Marc Andreessen at Startup School SV 2016

Okay, so, uh, up next we have a fireside chat with, uh, Mark Andreessen. Uh, come on out, Mark! I think he's right back. [Applause] There, they know who you are. No introductions needed. So, uh, thanks for coming back. This is, uh, your third startup school, and I think maybe you're the only person who's been on stage both as a founder and now as an investor.

So, uh, I won't go through your history, but what I'd like you to kind of talk us through is how you came to the decision, after starting a bunch of tech companies, to become an investor. Uh, there are a lot of venture funds that exist, so why and how is a16z different?

Yeah, so the big thing we—we came up—Ben and I came up in the 90s, uh, in the industry, and at that point, um, venture—the good news is venture capital had professionalized. And so, you had these great venture firms, like Perkinson, Benchmark, and SEO, others that had gotten really good at what they did. Part of the professionalization was they had tilted very strongly away from founder CEOs and towards professional CEOs.

Um, and, uh, there were good reasons for that that we can discuss. Um, but our view of the industry is that not all, but many of the best companies in the industry are run by their founders for a long period of time. So, IBM run by the Watsons—two Watsons, actually—uh, father and son; Hewlett Packard run by, uh, HP and Packard for 50 years; many other, you know, many other examples—Microsoft, Oracle, uh—many others.

Um, and so the model had just kind of drifted away from that, and the kind of default move became bring in the professional CEO. What do you mean by a professional CEO? Professional CEO. So, um, the basic thing is, well, there was what they would—there's what they would say, which is what they said to us at one point, which is, um, “Oh, so when are you guys going to hire a real CEO?”

Um, and so, you know, there’s sort of the—the general version of that answer. Um, the specific thing, um, when they say professional CEO, generally they talk about somebody with a lot of business and a particular sales experience. Uh, so, somebody who's very comfortable with taking something, you know, taking something that somebody has built—engineers have built in the lab—and then taking it out into the world.

And so, it's everything: raising money, marketing, sales, finance, um, you know, organization management—all the things involved in building the business and the company, uh, around the product. Um, and so, we just have, you know, we had this idea that if we sort of want to tilt things back a little bit—now, we're not religious about the founder CEO.

And there are a lot of founders who don't actually want to be the CEO or want to partner with a great CEO. And then there are also founders who can't be the CEO. And so, you know, there's sort of no—it's like to say, there's no concept of CEO tenure, right? CEO is not a guaranteed position. Like, you do have to perform, and if you don't perform, by the way, it's not the VCs that are your problem—it's that your company collapses.

And so, you know, it is important to perform, but we did believe in the generalized model, um, that founders are really important and that the founder CEO is a good model when you can have it. And so—and just what we observed was there was not, at that point, a venture capital firm that really had developed an operating model around how to do that.

And so, we decided to start our firm, and then we decided to really bear down on two things, which has been our big areas of focus for the last several years. So one is all of our general partners, um, have built and run companies before. Um, and so, um, we like to talk about the importance of having people on your board who have been through the struggle. Um, Ben, in his book, uh, “The Hard Thing About Hard Things,” he talks about the struggle being just all the stuff that goes wrong when you're trying to accomplish something big.

Um, I call it the—he's more polite than I am, um, but it's just all the stuff that happens. And you know, we could talk for hours about that, but to have somebody on your board who's really been through it. Um, and so, all of our— all of our GPs, by definition, uh, fit that criteria.

The other thing that we did—and this is a huge part of the firm for us today—is we built a whole operating model around basically—we call it giving the founder superpowers—basically the equivalent capability give a founder who’s not operated in business before the ability to operate like a professional CEO would be able to if they just walked in the door.

And a lot of that has to do with network. Um, and so the network of all the customers that you're going to sell to; the network of all the later-on investors you're going to raise money from; the network of all the big industry companies that you have to work with; the network of all the reporters, uh, that you want, you know, to get your story out; the network of all the engineers that you need to recruit; the network of all the executives, uh, that you need to recruit; the network of all the regulators and policymakers that you need to deal with in a lot of these businesses.

