Y Combinator in the Age of AI | Ep. 43

[00:00] I think like AI and like a big change like this, like favors younger people and for all sorts of reasons. So I get it. For what it's worth, I'm a late bloomer. I did YC when I was 27. I was 26. Yeah. So hey, late bloomers. High five. The old guys. All right, guys, I'm really excited to be here with you. Thanks for thanks for all doing this with me. So I guess just to start, when did you all first go through YC as founders? I think I did it first. I was summer 2006. So the third batch ever. I did winter 2007. [00:28] So six months later. And then summer 08. Okay. So a long time. Yeah. It's like pretty close to the beginning. Yeah. So I guess where I want to start is like you all have in many incarnations seen, you know, what YC has been like. [00:41] as founders, as partners, you've sort of worked outside of YC, inside of it. What has changed? And so I guess maybe the lens I want to ask this question through is, what was the value proposition to founders in 2006 versus 2016, 2026? Like, what has changed? What has stayed the same? Yeah, what was it like in 06? The most surprising thing to people from the outside, I found, is actually how little has changed. And I think that's kind of by design, which is like the thing that Paul Graham created, [01:11] did in the early 2000s, it was great. It was a great product. And as you know, when you have a great product, you're like, don't fuck with it. I know there are some things that have changed, but I would say in broad strokes, it's much more the same than it is different. - How would you capture the essence of that product? Like to a founder, if you had to boil it down to like two or three things, what is that product about?

[01:32] So he recently worked on a redesign of the Y Combinator website. Actually, Gary, [01:36] We actually tried to write it down, possibly for the first time. Do you want to explain this? This is really your words on the website. Oh, gosh. I mean, it was team effort, but like. [01:47] Honestly, we tried to return to, you know, what is the founder's experience? And then we found all of these old photos of, you know, actually, your brother is like the first one on there. And it's like him with double pop collar. Yeah. And then we'll pop that up on the screen. And then you start flipping through and it's, you know, Patrick and John Collison. It's Brian and Joe and Nate at Airbnb. And you see them so young. And then you see them, you know, ring the bell. And it's like. [02:17] Basically, what we're trying to create is Disneyland for transformation from startup founders who are just starting out to literally the people who make the companies that really matter. It's actually a social process. All reality is socially constructed. I remember when I found YC and then came to my first startup school. It was kind of like being a fish out of water and then jumping into water. [02:47] who were builders who like were earnest and uh you know basically when you get into yc like we take people who are earnest and technical and then at the end of that process hopefully they become formidable and so when you go to the home page that's what that's about it's like yc is like a transformative process yeah it's like hoffman or something it's like this it's not that new agey it's actually very chill it's more like um you know

[03:15] I don't know how you would describe it. I'm a Harry Potter fan, so I prefer to think of it like Hogwarts. That's good. Yeah. I mean, I kind of felt like when I did in Winter 16, and I felt like starting a company is such like a weird, odd experience. And in some ways, having this group that normalizes what it's about, where you're like surrounded by other people, where they like talk about like, here's the new language and the new set of things you should be thinking about. It almost like makes the strange dream that you're in like. [03:40] you'd like get calibrated or something like that. Totally. So I thought that was a big part. And then I guess also there's always been the thing of like, there is some stamp of approval. I think that's like hard to, you know, sort of underweight. I'm curious how you guys think about that with like outsiders in particular. One of the things that seems different to me, at least, is now more than 10 years ago, and I'm sure more than 20 years ago, like the ecosystem is like an understood thing. And like a lot of founders, like, you know, they can read a lot of stuff [04:10] known in general, but I still feel like there's this thing that YC can do, which is take people who maybe aren't yet in that vortex and [04:19] identify talent somehow and bring them into the vortex. I should be curious if you guys like spend like active cycles thinking about that all the time. I mean, actually, I'm in the middle of my very intense addiction to cloud code and codex and like actually using this stuff is pretty wild because I'm. [04:38] Like I basically recreated my 2008 startup. It's like 70,000 lines of code. I did it in about like, I don't know, 90 hours over two weeks because I have a full time job and like also like trying to raise kids. It just like really compressed a lot of my sleep. But at the end of it, I have a code base that is better than what it took like five engineers and me taking antinarcoleptics to like build for my YC startup. That's crazy.

[05:08] November, end of November when Opus 4.5 came out that, uh, [05:13] you know, [05:14] I heard about it. I was like, hey, what's going on? Like the most interesting thing for us is like we've been talking about it for years and we've been using it for years. And then it wasn't until really even December where AGI is here, guys, for code. Like, I feel like I could create like in 80 hours something that I could not create with five million dollars and five engineers in two years. [05:44] identify as greatness if before this moment you were like, we need to find great engineers. Like what is the what is the thing you're looking for now? How has that changed? And then my second question is like these startups, you know, like the advice has been, you know, write code, talk to users. [06:00] like the, [06:01] What's what is that now? Is that different? It's prompt and talk to users. Yeah. Like there's a lot less than prompting, I guess. So we just like literally this happened. And then all of us have been sort of collectively like Harge did a project like Jared's coding all the time. Our team is all using this stuff. And then we realized, hey, why don't we actually put it in the application process? [06:31] and we're like starting to figure out how to process it. And we put it in a security sandbox because we figure it'll be prompt injected very quickly. You can tell a lot about whether someone can build just from like how they prompt the agents. And it's like, do they know systems? You know, for YC, you get a T-shirt that says make something people want on day one. It's like, you know, if you look at a resume, like you can sort of guess at whether they can make something. And then you really can't tell whether they can make something people want.

[07:01] their GitHub and you can maybe see if they can make something. The only way you can really tell if they can make something people want is they did it already. Right. And there's not exactly like a GitHub for prompting, I guess. Yeah. I guess that's what you're asking people for. But you can tell. Do they use plan mode? [07:17] Do they think about systems? Like, you know, are they prematurely optimizing? Are they overengineering? Like, you know, are they, yeah, basically like what is a feature to them? What is the complete release? Do they think about like the edge cases? Yeah, how you do anything is how you do everything. My favorite Steve Jobs quote, I mean, he talks about like, you know, if you're a carpenter, you can tell. [07:40] other really great carpenters. And we don't look at the front. Everyone looks at the front and we're like, oh yeah, this seems good. What a great carpenter does is an artisan looks at the back of the cabinet [07:52] Because like the cabinet, the back of the cabinet, only other carpenters are going to look at. And so I feel like that was one of my favorite things. Like when I funded companies that initialized, when I was a partner at YC the first time around, like that was my number one thing that I loved. Just like, is there game recognized game? Like, you know, would this person, would I go work for this person if I weren't doing the thing that I was doing now? This is how you spotted our pivot with Instacart, right? Yeah, he came in and honestly, like it was craft to get.

