{"url": "https://www.youtube.com/watch?v=y-otm4RJHyc", "page_url": "https://www.youtube.com/watch?v=y-otm4RJHyc", "title": "What It\u2019s Really Like to Start a Startup - Sam Altman, CEO of Open AI | WashU Engineering", "show_or_channel": "WashU Engineering", "date": "2016-02-15", "duration_s": 3576, "type": "interview", "speakers": ["Sam Altman"], "confidence": "high", "notes": "WashU Engineering", "people": ["sam-altman"], "id": "2016-02-15_washu-engineering-what-it-s-really-like-to-start-a-startup", "audio_duration_s": 3576.4186875, "worker": "jc-w20", "transcribe_s": 143.6, "model": "large-v3", "segments": [{"start": 1.68, "end": 28.22, "text": "This is terrific. I'm glad so many of you can make it. My only job is to quickly welcome you folks to engineering. I'm Aaron Bobick. I'm the dean of engineering. How many people here are not engineering students? Well, welcome to the cool part of campus. That's great. It's great to have folks getting down here. And the only other thing that I'll say is one of the reasons I'm so thrilled about this event is we've been pushing..."}, {"start": 28.22, "end": 51.46, "text": "Engineering to help get much more connected to the notion of entrepreneurial activity in general. And sometimes that involves starting companies. Sometimes it just means students starting to entertain the idea that working at a company that has fewer than 1,000 employees is OK, maybe fewer than 100, maybe fewer than six. And that there's a total continuum spectrum between, you know,"}, {"start": 51.46, "end": 79.82, "text": "Drop out after your first month and start a company. Not recommended. To getting engaged with, where's Steve? Steve Lindemann, you're right here. Getting engaged with Idea Labs, where you start doing and understanding what that is. Up to one of our colleagues took a job with a regular company. I think it was Raytheon, maybe? And decided he didn't like that. And he joined this little company. It only had eight employees at the time. And it had this funny name called Google. So he was the eighth employee there. And that turned out to be, shall we say, a good idea."}, {"start": 79.82, "end": 112.69, "text": "So understanding where you sit on the spectrum, I think, is a great thing. So without further ado, I'll introduce a man that you all are beholden to. That was a joke. Beholden Thorpe. Right. Anyway, Holden Thorpe, the provost. Welcome, everyone. Great to see you here. Thanks, Dean Bobick, for hosting this. And thanks to Sam Altman and Dave Messina for coming to talk to us. It's great to see such a huge and enthusiastic crowd."}, {"start": 112.69, "end": 138.3, "text": "Given all the work that the university is trying to do to enhance interest in innovation and entrepreneurship, anybody who wonders whether they're working or not, seeing this room full of people would be a great testament to the vibe that we're getting going here. And so thank you all for that. Y Combinator, you probably know, is a Silicon Valley accelerator that helps founders in the early stages of their startups."}, {"start": 138.3, "end": 167.76, "text": "And has launched companies such as Airbnb, Dropbox, Stripe, and Zenefits. All of those have worked through Y Combinator to get their start. And Y Combinator companies are worth more than $65 billion today. Maybe a little less after today on the market. Sam Altman is the president of Y Combinator, and that's a position that he's held since 2014."}, {"start": 167.76, "end": 196.18, "text": "And before he was there, he ran Hydrazine Capital and Looped. And before that, he studied computer science at Stanford University. In January of last year, Sam was named the top investor in Forbes 30 under 30. And these are the kinds of events where somebody puts on a suit to introduce someone that he knows will be wearing a hoodie. And that's a great thing about the world today. It's awesome."}, {"start": 196.34, "end": 235.5, "text": "Sam grew up here in Clayton, Missouri, and as a kid he spent time hanging out at the South 40. Sam, it's great to have you back at WashU, and take it away. Thank you. Thank you. You spend time at the South 40. I spend a lot of time at the South 40 myself. My wife and I are involved in faculty fellows at a dorm there. And you've got a room full of college students, many of whom are thinking about how do they tell their parents, you know what, I don't want to go for a job."}, {"start": 235.5, "end": 265.07, "text": "I'd like to think about some entrepreneurial pursuit. What kind of advice do you give to a college student who's thinking about a startup? Well, I think the good news is startups are sort of fashionable right now. You can tell your parents you're going to start one and they might actually be excited. I think this is a big change from certainly 10, 20 years ago. But even in the last few years, I think given the sort of uncertainty with all career paths, people are relatively supportive of their kids starting startups."}, {"start": 265.07, "end": 293.9, "text": "The general advice that I always try to give is that just deciding you're going to start a startup for the sake of a startup doesn't usually work. It's much better to have a particular idea, something you want to build, and when you're in college is a great time to figure out what that is. So I think this sort of thing where you don't think about it at all and then you kind of graduate college and say, well, I'm going to start a startup, and you have to sit around and come up with an idea, that's really hard. But if you have an idea that you really love, then I think"}, {"start": 293.9, "end": 322.56, "text": "Startups are really good ways to make that happen. There's this way in which the greatest invention of the last few hundred years was the limited liability corporation because it allowed people to, big groups of people to collaborate and share incentives and that's been really transformative for making things happen in the world. And so if that's what you wanna do, I think a startup is often a really good way to do that."}, {"start": 322.56, "end": 350.64, "text": "The provost started out by talking about the sort of troubling times that we're in the stock market. What are three pieces of advice that you give your founders at Y Combinator? Well, I'll first comment on sort of troubling times in the stock market. There's a lot of noise about this. There are a lot of people that will try to talk you out of starting a startup right now because it's a bad time in the economy or maybe you'll become one if the stock market goes down like it did in 2009 or whatever that happened. Here's the thing."}, {"start": 350.64, "end": 378.98, "text": "Any company that goes on to really transform the world and be worth billions of dollars and have something that hundreds of millions of people use will go through multiple up and down business cycles. And so there's a lot of people that talk about, well, the stock market's down today. What does that mean? But if you think about the companies that really make lasting impacts, it takes them 10, 20, 30 years to do it. And they have weathered multiple ups and downs to the stock market. So you should not worry about it. And in fact,"}, {"start": 378.98, "end": 407.44, "text": "I actually think it's a benefit. The very best companies tend to get started in economic downturns. When the stock market's going crazy, when people are flinging money at anything that sounds like a startup, the only thing that is cheap in that environment is capital. Everything else, office space, employees, anything you could possibly need is really expensive. So actually, when there's choppiness in the stock market, we try to turn up our investing. And when other investors are doing less, we try to do more."