[0:00:00] Good afternoon, everyone, and welcome to the Innovative Leadership Series. We're very fortunate to have Sam Altman join us today. For those of you who aren't familiar, the Innovative Leadership Series features business leaders from South Carolina and across the country who cover topics related to innovation, leadership, analytics, technology, and entrepreneurship. Events are free and open to the public, and they occur twice per academic semester to kick off our residency weekends for the working professional MBA and entrepreneurship.
[0:00:28] and MBA in Business Analytics students. As I mentioned today, we're very fortunate to have Sam Altman join us. Sam is the former president of Y Combinator until 2019, is currently the CEO of OpenAI, and was the co-founder of Looped, a location-based social networking app, which was part of YC's first batch in 2005 and acquired by Green Dot in 2012.
[0:00:53] Prior to taking over as Y Combinator's president, Allman was a part-time partner at Y Combinator since 2011. And for all of our students, you've read a number of his essays on startups and technology, as well as his blog. So without further ado, Sam, thank you so much for joining us today. Thank you for having me. So I'm going to go ahead and pin you here, make sure everyone can see. All right, we're all set. So I sent you a list of questions ahead of time.
[0:01:20] Given the crowd that we have and kind of the Zoom beeping in and whatnot, what I think would probably be easier, depending on if everyone has their video on us, I'll just go ahead and ask you the first set and then we'll kind of see how it goes. Does that sound good? Great. Awesome. So a first question came from one of our current students, Allie Masters, who wanted to better understand, you know, your reasoning for dropping out of Stanford and getting started in entrepreneurship.
[0:01:45] What was kind of the catalyst for that? Was it all you? Was it parents? How did that play out? No, my parents were against it. I actually wasn't planning to drop out. It was sort of this gradual process where I just kept extending my leave. But I just started working on a project and fell more and more in love with it. And then it kind of snowballed. But I never have made a lot of plans in life. I didn't have one this time either. I just fell in love with something I wanted to do.
[0:02:14] Did you have peers or others encouraging you or was it just the drive from yourself? You know, this was like a long time ago and it was sort of after the dot-com blow up. And it was this, you know, this was like 2005. And so it was this period where it was really uncool to drop out. So I definitely didn't have anyone encouraging it. Understood. Thank you for that. And so, you know, looking at the second question, what got you into entrepreneurship? How did you start Lube? Why did you start it?
[0:02:43] I mean, what was really driving you? I mean, this was one of the first network location type apps out there. I remember being one of the first users a long time ago too. What was really driving that whole situation? I fell in love with this idea that startups were the best way to get things done in the world. And I believe that even more strongly now than I did back then. I kind of had like this idea that it's like, it's sort of cool. You can just like...
[0:03:12] Do something. And you don't have to play anybody else's game. You don't have to slowly rise through the ranks. You can just say, I'm going to do it right now. And we're going to use these forces of capitalism and alignment of interest to get this project out into the hands of a lot of people. And as soon as I saw that work once, I think it...
[0:03:40] You only have to learn that lesson about how powerful that is one time, but it's sort of hard to imagine until you do it. And so then after that, I decided I wanted to like do a bunch more work with a lot of startups. And then even after that, I was like, wow, I think you can use startups to do these like very big projects that governments used to do as well. Exactly. Exactly. So having done looped and a bunch of other things, you know, what do you tell your younger self to do differently? And I mentioned this to you before, like risk calibration you've written about and talked about in the past. And I think our students really need to hear this.
[0:04:10] We talk a lot about affordable loss, what they're willing to lose. How would you describe that or looking back, doing things differently? I'm pretty happy with the trajectory my career has gone on. It's not all been good, but I've learned something at every stage. And I wouldn't want to, even the hard parts or the bad parts, I wouldn't want to trade off. I do think this point about taking risk and understanding risk and how wrong most people get that is really like...
[0:04:40] A very big deal. I think in some sense, where you hear people talk about risk the most at the end of their careers, the things they didn't do. I was afraid to do this. I stayed on this path. I didn't want to take a risk here. And I'm always a big believer in asking, what's the worst that can happen here? I start this thing and I get embarrassed. OK, I'd rather not, but I can make it through that if it fails. Or I start this thing and I like, you know.