[08:22] you know, an iPhone app that was a demo app, Cloud Code didn't exist. So he actually had to build it himself. And then the way I could tell is, you know, turning the cabinet to the back was, he could scroll it and it scrolled really smoothly. And that wasn't true for like that era of iPhone apps, like the App Store and the iOS app was brand new. And then I feel like you guys do this all the time. I mean, to your vision of like, how does it change the way we might think about picking founders? I think it's going to just expand the net versus necessarily change it. I think we're [08:52] absolutely want to fund genius engineers who aren't necessarily using flawed code. There's going to be a Patrick Collison of every era and you want to fund those people. The way we've talked about it internally is that this is probably an era where we might find more Parker Conrad's. Parker, when he applied with Xenefits at least, like 2013, I remember the application extremely well written, but single founder, kind of not [09:15] in central casting, like kind of like not quite technical, like technical enough. I think he had like a demo, but I didn't have a CS degree. He wasn't like sort of like the traditional YC archetype. As soon as I think actually the only reason we interviewed him is at that point, his previous company, Sigfig, I had used it. It was like a personal finance app. It's like he's pretty capable of doing stuff. Yeah. In the interview, it was clear immediately that he like deeply understood what he was talking about. Super articulate, clearly like a really strong product thinker.

[09:45] And then obviously, Zenefits had its thing, but then Rippling is obviously huge. But I suspect that there's probably just like if Parker hadn't got his co-founder during the batch, it would have been a totally different story. And I suspect now like Parker of today is just in called code by actually building quite sophisticated applications. I guess it's also interesting because like Parker is one of the most probably like commercially successful. [10:08] intelligent people [10:09] There is. And I guess coding applications are probably making... [10:15] people with that shape increase their advantage. Obviously, I'm sure the real technical breakthroughs are always going to be really important. And this is markedly better than just looking at a resume or really, I mean... [10:28] Yeah, I, you know, when I first came back to YC, like I read all the feedback, like, you know, we're on our slash Y Combinator, too, with our Anons. And we know, like, you know, what people are saying is like, yeah, like, we should pay attention to this. And frankly, like, we need to do better than that. Right? Like, just because someone, you know, Parker did go to Harvard, but like, actually, we shouldn't fund someone because they went to Harvard, we should fund someone because they actually understand the user and are super unbelievably tenacious. [10:58] product, any vertical, any like set of customers you could have, like, [11:04] the most important thing is, you know, agency and taste. Like agency is, I see this person and they have this problem and I believe I can solve it using technology. And then taste is like, oh, well, let me actually build the first version and then get it in the hands of people. And did it work? And like, oh, how did it break? Like you go all the way into the weeds and it's like this incredible amount of like, imagine like this, you know, hundreds of thousands of lines of code.

[11:34] the tireless ceaseless gardener that is like necessary to like have you ever been to like the Baha'i gardens in Haifa is the most beautiful gardens I've ever been there's not a single weed there's not a leaf out of place it's like I think about like the best products in the world the best experiences in the world are like that set of gardens and it's because people like it's actually religion and like they really really care actually I'm interested on that topic I feel like um a [12:04] Yeah. [12:04] change a little bit over the years is I think it used to feel a little bit more like you could just ship something broken fast and then iterate from there. And I still think that's a good general mindset. But I do feel like the bar for products is just so high now. The quality of a software product at Series A seems really high to me lately. I don't know. How do you guys advise people as [12:34] week while also running YC at the same time. I feel like the bar for what, you know, two founders working on their idea full time could do before they interview with YC, like should be a lot higher. And we see this in the batches, like a few weeks into the batch, we do some of the groups do what we call we call it product showcase. Now you just like get up there and like demo quick demo of what you've built so far. And it's just like every six months over the last three years, it's like the bar for what you should demo even a few weeks into the batch just keeps going up and up.

[13:04] a company ought to be able to pivot more times than they used to be able to like company ought to be able to try stuff see if they get traction in a few weeks if they don't like it move like are you seeing that happen i mean i know a lot of this is new but do you feel like that's going to be happening is it happening absolutely i think we're seeing companies try many things during the batch do you advise in that direction i felt like historically there was um the yc school of thought [13:29] I interpreted a little bit more as like ship fast, iterate, you know, see where you're at, go from there. And I've interpreted versus in contrast that Keith Reboys School of Thought was like, you're a movie producer, like dream your movie and then like ensure it happens and like don't let anybody say no. I feel like the YC job is so much more in sort of almost like the psychology therapist end of the spectrum on this stuff. And I think when it relates to ideas, at least for me, a lot of it is more going off the vibe of the founder. [13:59] blanket like should you pivot quickly and give up on this idea or stick with it for a long time it's just like usually when you're meeting with founders you can just tell like if someone's like been working on something like they was never seemed excited about it and like two weeks in they're still not very excited about it it's like hard to tell them oh you just need to like persevere [14:17] going, it's usually better for them to find the thing they have the spot about. And what do you think? Yeah, I agree. I think an anti-pattern for founders who are pivoting is they have no existing prior on what a good idea is. And they're hoping that the outside world will tell them what a good idea is. And so they've launched five totally different things or five totally different groups of users, hoping that one of them takes off. They usually don't. Typically, what I'll

