Sam Altman on Y Combinator and What Makes The Best Founders | Blitzscaling 02

This is session two of Technology Enabled Blitzscaling, a Stanford University class taught by Reid Hoffman, John Lilly, Allen Blue and Chris Yeh. This class features Y Combinator president Sam Altman. This podcast has been produced by Greylock Partners. For more podcasts, class notes, slides and videos, please visit greylock.com can you talk.

A little bit about Y Combinator and what it is and how it works?

So we fund startups, not only that, but mostly that. We sort of look at our role as enabling as much innovation as we can in the world. Startups are really a good way to do that. We fund about 250 startups per year in our core program. And then we have a new thing called the fellowship program that we're going to scale up much beyond that, which is one tenth of the money and less advice, but we think is that we figured out a way to sort of build this, made a company that helps startups a lot. And so if you're part of YC, we want you to have a 100x leg up on where you would be otherwise. And we try to give advice and connections and access to this community where we can offer things that just really help startups. So, you know, we try to figure out the advice that really matters. But mostly the thing you get is this alumni community. There are now more than 2,000 people that we've funded. People feel a very strong affinity to yc. They work with other YC companies, help them find, fundraise higher, buy each other's products, whatever. And we really sort of try to make that a tight community. So we're somewhere between a product company and a traditional investing firm.

I think something happened maybe three or four years ago, maybe four or five years ago when people started really, really identifying as being part of YC and part of the fact. But you dropped out of Stanford.

I did, yeah.

What year did you drop out? 0505. I mean, what year were you at Stanford?

Oh, I was going to finish in three years, so it was sort of my last.

And you had done the CS core, so you knew how to code. And you dropped out to start looped?

I did.

And did you start looped and then go to yc or did you talk about that?

Yeah, so I wasn't called at the time. It was a project I was working on with some friends. And then I'm pretty sure that we started. We went back to try to figure this out. I'm pretty sure we started working on it just about the day that YC announced that it was going to Be this thing. We didn't think much about it then. We didn't even hear about it right then, but worked on this and then heard about it in detail and really seriously considered it. The day before applications were due, this was like, I think it was like April, middle of April, something like that. So we applied right then and flew out there to Boston. And you know, it was like, at this point the world of investing has really changed, I think partly because of YC. But in 2005 it seemed very mystical. It seemed like people really wanted business people running these companies, not engineers. And there was all these things and if you didn't know the right words didn't work and that people really wanted business plans and financial models and whatever else. And so when we got to yc, it was like, these are our people, these are engineers themselves. They care about the product, not the business plan, not the financial model. And I think it actually took the YC partners maybe two years to figure out what they were onto. But to the founders they funded to the eight companies, that first batch we knew right away and we were very confident this is what the future of investing was going to look like because we had seen it the other way. And in the 10 years since then, I think it's really changed. I got involved as a part time partner in YC in 2011 and at that point YC was still fairly small, but even by then the environment had tipped in favor of the founders.

Yeah, I think that what Sam says is really right. I was here in 1995 is when I left and there was nothing you could find out about how to pitch VCs. I didn't know VCs, didn't know investors. By 2005 you had guys like Tom Byers and stuff in Tina select who would at least connect you up. Right.

But yeah, they're sort of. I don't think it was that effective, honestly. I think that if it were, you know, like YC never should have been able to be as successful as we've been able to be. I think the sort of, the last generation of investors just sort of dropped the ball. That's great. I'm happy about it. I think the new world is way better for founders.

But it was kind of a crazy leap of faith you made to fly away from Silicon Valley to Boston. But maybe it didn't feel like that.

No, it was, it felt a little crazy. But like, you know, we needed money and they were giving us money and we needed advice and they were giving us advice. Okay. I was not excited about it. I was actually just in Boston and I went, I sort of walked by the apartment I lived in that summer and I was like, I was so happy not to have ended up living in Boston.

Boston's a fairly nice place, but it's not here.

It's not here. Nowhere else in the world is here. It's really like the, I mean, I think this happens to like a lot of Stanford students. It's really hard to like ever leave. And so we get sucked into this like vortex and we can never move more than 100 miles away from campus.

Yeah, but that's one of the theses, one of the cores of the class. And no surprise given Reid. Right. It's like networks are kind of everything. Networks of information, networks. And so you know Elad Gill, who founded Color and Color Genomics, he wrote a blog post about this, that YC is a network, affects business, which I don't think anybody knew 10 years ago.

I don't think anyone knew 10 years ago. I still don't think people understand how much of the case that's going to be. You know, there's this statistic that people ask me about a lot. There's about 2,500 accelerators in the world. There are 8 billion plus dollar companies that have ever been created in any accelerator. And all eight of those are nyc, which is like not an obvious statistic that that should be the case. And people ask why? And my answer is always that there is actually a network effect at this stage of funding. It doesn't exist at the venture stage. I don't think as much. You know, there are five or ten really great firms, all of which do very well. But we haven't been very fortunate to pretty much dominate our stage of investing. And the difference between the number one and the number two at our stage looks more like the difference between Facebook and my yearbook. I don't even know what the number two is anymore. And I think the reason for that is because it is really a network effect business. The value you get from yc, as much as I'd like to think it's the advice from my partners is the, the sort of the network you get of YC alumni and other companies.

Can you talk about how that plays out? Because I talk with a lot of founders who will ask other founders for help or founders from a couple batches before and it's a little bit like hbs. You used to be an alumni of hbs. You could pick up the phone and anybody else would answer.

One of the Things that makes our stage of investing special is we are the first investor in these companies. And you feel. I remember this as a founder myself, the loyalty you feel towards your investors goes down exponentially with time. So the first people that take a bet on you, you really care about because they took a bet on you. When you were a few people in an idea, your A round investors still bought stock before you, an obvious success. Your C and D round investors, yeah, they had to think you were smart, but they're mostly just investing off of a financial model. And then by the time you get to public shareholders, they don't care about you at all, and they will sell your shares if your earnings missed by a penny that quarter. So because we're the first investor, I think the founders feel a stronger bond with us than subsequent investors and thus are more willing to sort of like, give back to the community. Also, now that that flywheel's up and going, because people help other companies. I don't mean this to be a YC commercial, so maybe we get off this topic soon because people sort of like help other companies during the program so much they feel like they need to pay that on. And, you know, then we do things to help the community. So we do a. We take all our founders camping once a year, and people just hang out and they talk about how things are going. And you have peers, you have other people. It's very lonely to be a founder. So when you have other people you can talk to about, like, man, this is really hard. That's actually really important. It's not obvious how important that is. We have, like, spaces that the founders can come hang out together in, like, office space. We also have online spaces, private social networks for the founders. So we really do try to cultivate this community, and that is what gives us sort of the network effect.

It's not really commercial. You know, the whole point of this class is that Silicon Valley has these big companies that get very big, very fast, a lot by knowing how to step on the gas and how to use the network that's underneath them. And it just occurs to me that YC is a very good example of that. And you get punched on the gas recently in a bunch of different ways. Okay, so one last thing about yc, actually, that's. So one more thing. So you are a founder and CEO of Looped. Yeah. But you were sort of an outsider coming into. Not an outsider exactly, but you're a new partner and you took over as president at.