[0:05:07] I'm like totally out of money. Like, okay, I can go get a job at Google. So I think one of the best things I learned early on, and this is for all of the faults of Silicon Valley, I think this is one of the best, is that if I did something and failed, it didn't have to be the end of the story for me. I think Silicon Valley does not celebrate failure, but it tolerates it. It doesn't.
[0:05:38] It's okay. I think that really changed the way I thought about risk before coming here. Speaking of false Silicon Valley, I'm just going to throw one out there for you. When we look at Elizabeth Holmes, Trevor Milton, Guilty, does that change the entrepreneurial mindset in Silicon Valley at all with these great big ideas or did what they do, was it really wrong?
[0:06:06] It's super wrong. That was like full on. There's like a difference between fraud, which is really bad and failing, which like failing in good faith, which is okay. So like it's fine to like start a blood testing company, raise a bunch of money and have it legitimately not work. Like we're venture investors. That's what we sign up for. That's the model. Like no one will hold that against you. But when it comes to like, you know, really serious intentional defrauding, obviously that's terrible.
[0:06:35] One of the things about Silicon Valley is I am pleased that that is as rare as it is. Part of the reason that these stories get so much attention is they are the exception, not the rule, in an environment where it'd be easy to do it a lot more. 100%. Yep. Thank you. This is fun. I'm going to keep going here. So you've been on both sides of the table, entrepreneur, YC president.
[0:07:05] Which one do you prefer more and why? Or what did you learn about the other side of the table that you don't really like and vice versa? Man, they're both great. They've both been totally amazing. I think I like running companies more than being an investor. But being an investor is also really fun. At YC, I got to work with thousands of companies, many of which have gone on to be significant.
[0:07:37] Build products that I love and use every day that have sort of done really interesting things for the world. It's a very special, running YC is like a very special job because you can just like wake up one day and say like, I want us to like fund Fusion and every Fusion startup in the world will come to you. And you just get to like make it happen. Or like, I want to start an AGI effort and like all of the good AGI researchers in the world come to you. These are the two things I work on now.
[0:08:03] And they both happened because I woke up one day and was like, we're going to do this. And so that's a very cool thing about the role. But the excitement of being in the arena doing a company is, I would say, for me personally, I like it better. Yeah. And then you get to see deal flow from, like you mentioned, all over the world, anything you want to see. Was that the fun part, I guess, was the creativity and seeing what people were coming up with?
[0:08:31] But really, you want to own it and do it yourself? The most fun part, I mean, there are a lot of fun parts, but the most fun part is watching how quickly founders can go from like having no idea what they're doing to running these amazingly effective and big companies. So like helping people develop from basically like nothing to running these like, you know, $100 billion businesses and getting to like help and coach and play a small role in that journey.
[0:09:00] That is the most fun part. The second most fun part was, as you mentioned, getting to just see so much of the future. A thing that I always believed in is like, if we have strong ideas about what to fund, we're going to be years behind. And if we just find the smartest people we can and fund whatever they want to work on, that always has done us right. And so getting to see that breadth of innovation across all these industries and around the world, that's also awesome.
[0:09:28] I'll add one more question on top of this. When you look at all the deal flow and all the people that you've come across and all the different walks of life, I do this too in a small subset with teaching entrepreneurship. How do you predict, can you figure out who you think is going to be the ones that are going to make it? I mean, you've seen thousands of batches of YC companies. Is it truly just, what is it?
[0:09:49] You do get good at that. No one is 100% good at that, but you can get surprisingly good. One of the things that always shocks people about YC is that we only interview these companies for 10 minutes and we do as well as we do. It's not perfect. We have a lot of false positives, but we don't have that many false negatives. I think that is surprising to people that you can tell that much.
[0:10:19] In a 10-minute slide. Yeah, 100%. I always tell people you can't predict tenacity because you can work out of a bad situation. Just some are more willing to do so than others. So this is great. Yeah, most of the questions that I would ask in an interview or in a YC interview were about that because you can tell like intelligence and effective communication and sort of...
[0:10:43] A lot of other things very quickly, but without, just from like how the conversation goes, but I would really drive into the stories of tenacity and resilience. And I would like look for any evidence of that, whatever life circumstances someone had before coming, you know, to our interview room. I'd look for like, you know, the sort of like four standard deviations of tenacity. Excellent. Thank you, Sam. So, you know, Y Combinator,
[0:11:12] You know, seemingly agnostic toward industry as you've gotten to bigger cohorts. And with this rise of these industry-focused firms and accelerators, do you think that's beneficial, harmful for entrepreneurs? What's your view on all this? I mean, I think accelerators, with the exception of YC, like I'm very biased here so you all can discount this heavily, are very harmful to entrepreneurs.