[14:47] Find an idea that they actually care about and then see how we can like turn that into a startup idea. One of the things that I've noticed, which a lot of people have noticed, this isn't some big insight, is that the sort of medium startup NYC batches is a good indicator of like trends that are upcoming. Like I noticed a couple of years ago before I was like there was like one batch where like all of a sudden like half the companies were AI companies. And the next batch, it was like 75 percent. And then it was like it all looks like AI companies, basically, except for, you know, hard tech. Yeah, hard tech, too. [15:17] YC is the hard tech of hard tech. What is the trend right now that you are seeing a lot of that you think YC might be particularly attuned to early? I don't think there's any strong trends yet. It's all just AI right now. It's all just AI right now. I'd say some things that are smaller trends, it might be glimmers of the future, just in the current batch. [15:47] to a generation of people. Staple coins and crypto stuff might be another interesting non-AI trend to talk about. Do you have a sense for why Calci and Pelt? Why are these companies, because I agree, they've captured something. What is that, do you think? I feel like there is a generation that has been very motivated and interested in it and is clearly taking off. Do you have a feel for what that thing is? Anytime there's a regulatory change, these things were in a gray area

[16:17] it's green light. So, dude, everyone's going to do it. Like there's, you know, capital flows to it. You know, capital is required for building consumer businesses. And boom, you got a consumer business growing super fast. Like you get more capital, the flywheel happens and you got another door dash, which is great. I mean, capital, you know, capital as a bludgeon works there. Yeah, really, really well. Yeah, it's interesting. So but what's funny is like, it's not clear to me capital as a bludgeon works as well with AI companies anymore. What do you mean? Like you [16:47] to have like the thousand people company anymore. - It's interesting. I still feel like I'm waiting for that. Where like, I know there are obviously some companies, like backed YC companies that are like, I'd love to, like, you know, I invested in it, I'd love to invest more, but they're super profitable. I'm like, okay, that's great. I'm very happy for you. - Yeah. - But there's also companies raising bigger rounds than ever and consuming crazy amounts of capital. And in some ways it looks even more capital consumptive than ever before. And I don't feel like I have a good mental model to square exactly why that's happening. - It certainly seems like it's easier to get [17:17] to like a million or $2 million of ARR without hiring anyone. We're used to seeing this. We get so many investor updates and usually we're used to seeing like, we just hit a million ARR and we had like 10 people and now it's just like consisting and we haven't hired anyone. So that's like new. I agree that like the step after that though, it seems like [17:35] the bees are bigger than ever. They're huge. Yeah, right? Well, I mean, venture is contracting a little bit. It's like not on a dollar basis. On a dollar basis, like there's a quote unquote, flight to quality. And then one of the things we've been talking about internally is like the world is actually increasingly full of these mega funds that are friends and they do great work, but it's like more and more dollars behind fewer and fewer people. And so fewer firms. Yeah,

[18:05] Capital is a Bludgeon, like the fixes in, which is great. Like often like for YC, YC is sometimes the number one, number two and number four, you know, of any given like vertical SaaS space. And we're in the billion dollar one. We're in the half a billion dollar one. We're in the Arm the Rebels one that works with the partner. Like this happens over and over again. I mean. So do you think in this environment then is the like king making meme, is that like more true [18:35] environment? Basically, if the founders are good, like the capital... [18:39] like helps them get there a little faster but i don't believe that like capital as a bludgeon especially when things move so quickly like you can end up having capital and then like moving like [18:50] fast in the wrong direction. Like Harch funded Giga. Yeah. And these guys beat, I mean, we love Brett, we love Sierra, but yeah, like Giga, this sort of 10-person team beat... [19:04] all the incumbents for things like DoorDash. I mean, the back to back to back, it's like they have the best tech. And I think there may be like still under 20 people right now. And Jack, speaking of like insiders versus outsiders, that's such like a classic like insider versus outsider story. I mean, what were the backgrounds of the Giga founders? Oh, like they were two, they got IIT in India. They were still in India. Yes, they're in India when they were in front of them. Yeah, they actually make it out here for the batch for visa reasons. But they were just like

[19:34] students. They had done their sort of, as undergrads, they'd done sort of PhD-level research in fine-treating LLMs before, like, everything really took off. So they were just, like, clearly exceptional now. Yeah, I mean, to this point about, like, the king-making stuff, and not to, like, talk around book too much, but, like, you know, like, Logoro was a, you know, startup that was, like, [19:52] coming behind something that was seems really established and they're based in Europe and then they went through IC and they're doing great. And so it doesn't seem like stuff is like locked up. [20:00] immediately and that's a good example of like the price like even if you have the capital you might not build like the right product off the bat and then things change um i think gary's getting better the model's getting better so so like back to back to back like every single year the models are getting better you know the rumor for harvey is that like you know you might they might have spent a bunch of money a bunch of the vc capital on fine-tuning models that are like not better than the frontier models i mean you don't have a crystal ball you couldn't have guessed that that you know [20:30] It was just an idea that maybe that might happen. Of course it did happen. And then now we're sort of in this situation where like, you know, if you have hundreds of millions of dollars sitting in your bank account, you're tempted to use it. The models thing is so interesting to me because one of the things that has struck me is that a lot of the source. [20:47] of product market fit actually exists outside the startup delivering the service. And so it's, you know, like you take an amazing founder like Max, you put them in a market like legal where there's just a lot of uptake for some set of reasons. And then it's like you have this tailwind outside the startup that's actually driving a lot of the like aha moment, not to take away from anything that they've done. This is the case for all these startups. And so that just sort of changes some things. I think this is from the family's perspective, at least and a reason why they're raising the bigger.

[21:15] B rounds. I mean, there's always that case of just like bank the money while you can get it. But it just feels like the surface area for the products is bigger than ever. And they are like still fundamentally constrained on like the number of people they have to go and execute on things. And it feels like we're now SaaS era was sort of like you build one core feature and you hit product market fit with that and maybe just run with that for a few years and then add on things. Whereas now it's like even within the batch, people are trying to like add like more into the product and have it do more. [21:45] should be doing and they have competitors and the competitors are moving faster. So haven't as yet seen that, like once your post product market fit, it doesn't feel like the people won't need to hire as many people. They're at least not acting on the belief that they need to hire as many people. Yeah, I would say one thing that doesn't seem automated yet is sales, for example. Like, it seems like you still a lot of people thought that it would be automated by now. It hasn't been automated by now. Support has, but sales hasn't. And so like you still need a lot of salespeople. Seems like, you know, engineers are way more effective. But if you've got the capital, [22:13] 50 engineers are still going to be better than five engineers. So I think some of these things just haven't necessarily put out intuitively. It's great for everyone using the products. It's just like the bar for like what you expect out of like the products you use just keeps going up and up. Yeah. It's awesome. There's also a lot of these markets that are so genuinely blue ocean and like they look like good ideas and they are good ideas and it's totally new. And so the result of that is 50 startups doing something similar. Yeah. [22:39] I think that's a good thing for end consumers. Obviously, you guys have a lot of companies that are in that situation, like we all do. I'm curious how you...