And so, we now have 85 full-time professionals in the firm that build and run these very large networks on behalf of the portfolio. And so, a founder works with us. Uh, Jeff—uh, Jeff Wo, the founder of NutriBox—described it as, “You plug into the Matrix. You just kind of instantaneously get this upgrade as a new CEO or even an experienced CEO. You just get this big upgrade where all of a sudden you're plugged into the world, uh, at scale.”

Um, and that—and that's working pretty well. Did—did you think about this back in 2006, 2007, before the firm started? Did that evolve as you, as you kind of built out the firm?

Yeah, so that's all evolved, and we've had, you know, a whole bunch of people join the firm that have helped us figure this out, um, and that run all these—run all these—all these teams and networks for us, um, that have really helped us advance our thinking. But the core idea was there, you know? Yeah, sort of a16z was a startup, um, and you know, having done startups before, we were—we were big fans of thinking it through and having a differentiated strategy.

And so, we spent about a year and a half planning this out before we started the firm, and these ideas were at the core of it. So, uh, a lot of first-time founders in the audience, and kind of raising money—, as we just did kind of pitch practice here, uh, is a little bit of a black box. What is the—just very precise way that, you know, what's a process that you use to work with companies?

How do they get a hold of you, and what actually happens after the first email?

Yeah, so I would really separate, uh, fundraising— and this has gotten a lot more complicated in the last few years in good ways. Um, I would separate fundraising of fundraising from, um, angels, seed investors, um, uh, accelerators, um, uh, you know, sort of accelerator rounds, pre-seed rounds, seed rounds, seed extension rounds, post-seed rounds. Um, I'd put all those in one bucket.

Um, and there—and that's a—you know, that's a process that the—the YC plugs you into, and demo day is organized around that. Um, and, uh, your, you know, your friends who have had a startup will have been through that, and so there's—there's a whole art to that, and that's—that's a lot of what gets talked about.

Um, then there comes a time when you engage in a different kind of fundraise, um, and what I would call an institutional fundraise. Um, and I think that really starts with a formal Series A, um, or anything equivalent to a formal Series A. Uh, it also applies, by the way, if you're going to go try to raise money from, you know, a corporate investor. Whatever, it's an institutional process; you're dealing with somebody who's running a highly professionalized firm, you know, with real fiduciary responsibility back to their LPs. You know, their firm, you know, a lot of venture firms have been in business for 30, 40, 50 years.

There's a way of doing things, and things just get more formal. Um, and so, I—I think that that’s the first really important thing, uh, to think about. Um, you know, when we—and we're actually schizophrenic on this because we do a lot at the seed level, and when we invest at the seed level, it's a lot like dealing with any other seed investor.

It's, you know, are you smart? Do you have a really good idea? Do you have a great deck? You and let's see how it goes. Um, but when it's a five or ten million dollar round or B round, then things get more serious. And so, when you mean serious—so, uh, let's say I am raising that $10 million round. Uh, and I get to give you an email and say, “Hey,” and then I come in, and what happens? Is it just a series of meetings and then diligence, and then you have a big check, like in those game shows?

Yeah, exactly, exactly. Exactly! It shows up at your house when you're least expecting it. Um, so, um, really, really impressed the spouses.

Um, so, uh, so the first thing is how do you actually get in? And so there's—there's actually an argument that kind of plays out on—on this topic in the valley, which is the importance of what are called warm referrals or warm introductions. And so, the way that most of the top venture capital firms work is basically they’ll take you seriously if you come in introduced by somebody they’ve worked with before, and they won’t take you seriously if you don’t.

And there’s an argument that plays out in the valley, which is that that’s sort of unfair and unreasonable. And you know, what about all these—you know, you guys talk about all these founders all over the world, and like why can’t they have access the way the people who are already connected can? Um, the argument in favor of the warm referral is that it’s the first test.