Well, I had been involved in YC in some sense or other. Like from the very beginning, even before I was like officially a part time partner, I sort of did that as just for fun. Like I would advise companies in subsequent batches, close friends with Paul Graham all the way through, just hang out and help him. And then I became a part time partner officially in 2011. I also. So my company got acquired in 2012 and while I was there I was sort of like looking for something to do on the side. So I raised a venture fund, which was really important actually because I learned that I did not want to be a venture investor. No offense.

No, no, not at all. I don't love the class either, to.

Be honest, but I really hated it. I was like, you know, I feel like I. There's no way I can say this next part without offending you, so I'll just apologize. I felt like I was like a public equity picker. Like I was not actually adding value to the world. I was allocating capital, but there were a lot of other people allocating capital in the same space. And so I was trying to like identify the best companies a little bit sooner and do a little bit better of a sales pitch to get them to take my money. Because the best companies today have like a huge amount of people trying really hard to give them money. And it felt really different than the, you know, little bit of seed investing I had done before. I had much money in 2009, 2010, and this was when I was writing $15,000 checks and the founders had no other options. And if I didn't invest, maybe these companies didn't happen. And I realized that that was actually where I felt like I was having an impact on the world. So I had learned that I did not want to go be a venture investor. I did not want to be a public equities picker. I did not want to be competing to get my dollars into companies that were already going to be successful. I wanted to actually enable new companies. And so as I was thinking about what I was going to do next, I sort of retired for a year and did all of the normal stuff. Then it was, I want to run some sort of company and I want to pick the job that will have maximal impact on the world. And that was how I ended up nyc.

For the record, I kind of agree. I think that venture investors are not as important as early, but I think there's a lot of work that happens that's a little different.

There's a huge amount of work. A great board member, and we tell our Companies this all the time. A great board member in your series, Aaron, is like the most important thing to optimize for. And we at YC ourselves are not good at that. And if you don't get that, it's very hard to build one of these multibillion dollar companies. But that's not what I was good at, nor did I what I really wanted to do. So if I was going to be a later stage investor, I was going to be one of these hands off later stage investors.

And so what has been president of yc. How's it different than when you were founder and CEO of your own thing?

Well, it's not that different. It's much more like being a CEO than being an investor.

Oh yeah, yeah, yeah.

I guess.

I mean, when I was CEO of Mozilla, it was really different than I was CEO of my startup. Because being a founder and CEO and growing the whole thing, it's just a little bit different relationship. I found it was a little bit different relationship with the organization.

You know, there's this weird way in which I. It's very hard to like take over from a founder. In any case, it's very hard to be a new CEO. But YC was very small when I took over. You know, we've maybe tripled the number of partners, maybe more since I joined. And because I had been around a while, like I was literally the first check I was there for the final thing of yc, first check YC ever wrote. And so I had sort of the credibility with the organization. I had the muscle memory that it was not a particularly traumatic, you know, these sort of leadership handoffs in sort of relatively high profile companies are usually much more traumatic than this one was. And I think the reason for that is just because I knew the organization so well, I'd been involved for so long that I was like an internal candidate or whatever.

Yeah, I mean, I remember Paul talking a lot about how you and he were very similar in outlook. He hadn't found anybody who thought quite like you did compared to him. I mean, Reid talks. You know, I came in and took over from Mitchell, but it took a long time to develop that relationship. Reed talks a lot about Jeff taking over him and he credits the idea of the second founding of LinkedIn. In a lot of ways, YC feels different to me now than it did a couple years ago. Before you took over.

Well, I think for successful new CEOs, it has to be a refounding. Every time people talk about that, like it's a rare Thing, I think every time it works 100% of the time, it has to be sort of a refounding of the company. So, yeah, I think we've done some different things. It probably looks more different on the outside than it does on the inside. You know, on the inside, like, we just do the same thing. We sit around every day and we try to figure out how to find the best founders in the world and how to help them and how to build YC into this sort of platform that other companies can sort of get big on top of. And, you know, like, we do that in different ways and we have some big new ways coming, but internally it feels sort of. Because the mission is so clear, because everyone knows why we do what we do inside, and if we change the how we do it a little bit here and there, it actually doesn't feel that different.

That's cool. That's neat. So why don't we. As advertiser, what we talked about is like, Sam sees thousands of startups, sees many more founders, helps hundreds of startups, give or take, pretty roughly. So let's talk about team for a little bit. So what consistent qualities do you see in the best founders and companies?

This we talk about a lot. We actually have tracked this over years and we rate founders on different qualities, and so we see what seems to matter in no particular order. The sort of ones I'd say are incredibly important. Clarity of vision. Can the founder explain exactly what they do and why? If the founder can't explain it clearly to us, then, A, they're never going to be able to recruit, hire, sell, talk to the press, anything. And B, it means they're not the kind of person that is a really clear thinker in general, and that's just so important to a business.

And clarity around what? Cause a lot of people, a lot of times you don't know exactly what you're starting when you start, right?

Well, you know, what you're gonna go work on in the next few weeks and, you know, kind of why you're doing what you're doing. So, like, you know, with Airbnb, for example, when they. When they applied to yc, they were doing this thing that seemed a little bit hokey to all of us, where you stayed on a stranger's airbed while they were still there. And none of us wanted to do that, but they knew exactly how to do that. They could explain exactly how it worked, why they were doing it, and the vision for the company was that we want, like Brian told a story of like, well, this is how my grandfather used to travel. And, you know, it's a lot better to, like, stay in a local's house and with a local than it is to stay in sort of like, you know, the four seasons of wherever, which all look somehow exactly the same with one decor element from whatever the city is. So the vision was so clearly articulated, it then took a little while to evolve into exactly what Airbnb is today, which is where you discover that you can just rent out the whole house, the whole apartment. But he was so clear on why he was going to build this and what the end state vision was, and then the short term, like, what he currently had to do to make it work. So that's one, two is we look for people that are sort of, like, very determined and very passionate about what they're doing. There are founders who don't take no for an answer, and there are founders who just bend the world to their will, and those are the ones we want to fund. And then there are founders that every time they run across a small impediment, just turn around. And fortunately, you run over so many impediments, run into so many impediments every day, that if you're the kind of person that just turns around at those, that's really a problem. This, and you also have to really believe that what you're doing is important. The best companies are always somehow mission oriented, even if it's just that the founder loves the idea of building a big company in this space that he or she knows really well. That element is always there. So determination and passion together, that intersection is really important. Obviously, raw intelligence, that one is obvious, the ability to sort of get things done quickly. This is not quite the same as determination. There are determined people who are still very slow movers, and that works in a lot of things in life. It does not work for founders. It's not entirely accurate to say that speed and quality of decision making correlates exactly with startup success, but it's not like a bad first approximation. Being quick and being decisive and getting things done quickly. If you look at our data, that would just correlate almost exactly with all of our successful founders and other founders that look on paper like they should be really successful but fail often are missing this one trait.