}, {"start": 407.44, "end": 435.04, "text": "And if I were going to start a company, I would look at a time like this. If it does sort of turn into a worse than it is right now, it's a great time to start a company. So I wouldn't worry about that. If you're planning to do this for a long time, you can basically ignore the ups and downs of the market. One of the best things about investing in startups, having basically all of your net worth tied up in startups, is that the price only quotes every 18 months per company."}, {"start": 435.04, "end": 462.99, "text": "So you don't have to sit there and watch the ticker of Apple stock and stress out every day as it goes up and down. The advice that we give to startups, the most important thing, if you can only say one thing, is to make something that users love. If you think about the companies that go on to be huge, they start off with some incredibly passionate customers. And so this is more important than anything else. This is more important than"}, {"start": 462.99, "end": 491.55, "text": "Your legal structure, your business development deals, or how nice your fundraising deck looks, or anything like that. If you think about the really iconic companies that have been started in the last however long you want to think about, you probably heard about most of them for the first time from a friend who was raving about how great the product was. And then once you started using it yourself, you probably used it a lot. Facebook and Google are the two examples that stick out in my mind where I can remember."}, {"start": 491.55, "end": 520.27, "text": "Someone telling me about them, using it for the first time, and being like, wow, that product is incredible, and I just use it all the time. And so what we look for is evidence of that. We look for startups that have users that are that passionate, that they tell their friends about it, and then once people use it, they really use it a lot. So engagement is the thing we look at the most, and we tell founders to focus on the most. More important than trying to optimize like crazy for top line growth."}, {"start": 520.27, "end": 547.71, "text": "It's just do your users come back and use your product every day? Or if you are developing a cancer drug, does it work so well that doctors are beating down your door to get more of it? So that is by far the most important thing that we tell companies. Pick two more. The second one I'll pick is choose your co-founders carefully. The number one reason that our startups fall apart is that they fail to make a product that users love."}, {"start": 547.71, "end": 576.7, "text": "But the number two reason is that co-founders have a falling out and that often kills the company. It's sort of amazing how little effort people put into making such an important decision as who they're gonna start their company with. But one thing we've observed in our data is that founders that don't know each other for a while before starting a company tend to not work out. And I'll give one little data point on that. We now have a rule against adding"}, {"start": 576.7, "end": 605.22, "text": "Co-founders to companies after YC funds it. But we used to not. And so what happened is a lot of people would come in, they'd get into YC, and then somebody else who really wanted to also get into YC would say, oh, there's this backdoor, I'll become a co-founder and I'll get into YC. At the height of that, in one batch, out of 60 companies, 12 of them added a co-founder after we funded them. And one year later, all 12 of those companies, the co-founders had split up. So you really want some shared history and friendship before."}, {"start": 605.22, "end": 633.81, "text": "This particular company where you work together and you know each other and so that if things get hard, you'll have this obligation of support to each other. I think that's really important. And then the last thing that I'll say is the best companies move the fastest. And everyone has some explanation for why they're different than that and why they're doing this really hard thing and we just don't understand and they have to move slowly and we just don't get it."}, {"start": 633.81, "end": 662.0, "text": "And those companies always die. And the companies that, for whatever industry they're in, adjust to being the fastest of their peers and iterate on their product the most quickly, those are the best. The software companies measure their iteration speed, basically build something, get it to a user, and then make changes in hours, or some of them in tens of minutes. And I think that correlates really well with success."}, {"start": 662.06, "end": 694.48, "text": "So how do you draw on your experiences from Loop as you do your daily job at Y Combinator? At this point, my job at YC is more like running a company than being an investor. Relative to investing firms, YC is like giant in terms of the number of employees and the number of organizations we have. Relative to product companies, it's not that big. But most of my time is actually spent"}, {"start": 694.48, "end": 724.1, "text": "Running the company. And for that, certainly, that's one of those things where the only way to learn it is with experience. Tell us about the YC fellowships. You know, when's a good time? Okay, so we have this problem, which we created, and it's a great problem to have. But, you know, we have become pretty well known. And we have 25,000 companies per year applying to us. And that's growing at a rate of like 30%. And many of these are really good."}, {"start": 724.1, "end": 751.6, "text": "And that is more than we knew how to fund in our current model, which is very hands-on. We work super intensively with the companies that we fund. But if you believe, like we do, that A, startups are the future, and there are going to be a lot more of them, and B, that the value of YC is that there is this network effect, and we have this community that makes all of our startups 100 times more likely to be successful, then you don't want to just say, well, we're not going to grow."}, {"start": 752.18, "end": 782.48, "text": "So we started thinking about, okay, right now we fund maybe, well at the time, let's say we're funding 200 companies a year, now we fund maybe 350. How do we 10x that? And our current model was not gonna scale to that. And where we were seeing the most growth was on very early stage applicants that seemed really smart but had still a very unrefined idea. And so we thought, well wouldn't the perfect sort of solution to this"}, {"start": 782.48, "end": 811.18, "text": "Would be if anyone reasonably promising who wanted to start a startup could click a button and we would wire enough money to their bank account that they could live and build their first version. And then we can give them some advice. It'll have to be mostly sort of mass generated, but enough to get them going. And then we can build some tools and some software so that they can have some of the community experience, virtual office hours, like online stream talks, that kind of stuff. And if we could do that,"}, {"start": 811.18, "end": 841.15, "text": "To anyone around the world, don't require them to move. That would let a lot more people be entrepreneurs than can today because you still need a pretty good life safety net and a reasonable amount of money to start a company. So we started doing that. We ran our first trial last fall. We're doing our second one right now. We still have a few more things to iron out, but probably after this current one or maybe one more, we'll be ready to really make this a full program and scale it up a lot."