[22:49] you know, in this competitive of a market moment when every startup's got a ton of competition, like, does that change anything? Like when you're working with like specific companies, do you find yourself saying, you need to go faster, you need to be thinking about something differently? Does it update anything when we're in this type of environment? [23:05] Me? [23:07] I felt like Gary was holding back an answer. I'm hoping you guys have a good answer. I can answer my own question. [23:19] especially during the batch we're so focused in on the like is this even like is this worth investing like another two weeks of your time in the units of time? Is there a glimmer of product market there or not? Is there anything? And that doesn't really have to do with competition. Yeah. At all, exactly. And so I feel like we actually spend the vast majority of our time talking about competition telling founders not to worry about competition because it's all founders who are like, like imagine if the Lagora founders had like not launched Lagora because they looked at Harvey and been like, oh, it's over. That's what we see a hundred times a batch. [23:49] is that story. And so it's just being like, don't worry about it. Like, just out-execute them. Yeah, I think we always just go back to the make something people want. And that, like, it says make something people want. It doesn't say, like, do a market map. Yeah. Based on what, you know, perplexity tells you. You could add that. You could say, make a market map and then make something people want within it. That'd be, we should make an April Fool's t-shirt that says that. That'd be good. It's just like, make a market map, then make something people want. It's like, what the hell are you talking about? Like, that's definitely not how you do this. So, like, let's say,

[24:19] on customer support like i'm pretty like that market is not saturated like i'm pretty confident that a good team that comes into customer support and is like i'm gonna go find some more customers like they could do it and so basically the view is just like hey if you can find customers and you get it going like don't even think about who else is out there just go i think that's basically i mean like it's like go out and get customers and if you have good competition then like you have a hard time getting customers it's like and if i try and launch a new payment processor i'm going to run into stripe and like it's going to be hard for me to grow really quickly and think [24:49] I'm curious how much you guys think [24:52] at like a macro level about stuff. Because I know like the most important thing, which is I believe obviously correct, is to think about the micro, getting this startup... [25:02] off the ground and going and then things can go from there. I am curious about some of the like macro things like one of the ones that comes to mind for me is like the recent, you know, trend in public markets about sort of like SaaS multiples just getting totally hammered. Do you guys feel like SaaS is dead? Does that resonate for you? Do you see anything different in the companies you're working with? Is SaaS dead? I mean, I think it's dead. [25:25] The thing is, if you run a SaaS company, you don't have to be dead. All you have to do is embrace Cloud Code or you have to embrace, you know, [25:33] top to bottom, gentic view of like how everything's going to work. Like put it this way, like the same week that I personally realized that everything was different. You know, I funded a team from Meta Superintelligence who had left and they were pointing out like, you know, Meta has 20,000 people working on reality, you know, Alexa has 20,000 people. And it's like, and I thought about my experience. It's like, I didn't even have like 20,000 people. I had five people. And like,

[26:03] I knew what the architecture was. I knew what I wanted to build. And then I had to farm this out. But then I had to have meetings and I had to, you know, come up with a doc and then other people have other opinions. And we have like five meetings about like the architecture and like we argue about it. And then, you know, two weeks later, like maybe something happens. But if you're in a big company, it's like three months later or something maybe happens. And it's like, look, like we don't have to do that now. Like we could just try both, like go into plan mode. [26:33] Just do it. And then literally like an hour later, we will have something done that would have taken like two weeks, two or two months or sometimes two years for if you're not a tech company, and you know, you're an incumbent, like it's like two years or never that you would even make that decision. Right. So like the speed of making that decision, like how decisive you can be, like, honestly, I think like going back to the transformation thing, like that's actually the thing that I feel like I learned at YC. [27:03] It's just that I was employee number 10 of Palantir. We were moving – I was sleeping at the office. The big difference was realizing instead of getting 20 basis points of Palantir, which now is actually an astronomical amount of money, by the way. We didn't know it at the time. We'll run some math. We'll put the math up on the screen. It's in the billions, I think. That's a thumbnail. That's perfect. Basically, I wanted 97% of the company that I started.

[27:33] Bye. [27:34] I went from a place where it was already fast and then being the founder and the CEO, like YC sped me up even more because you're in office hours with people. And it's like, oh, man, this person actually grew 10 percent this week, 20 percent this week. Like, how did they do it? I need to do it. And I think all of this is an accelerant, like Cloud Code and Codex and being able to, like, make two years worth of product progress in about two weeks. How could that not make YC more insane? [28:04] Like the amount of things that you could try and do. It's like, honestly, you could do two years of work, realize that like actually nobody wants this or there's too much competition and you throw it out and then do it again. And the thing is, like, that's not throwaway. You learn something. You also got better at like using these tools. And then you went out and you get another shot. And so what's funny about like like seed is you could think of people raising like two or three million dollars. You know, that used to be Series A, by the way, which is hilarious. [28:34] Yeah, yeah. I was like, oh, that's kind of a small seed. It's like, are you serious? That's so much money. This is outrageous. You don't even need this money. It's crazy that you can go so much faster. I actually, I do want to come back to fundraising advice because I feel like it's gotten into an interesting place there. But the opposite question of the is SASTED? [28:51] is what do you feel is not AI but safe from AI? Like, are there areas where you feel... [28:57] that you think are sort of insulated from this mega trend? [29:02] that you're like happy to back without fear. I think the obvious one here, so just like marketplaces that are like aggregating people. I think Airbnb is like very safe. I think one not safe. I think there's a bunch of marketplaces since DoorDash is totally safe. It's going to be a very clear one, right? One of the things we've been talking about is what the agents want, what the coding agents tell you to do turns out to be like itself a really big