And how do you judge for that?

Well, this is one of the advantages of operating the scale that we do. We look at 20,000 companies a year, something like that. So we just get so much data that you develop over time. The interview questions to Ask. And the responses that correlate with different things.

Yeah. Your determination. Have you read the book the Martian?

It's like eighth in my book stack right now.

It's a pretty good book to read. It's this novel about a guy gets marooned on Mars, but it's all problem solving. It's like, okay, I've got to solve this problem, go do it.

And.

And then the side effect of solving that problem means this other thing comes wildly undone, which is a very startup. It's the best book about startups I've read in a long time. That's the guy on Mars.

The trick to being a great founder, at least in the early stages, and honestly this kind of goes on forever. Is your ability to be presented with a problem unlike anything you've seen before and solve it very quickly. Yeah, that's right. And that, you know, like there are some people that love that and some people that hate that.

You talk about solo founders versus two person teams versus bigger.

We prefer two or three person founding teams, but we are not religiously opposed to solo founders and we funded a number of solo founders that have gone on to be successful. Although in all of those cases they got a co founder at some point after we funded them. So I believe, like Dropbox and Instacart at the moment we said yes to funding them. Didn't have a co founder, but either before our program started or quickly after, both of them did. But we're not like totally opposed. The issue, the two issues, one, there's like a lot of work and it's really hard for only one founder to do it. And two, is that the like the psychological toll of being a sole founder? Even people that think they're really tough manage to underestimate this. So look, it's definitely possible. Although if you made a list of the sort of 10 billion plus dollar tech companies and looked at how many of those didn't have co founders by say six months in, it's a pretty short list. I'd guess. It's definitely worse to have a co founder than to not have a co founder than to have a bad co founder. So what we really hate is when people like bring on a co founder because they think we'll like it, that ends in heartache 100% of the time. Literally 100% of the time. But I think start working on an idea without a co founder but never stop looking for one. If you're a solo founder and the biggest companies seem to generally have had two or three founders, I think Almost always.

I actually can't think of a single one that had a real sole founder. The psychological toll that Sam talks about is real. It's a lonely job, but also it's people telling you you're a moron a bunch of times in a row and you just have to mostly not care. It's a lot of things that are hard, and it's just a little bit easier if you have someone else to talk to.

You know, ideally, when you, like, everyone's going through these, like, ups and downs, sort of this, like, wave, and hopefully with two founders, you're not very often both at the bottom. So one of you sort of like, cheers the other one up. In theory, that's how it's supposed to work.

You'll both hit bottom occasionally.

It does happen. Yeah.

And what do you think about. What's your Dan say about teams of three or four or more?

Three is certainly fine. A number of companies that have had three founders and gone on to be very successful, like Airbnb, which is the most successful company we founded. Four is harder. What happens in practice with four is usually someone ends up leaving in the first couple of years and then it's a team of three or two.

Can we talk about that? So what happens when founders aren't getting along? How do you tell. How do you encourage people to work through disagreements and different point of view?

You know, we all co founders get into big fights at some point. They don't all come tell us about it, but many of them do, and we try to help them work through it. But there have been plenty of times where he said, look, this relationship is done. One of you needs to move on.

Yeah. And then do they. Do you find people are conflict avoidant? Do you find people will run into each other? Is it all different?

Yeah, it all depends. I mean, everyone has their own personality style, so there's no one pattern there.

Yeah. Yeah. Did you have a co founder, Luke?

Yeah, we actually, I mean, we had plenty of fights, but we got along really well.

Yeah, we went through. My co founders and I went through phases where we got along, and then after I left, we went through a phase of a year or two where we just didn't really want to hang out much.

Wow.

We got better after that. It's better now. It's an emotional thing. I mean, you're putting everything into it for a little while. Okay. And then what do you think? So let's talk about diversity for a second. Not at the ecosystem level, but at the founding team level. How do you think about diversity at the first two or three or four or five people?

Yeah, I think that it's a clear win. I think that you should never force, you should never hire a team member just because diversity, but diversity backgrounds, diversity of perspectives, that's really good. The only bad kind of diversity that we have observed in startups is diversity of vision, where founders have a very different belief about the kind of company they want to build. Not so much what the company's going to build, but the kind of company, you know, what kind of culture are we going to have? How are we going to make decisions? How are we going to decide whether we sell or not? How are we going to decide who we hire? How are we going to handle disagreements? So if you have founders that don't agree on that, that is the bad, that is the killer, usually kind of diversity.

I think Keith Rapaway would say, or he asked it on Twitter. He thinks diversity in starting teams is bad. I tend to agree with him because I think if you have differences that are too acute in how you do things or how you process things or how you work, that makes you slower at the beginning instead of faster able to solve problems. Is that match what you think or not?

Well, again, it comes to diversity of vision. If people want different things, if people have a different place for where they want. The two times you get conflict in a startup is when people want different things for the company and when they want the same things for themselves. Those are the two areas where we see the most conflict among the founding team. So if you have, you can have two people that have very like, you know, like a harder charger and sort of like a softer person, and you have conflict about the kind of company you want to build. Or you can have like two people that have exactly the same background. You know, they like grew up in the same environment and they both want to be the CEO and they both want their face on the magazine cover, that conflict is actually the worst one that gets really nasty. So both of those things are bad. What you really want are complementary teams where the people have the same vision for what kind of company they want to build and how they want to get there. And they bring very different skills to the table. So for a long time, the classic co founding team was one very strong business person and one very strong tech person. And that worked really well. A lot of times now it's shifted more towards two really strong tech people. That works a lot of the times it really does. I think that may be better overall. But when Those people want exactly the same thing. Then there's a lot of internal conflict. But again, diversity of skill sets, diversity of background is great. And I think that's worked a lot of times. Steve Jobs and Steve Wozniak, it's true they're both white guys, but within the realm of white guys, very diverse people.

That's right. I think it's interesting to think about co founders. A lot of times what I'll see is two technical people, but one decides to be the product owner and one decides to be the technology owner, and that works out fine, too. It's just, I think understanding what you each like and what you're good at is the thing. And so what about. So say you've got founding teams, they start to put together a demo. They have to get a little bit of momentum, and they say, well, there's really a lot more work than we thought, so we need to go hire somebody. How do you help people figure out how to hire? Because a lot of increasingly people not only have not hired people, they've worked for somebody.