[0:11:39] I don't know what the current numbers are, but I remember when we calculated it when I was running YC, YC had something like 120% of all of the profits of all accelerators. Like we made money and everybody else lost money. And I think a big part of that reason is most accelerators give very damaging advice and requirements to founders.
[0:12:04] One of the things that made YC different is we, and I think the world has adapted somewhat more in our direction now. So, you know, this is probably years out of date, but like we would never tell founders what to do. We would never say you have to do this or we're confident about this, or, you know, we want you to do this thing that way or whatever. And other accelerators would like basically try to run the companies for the founders and make the founders do what they wanted. And I think that's just like terrible. The industry focused firms,
[0:12:34] We're the worst of this because they had these strong PCs and they were trying to get founders to work on their pet idea or whatever. But I think YC has been very different than any other accelerator, at least historically. Again, maybe it's changed now. And that was, I think, really important. But I always tell people, if you're going to do any accelerator or any investor, make sure they're going to be there to support you and not to be your boss.
[0:13:03] Because that just ends badly for everybody. Yeah. And that's the balance, I think, between the accelerators now and the rise of what I would call venture studios popping up in a lot of number of cities where, like you said, they're directing your efforts or directing your efforts. I personally would not be excited to do that. Yeah. Yeah. Okay. So the one that got away, right? This is from Artie Cressman here. Was there an ever...
[0:13:28] All of my biggest mistakes have been not investing more in companies that I made an initial investment in and I knew were great. So, you know, there's like, yeah, there's like some company that I'm sure I'd pass on that I shouldn't have, but that's okay. Anybody can make that mistake. And I don't like hold it against myself. Those are just hard.
[0:13:48] The stuff that annoys me is where I had inside information. I knew the founders were great. I knew the company was doing well. And I did not fight to put more money in each round. You've got to keep those programs. Many millions of dollars of mistakes in there. And it was just like a full-on own goal I shouldn't have made. The thing I always tell people, which is you've got to have a lot. When you have real conviction and inside knowledge, you've got to act on it.
[0:14:19] Given how strongly I believe that, when I've missed the mark there myself, it's particularly embarrassing. The courage of those convictions that you had early on had to continue. It's just the doubling down always kind of freaks it out. That's right. So Vivian asked, next steps for WorldCoin? And maybe for those people that aren't familiar, can you give a quick background on WorldCoin and what your hypothesis was, what we were trying to achieve with it?
[0:14:46] So WorldCoin, I think I'm the chairman of it, or at least I'm on the board, and I helped get it started, but I'm not that actively involved. It's basically a UBI cryptocurrency that's going to try to use iris scans to identify everyone on Earth and give them a new cryptocurrency. But my main projects are OpenAI and Helion, so I know a lot more about those. I believe the next step for WorldCoin is just to launch, and I think that's coming up in a few months. Excellent. Okay.
[0:15:15] All right. And then, you know, a lot of your writings and things talk about luck and how to enhance luck, how to make luck, how do you factor luck into strategy? You know, what if someone told you, hey, you just got lucky? I mean, how do you respond to luck and how do you view luck? I mean, I for sure have gotten extremely lucky, but I think that luck is a random variable. There's not a ton you can do about it.
[0:15:46] A thing that I think I have done well is just to try to take more shots on goal than anybody else. I assume that luck is a random variable. I think I have had more than my fair share of it for sure. But I have also gotten unlucky a lot of times. And again, that's how it's supposed to go. So I think you can learn this very important lesson. Everybody learns it at some point in their career. The earlier, the better.
[0:16:15] Which is like, luck goes both ways. And when you get lucky, you don't want to feel too great about it. And when you get unlucky, you don't want to get demotivated. That's the hard one. So if you do something that doesn't work, you just like try the next thing with no loss of enthusiasm. And no embarrassment and no, you know, questioning of your own ability. And like the luck will eventually come. And you just want to play the game with as many shots on good shots on goals you can get.
[0:16:44] Yeah, it's not like gambling where the house wins over time. You're going to eventually get and pick up your share, fair share of it. Yeah, the odds are in your favor, but the variance is huge. So there's lots of shots. Lots of shots. So speaking of shots on goal, Michael Lawrence student asked, you raised a billion from Microsoft and later stated that heavy funding would be required. You know, this is for OpenAI, so you can talk a little bit about the hypothesis with OpenAI.
