, I basically think a lot of it ends up getting pushed to consumer and business surplus, as it should. Right now, it's just like, wow, the surplus is so good, even if it's a third the cost of an engineer. [42:14] for a while, you know, before AI, it felt like in order to come up with a good idea, you had to be like kind of non-consensus. At the moment, it feels like there's actually a lot of ideas that are both like consensus and right. And then the reality is then you get like, you know, a lot of competition and it floods and they kind of all do really well. So I'm curious to see how it like plays out. But there's a few of these spaces like... [42:36] support and code gen and these things where it's like, [42:38] Everybody knows it's working. They are indeed all working. And, you know, that's just not even close to the market demand yet. The last topic I was curious to get your thoughts on are that there was a good post about like agency being more important than intelligence now. And, you know, intelligence is like abundant. And so like the thing that's left is agency, which I want to believe this, you know, like that's the model of the world and that like, you know.

[43:02] They'll never come for our agency. And we'll see if that's true or not. But I'm curious, as you've talked, you spend a lot more time than you probably ever have with AI researchers and you think about this. So I guess, do you see it that way? Do you agree that agency is going to be the sort of last human resource? And then do you think if so, does that... [43:22] impact the way that you think at all about like the types of founders you're investing in? I'd focus on being hot and funny. That's good. That's high age. No, I don't know if I have like a... [43:34] A good answer to this question. [43:39] I guess I never particularly thought like intelligence was enough anyway. [43:43] Right. And so like when like it sure helps. Right. When you think about the founders that we select, like without being over a particular bar, like all other things considered, like, of course, we're going to back people who are really, really smart. I was on this campaign at my prior firm to work. [44:03] try to be more specific about how we talk about people. Because it's like very easy to say like, oh, the person is special. What does that mean? The thing that we most look for is, besides intelligence, is kind of like force of will, and the ability to like take a point of view and be right. Right. And so I don't know what that is. Like maybe, like maybe AI is going to have better predictive power than founders about the future of different markets at some point, like the ability

[44:33] what you want. Like, I don't know if that's how you think of agency, but like that doesn't feel in grasp of pure intelligence anytime soon. You have three kids like me. Are you thinking about their education at all or what you think they need to learn or what you're teaching them, you know, in this? [44:49] as you're thinking about what's obviously happening now? Of course I am. I don't know if I'm more like doing anything about it. Our kids are young enough that I think about like behaviors as much as like knowledge for now, right? Like frustration management, the ability to concentrate, right? The ability to upskill yourself in something. I think those are going to be important no matter what. For now, like maybe if you just assume that like information retrieval will get better. [45:19] very convinced that you still want to build like a certain types of reasoning pathways in kids brains before they get too much older and the like purest forms of reasoning are essentially stem reasoning which is funny because we seem to make be making like very good progress on coding and stem reasoning in these models today but i um i still think like the ability to like structure a [45:49] that seems useful and your ability to use even all of these tools will be better off for your ability to do structured reasoning. And that seems like, you know, there's a reason, you know, mathematicians don't tend to prove a lot, like after the age of 25 or some really sad number, like those pathways get built. I guess I like subscribe to that pretty mainstream view. But I also think like I am willing to be like very flexible about what education looks like for them

[46:19] Stanford. It might not be, you know, traditional education. Right. All right. Well, Sarah, thanks for doing this. I really appreciate it. It was so fun to hang out.