Yeah, look, I think hiring has always been hard and always been sort of like fraud with peril. In the current environment, it is harder and more fraud with peril than ever before. And so there's more and more of a premium on being able to do a lot without hiring. So we have noticed this trend for a long time where our best startups hire the least and our worst startups are the ones that want to say, I have 10 people working for me, I'm really cool. And they go off and hire. But if you have employees, then you have a burn rate. And in today's salary market, it's a very high burn rate. You have more organizational inertia. It gets really hard to turn the battleship when you have a lot of employees. If it's just like the two or three co founders, you say, hey, this thing's not working. We're going to do this other thing. Okay. You have internal politics, you have people thinking about their career trajectories. You need all these things once the company is really ready to scale, before the company is really ready to scale. This just creates this, like, there's this potential energy in the company of these smart people that want to do something and they're not exactly sure what to do, and you don't have enough for them to really run at full speed, and it just causes problems. So Airbnb, I think they took like nine months to hire their first person. Stripe took about that long, too. Dropbox Took a very, very long time. I think companies always want to hire because they think it'll make them more effective and they think it'll like give them status. It does give you status, but it definitely makes you less effective. So we really like the companies that figure out how long they can wait to hire their first non founder employee.

It can be agonizing because you've got only a couple people and you're trying to do work and it feels like you're making very, very slow progress. And you just said, well, if we just multiply this times five or six, things will go faster. But a lot of times especially it comes back to what Sam was talking about around clarity of vision. Like it takes you a little while to figure out how to really be clear on some of the important things. And if you hire people before you clear, you get eaten up on communication effects and like the number of interconnections and like Sam said, people wanting different things. Right. That's interesting. So and so given that you flew to Boston the first time.

Yeah.

How do you think about Bay Area versus other places now?

I think the Bay Area is the best place in the world to start a startup and I would definitely start a startup here. The costs are the one thing that could break that.

You mean just dollar costs?

The dollar costs of living here. When you're a startup and you're trying to survive as many months as you can, because if you have a linear amount of money and an exponential growth of users and revenue each extra month, you can survive a long time. And so each 20% increase in the cost of living is really bad. I think it's the best place to start startups. You know, you, since you mentioned sort of diversity earlier, one thing that has been odd to me is that the Bay Area feels to have gotten less diverse in terms of founding teams in this classroom, 15 women, maybe like two black people, like that is horrible. That's like a shame. And this is the startup class and Stanford is much more diverse than the people in this room. And I don't know what it is about the Bay Area that has gotten, that feels like it's gotten less diverse relative to the rest of the world. But you know, when we, because we spend a huge amount of time traveling, we fund people from all over the world. More than half the people we fund are not born in the US and I think there were 40 something countries represented last time. The rest of the world is certainly more diverse than the Bay Area when it comes to startups.

Yeah, I think so. Too. I think we've gone backwards. I'm not quite sure what to make of it. We were a little surprised in this class. On the top of the funnel, we just didn't see a lot of gender diversity at all. And so that makes me think that next time we do it, we need to really focus on sourcing more, which is obvious. We know that in our startups and it just kind of seems crazy that we didn't just apply it here. But I think that's one of the problems. I suspect we have a bunch of nationality spread here, but certainly not gender spread. Yeah, it's certainly a problem. Not sure what else to say about it. You guys have done some things to.

Look, I think that everyone. Diversity is actually not as hard as everyone says it is. You know, less than half of our partnership is white. White guys, great. You know, that means we'll fund really diverse people, too. We fund, like, we just track. We make sure that we reach out to as broad of a set of people as we can, and then we make sure that we fund them plus or minus the rate they apply. And so I think everyone. Well, we're both white guys, so I'll say this. I think white guys say the problem is much harder to solve than it actually is. I think it's just. I think, yeah, we can solve it.

Yep, yep, we're definitely white guys. All right, these are interviewing skills, Charlie Rose level. So this is like class number two. So we're still working through it. So let's talk about idea for a little while. I mean, I would talk about team forever, but let's talk about idea. So maybe PG says the key is to find good ideas that actually seem bad.

Could you. Yeah, this is actually a Peter Thiel ism, not a PG ism, but I think it is the central. Like, if you think that what you're looking for are undervalued assets, then it obviously makes sense that you want the really good ones that look to most people like they're bad, because the really good ones, everyone will be going after and they'll be super competitive. And the really bad ones, obviously you just don't want to fund. So this idea of great ideas that look bad certainly has what has led to outsized returns, investing in startups and starting startups. And it's harder to do than you think. I think we're all much more affected by trends and what other people like than we'd like to believe. Leave. So it's really hard to have a lot of conviction in an unpopular idea as a founder or an investor. It gets easier with time on the investing side because you realize that's where you make your money. And also what people say about you being stupid, you don't care about it as long as you keep winning. But it is the temptation. I like to think of myself as an independent thinker and I still find the temptation to just like whatever other people like very strong. So it is really hard to do this. You don't want to be an investor that's investing in a derivative of the thing that was just really big. So in 2006, 7 and 8, most investors were trying to invest in the next Facebook. And then in the 2009, 2010, they really got crazy about trying to invest in the next Facebook. In 2009, 2010, the two companies that mattered to invest in as a venture investor were Uber and Airbnb. These look nothing like Facebook when Facebook was raising money in 2004 or 5, whatever that was 2004, I guess most in Google had just gone public. It was about to go public and there was a huge amount of focus on ad tech businesses and even some search businesses. And that was exactly the wrong thing to be invested in. All of those companies failed. And what you should have been investing in was Facebook, which looked nothing like the giant winner before it. Now a lot of investors are rushing to invest in Uber. Like things. This idea that I push a button and something happens. So everyone's trying to say, well, we're going to be the Uber for food, we're going to be the Uber for like dog walking.

That's real.

And the next Uber will be a $100 billion company. But the next hundred billion dollar company is very unlikely to be any of these thin verticals on top of Uber. It'll be something totally new. So while everyone else is starting those, you should ignore all of those ideas. They're going to be brutally competed away from you and they have very, very bad economics anyway. And what you should instead do is say, hey, what is the next hundred billion dollar company that has nothing to do with the last. What is the shift in the world? What, what has shifted in the world to enable a giant new company that didn't exist a while ago and then go start that company. And look, here's the hard part of this. It's going to look bad. It's going to look bad. So you have to have internal conviction that it actually is good. And a lot of people are going to tell you, well, why don't you just do an on Demand company, because that's what everyone's doing. And you have to just have the courage of your convictions to keep doing this unpopular thing because you understand where the world is going in a way that other people don't.

And how do you do that? How do you develop internal conviction?

Well, there are a few tricks for this we've learned. One is that once you start, once something really starts to grow and you can tell that users really love your product, it gets easier to sort of not worry about what the haters say. And so the quicker you can launch and the quicker you can have a set of users that become dependent on your product, the better you'll feel. So Uber, once they launched, a lot of people made fun of them, but a few people used it all the time, even when it was just like the expensive black cars. And so while the rest of the world's making fun of them, they knew they were onto something because they were growing like 10x every year or whatever. It was more than that, Definitely more than that. And so I think once one of the advantages of launching quickly is that it will give you conviction even when the rest of the world is telling you this thing is like, bad. So that is important to do. I think another thing is if the reason that you have conviction is a shift in the world that people who have been doing the same thing for 20 or 30 years have missed, and you can identify precisely what that was. Go back to the Uber example. This idea that everyone had an Internet connected device in their pocket and no one did three years ago was an obviously huge change, that unless you were really paying attention, you could have missed how important that was. I think if you have, if you can rationalize, if you can explain to yourself why you have this conviction because something in the world has changed, that helps.