[0:17:13] How does a sizable infusion from a strategic partner like Microsoft set you up for following rounds of funding or maybe an IPO? Any disadvantages when you look at that? Look, we have been thrilled with the partnership with Microsoft. They don't always go this well, but this one really is. I suspect we will need to raise tens of billions of dollars. But my hope, my first choice would be to raise it all from Microsoft. They're a great partner. It's a very complimentary relationship.
[0:17:43] And we think the same way about AGI. So I'm not excited to go public. I would, I mean, we will if we have to. But if we can raise it all from Microsoft, which is one of the few entities in the world that could actually do the funding at the level we need if they wanted to, then that would be awesome. And I think there's like, you know, like all of the best partnerships that we just like, it's all about sort of...
[0:18:12] Shared goals, trying to make sure we're successful in a good relationship, and we'll just keep going at that. So talk a little bit about OpenAI, hypothesis, AI. Tell us kind of where you think the future is headed in that space. And a lot of our students are here studying business analytics, and it's not so much the number crunching, it's the interpretation of the results, right? That's what we're trying to teach here. So share with us kind of your thoughts on what this future AI looks like. You see it a lot in the business and B2B environments, accounts payable, accounts receivable, OpenAI, that kind of stuff.
[0:18:42] But what direction are you guys headed at? I mean, we have figured out how to make legitimate intelligence in a computer. It's not a parlor trick. It actually learns. It actually can combine concepts in novel ways. It can figure out new things. And this, you can see this now starting to really take effect with some of our products. We have something called Copilot with Microsoft.
[0:19:12] Developers that use that, it's now right in half their code. There's an AI doing half of people's jobs. We have another thing called DALI, which generates images. And there's a lot of benefits to that. It's generating tens of millions of images a day for people. But that's doing all of Illustrator's jobs. And then you have our text models like GPT-3 and beyond that are doing a very wide array of tasks. But we have actually figured out how to put...
[0:19:40] Human-like cognition into a computer. And that, I think, will be the most significant shift, probably, that we see in our lifetimes of the technological landscape. I think this will be bigger than any of the great technological revolutions of the past. Like, this will be bigger than the Industrial Revolution. It will change every aspect of society. And I think it's going to be, on the whole, just tremendously good. But it will certainly be a disruptive time.
[0:20:07] And so, you know, knowing that this is coming down the pike in terms of, you know, the future, five, ten years out probably, right? And it's going to reshape the landscape of the corporate environment. What should students be studying, learning, thinking about today in anticipation of what's to come? Because there are, like you said, Illustrator will be gone. Other things might be replaced. What's someone to do?
[0:20:34] I honestly don't know. And so my somewhat evasive but true answer is in a world with this much dynamism coming, the best thing to do is to just be very good at rapid adaptation. And like I have myself been trying to think about how I can get better at that because I think the landscape is going to change so fast and in such unpredictable ways. Agreed. Same with higher education. We can both agree on that one.
[0:21:03] 100%. So moving on here, a few years ago, you were really supportive of UBI, universal basic income, you know, writing that you're fairly confident at some point in the future, as technology and traditional jobs and whatnot, that it'll be on a national scale. Has that view of universal income changed since COVID? Or we still kind of on that belief track?
[0:21:26] Even more excited than I was then. At OpenAI, we're running the largest basic income study ever, and it's going really well. I'm very convinced it's an important thing. And I think there's no way this is a sufficient change, but it is necessary. I think it'll be one part of a package. In a world with true superhuman AGI, where there is a computer in the sky that you can ask anything and it can do it.
[0:21:55] Far more effectively than all of humanity put together. I think capitalism is going to break in its current form. I think redistributing money and wealth will be important, but not the most important thing. The most important questions will be, who gets to govern how such technology is used? How do we do that fairly and globally? Who gets access to it? How do we divide up the time slices? Because it will be the most important resource. What do we do to find?