[29:32] written to prompt inject cloud code to force it to use you i'm just joking like it's not clear that you can do that but like if you could you would because it's that powerful like people will just i need x and then you know what's the best thing on the internet to do that and uh actually that's really powerful yeah i think you might say like maybe the sass thing is like even within sass you might say like things that feel like they are [29:56] essentially databases or systems of record things like rippling feel like they're going to be in a good spot and then things where they like the moat was around the number of integrations they built or like [30:08] data connectors or that kind of stuff, which you can now just like code in 30 minutes is brittle. Do you think it's system of record that makes payroll sticky or do you think it's touching money? Touching money, regulatory. I mean, once it's you, I mean, you have systems that are working, you don't want to touch that unless you have a really good reason. Yeah. Because like one of the ones that comes up a lot right now is like, are CRM safe? You know, like our old, which I don't have any, I don't have actually a particular opinion to, but it just seems like an interesting question where it doesn't exactly touch money, [30:38] includes a lot of information that's important, but like that information also now can like live in email or somewhere else like that. So, yeah, I think that I think Salesforce is probably screwed. Like, I feel like there have been so many attempts to do the Stripe strategy of you get the startups in the YC batch using your CRM because everyone hates Salesforce. But no one could ever really grow into a big company because at some point your head of sales is like, no, I need like these reports. To your point, you need the integrations. Yeah, exactly. And now, like, that's all just going away.

[31:08] a YC batch that sells to all the other YC startups and investors will say, oh, it's not going to grow because no one can really compete with Salesforce. Also, all its customers are YC companies. Yeah, exactly. It'll never work. [31:23] What about hardware? Is hardware safe? Hard tech is just hard. The moat comes from it being hard to source, hard to make it work. It's just another muscle. It hasn't exactly come for atoms yet. [31:38] Yeah, robotics is still a little ways. We just need more. We need ASI. We have I think we have AGI now and then ASI is coming like super intelligence. It's just clearly just around the corner. You think it'll just be like a when we know we know or you think there will be like a thing. Yeah. [31:53] that to you would say that like that [31:55] new moment is here. I mean, I think like we have limited versions of ASI right now. Like I saw it talking together. That was crazy. Yeah, that was crazy. I mean, that's a great example of like, that's the first example of like swarm intelligence. Yeah, I was like, oh no, it's bad. Like in AI research, that's like swarm intelligence is like this huge field, but like this is a huge validation for that field because, you know, it's kind of an interesting question. Like, [32:25] big AI labs, they're like, yeah, we're just going to build the God model. You know, it's going to be mega big and, you know, just think about like Dr. Manhattan or something. And then, you know, biological systems and even human society are not modeled that way. It's like we have lots of people with lots of diverse hardware, a lot of different opinions about all kinds of things. And then you sort of come together and see what sticks. And that's what like research is, for instance. So,

[32:55] level intelligence is actually a very interesting thing. That's like just the beginning of that. That just happened literally last week. Is there a type of [33:03] project or... [33:06] pathway for a startup to build that you're not currently funding that you'd like to. And like the example that I'm thinking of when I'm thinking about some of these hardware companies or projects at the beginning that I don't think can start with a million dollars, like some of these genuinely do need to start with a million dollars. [33:22] 10 or 20 million dollars. Is that something... [33:25] that [33:26] you think about is that something that like is that would would a divergence like that ever be worth it to yc or is it like we don't need to back every single type of company of all time we would like to back every single type of company that's actually the kind of thing i was curious about i was like is the conversation when you guys are thinking about like [33:45] YC growing and all that is it like we have our style and we like to just get everything within that or it's like no we'd like to back every company of all time we're generalists and then the smart I mean YC funded Coinbase when crypto was like the weirdest thing but you know. [34:00] Brian Armstrong was in the risk team at Airbnb. He was in the anti-fraud team. And he was already in the family. And then he said, oh, well, how do I start a company? This is clearly the way to do it. And then what's funny about it is you could start 20 other Coinbase competitors, but all of them died. Because when you're early, it doesn't matter. It matters more who's the person and what do they believe. And then that person goes on and creates the future. So being generalist is an

[34:30] truly the best. Is there a slice of the market or a type of company or founder that you feel like you want better exposure to that you're like actively working on? I mean, coming back to YC, one of the things we realized is like we have a huge media presence. But on the other hand, like on its own, if you just watch the YouTube channel, it's like, oh, this is like something in the sky. I heard it's, you know, one percent acceptance rate. Like people just kind of, you know, think [35:00] people either know someone who did YC or they met a partner directly and at an event or they actually can't be like we actually have to be in the community. We can't just be like in the sky on the Internet. Right. You can't just wait for apps to come in. So, I mean, Jared led this like you basically. [35:17] you know, [35:18] got us to like how many college campuses last year? Over 30 college campuses. Yeah, so we have a huge like boots on the ground effort now to go and talk to undergrads everywhere. We just got back from a big trip to Europe. We're going to India in the spring. To also go back to your point though about like groups of people that we would love to see more of, we have a big effort this year to do what we've done with undergrads over the last two years. I think it's been pretty successful and to expand that to grad students and people who are more like Brian Armstrong's age,

[35:48] in a good way, like they're unbelievably impressive. But YC founders are young. And they seem like it's like in recent batches have trended even younger, potentially. They have. Yeah. And it makes sense. I think like AI and like a big change like this, like favors younger people and for all sorts of reasons. So I get it. For what it's worth, I'm a late bloomer. I did YC when I was 27. I was 26. Yeah. So hey, late bloomers. The old guys. That's the old guys for us. Many of the biggest YC companies were started by founders in sort of like mid to late 20s. [36:18] That's old. Yeah, super old. You should definitely quote me on that. There are tons of old people. I think it's 26 year old. Yeah. We started with going back to colleges. And I mean, this was kind of driven by them, though. I mean, I think big tech stopped hiring. [36:37] And then simultaneous to that, like we have a real vibe inside the batch sometimes among the young founders that is the last time to participate in capitalism, which I definitely think not. [36:48] It's like a powerful idea. I think a lot of people feel that way. Why is that? I think there's probably a lot of people who are like, if AI is going to actually stay on this trend, what are we possibly going to be better at? So I got to do it while I'm still better at something. I feel like that is so short-sighted. It's unbelievable. Ryan Peterson always talks about, don't you think that human capacity for desire is virtually unlimited? We have a God-shaped hole in our heart. We're going to want more and more stuff.