Can I tease apart something real quick? So you said, you know, launch as quickly as possible. As quickly as possible is like a complicated phrase because to diverge for a second, you're on the board of two companies building fusion reactors. Nuclear reactors, sorry, nuclear. Yeah, they're going to take a little while longer to launch than.

Well, they're gonna take a while to launch their final product. But they break it up into small internal products that are measured very precisely and they know exactly what they're trying to improve. 10% every week. When we started doing the hard tech companies at Y Combinator, a lot of people. This is, again, this is a good courage of your conviction story. The prevailing wisdom was that we were gonna be horrible at this because we didn't understand these companies, and they take years, and they need a different cycle than the software companies. And what we thought is that actually what hasn't worked with these companies is they don't look enough like Internet startups. They actually need to operate with the cadence and speed of an Airbnb or a Reddit or whatever. And that was the problem. And so we brought them into our same program. And no one likes to feel like they're losing. If you're around companies that are measuring something and it's going up 10% every week, after a little while, you're like, well, fuck, my life sucks. I'm going to figure out something that I can measure and I'm going to go faster. And that works. So we have to work hard with the hard tech companies to figure out what their core metric is and what they can get 10% better at each week. But a company has got to get better at something 10% each week in the early days. And if you compound, if you can have this product cycle where whatever you're building gets 10% better every week, and you project that out, you do that for three or four years, and then you're like a really big company, that'll.

Be a lesson to carry. Compounding interest is pretty good. 10% a week is a lot. But how will you know whether this is right or not? So your thesis about building hard things make them more like software companies. It's going to take a little while.

To figure out if it feels like it's really working. We won't know for sure for five years. Look, the Fusion company works. No one ever says anything again about whether or not YC can invest in hard tech companies. But we do need some of them to work right.

So you just won't know for a little while. Okay, can we talk about. We talked about a bunch of these. Somebody in the class asked, who can you safely ignore when they tell you your idea is bad? Who can you not ignore? Truth is, you can ignore everybody.

If you can ignore everybody except users. If your users are telling you your product sucks and you can't find any other users that think your product is good, that is the constituency to listen to.

Yep, that's a good answer. And so what about startups who pivot? How do you know? How do you decide when to do that?

Look, I think Silicon Valley really gets the pivot thing wrong. I think it's like, I try never to go to startup parties, but every once in a while I forget why I don't go to Them. And I go, and the last one I was at, I was really struck by how much people were like, bragging about pivoting. This is basically bragging about failing. And I think the attitude towards failure should be tolerant. But, hey, that sucks. Let's not do that again. And I do think we sort of over glorify failure. We try to learn too much from failure. We talk too much about how great failure is. Failure is still bad. It's important that it's tolerated. It's important that it's not held against you. But it's certainly not something to go out of your way and celebrate. And I think the pivot culture sort of a little bit is like, people brag about how many times they pivoted. The only kind of pivots that I have seen work are the following. One, it's what the founder wanted to build all along anyway. And two, it's what the founders discovered while they were building their first thing that didn't work. So, for example, I'll give one for each Instagram. Kevin Systrom had done this other thing called Bourbon. It was not working. He went away on vacation and built the thing he actually wanted to build. He loved photography, but at this point, the iPhone cameras were not very good. And so you needed filters to make them look intentionally bad instead of just bad, because the sensor was bad. So if you make it look like this old cool look, that's okay. The sensor doesn't need to be perfect. It was brilliant. So that was what he actually wanted to build, and that was what he's actually passionate about. Those kind of pivots work. The other kind of pivots that work are the Airbnb founders, for example, were building this other startup totally out of money, not working at all. No one would give them money depth of the financial crisis. They actually had maxed out so many credit cards. You know those little, like, plastic sheets you put in binders that hold like nine baseball cards. They were using those to hold their credit cards. They had to max out and keep track of all of them. That's how out of money they were. And their rent was due and they didn't have any money. And they thought and thought and thought. And the only thing that Brian had was a little bit of extra space in his apartment. He had some number of dollars in his bank. Rent was more dollars than that, and it was due in 10 days. So he rented out his spare room for the amount of money of that delta. And, you know, like, the crazy thing of all this is the Airbnb started as an affordable housing company. So in the current, like San Francisco, little piece of history, but I think that kind of pivot works as well. Other than those two classes of pivots. The normal class of pivot is where founders sit down, they say, our startup is not working. Let's sit in front of a whiteboard and figure out our new thing. I have never ever seen that work and I have a lot of reasons, I think, why if you're in a rush to think of an idea, you're less likely to come up with a good one. If you already have investors, you have to have a plausible sounding idea to go back to them with. It's okay to have a bad idea, a good idea that sounds bad to start with. It's harder to pivot into such an idea. So people try to find these ideas that sound reasonable on the pivots.

But yeah, I would say that that investor thing I think is real. I think investors do put there's some sort of weird social contract where you feel like you have to go put together a plan that works on a deck and that kind of stuff. I think the more thoughtful investors actually would prefer you just to say, I just don't know what to do than to say, here's a shitty idea that I'm going to keep working on. Even if you can wrap it up in a way that looks nice.

What I tell founders when they're thinking about founders, to have a startup that's not working unless you have something, a pivot in one of those two categories that I just mentioned, shut the company down, return the remaining capital to investors, go travel the world and wait until you have an organic new idea you want to work on.

Slack is probably the most famous pivot lately. And Slack started as a game company. They made glitch and what they realized is that nobody liked their game, but their internal developers really liked using this IRC client they had built that became Slack. Right.

So when you build something that you discover other people need along the way, as one of these stripe was even a version of this. Patrick had been building this encyclopedia app, this iPhone app, and realized he couldn't take payments. So sometimes you discover organically this other thing that people like better.

Yeah, internal tooling is often a pattern like that. And then there's a great essay you shall also read by Joe Krause, who started Excite and Jot Spot and a couple he wrote about. Basically, people say they like learning from failure, but learning from failure is kind of stupid. Learning from success is the thing. But like Sam says, you should tolerate failure. But there's not a lot to learn from failure. There's a lot to learn from being in a successful. You learn better, you learn more good patterns.

Well, there's a lot to learn from failure. I'd say there are more important things to learn from success.

Yeah, that's fair.

Here's my one piece of career advice. If you're going to go join a company, go join the most successful company growing as quickly as possible that you can find. Because that is how you will the lessons from seeing a really successful company for a few years sort of scale. I think that is the only training, that's the only pre startup training is actually valuable.

I agree with that 100% also. There's so much to do and there's so many holes that you get put in places of responsibility that you're not nearly qualified for. But it's the only way to really, really learn. Come on products, you say you know, a lot of what you write about, not what PG writes about, is like focus on product, product, product, product, market fit. What does focus on product mean to you?