[0:22:21] Meaning and to spend our time in a fulfilling way, which again, I'm optimistic that we can do much, much better than we do today. But we'll have to find new ways to do it. Yeah, this is great. So, you know, you've talked already, but I think the most radical belief you have about the future, whether it's AI, UBI, where does that all sit? Maybe there's a third one you want to add to the list? I mean, I think the conventionally most radical belief I have is that
[0:22:52] This is the decade where the price of intelligence because of AGI and the price of energy because of fusion are both going to trend towards zero. And those are the two fundamental constraints on society. If you can get both of those driven towards zero, you can have an unbelievable increase in quality of life possibility, everything for everyone. And I think in some sense, most people don't believe either of those. And so it's very radical.
[0:23:20] In some other sense, the most radical thing I believe is just that entrepreneurship is the most magical force in the entire economy. And the fact that those two mega changes are going to happen by companies that didn't exist a few years ago, and that anyone can just go start those. It's extremely hard. It's extremely painful. There's not a lot to recommend it from a lifestyle trade-off perspective. But from an ability to have impact.
[0:23:49] The fact that startups do these things now, that's a radical belief that I'm very excited about. This is great. So I can't solve fusion. Maybe some of our Clemson engineers can, but we can work on education. I tell my students all the time, there's three things on my phone, New York Times, Wall Street Journal, and Hacker News. And that's literally what I read daily and religiously. What about you? Yeah, I'm on it all the time.
[0:24:17] I'm always flipping. So what's on your list? Like, where do I get my education from? And we're in Greenville, South Carolina, right? It's not Silicon Valley. We don't have this network effect per se. So where do I go to get, absorb that content? Where else do you look? I read a lot of Hacker News, a lot. And, you know, everybody finds the ways that they learn.
[0:24:47] My two are reading books and papers and then talking to people that I want to learn from. So that's what's worked for me. But I think other people have very different ways of doing it. And you've got to find what works for you. Yeah. All right. Last but not least, you're still pretty young. What do you want your legacy to be, whether it be politics, whether it be AI?
[0:25:10] I will go think about that in 50 years. In the meantime, I just am going to try to put one foot in front of the other and keep working. Awesome, Sam. Well, I'm not going to take up too much of your time. Thank you for joining us. I also want to make a little shout out here. Carmen, thank you for putting this together for us and making this happen. Do you want me to do the questions in the chat real quick before I have to run? Yeah, if you've got time, sure. Fire away. In terms of institutional funding and partnership,
[0:25:39] Ships being more beneficial than traditional VCs. It totally depends on the particular investor. I used to have all these rules about this. And then I realized that the contract, as in most of life, doesn't matter that much. The structure and the overall things, it's all about the specifics. And it's if you find this partner that you think you want to be in the trenches with, that gets you and that will give you reasonable terms and is a good person to work with.
[0:26:06] Make the decision about who you raise money from entirely on the individual, not the sector or the structure. It's one of the most important choices you make early on. In terms of the continuation of human creativity with the progression of AI like Dolly, if you asked experts in the field a few years ago, they would have said first AI is going to come for the blue collar jobs, truck driver, whatever, then sort of the low end of white collar labor like a, you know.
[0:26:33] Paralegal or something. And then the hardest stuff, like a mathematician or a computer scientist. And then finally, last of all, if ever, but maybe not because maybe it's magical and human, is creativity. And now we can see that it's going exactly the other way, which is sort of interesting. I think there's like a bunch of reasons for it, given the shape of the technology, particularly that it's not very robust and creativity can handle errors in a way that truck driving can't. But creativity is, I think, the first thing where AI is doing all of the work.
[0:27:03] Humans are finding amazing new things to do in the same way that artists, once the camera came out, found amazing new things to do. But it is like visual art and soon other kinds of creative work is forever changed, no question. This is a new tool, I think, on the order of a camera for visual artists. And what it means to be an artist is going to be really different. And it's going to be how effectively you can use these tools to do amazing things.
[0:27:32] On nuclear energy, I think if we're in the fission world, we have a lot, a huge PR battle in front of us. We should still do it. And that's the biggest thing to overcome is the PR battle and the associated risks and helping people about those. But if we end up in the world of fusion, it's very different. And I think the most important thing is that we call it fusion and not nuclear fusion. If you pull people on fusion, they really like it. They want it. They understand that it's safe and different.
[0:27:59] And the characteristics are very different. If you call it nuclear fusion, it's a surprisingly different response. All right. Thank you for having me. Thank you so much, Sam. Thanks for joining us for the Nuclear Leadership Series. Everyone, it's class. So enjoy the rest of your afternoon and residency weekend. Thank you again. Bye-bye.