[37:18] do it now. Like I was just thinking about the turn of phrase, like I'm sure you've been in business meetings or like making decisions about products where it's like, whoa, whoa, whoa, let's not boil the ocean now. And it's like, I love that term of phrase because I've said it a lot. I've used it to like justify not doing things. But like in the age of intelligence, you can just do things. Why not? Like, I mean, maybe not boil the ocean, but let's boil a few lakes. Why not? Right. Like, actually, this is sort of the moment. And so when you connect that to what Ryan says, [37:48] It's like, that's what that would look like. If you're an investment firm and you beat the market, you're like 20% net IRR back to back. It's like, what does this stuff mean? Does it mean that we're going to fire all of our analysts and we're just going to have the AI do it? And this one person who runs the firm is going to make all the money. Why would you want to do that? Because your competitor isn't going to do that. Your competitor is going to say, you know what? We have AI now. I want 50% net IRR. That's the thing I have not understood about, oh, we won't need capital. [38:18] years if your competitor has 50. Yeah. It doesn't make unless they can't do anything productive with all the agents. It doesn't make any sense to me. Yeah. I think it's just like, yeah, let's boil a few lakes first and then we can boil the ocean. Like I think, I mean, I'm not serious about that. Obviously, there are limitations to it, but like the the invective against AI and this idea that like society is going to fall apart is so extreme that I'm like, maybe that is like we need some

[38:48] mean on that it's been just like a weird thing watching sort of the efforts to both sort of like manage but also like impact sort of like societal like understanding of like what ai is and i feel like there is still a lot of like fear embedded and like i think outside of san francisco you go to other places in the country i think like you know i'm from st louis like i don't think everybody trusts ai fully and you know whether that's right or wrong like i think it's um i do think it's important to like get out of our little bubble sometimes on a lot of this stuff just to know [39:18] the world's out a bit more. That's for real. It's interesting, though, it feels different this time because it's clearly in our little bubble here, everyone's all pro AI. I would say my other extreme, I have parents, immigrants, don't speak great English. My mom's totally addicted to ChatGPT because now she can just do all this stuff that she needs, she can just send letters to reply people in a way that's super empowering for her. It's like the kind of middle, like the people who are threatened by, like, is AI going to take my job? [39:48] threat to sort of knowledge work or white collar people like usually i feel that people who are actually not early adopters but tend to get on the train immediately after sort of the san francisco yeah train takes off are now resistant to it and it sort of bypassed them and has gone to like the other end i mean i sympathize with that and like as you know yeah labor that became management and capital i like totally feel that and i think it is actually about you know in the marxist sense it is actually about like where you fit in there and then my argument would be like it has been

[40:18] to become a founder, which is like management that becomes capital now than ever. It's way more possible. Yeah. I think that like that fear of... [40:27] for workers about what's going to happen, that is actually something that management and capital has to take. It's a responsibility. I think so, too. I think it's very easy to be like, oh, in the arc of history, there's always these new technologies and people find new jobs. It's like, well, yeah, but there's a lot of structural unemployment that happens in the middle of that. And that's real. [40:57] just will never change. They will never get better. There's I mean, some of it is in the context of like no competition. One of the things we learned from hard tech companies, for instance, is that like it's impossible to get like a certain block of metal. Like you have you know, you have to get it fabricated or like smelted in China. Like America's like lost the ability to do it. Right. Like when you have a market that is so broken that you can't get it inside America. Like how did this happen? Right. And so I would go back to, you know, capital and management and say [41:27] imagination, right? Like, we're not here to like, just continue to do like, let's shave off like one or 2% every single year, and like increase our net profit. And that's it, right? Like, we need to think way more about how can we use this technology to radically change how businesses work, how like what products are, how much better could they be? I completely agree. And I think if capital management don't, like take some responsibility there, like, there's not just

[41:57] There's also like the whole world getting more expensive. Home prices are going up while there might be like pressure on wages because you can do it more cheaply with AI. And so you have like inflation happening with wage pressure at the same time. Like I do think it's like a more real thing than like our echo chamber once like and again, totally for it, too. But I think it's like I think it's an important thing. I mean, that's why we've been so vocal about this idea of little tech. [42:27] for startups to be able to actually train AI models, to be able to enter markets. And, you know, frankly, like I know we have lots of friends at, you know, Apple and Alphabet, and, you know, we have huge respect for those companies. And then once in a while, like you'll see in the press, like, oh, yeah, like we submitted an amicus about Apple and Alphabet, not because we hate those guys, but because actually we need tech to allow new startups and new entrants to come in. And so to [42:57] way more. [42:59] more aggressive about like what our products and services should be and can do. And then we need markets that allow those people to actually do [43:08] you know, exist, thrive, hire lots of people and create new jobs. Like that's basically like, you know, a lot of people are like, "Oh, I work in tech. I don't know how to do this." And it's like, look, it's actually abundance. Like we actually have to build again. Like we've become a litigious culture. I'm sure you've read this book, "Breakneck," you know, like China versus the US. It's Dan Wayne's book. It's incredible. It's really good. It's basically, you know, we built a lot in like the 40s, the 50s.