}, {"start": 841.15, "end": 870.8, "text": "And it is something that we would like to scale to thousands of companies a year. And the early results, I'm thinking the first version was a little bit rocky. We learned a lot. We're in the middle of the second right now. It seems to have gotten a lot better. But it feels like we're onto something here, and this will be a program we keep doing. So you mentioned earlier that one of the bits of advice you'd have for the group out here is have a passionate customer base. Yeah. And you talked a little bit today about operating at scale."}, {"start": 870.8, "end": 901.71, "text": "And you guys are doing some innovative things to try to find, you know, to focus on problems of which you're passionate about, big problems that impact the world. Talk about, you know, research and the labs. Sure. So we have always viewed our kind of role in the world and our mission as we want to just increase the innovation output of the world. And that is good because it's intellectually interesting. It's certainly more fun to work on new things than things other people have already figured out. But also, if you sort of like,"}, {"start": 902.83, "end": 930.53, "text": "Well, a lot of people have different answers for how we're in the situation in the presidential race that we're in right now. Whatever you think about that, it is certainly unusual. And one of the things that I believe is that that's what, if you have a low growth environment and you have a democracy, then it's very unstable. Because everyone wants their lives to get better every year. That worked in the US. You had like 100 years of territorial expansion, 100 years of incredible innovation."}, {"start": 930.53, "end": 957.82, "text": "And so you had like 8% GDP growth for a long time. Now you have like sub two and it gets like really zero sum. And so I think the only way that you kind of stay in a stable configuration is if you can have a lot more innovation and everyone's life can get better every year and you have growth that way. So that's what we like to do and it's just fun to work on. 80% probably of the innovation that we think we wanna support."}, {"start": 957.82, "end": 988.06, "text": "It does actually work in a startup. Startups are incredibly effective organizing vehicles for groups of people. But there are some things that don't, for whatever reason, fit into a startup. Either because they're never gonna make money, or because they have a very long time horizon, or because they are something that should be owned by the world as a whole. And we still wanna be able to support those. And we have a number of, or I have a number of problems with the existing"}, {"start": 988.06, "end": 1016.03, "text": "Research institutions in the world. So I thought it'd be good to create an alternative. And so now we have this thing where it's a nonprofit. So all the IP we develop is free to the world. And we fund things that we think are important. The first thing that we did is OpenAI, which is a group that's trying to develop superhuman AI and have that be as good as possible for humans. The second one is a basic income study. What happens if you just"}, {"start": 1016.03, "end": 1043.62, "text": "Give a guaranteed minimum income to people. Are they happy? Are they fulfilled? Do they work? Do they not work? Does it help their personal balance sheets? Do they contribute to the world? We're going to do something else soon with education and future computing platforms. We'll probably do some stuff with biotech. But yeah, it's been a lot of fun. So you mentioned your goal is to increase the innovation output of the world, which is great. It's lofty."}, {"start": 1043.62, "end": 1073.82, "text": "Talked about research institutions that are existing. This is one of them. So what advice would you have for us collectively as an institution of education and research on how we can help impact the innovation output of the world? Yeah. I think the thing that gets built up in institutions, which is bad, and which we really try hard in YC, especially on the for-profit side, but also the research institution side, to avoid is as things become more prestigious,"}, {"start": 1073.82, "end": 1103.12, "text": "And more successful, people are willing to take less and less risk. And so I think like, when I was in school and I was doing research, one of the things that I noticed is that people, if you were gonna work on something slightly heretical, that was like a big upside, big downside kind of situation, and it didn't fit the mainstream popular ideas, you had to be careful not to talk about it that way. And that seemed really toxic."}, {"start": 1103.12, "end": 1130.59, "text": "And if you wanted to go explore something that seemed interesting, but you didn't know much about it, but it was sort of this happy accident that was happening, but it was not on your grant proposal, you couldn't go pursue that, because that would be really bad. And that also seemed really toxic. But I think what's important is to let happy accidents happen, have a prepared mind for when things come up, and..."}, {"start": 1130.59, "end": 1160.5, "text": "Have an environment where people can work on things that are considered unfashionable. One of the things that we learned when we were starting our AI group was that many of the top academics actually believed that human level AI, general true sci-fi AI was coming in a short period of time, 20 years, 30 years, some thought less. But they could never say that out loud cuz they thought that'd be career suicide. And so how crazy was this, right? You had the best academics in the world that had a deep belief related to their work."}, {"start": 1160.5, "end": 1192.99, "text": "But they couldn't express or work on, because they thought it was too unfashionable. I've got one more question, then I'm going to invite Dave up to work with you guys and other questions from the audience. And that is, I often say that failure is not an option, but it's a distinct possibility. So would you talk to us about the role that failure has played in your successes today? Yeah. I think sometimes Silicon Valley gets a little bit too into failure glorification. People brag about how many times they failed. It's still bad."}, {"start": 1192.99, "end": 1219.66, "text": "And I can tell you from personal experience, it's still unpleasant. But I think Silicon Valley tolerates failure. And if you do something that doesn't work, it's not held against you as long as the next thing or at some point something works. I think one thing that's really good about Silicon Valley is you are judged on your best success and all of the other failures kind of get forgotten. And in some cases, great things happen out of failure."