I mean you have, you know, maybe if you're really good and really dedicated, you can work like 70 really productive hours in a week. And so you have 70 hours to allocate. And there's a question of how you allocate those. And when we say focused on product, it means that you put almost all of those into talking to users and creating product and making sure that you have some users that really love your product. Most startups face a choice in the early days. Either I'm going to get a lot of users that sort of are somewhat ambivalent, they like me, but they don't love me. Or I have a small number of users that really love me. Every really successful company I've been involved with has been in that second category. So focus on product means focus on love, not like. And spend your time writing code and talking to users. You need to get the product to the point where some people would be really, really bummed if it went away.

Yep. So this is going to be a funny question because I think YC answers change over time. So for enterprise companies, how do you help them think about how to get their first customer, their first 10 customers? Like with YC, what happens now is that a lot of YC companies become customers. Yeah, but if you don't have that network, how do you think about it?

Well, I spent all my time working with YC companies, so I Have a really easy answer for the enterprise sales companies because they always do get their customers among the YC community.

That's something that's shockingly uniform now.

But I don't really know what you do without that. Honestly, I just don't advise non YC companies anymore.

Yep, fair. Okay, we'll keep moving. Means I'll have to answer later in the quarter.

I think it's really hard to do that without a built in network of some sort.

It is. I mean in Greylock we started a lot of things from scratch and we help introduce people to. You have to find sort of an enlightened customer who understands what it is to work with a startup. YC companies kind of by definition understand that, but I think it's almost always network based no matter which network you use. Okay, let's talk about something else. So like you wrote a little while ago about the post YC slump.

Yeah.

You get the demo day and then things change. Can you talk about that? And this is. It's a broader. The reason I brought it up is it's a broader phenomenon. It's not just yc.

So startups work really well when the founders have the gas pedal all the way down to the floor and you know, yes, a lot of internal things are broken, but you just keep growing and you keep winning. That feels good. And then the hard part is at some point you have to go clean up the technical debt, the organizational debt, the cultural debt, whatever. And founders always think, okay, I'm just going to stop worrying about growth entirely and fix these other things. And in theory that sounds like it should work, that you can focus on one thing or the other. But the issue is if you take the focus off growth, this. A board member of mine used to say that sales solve all problems and then later he modified it to growth solves all problems. It's not entirely true, but it is the tail end of that is really valuable. And so if you stop worrying about that to go fix other problems. The issue is as soon as you stop growing, you get a lot more problems. And so you have to figure out a way where you maybe take 10% of your bandwidth off of worrying about growth to work it to solve these other problems, but not 50% or certainly at 100% because this is another reason why having employees is a problem. Employees can sense a change in momentum in the startup faster than the founders can usually. And you just like the prime directive of running a startup is never lose momentum. And it's so easy to do that. It's tough. When you start a startup, you're signing up, you know, you are signing up to like work all out for 10 years. And that's just really, that's really hard. Like, you never, you know, you can like go on vacation, but you never get to stop thinking about your company. You can, you know, like go hang out with your friends, but you never get to stop thinking about your company. There are new crises at every stage and they only get worse with time. So it's just like that's really hard to sign up for.

One of the other ideas you put in that post was the study of fake work.

Yeah.

Can you talk about that?

The short version of this is like if you imagine a movie about startups, actually we can do a more real example. If you watch there's a reality TV show going on about startups right now. It's horrible. But if you think about the sorts of things that would make it into a reality TV show about startups, the things that would make for good tv, do none of those. And do the kind of work that would get cut on the editing room floor, which is sitting in front of your computer or at your lab bench building your product, going off and talking to your users and saying, how can I make this better? And that sort of work makes for a very boring tv. It is very boring to watch someone Write code for 12 hours in a day. It is very boring to watch user interviews. It's very boring to watch sales meetings. It is much more fun to watch whatever they put in these movies and these reality TV shows. But that is not actually how you win. It may make it feel like you're winning. It's very fun. Some founders find that they really love fundraising because it's sort of like you get to manipulate people. It's very high stakes, feels important. But fundraising, while sometimes necessary, like, no company has ever become great because they were great at fundraising. At some point, you either build a great product that users really love and you figure out a way to grow, or you don't. And many founders that we come across that we don't end up working with, they're great at everything except this. And they spend all their time on everything except this. So they choose the best lawyer in the world. They have a beautiful logo, but they never make something that people love. And you can get everything else right. And you can fail at that. And you will fail.

Yeah. When I made the point last week or Tuesday that Reed is better than anybody I know at triaging, I mean, he has an above the Line and below the line. He does not give a lot of stuff below the line. It's just above the line. And he'll pay attention to all of it all the time. But below the line just doesn't matter. So you have to learn how to let fires burn a little bit.

Yeah, you need to trade off getting the big things really right and be willing to let all of the little things be a little bit wrong.

You have a question?

I do. Could you talk a little bit about these distractions? What pulls founders away from growth and product? Different founders fall in love with different areas. So hard to give a general answer. The visible example that you can see as an outsider is founders that fall in love with their public image in the press, people that really care about getting on the 30 under 30 lists or going to the conferences or whatever. So that's a big thing. Another actually is just like going to networking events. You sort of somehow think in the abstract sense these are going to help you. I fell victim to this briefly because all investors have these CEO conferences and they're like, oh, you got to come. It's great. It's going to be really fun. You're going to meet the other portfolio CEOs, this, that, the other. OK. And then you get invited to Slum Valley and it sounds really great. It is cool. And you're like, oh, it's only five days, I'll go do it. And then you talk to these founders and you're like, you were out of the office more days than you were in the office at networking events. And when you say it to them, it's obvious then that, like, oh, yeah, that probably isn't great, but it just sort of. You get there gradually and you get away from the things that actually matter. So that's a common one.

What are some of the things you see your best founders screw up frequently?

The number one mistake that great founders make and everyone in this room that becomes a great founder is going to make this mistake on. I'm just going to mention it. I think it's unavoidable. Is waiting too long to fire bad people. I don't actually think this is a teachable lesson. I think you just have to make the mistake and learn it. But it is the universal biggest mistake that otherwise really great founders make.

And why do you think it takes? Why does that happen?

I mean, firing people is like the worst part of running a company. So it's just like, it sucks, right? You made a mistake and now you have to really negatively impact someone's Life. So I'd actually worry if there was someone that had an easy time firing people.

Yeah. My reflection, I was slow to fire people in both contexts where I was CEO is that you start saying, well, this person's not succeeding, but maybe it's my fault. So maybe I made a bad call on the hire, but now it's my responsibility to fix it. And so maybe if I invest a little more, if I help him a little more, or maybe if I create a different context, he can fix it. I think that good founders tend to. To think they can fix things. People, issues like that are tough.

The one thing I've learned that may be helpful here is that you actually hurt the employee more by keeping them in a position for longer. You know, if you fire someone two or three months in, they never put it on their resume. It's just like it's a null. That never happened. You know, if they work there for like a year and don't excel and then you fire them, it actually, it hurts them more.

Yep. And they know too, when they're not succeeding. They know it. Let's shift over for just a second. So machine intelligence.

Yeah.