[43:38] And then sometime around the 60s and 70s, you know, six, seven, I mean, they literally stopped building. And we've been in this sort of, you know, we can't build high speed rail. It's insane. Like, why? Because like, basically, we're totally a litigious culture that cannot like get out of its way. Yeah. I saw some Peter Thiel talk about like, we built all this stuff till the 70s. And then for like 50 years, nothing happened except computers, more or less. And even right now, most of the revolution is in computers, which is great that it's better than nothing. And obviously, other stuff is happening, too. [44:08] but it is you look at you know what China does like standing up a city in no time it's like [44:15] They're really good at robotics. [44:17] What are we doing? On this sort of just like societal topic, obviously you guys are very engaged with like the city and the state and things like that. Some more than others. Some more than others. Yeah. Maybe like my one question on this is like, what do you think is the posture that San Francisco and California need to be taking? Like, what's the most important thing? [44:40] posturally that we need to be taking i mean this is gary's area gary how do we fix california politics and there's a plan like i'm kidding but like actually he does yeah i mean uh matt mahan just uh you know announced his uh race for the governor i think that he is the perfect example of someone who is not virtue signaling he's like you know he built more than 1400 homes in san jose the year before that he hadn't passed all the legislation he wanted so zero market rate

[45:10] before when he came into office, he reduced homelessness by 20 percent, like more than a thousand people came inside and got treatment and recovery because he actually supports treatment and recovery. I didn't want to be involved in politics. But when I saw that my Asian American grandpas and grandmas couldn't walk down the street without being assaulted and killed, when I saw like people like me when I was [45:34] you know, 16, 18, like, you know, I wanted to participate in tech. I knew I wanted to be an engineer. I didn't know that I would get into Stanford. Like... [45:43] I wouldn't have been able to do that if I didn't have algebra in like... [45:47] Middle school, public middle school. Like my kids, like, you know, my kids go to private school, but I went to public school, you know, and we should have a government that like doesn't require you to be rich to become a startup founder or good at math even. Like, I mean, this is like, how do we get this bad? Well, I'm glad you're fighting for it. It's very important and it's not pleasant. And I see you fighting on Twitter in a way that I would not have the stomach for. And I'm glad you're doing it. Yeah, I appreciate it. Yeah, no, I mean, I think for people watching, it's like, look, I'm not going to get all the takes right. [46:17] here when I don't get it right. But on the flip side, it's also like have the debate. Yeah. Let's actually talk about it. It takes a lot of courage to say stuff that you think and you know you might be wrong about and you might get a big blowback on it. Or even if you have it right, I don't know if we're going to get mad. San Francisco and California got this bad because the people who... [46:37] I mean, they run businesses, they have jobs, they are trying to raise their families in California. It's like all of it was so big and so scary that we stopped paying attention. And San Francisco is on a better path because we started talking about it. At dinners, we started talking about it. Hey, did you hear about so-and-so? They got assaulted. Did the police actually show up? What happened with the judge?

[47:07] Francisco for whom like, you know, [47:09] I was looking at the records, like there are 100 cases and like 90% of the cases, this judge, she [47:17] basically just [47:18] like dropped it on the floor at an extreme rate. Like, I think it was like three or four times higher rate of just dropping cases on the floor purely for ideology. It's like unbelievable. Like, how is it? How is this person serving the people like they're not? And then the thing is, there's a reason why we elect these people like there is an election coming and, you know, [47:38] We have to make sure that we hold these leaders to account. Totally. No, it's great. I mean, it's super important and it's like not easy work. [47:46] People got someone's got to do it. And I'm glad you're doing it. I appreciate it. Yeah. OK. Politics tangent aside. Although I could talk about that the whole day. I know, right? But we'll do the politics episode. We'll do it later. Yeah. Going to get you and some politicians on here. One of the things that I'm curious about that has been an obvious topic for me on other episodes has been like mega influx of capital. [48:06] into venture. Is that [48:08] a positive or a negative for you? And in what ways do you experience it as both? I think it's mostly positive for us. Why I see it at its best is not competing with other venture firms for deals. Why I see it at its best is convincing people who want [48:23] didn't seriously think about starting a startup to go for it and then being their first believer yc only works if there's a large pool of downstream capital that can then fund all the subsequent rounds for those companies and so i think actually yc does best in those environments yeah we're a managed marketplace so we need like as many great investors as possible and actually what's funny is like in that managed marketplace if uh

[48:47] you know, the sort of supply goes up, like we need to go out and find the best possible people. And then we're actually really good partners to the rest of the industry. Like VC can, you know, 2x, 5x or 10x over the next 10 years. And we will meet the demand. With companies. With really the smartest people of our generation. And if that happens. So you're saying you feel limited by the amount of capital still? How about this? I mean, capital from VCs like you, or VCs like our friends. Stop. [49:17] You got to do the work, show up, don't disappear, don't be a dick, do no harm. When I first came back to YC, one of the most interesting evolutions of how to deal with investors was that most investors are actually bees. [49:35] like B or B plus, like you should be so lucky to have someone who like does not mess with you. That's really good. And then obviously, yeah, I mean, basically, yeah, damage the company, right? If there's someone who, you know, has a great network or, you know, can make the Keynesian beauty contest happen for you and your company and you're around, obviously, they're investors and yeah, they're going to catalyze something crazy for you. You know, you should work with it. Last time we were here is like, yeah, if you can get like Keith Raboy to invest in your startup [50:05] give you like 20 million dollars you should probably take that money man like you know do it right yeah but for everyone else it's like you gotta you know find people who sometimes you gotta do b plus capital it's okay you know well it's sort of interesting because you know you made the point which i agree with which is that there's despite the total volume of dollars going up there's like in some ways a consolidation of the number of venture firms or players or things like that i don't know if it's like more companies getting funded or if it's just a lot more dollars going to just like a couple breakout successes i guess both of those help i mean we actively

[50:35] think about we need more a and a plus uh you know investors like humans doing the work exactly um we need i mean and i think they're all going to be yc alums like you i mean i think you're great uh ili sukar is incredible dan levine at excel like you know iris now uh runs the seed program at gc and he's incredible like i you know i fully hope and expect like the top like [51:00] 80 out of 100 spots of the Midas list in the next 10, 20 years will be always sealed. That's funny, you're like funding VCs like just on the 10-year delay. Yeah, I mean, Liz, Liz at first round is killing it. She's incredible, right? So I think that's like, you know, [51:14] scratching the surface. I think we're going to have dozens of the most legendary people, and they're all going to be YC alums who deploy all the world's capital. Increasing the number of good startups in the world is, I would argue, just the core founding principle of YC. When I first started working at YC in 2010, it was the first thing [51:33] um pg said to me it was actually clear why he wanted to hire anyone like what like this seems like it's a nice like family business two of you working on it seems great um and his whole point was that there's like the conventional wisdom in the vc industry it all comes out of andy ratchleff's research like there's like only going to be 10 companies per year that will go on to reach like 100 million dollars in revenue and be significant and ipo at some point and you just have to be in one of those 10 companies every year otherwise you may well just like stay at home and pg was like [52:03] I don't want to go out and fight for those 10 companies. I just like, we're going to make more of them. Yeah, he's like, let's make more of them. Let's see if we can make two of them. Like, why is it 10? It's just like such an arbitrary number. And I think like clearly over the last 15 years, that's proven to be true. And I just feel like everything we do here is driven by we want that number to go up. What are the bottlenecks? Are the bottlenecks like more founders? Let's go find more. Yeah. And you went to four batches. That was great. Yeah. That was exactly a good example of it. It's like that was a bottleneck. Like you shouldn't have to wait six months to do a YC batch.