}, {"start": 1219.66, "end": 1247.89, "text": "In general, I think pivots are bad, right? If you fail at something and then you have this company and you have to rush to something else, that usually doesn't work. You're under too much pressure. And speaking of having to pick non-heretical ideas, if you have a shell of a company and investors put in a lot of money and your first thing didn't work, you cannot go have a crazy idea next. It's gotta sound really good. And the startup ideas that sound really good fail for a bunch of reasons which we can talk about later."}, {"start": 1248.08, "end": 1272.18, "text": "The biggest success so far in our portfolio, Airbnb, founders had been working on a different startup for a couple of years, and it was massively failing. And it had been failing for a long time. So they were out of money, not only were they out of money, they had been taking out credit cards that you can get as like a student with a $500 limit, using the whole limit. And then, you know those like,"}, {"start": 1272.27, "end": 1299.68, "text": "Binders that kids put baseball cards in. They have those like little pages where you can put like nine cards, three by three on a page. They're just putting their credit cards in there. When I met those guys, they had like many pages in a binder full of credit cards that had been maxed out of their credit limit. So they can't get any more credit cards. Really out of money at this point. And they have rent due in like 10 days or 12 days, something like that. And started continuing not to work. Brian Chesky, like pretty crazy person."}, {"start": 1299.68, "end": 1327.09, "text": "Not gonna give up eating cereal from the dollar store at this point. I wouldn't advise any of this. This was a real mess. And they had rent to and they had no idea how to pay it. And they were absolutely failing. And out of that, they got the idea that well, this one thing we have is space that we're not using in our apartment. And so maybe someone will pay us for that so we can pay our rent."}, {"start": 1327.7, "end": 1366.08, "text": "They did, and they made rent. And they did it again. They were able to pay down some credit cards. And out of that, they realized that a lot of other people have this problem. And so I think you see that pattern where people fail at one thing, but in the process discover another. And that is the kind of pivot that works the most often. Thank you very much, Dave. Nice of you to come up. We'll move into the question and answer period. Hi, everybody. Do I ask you questions now? Let's try it. All right."}, {"start": 1366.08, "end": 1394.42, "text": "That was going to be my first question, but you go ahead. Great, okay. I'm Dave Messina. I'm the COO of Cofactor Genomics. We are a biotech company located here in St. Louis. Our building is over in the Cortex District. We were founded in 2009. Me and my partners were all, talk about a bad time to start a company in 2009. Yeah, that was a bad time. Yeah. Been working out pretty well so far."}, {"start": 1394.42, "end": 1423.47, "text": "Me and my partners all worked together on the Human Genome Project as scientists here at WashU. And what we're doing is we're using RNA to diagnose disease. So we are developing molecular diagnostics around RNA. So most of you have heard about genetic tests and diagnostics using DNA. DNA is the blueprint of life. And DNA is great for some types of diagnostics. But in other ways, it's not so great. So DNA doesn't change over the course of your lifetime."}, {"start": 1423.47, "end": 1452.75, "text": "So the DNA you have at birth is the same DNA that you have today. And so for something like 95% of diseases, actually, DNA does not really make a good diagnostic for this reason. RNA, on the other hand, changes all the time. So it's what you can think of as the active form of genes or DNA. It's in every cell in your body. And so it changes in response to your diet, exercise, infection, and what we're interested in."}, {"start": 1452.75, "end": 1480.64, "text": "The changes in response to disease. So what we're doing is measuring levels of RNA and which RNAs are characteristic for certain disease states. And so we can, we believe, find the presence, see the presence of disease before you're symptomatic, before your doctor normally could. So the vision is that right now you go to your doctor. He's already drawing a tube of blood to check your cholesterol level."}, {"start": 1480.64, "end": 1509.33, "text": "He could send another tube of blood to our lab here, and we can screen you for the presence of disease before he'd be able to see it otherwise. Actually, I'm going to ask questions about YC, but I figured I'll talk about YC for one minute in case people don't know about how we work. So we have four programs, but the main one that we're known for, which we call our core program, we fund about 125 startups in a batch. We do this twice a year."}, {"start": 1509.33, "end": 1538.26, "text": "They come from all over the world. We had 50 countries represented in our last batch. And we invest $120,000, and then we work very intensively with the company over this three-month period. So we have a dinner once a week where we bring in guest speakers. Actually, speaking of Airbnb, the Airbnb founders spoke last week. It's mostly founders, some investors, and we just ask them to tell their story. And it's off the record, so people tell, like..."}, {"start": 1538.45, "end": 1568.37, "text": "Crazy stories about, Dave can vouch, about like what really happened in the course of their startup. We do teach some things but not much. So we have maybe like five 30 minute sessions over the course of the program where we'll do like a crash course on PR and marketing, sales, hiring, raising money, whatever. And then mostly we give one on one advice. So most of our time is spent on the biggest problem and how to solve it."}, {"start": 1568.37, "end": 1597.89, "text": "And so it's meant to be a very advice heavy program. Each company gets assigned two partners, and we work with you a lot. At the end of the program, then we have something called Demo Day, where you present to investors, and hopefully you raise money. Some companies do, some companies choose not to, and some companies don't get offers. And then after that, YC kind of goes on for the lifetime of your company. So we still have, we have about one event a month for alumni."}, {"start": 1597.89, "end": 1625.98, "text": "We have sort of a little coffee shop kind of thing in San Francisco that alumni can come to. We do a big event once a year. And we have a bunch of software and email lists to help the community stick together as well. We talked about the fellowship. We talked about the research lab. And we have this later stage growth fund, which actually Washington University has partnered with us in, which is awesome. And we use that to invest in our company as they start to scale and need more capital. That's been great so far."