So Sam's very, very outspoken. Everybody should read Sam's blog. It's very, very interesting and broad.

That is actually not true. My honest advice, it was started as a thing for me to like, practice writing.

You should read Stanford's blog. You've been writing about regulating and sort of directing AI, but I don't want to talk about that so much as. What do you think the interesting opportunities are around machine learning?

Well, I think in the short term there are all these vertical machine learning companies that are doing an incredible job at optimizing far better than humans can, some particular industry. And I think now you're seeing this get extended to robotics, computer security, trading, lots of other things. So I think we will have a 10 years in the sun of wonderful domain wins, where machine learning in a particular vertical has great results. And then someday someone is going to discover the algorithm behind actual intelligence and creativity, and that's going to go one of two ways.

And so you talk about vertical AI. I'm going to try to avoid that. I think that's a good thing to talk about, but I think not as germane to the class. Do you think it's all vertical AI? Do you think there are horizontal things that people should work on now? There are horizontal opportunities, yeah.

I mean, people, like a lot of people are trying to develop the first sort of superhuman general intelligence. So, yes, I think that's a cool thing to work on, hopefully. Yeah, I do think people should work on that.

Okay, let's talk about areas beyond just consumer enterprise that you're interested in. One that you're not, you don't seem particularly interested in is bitcoin. You want to talk about what you think about bitcoin?

I don't actually know where this myth that I'm a bitcoin skeptic came from.

Well, I just think you said. I think you've said a lot of places you haven't seen. What have you said?

Well, the thing I get in trouble for is whenever I'm asked publicly if people should buy bitcoin, I say no, which has been the right prediction. It's gone down after every time I've said that someday I believe it could work. And if I think so, then I will change my advice on whether or not you should speculate on bitcoin. But I think that if we go back to this metric of you need some users that love something. How many people in this class have ever bought something with bitcoin at all? And how many of you have done it more than five times? And how many of you have done it as a daily habit at this point? That is the entire problem with bitcoin. Yeah.

No compelling use for reason.

People don't love it. It does not become part of people's daily lives in the way that the really successful companies always have to. It may, it might. I hope it does. I think it's cool. And I believe that, you know, I believe that it has the makings of something that could be huge. And I think it's interesting for people to work on. What I have said is I wouldn't go buy a bunch of bitcoin right now because I don't think it has yet found product market fit. And I also don't invest in startups that I don't think have yet found product market fit beyond the YC stage. So that's all. I just think it needs users who really love it. Right now, the only people that really love bitcoin are the people that are expecting to get rich on it because they bought their coins early.

I'm not sure that's quite right. I think there are a lot of computer scientists who think it's a pretty.

Amazing blockchain, is an amazing technology, but that this idea, like even those people don't say, well, we're going to buy everything with bitcoin. They just say that this idea of a distributed ledger is awesome, which that I 100% agree with me too.

Yep. Yeah. Bitcoin as a commodity is a little bit terrifying right now. You want to talk about Helion and upar?

Sure.

Why you, why you got involved? Those are the only two committees you're the chairman of, right?

Yeah, yeah. Well, I hate being on boards, so I try not to do that.

It's a lot of work and a little bit boring.

I got trapped on the Reddit board, so I'm also on that one. But that was not intentional. I think if you could pick one thing to do that would help the poorest half of the world the most tomorrow, you would pick friendly, superhuman machine intelligence, and that would be choice one. I think the second thing that you could do would be really cheap, safe, clean energy. I think if you look historically, the quality of life and the cost of energy correlate almost perfectly through time. And every time the cost of energy comes down a lot, the quality of life for people, especially like the poorest people in the world, it really goes up a lot. So I think the problem of solving energy is so important. And this leaves out all the climate change issues, all of the wars, but basically I just made a list once of the problems that I cared about and I thought about the underlying issue. And you can do these things like health, economics, like war. War is basically get fought over energy. We say they get fought over other reasons, but they get fought over energy. And so it's just like if you think about the optimist, people need sort of, or I need this optimistic version of the future and a version of the future where energy is a penny a kilowatt hour instead of 8 cents a kilowatt hour, I think really transforms a lot of these things that I care about.

So you're involved in both these companies.

And they're both, both these companies. One is a fusion company, one is a fission company. I think that there's already a lot of good stuff happening in solar. My view of the world is that it will be 80% and 20% of Earth based nuclear and solar, I don't know which is which, but there's already enough people working on solar and almost no one working on the nuclear side. So I think I thought it was more important for me to focus there.

Yeah, the cost curves on solar seem pretty up pretty good.

Yeah.

Okay, so I think you have to bump at 5:30. Is that what you said?

I can do a few minutes after.

Okay, well, so I have one more question and then maybe we'll open up to a few more Questions in the class. So my thing is, when you turned 30 a few months ago, you wrote a thing and you talked about the days are long but the years are short, which is my favorite saying. But I didn't find that saying until I had kids. So I'm recording now. Time speeds up even more when you have kids, but I believe that given the days are long and the years are short. How do you think? What did you mean? And then how do you think about your time and urgency now?

Yeah, look, I think, you know, when I was 19 and when I dropped out of college, I sort of felt like I had infinite time. And the only kind of. And someone said to me when I was maybe like 23, that the intensity that you need to work at a startup, you can't keep that up forever. Maybe you have 20 years of your life where you can work as intensely as you need to be a great startup CEO. And you can do that from 22 to 42, you can do that from 28 to 48, whatever. But if you believe that, then you have a limited number of shots on goal. And the biggest thing to think about then is just opportunity cost. If I'm going to start this startup, it's going to take two years if it fails, three years if it fails, or 10 years if it's successful. And I have what is actually an incredibly short window of how long I can be a great startup CEO. I think I actually worked too hard on my first startup and got too burned out. And thus the thought of doing another one now is just sort of terrifying to me. But I think the mistake is to think of life as 100 years long and to think about when you're thinking about what to work on and think that I have this 20 or 30 year period in which I will concentrate a huge amount of the output of my professional life. And that's actually not very long. That's a small number of startups. And, you know, what do I believe in the most? What do I think is the most important way I contribute to the world that I'm gonna find most satisfying and how to spend my time on that.

Yep, that's excellent, excellent advice. Questions for Sam. You can have a bunch. So why don't we start back here and you just need to tell me.

When you need to go.

Sure.