[52:33] like what was hard about that what did you have to like change if anything [52:36] Was it just you didn't get as much time between to like source new companies or maybe this is in the weeds? But like one of the structural changes we made at YC since Gary came back is to sort of decentralize YC, I would argue. Like it was pre-Gary coming back a lot more centralized. There was sort of like a team that sort of ran the batch and sort of things were more top down. Decentralizing decisions, batch operations. [53:06] companies and then we club together to form a group that runs a batch. But we're just like more nimble. Like we can just sort of do things in a way that wasn't as easy before. Yeah. And plus, we have 15 people now. This is like the most number of partners we've ever had. We have 15 visiting partners right now. So it's actually like something like 30, 31. Well, is there an artist total right now? Do you feel that there's a limit to those numbers or do you think you could could you double the number of YC partners and therefore double the number of [53:36] funding four times more companies than you are today? What would you need to overcome still? Well, a cool thing about [53:41] the way we see this structure now. I think this is something that's often misunderstood from the outside. I think when... [53:48] Mm. [53:49] Most people hear that a YC batch is 200 companies, but they imagine the batch experiences is like a room full of like 400 founders showing up every day. It's actually four batches of 50. Extra haul. Right. But it's it's actually more like each partner is running their own like autonomous mini YC. And it's, you know, 30 ish companies. We call them pods. I mean, I had my whole pod over to my house for dinner last night because like that. And it's approximately the same size as YC was when Gary went.

[54:19] or eight simultaneous 2008 batches with PG. Exactly. And we have 15 PG's now. [54:26] Yeah, and so with that kind of structure, there's no inherent change that has to happen if you go from 15 to 30 of them, because it's fully paralyzed already. Yeah, and one of the ways I've perceived YC change... [54:37] over the last several years or something was there was this period, I guess, when my brother was here, where there were a bunch of new projects. And then I felt you guys kind of, you clearly went back to the core and then did more of it. I could feel that very clearly that it was like, this is our thing and we're going to refocus on that, but we're going to make it more and bigger. And I think that's been super successful. And you guys have clearly throughput more companies. The companies are awesome. [55:07] to me are there initiatives outside of that core that you are interested in with the new market moment or are you like we're gonna just do more and more of this distilled core or it does depend on the day well first of all i think that's a good overall description of like how yc has changed especially since gary came back is like refocusing on the core that's been sort of like the high level of the last three years but yeah we do have some big projects in the works i'm not sure which [55:37] Yeah. I don't know. It's not like we have like [55:41] new initiatives that don't feel core. I still think it all falls under the umbrella of like the core thing is how to create more good startups. Yeah. And like keep pushing down on that. And as Jared mentioned, like fellow is just an example of something we just launched last year, which is like, what's another bottleneck? There's more founders. We want to meet the founders earlier, like.

[56:00] college students before they're ready to do startups. And so we started offering just like grants and like community to like exceptional college students that we feel like maybe they're not ready to start right now, but they might be in a year or two. That's a good example of like widening the base even further because like, yeah, it's almost like you could like narrow the top and you could say, you know what, we're going to actually [56:19] put a bunch of money into growth. You know, obviously you did that, but you know, the other way to go is you could widen the base and you can say, actually, we're going to try to, [56:25] to have some engagement with 10,000 founders or something like that. That's basically what we think a lot about. Yes. Yeah. Because amount of follow on capital is not really the bottleneck right now. Really, the bottleneck for us now is getting more great founders to want to do startups and to do YC, which is what the bottleneck should be. Like if the bottleneck is anything else, something is weird about about about the world. That's that's how the bottleneck always should be. For a long time, YC was just struggling so much to just scale the operational side. We're here like a post product market fit company that like couldn't keep up with [56:55] demand. And so instead of like focusing on growth, how do we get more users? We were like YouTube when it was like scaling by like 2x every month and just everyone was just like trying to figure out like how to keep the site up. I was certainly the first few years of me at YC was just like trying to figure out like how to not fall over. But now that we've like really got the operations down, we can go back to focusing on what we should be focusing on, which is like, how do we broaden the base? How do we get more great people into our ecosystem early? How do we inspire more people to start companies? I mean, one of the [57:23] most fun things from, I mean, I pinch myself, but like I get to, you know, I had left YC and then now I get to, you know, go hang out with PG and Jessica again and hang out with Brian Chesky who's on the board. And then one of the directors from the board that's awesome is, um,

[57:39] you need to make sure that we're having fun. And that was true back in 2011, 2012. Laughter test. Yeah, yeah, yeah. Basically, that's sort of the directive we got this year that's not a directive at all. It's like, [57:53] you know, awesome, actually. It's like, if we're not having fun, then we're doing something wrong, actually. And I remember just being around YC and the partners and PG and PB and Jeff and all these, you know, it was just very, very hilarious all the time. It was unbelievable how weird startups can be and like, can you believe this thing happened? And then I feel like our partner lunches are that again, which is really fun. It's hilarious. And that's so important, having a partnership like [58:23] trusts and enjoys each other and respects what people think like that's just got to make the whole experience so much better it's why everyone's a former all the partners are former yc founders so it's like it feels just like a little bit more than a job for everyone like everyone's got sort of a pay it forward yc changed their life in some way yeah like just adds to the good vibes i think yeah it is obviously an iconic and very important institution and you guys are doing a great job running it and so i'm sure it is both like it feels like a heavy responsibility but i'm glad [58:53] Thanks for doing it. Really appreciate you making time. Thanks for having us. Thanks, Jack.

[58:59] So when are you going to be a YC partner? [59:03] Keep checking my inbox.