}, {"start": 1625.98, "end": 1654.32, "text": "One of the things that we heard from our alumni was that although YC has done a lot to make the early stage environment better for founders, late stage investors are still difficult to work with. And I think even in our first six months, we've been able to kind of push that environment to be more founder friendly. So that's about YC. Our next batch is next summer. I think applications for that are due in like a month. I hope many of you decide to apply. I'm happy to answer questions about that later."}, {"start": 1654.32, "end": 1681.92, "text": "Okay, I'm gonna ask you questions. This is fun, I never get to do this. Here it is. What was your experience in YC? So it was really great. I mean, obviously you'd expect I'd say that, but it was actually way better than, we already had high expectations and it exceeded our expectations. So I can tell you for us, we had already been in business for quite a while, right? So we were founded in 2009, we went through YC last summer, in the summer of 2015. What we found was that a lot of"}, {"start": 1681.92, "end": 1710.16, "text": "The expectations we had about our fit for YC and YC's interest in us were way wrong. And so if there's one thing I could tell all of you tonight is please don't disqualify yourself from applying to YC or indeed from starting a startup because I think there's a reasonable likelihood that some of the, you may have some misconceptions about what that's like. So first of all, we were a biotech company, right? So we're not three guys with a software idea. So that was something that we learned that."}, {"start": 1710.16, "end": 1738.38, "text": "YC was actually very interested in and had been recruiting for heavily. So we were wrong about that. We thought we were too old as a company. So we had already generated a lot of revenue. And we thought, well, OK, we're too late. And YC is not going to be interested, or we might not get as much out of it. Also wrong. We got a huge amount out of YC. In fact, we found that there were plenty of companies that go through YC that had already generated revenue and that were further along."}, {"start": 1738.38, "end": 1764.82, "text": "Honestly, we also thought maybe we're a little too old. So, you know, we're not 20 year olds. We had all spent a lot of time, had careers already. We found that the average age of YC founders is something like 30 now, something like that? Yeah, 29 maybe. Okay, almost 30. So that also, we were totally wrong about that. What YC gives you is leverage."}, {"start": 1765.07, "end": 1794.61, "text": "I think that's maybe the purest answer I could give is that YC helps you to go farther faster and actually challenges you to do that. So like I said, we were already doing pretty well, but having advice from people who know a lot about how to make a company successful is extremely helpful because they can help you identify the area you need to focus on the most at any given time. Yeah, if I could just add to that."}, {"start": 1794.61, "end": 1824.3, "text": "The thing that we're experts in is how to be a good startup. So we will never know as much about you as a particular area, and so people are like, well, how can you advise a biotech startup? We can't really advise them much on the biology, but nor do we write code for the software startups. Really what we want to do is just say, okay, we have seen a four-digit number of startups. Here is the mistake you're about to make. It seems counterintuitive, and this is what you should do."}, {"start": 1824.3, "end": 1854.13, "text": "So we really do just try to do that. What do you tell people to get sort of the most out of going through YC? I think you have to be, and I think you guys are good at selecting for this, I think you have to be open to advice. You have to have a strong point of view. I think Marc Andreessen talks about strong views weakly held. So you have to believe in your vision and know what you're trying to do. But you also have to be open to new information and to taking advantage of that. I think that..."}, {"start": 1854.13, "end": 1880.38, "text": "If you are not in a position to change your mind often, you're probably going to make more mistakes than you would otherwise. Other than advice, what turned out to be most helpful about YC? So the network. And when Sam talks about not just the partners at YC, and there are probably big 20, yeah, almost 20 partners now, so they have expertise in..."}, {"start": 1880.38, "end": 1908.24, "text": "Some partners with specific expertise in biotech also, by the way. But the thousands of companies that have been through YC, because it's so easy to connect to them. So that was one of the things that really surprised me. I'd be talking with one of our advisors, and they'd say, oh, hey, you might want to give a call to so-and-so from this company. I could send an email, and we would be on the phone the next day. People in the YC, part of the philosophy is to be helpful to each other. Here's the bad news about that."}, {"start": 1908.34, "end": 1938.69, "text": "You may not quite know yet. There's this culture in YC of when you're in YC, you can anything you need. You have to pay that forward someday. I'm already paying it forward. I'm already mentoring one of the YC fellowship companies. I think this is, though, a general rule, which I think it goes outside of YC. But one of the things that I think is most important about YC is that people in general want to be helpful and want to help you succeed, particularly if you are passionate about what you're trying to do."}, {"start": 1938.69, "end": 1968.38, "text": "If you're starting a startup, that's something you've got to put a whole lot into. And I think it's a mistake that many people make that you think it's just going to be you doing that. It's going to take a whole lot of help for you to be successful at that, not just the people who work on it full time, but really leveraging all the resources around you. And so I think YC is really good at fostering that idea and pushing people to be part of that. And then beyond that, I'd say because YC has such a fantastic track record that"}, {"start": 1968.38, "end": 2003.57, "text": "It's a good signal to be a backed by Y Combinator. So certainly for us, we had been trying to raise money for a little while and that was a game changer for us to be able to have the support of an organization that people believed in. Can we take questions from the audience? Let's do it, yeah. You can answer questions on any topic. You can direct it to both of us either. Yes. The question is, do we care if the company has a good name or not?"}, {"start": 2003.57, "end": 2032.03, "text": "For enterprise companies, no. For consumer companies, yes. There is a famous observation from Peter Thiel that you can actually predict the success of a consumer company by how good the name is. I think that a little bit selects for the founder, and good founders come up with good names, but I think it really does matter. And the really horrible names you hear sometimes that are unpronounceable or unspellable or whatever, for consumer companies, that does turn out to be problematic. So we do look at that on the consumer side."}, {"start": 2032.03, "end": 2066.32, "text": "It won't disqualify it, but we may ask in the interview, like, are you open to changing your name? And if founders have a terrible name, we can explain to them . On the enterprise side, or like a biotech side, we honestly don't care much. Well, and the domain name is important, too, right? Being able to have that kind of... Yeah, we do care. Like, if you can't get the .com for your name, we usually suggest, but not always, changing. Yes?"