How do you fire people, especially if they're your friends? How do you fire people, especially if they're your friends? I think the most important thing about this is, like, put yourself in the other person's shoes. Treat them like a human and do anything you possibly can to make it look like a win for them. So, you know, I never. I never. Like, every person I've ever fired, I have found them a job before I fired. Like, I sat down with them, I told them we were going to do this, and then I found them a job before they left my company. So when they left, they announced that they were resigning, and they announced their new job, and it looked like that. And, you know, we would celebrate them on the way out again. I feel like if I hired the person, then I screwed up, even if they didn't work out. And so, you know, you sit them down and you say, like, look, this feels terrible. This is honestly the worst part of my job. This just isn't working. And I don't think it's in the interest for you to be here either. I really appreciate you as a person. If you don't, if they've done something unethical, then it's a different scenario. But in most of these cases, it's like, I really appreciate you as a person. I really value our friendship. I know this is going to be rocky for us, but I'm sure years from now we'll laugh about this and be close friends again. But this isn't working. And almost always they also will think it's not working. And then you say, but look, I want to do this. I realize this sucks. I realize this upends your life. I want to do this in the best possible way for you. So, you know, A, I'm going to be super generous with severance, and I always am. B, I'm going to help you find the job. And C, either you keep working here while you look for your job, or if it's really not working, you know, you go work from home for a while, but you still act like we still say you're here until you're ready on your terms, to announce that you leaving and what you're doing. And what I have found is if you just treat people with respect and you make it clear that you're firm on the decision, it's definitely happening. But you're going to do anything you can to make it smooth for them and be respectful to them, they'll work with you and they'll say, okay, let's figure out how to do this. And I think people really appreciate that.

What Sam said at the end about being firm on your decision is super important. Because if you come in saying, I'm not sure, sure I'm thinking about this it turns into a pretty complicated power dynamic very, very quickly, which almost never goes well. What Sam says, if you're clear and you're firm, then they have to process very quickly. They're in a new situation, and then you are constructive together.

The advice in the management books is that you have their last paycheck ready to go, and you have their termination letter. You have a bunch of legal documents for them to sign that you present in a meeting. And if you do that, that actually is horrible advice. I think if you do that, it's what the lawyers tell you. What the lawyers tell you is usually bad advice, but it's particularly bad in this case because it makes the person feel like it's a very adversarial thing and it's happening on your terms. What you want them to think is that this is going to happen. I have decided this is going to happen. But I want to hear the terms you want it to happen on, and I want to hear how we sort of position this to the company, to your next job, to everything else in the best way for you. And if you sit down and say, and you're firm, that this is going to happen. But you talk about how can I help you have this happen in the best way. We'll sort out the paperwork later. It just makes the whole thing feel not adversarial. Sure. The last guys we talked about, kind of the specialist going from generalist to specialist. How have you seen your founders adapt the transition as they get to different.

Levels of organizational structure?

I think that the best founders are generalists all the way through. I think that actually, like, you know, maybe you're like a specialist in a particular technology that you develop. But as soon as you become a company, as soon as you're no longer in product development, when you transition from building a product to building a company, you got to specialize in generalization starting that day and never look back.

Okay, so you talked a lot about.

You wanting to make an impact on the world. And so I'm wondering how, through things that you do individually and in your work at yc, we can work towards sort of changing this culture of tech from valuing things like Uber for pizza or Snapchat for puppies, to things that are actually going to make a positive difference for people outside of Silicon Valley. I always think the best way to do these things is just to go off and do them. You know, if we want to fund more companies like that, we can either go talk about it and talk about how do we do this, or we can just say, you know what, we're doing this and fund a lot more companies that do that. And we've just been doing that. The bad part is they don't get nearly as much press. So externally to yc, you don't see it as much, but we're super fortunate. All the partners at yc, we made plenty of money. So at this point, we want to work with the things we want to spend our time on. And so the answer of how we fund more companies like that is just that we fund more companies like that and we just say, this is important to us and we're going to do it. The good news is it's worked phenomenally well. It's worked over the last couple of years, like, far better than I've ever imagined. Two years ago we funded zero biotech companies. I think now we funded like 30. We have funded now companies that target their products to every area of the world, different people. And again, I think the answer is just say, okay, we're going to do this.

Okay, maybe two more questions in the back.

I return you to talk about the mistakes or distractions. There was a really widely read article by the deal professor in the New York Times about how liquidation preferences might be pushing up valuations, where it's at the detriment of founders and employees. Given you $8 billion companies at AYC, you're 59 across the world, all of.

Which are in the Valley.

Do you see fundraising becoming a distraction? Do you have a council? I know you're early stage, but do you ever counsel your startups and not take the money if you think it's getting too hot? We do all the time. We always tell startups not to go after the highest deal and sometimes we say, don't raise the money at all. You know, the only time that I think you need to consider whether a startup is overpriced is when you are buying equity in that startup. And that happens two times. One, when you're investing and two, when you're going to join as an employee and getting stock options. So that's when I really think about how much a startup is actually worth. Other than that, I think the valuation numbers are complete fictions, totally irrelevant, and I just don't pay any attention. The intermediate valuations mean nothing. All I care about is how much the equity is worth once I can trade it, which hopefully is not for many, many years. It's frustrating because, you know, the press, like all that I ever get asked from journalists is like, oh, how much are The YC companies worth. And what I say is like, I don't know, it depends what they're worth. When they exit the intermediate numbers meaningless, they're like, well, I need a number. So people get very focused on this. But I think you should just ignore it. And the important time to ignore it when you're going to work at that startup that's worth more than a billion dollars in your head, don't believe it's worth a billion dollars because yes, someone did invest with this liquidation preference that makes the sticker valuation meaningless. So what I think about these numbers is they don't mean anything. As an investor, I probably won't invest at most of these prices as an employee. You should not let the founders of that company or whoever's hiring you at that company convince you that your equity is worth x percent of the headline valuation, because it's not. And that's when it's important to think about. And then you really have to determine like the question of when you're joining a startup is not how much is the startup worth today, it's how much do I believe the startup will be worth in five years and what is my ownership percentage of that, but totally throughout the intermediate valuation.

Okay, one more question.

For those of us in this room who care about having a really big impact and a good impact on the world, would you recognize, would you recommend that we actually start a startup to do that or join a successful company? Well, not for the sake of starting a startup. I wouldn't recommend that. No matter what. If you have an idea that you think will really matter to the world and you think a startup is the best way to get that idea out into the world, which I think it often is, then you should start that startup. If you're starting a startup for the sake of starting a startup, that will usually end in heartache and I wouldn't do that. But I think joining a company that you really believe in and that is early enough you can really contribute to, I think the expected value of that is higher than starting a startup because so few startups go on to be really successful. Dustin Moskovitz gave a talk in my class last year about this very topic and I refer people to it all the time. I think he actually has a really great framework for thinking through that particular trade off.

That was the one in the first class.

One in the first class, yeah.

So we should.

We need one more.

One more question.

What difference do you see in founder who spend more time in school? Sorry, what difference do I do you see in founders who spend more time in school? Oh, what difference do I see in founders who spend more time in school? I don't think it actually correlates much either way. We've had success with founders that didn't go to college at all, that dropped out, that got PhDs. I think if anything, like, the extremes of, like, anything are bad. And so people that go to like, you know, 14 years of school, maybe those people aren't just like fast movers enough and people that don't go to any school, I think, like, I would have missed out on stuff had I not done some college, but within the kind of normal bounds. I don't think there's any big difference.

There's a lot of difference in how you want to live your life and how you got to think about what kind of person you want to be. Startups, I think it's not. I agree. It's not correlated at all. Anything else you want to say?

Thanks for having me.

You've been awesome. Yeah.