}, {"start": 2066.32, "end": 2095.73, "text": "I'll give you my brief answer. If you can stay away from the regulatory part, then do. Actually, we were talking about this earlier today. There's a reason why the regulatory environment exists, but it also does make it harder for you to move fast. And so if you can think of a non-regulated way to test your product and your idea, start there. Regulators are not bad people."}, {"start": 2095.73, "end": 2126.38, "text": "They're not trying to kill innovation. They're not trying to sit there and make you fail. And I think most people, like when I got on an airplane, I'm happy the FAA exists. Like when I take a drug, I'm happy the FDA exists. The US manages to lead the world in areas like aviation and pharmaceuticals without and be the most regulated environment. So I think people are too unwilling to work in a regulated area. But as Dave was just saying, I do think"}, {"start": 2126.38, "end": 2155.73, "text": "Given how important cycle time is for a startup, it's very useful to think of a way to get going with something that is not regulated or less regulated as a first prototype while you in parallel begin the full regulatory process. So you might see a diagnostics company start with a CLIA lab and do that for a year before they kind of go through the entire FDA process."}, {"start": 2155.73, "end": 2184.21, "text": "You might see a company that's developing new aircraft for passengers start with something with drones because that's less regulated. So I think it's always worth thinking about what can I do as a V1 that is lightly regulated, but I think people are too unwilling to start something in a regulated industry. If you look at the list of the most valuable startups started in the last 10 or 12 years,"}, {"start": 2184.21, "end": 2219.02, "text": "A surprising number of them are in a regulated industry of some sort. Uber and Airbnb operate in incredibly regulated industries, Tesla and SpaceX even more so. And it's clearly still possible. Do you think that's because there's more room for innovation in a regulated space because the existing players have become too high bound? Regulated markets tend to be really big. Yeah, the existing players are more calcified and less people try, like me personally in my life."}, {"start": 2219.22, "end": 2276.5, "text": "I think probably Paul Graham and Peter Thiel. Between those two guys, I probably learned everything I've learned about investing from one of them. What do you mean by? Yeah. I mean, not that much. I guess one thing is you go from having no feedback to having a lot of feedback, and that is hard because you've poured your life into this thing, and now maybe no one's going to use it."}, {"start": 2276.5, "end": 2307.17, "text": "Certainly a lot of people are gonna say it really sucks. Learning how to ignore the internet comments is a really important founder skill that people develop very quickly, but a very painful one to learn. I think the biggest change that happens is you go from doing this thing in a private way to doing it in a very public way. And now you are, sort of have this obligation to the people that use whatever you're building to answer their questions and help them make it better."}, {"start": 2307.17, "end": 2346.42, "text": "Talk to people. You just have to go talk to people a lot now. So I think the mindset shift that is important is going from solely product focused to being mostly customer focused. Yeah, I think the question is using events like hackathons to spur innovation. I have fun at hackathons. I did one recently. I had a good time. I don't think they're a good way to start startups because they end up being people that try to build something very quickly to throw it away."}, {"start": 2346.42, "end": 2386.32, "text": "And that's actually fine. It's good for showing people how quickly they can build things. But I think the way that innovation has been actually spurred for a very long period of time, hundreds of years at least, probably longer, is small groups of people sort of sitting together without other people around and believing something that most other people don't believe and then working super hard to make it happen. I don't think hackathons are the right environment for that, but they're fun."}, {"start": 2386.32, "end": 2426.1, "text": "Yeah. Yeah. Moving fast is not good for its own sake. And you do see a lot of startups where they've decided that if they just work harder than everybody else, they will win. And obviously that doesn't work. The trick is to work hard on the right things. But 90% of what most startups do is a complete waste of time. Actually, probably 99% if you aggregate it across all startups."}, {"start": 2426.1, "end": 2455.14, "text": "And moving fast may not look externally like moving fast, because you may take a while to launch. You may not release a ton of features. But the speed with which a new idea can happen, have a discussion made about it, a decision internally about whether you're gonna do it or not, and initial user feedback, I do think you see that pretty fast."}, {"start": 2455.14, "end": 2486.61, "text": "That thing right there across all really great companies. There are, companies may be deliberate. They may not launch most things they try. One of the things that was really striking to me is Apple does not release that many new products. If you think about how many products Apple released, say, well, there was a MacBook, an iPod, an iPhone, an iPad, which is kind of the same thing. And they don't release that many."}, {"start": 2486.61, "end": 2514.9, "text": "But what was interesting to me as I've gotten to know more people that work there is the speed with which they move internally as they consider a new idea, which they do all the time, they usually say no, or as they iterate on one idea. So if they're gonna make a new iPhone, they'll 3D print 300 very slightly different things and they'll iterate so fast on which ones are good and which ones are bad and this needs to be 1 tenth of an inch bigger and this thing sticks out too much."}, {"start": 2514.9, "end": 2543.78, "text": "They move incredibly quickly internally on new ideas, even though the external result is slow. You know, I hate the whole lean startup thing. Well, I think that a lot of the models are good, but the idea that, the thing I don't like about the lean startup thing is this, that you shouldn't have any ideas, you shouldn't have any conviction, you should just throw things, see what sticks, throw more things. The best startups are governed by this sort of"}, {"start": 2543.78, "end": 2583.28, "text": "Central organizing principle about what they're gonna do and they believe in that and that's gonna happen. That is the mission. And I think that the lean startup is somehow anti-mission in that it's just like have no ideas and just try to move really quickly and see what happens. And I do think that like you want some things that you just hold as religion that you, like that is why the company exists and then you're open to data and customer feedback and everything else. Yep."}, {"start": 2583.28, "end": 2612.53, "text": "So the question is about our thoughts about international expansion. I think the YC partnership visited 30 or 40 countries last year to meet with founders at events like this. Currently, our model is to go travel around the world and meet the best founders and then have them come to our program for three months and then go back to where they come from. The reason that's really good is during the three months, they get plugged into the YC network and they get to know kind of just all of Silicon Valley and Silicon Valley investors."}, {"start": 2612.53, "end": 2643.54, "text": "That seems to be working really well with sort of the exception of China where it's just logistically harder. We're not averse to opening offices other places and we may do it someday. But when we've polled international founders and said, would you rather have done this way or had us have an office wherever you're from, 95% of the time people say, no, it was so great to come to the Valley and don't change the model. So we do view founders as our customer."}, {"start": 2643.54, "end": 2681.9, "text": "The feedback has been so strong there. And there have been relatively few founders that we really wanted to fund but couldn't or didn't want to move to the Valley. So I think, you know, future's hard to predict. We're gonna grow a lot. We probably will open other offices someday. But the model works pretty well right now. Yeah. The socioeconomic background is wide except we don't get a lot of like truly poor founders that apply. I think"}, {"start": 2681.9, "end": 2711.31, "text": "One of the reasons we're interested in studying the basic income is you can't really, like people talk about startups being something where anyone can lift themselves up by their bootstraps and there is some truth to that unless you start like truly, truly poor and then it's impossible and you just, you know, you get screwed in the birth lottery, it's unfair and you have no chance or very little chance. So that's why we're excited about things like the fellowship and the basic income research that we're doing. Other than that though, the demographic"}, {"start": 2711.31, "end": 2739.89, "text": "I think matches the demographic of the world really well. Just because most of the world's middle class, we mostly get people from middle class backgrounds, or most of the US. But plenty of people on the extreme lower end of that middle class spectrum. And we do fund some really, really poor founders. And then we fund some really rich ones too. And in terms of social,"}, {"start": 2739.89, "end": 2782.51, "text": "Background, it really is all over the world. That's pretty diverse. Why don't you answer that, Dave? The question is, what are the similarities among founders at YC? They're all pretty driven. I mean, I think you pretty much have to be. But I also think that creativity is a huge part of success, particularly in a startup. So you have to be, Sam was talking about this earlier today, you have to be a generalist because the job changes so fast. So I can tell you,"}, {"start": 2782.51, "end": 2810.77, "text": "I have a PhD in computational biology. One of the things that I knew when I was working on a startup was that there's going to be some things about that which are really useful. There's going to be a whole lot of things that I'm going to have to learn. And so part of what enabled me to be successful in trying to build a company is because you have to want to learn new things, and you have to be able to change your perspective rapidly based on that information."}, {"start": 2810.77, "end": 2840.37, "text": "And generate ideas. And people talk sometimes about ideas being cheap. That's true to an extent, but I also think that being able to come up with ideas is an undervalued skill. I'd say there are the obvious things about smart and good ideas that we look for. Determination is super important to us. Optimists are super important to us. And people that don't"}, {"start": 2840.37, "end": 2869.98, "text": "Try to kind of like make excuses or give up when things aren't working are really important to us. And when you add all those things together, you get a group of people that's really fun to be around. And so like one of the nice things is that it's a community that shares those values and that's like a nice, that's a happy group to spend time with. You guys have like a no-jerks policy, basically. We have a no-jerks policy. No matter how much money we think we are going to miss out on the investment, we will not invest if we don't like it. That actually matters more than you might think. It's not worth it."}, {"start": 2869.98, "end": 2913.3, "text": "Spend time around people who are jerks. How many biotech startups have we helped and what's the ratio of success? I think we funded now maybe between 40 and 50 in the last two years. Way too early to talk about a success rate, unfortunately. We just haven't been doing it long enough. Many of them have raised millions of dollars or tens of millions of dollars and have incredibly promising results and happy customers and significant revenue."}, {"start": 2913.3, "end": 2957.38, "text": "And there have been a few small acquisitions, but I wouldn't be comfortable talking about our hit rate for another probably three years, something like that. In the back, yeah. So the question is if you have an idea but you don't wanna talk about it because a big company's gonna steal it and make it better. Okay, here is one of the things that takes founders a long time to learn. And that is no matter how great your idea is, no one cares. Everybody else is so distracted."}, {"start": 2957.38, "end": 2986.5, "text": "That you could probably go put that idea and exact instructions for how to implement it on Tim Cook's desk and take no risk. It probably is a great idea, but everyone's really busy with their own things. Everyone's lives are really busy. Everyone has their own plans. You can rely on that to protect you. Now, you can pick a few examples from history."}, {"start": 2986.5, "end": 3016.75, "text": "Steve Jobs walking into park that day where someone did get screwed for sharing their idea. But you really have to work to find them. And it happens hundreds of thousands of times every day with nothing bad happening. So it's hard though, because ideas feel like part of you and you don't want them to get stolen. And these big companies have lots of resources and you have none. But one of the things that we learned is that extreme secrecy among founders is a bad sign. You want people, you want to keep something secret for sure."}, {"start": 3016.75, "end": 3044.77, "text": "But you should be willing to talk about the broad sketches of what you're doing because you need that to recruit people, to get investors, to get customers. And you don't have to fear because the big companies are distracted fighting each other. They're not going to come do your thing. I think the corollary to that is that people are reluctant to go talk to people who they're worried about might steal their idea, who could actually give them a lot of good advice. So this is something that came up."}, {"start": 3044.77, "end": 3073.06, "text": "When I was at YC this summer, I was given the advice, go talk to this CEO of a successful genetics company. And my initial reaction is, are you crazy? I'm not going to share what we're working on with this person who could eat our lunch. And I got this great piece of advice from my advisor, which was, have you considered just asking that person, if you are worried, whether they are working on something like that, and if there is indeed a conflict of interest?"}, {"start": 3073.06, "end": 3101.41, "text": "Here is a personal example just to give you some comfort in this. We at YC talk about everything we do. We talk about how to operate. We give our best possible advice. I have given talks before to rooms of people about this many that want to start accelerators. And I say, if you want to start an accelerator, here is exactly what to do step by step, and here are the mistakes to avoid step by step. And people always say, are you crazy? You're giving away YC secrets."}, {"start": 3101.41, "end": 3131.06, "text": "And we are, and yet no one ever listens. There are 2,500 accelerators in the world. I just heard from some economics student that studied this that YC has 180% of the profits of all accelerators in the world. And we tell people exactly what we do and how to do it. People don't listen. Or people are worried that they're doing things their own way."}, {"start": 3131.06, "end": 3176.94, "text": "It was hard for me to learn this lesson of not fearing this, but I think in general, maybe don't give away the most important kernel, the cotter pin, but don't be afraid of telling people what you do. The question is what effect does being an entrepreneur have on your life? I am not unhappy not to be running a startup anymore, because it is a big life trade-off. And we try to be upfront about this with people. This is not the best choice for work-life balance."}, {"start": 3176.94, "end": 3205.34, "text": "You can do it. I mean, I think you can have a startup and have one or two other things you really care about, but startups are difficult to do a good job on unless you're willing to spend most of your waking hours doing them. And plenty of people make a super rational decision they don't want to do that. There is another kind of middle of the road option, which is what people call lifestyle businesses, and they always say that as a negative, but I think it's a perfectly valid thing. I know a lot of people, a lot of friends of mine have."}, {"start": 3205.34, "end": 3233.6, "text": "Businesses, they never want to take outside investment. They don't want to try to scale super fast. But they, you know, kick off a few million dollars a year in profit and they take almost none of their time and they go travel and do a lot of other great stuff and that's a pretty good life. But if you're going to do, like, the classic grow super fast, raise a ton of external capital startup, again, you definitely get a social life and friends and you get to travel some, but the startup is, like, always on your mind to some degree."}, {"start": 3233.6, "end": 3269.17, "text": "And that's a trade-off we try to be super upfront with people about, because it's something a lot of people don't want. Yeah, students that would be interested in working in one of our portfolio companies, please do, I'm delighted to hear that. We, on Hacker News, which is a website we run, news.ycom and air.com, a lot of our startups will post jobs. We also once a year host an event called Work at a YC Startup. The next one is actually coming up pretty soon, I think. And then also,"}, {"start": 3269.17, "end": 3308.51, "text": "Basically, every YC company, the founders' email addresses are their first name at thatdomain.com. And if you email them and you're like, I'd like to work at your startup, you will get a better response than applying on their website. So I would try that. Yeah, why do we start accepting nonprofits? Again, it comes back to the same mission. Our mission is to increase the amount of innovation in the world, and some of that is going to fit in nonprofits. So we're not..."}, {"start": 3308.51, "end": 3337.5, "text": "Trying to make money. I mean, we are, but that's not our central thing. We make more money so that we can fund more interesting stuff. And so sometimes a nonprofit fits. We originally had sort of a different program for nonprofits or a slightly different program, but we realized early on that nonprofits in general were pretty broken and needed to look a lot more like for-profit startups. So we just throw them in with a regular batch and they have to go through the same kind of thing. And I think that..."}, {"start": 3337.5, "end": 3375.33, "text": "That has been the best change we've made to the nonprofit program is not treating them any differently. Holding the same standards, making them operate the same way. We can probably do one more question. What if you've identified a B2B opportunity where it's really hard to launch at B1? This is the kind of question that it's hard to have a general answer for. This is the kind of thing we do during YC where we'll sit with you and try to brainstorm about what the right thing is. But it's generally to find some smaller customer willing to take a little bit of a risk and then work"}, {"start": 3375.33, "end": 3435.92, "text": "Being a founder."}, {"start": 3435.92, "end": 3466.08, "text": "What that means, but you can tell it quickly. We look for, if the company's been around for a little while, we'll ask about challenges the company's faced and how the founder has gotten around them. And if not, we will look for examples from earlier in their lives and other things they've done where they did that. Brian Chesky of Airbnb says when he's interviewing, the only thing he's looking for is people that have done the impossible in their lives at some point somehow."}, {"start": 3466.08, "end": 3493.3, "text": "And I think there's a version of that that's true for founders. The other thing that's really important, there's all the obvious founder traits I'm not gonna list, but the other one that's really important that people don't know is good communicators. So if you sort of are going to start a really successful company, you're gonna have to convince a lot of people before it is certain that you're successful that you're going to be. And if you don't know how to communicate well, if you're not good at persuading people,"}, {"start": 3493.3, "end": 3522.9, "text": "You will not be able to recruit, you will not be able to raise money, sell, anything. And so we really look for people that are good communicators and very persuasive. Do you have anything to add? I think all that's good. I think that part of the fact that you're asking the question is a good sign. But the formula is not complicated. I think forces of nature really encapsulate somebody who will succeed, who wants to succeed so badly that the obstacles are things to get around and that..."}, {"start": 3522.9, "end": 3558.72, "text": "You find a way to do it because you have to do it. And I think communication is a huge part of that because you have to convince people to help you get there. Okay, thank you all very much. Thank you."}, {"start": 3558.72, "end": 3569.9, "text": "Thank you, Sam. And I hear you get around here on occasion. Yeah. So if you see somebody playing on the South 40, it might be Sam. So once again, thank you. Thank you. And we look forward to seeing you again."}]}