Um, and it’s the first test of the ability to basically network your way to the investor. And basically the way the investor thinks about it is if you can’t figure out a way to network your way to a VC firm—which is of course in the business of meeting founders—then you're unlikely to be able to network your way into hiring, right, a great team or network your way into selling your product to customers.

Um, and so actually I think, I think that the role of the warm referral is misinterpreted. I think you just view it as, like, it’s the first test, and so I think it’s a test that you just want to pass. You don’t want to take any chances on that. The good news is if if you're connected to YC, if you're connected into the valley angel seed ecosystem, um, that’s a very easy test to pass. If you’re not connected into YC, then, you know, the thing to do is to get to the valley, right, or get into the network and get—you know, get into YC.

Good luck if you’re not connected to YC. Yeah, like YC, YC has been a huge—like YC has really widened the aperture for founders to be able to plug into this process, and I think in a very healthy way. Um, and so that—that works really well.

So that’s the first is the warm referral. Um, then there are two theories—one is going high, the other’s going low. Um, and, uh, you know, so one theory is—and you’ll hear this sometimes—is, you know, only ever talk to senior partners at the firms, and don’t waste any time with the more junior partners, the younger people in the firm because they can’t make the decision, pull the trigger.

Um, the counterargument is that the senior partners at these firms, um, make very little—very few decisions from scratch. The senior partners only get to invest in a few companies a year each, and—and they make very few of those decisions. If they're going to make a decision on their own without any work done by the junior people in the firm, it's probably going to be somebody they've already worked with before.

So it’s actually really hard from a sort of standing start to get just a straight yes from a senior partner. So the thing that works better and the thing that I would do if I were, you know, 22, 25 again, um, I think there’s a huge advantage to coming in with the junior people, um, and get networked into the firm and spend a lot of time with the junior people, understanding, number one, just like pre-flighting.

Like, how is this coming across? Like, how is this going to be received? Because they, you know, they can read the minds of the senior partners; they can tell you how these things are going to get received, and they can help you. Um, and then also, you know, we—frankly, the senior partners at a firm like ours look to the younger people a lot of like, okay, like what’s interesting?

And we count on the younger people to stay very connected, especially to the first-time founders. Um, and so I really encourage people to actually take the younger people in these firms seriously. Um, so from that, usually the process is a first meeting with one of—one of—one of those folks, which might then lead to a second meeting with a broader group of those folks, maybe with one or two general partners.

So you kind of have this kind of windup kind of period, and you’re—you’re kind of working through it, and are we talking about days, weeks, months? You know, it depends. Um, in a hot environment, um, it’s, you know, days to weeks, because you, as a VC, you assume that deals go down relatively quickly.

Um, like in—in the, you know, for the last several years, I think days to weeks has been the expectation. I will tell you, in the cold environments, you know, in 2003, you know, we’re in six months, nine months, twelve months, right? VCs said nothing but time, um, because, uh, you know, it was so hard for anybody to raise money for anything.

You could just let these things play out for a long time, but these days it’s—it’s days to weeks. We—we try to move in days to a small number of weeks. Um, then the—the big event is to then progress through, you know, one or two or three meetings, and then—and then the way you really know you're making progress is if you get invited to present to the full partnership.

And that’s usually, at most firms, a Monday meeting, and that’s a formal—like, that’s a full-on formal, and that’s like the general partners at the firm are there, like the whole firm is there. We’ll—we’ll have, like, sometimes we’ll have 40 people in the room for that. People come in, it's like presented to the army, but like we—we get everybody; then the founder actually just presents their company.

Yeah, and it’s—it’s a presentation. And—and so then this gets into another debate, which is like, okay, you know, you smart guys, like you go on and on about like creativity and all this stuff. Why do you want the founder to stand up there like, you know, like an idiot for, you know, 50 minutes, reading off of PowerPoint slides? You know, like why don’t we just, like, why don’t we have a conversation? Why don’t we do things casual?

Why don’t we—like, you know, why didn’t you do more research up front? Like, why do we do this formal presentation? And again, I think, again, I think that's misinterpreted. I think the formal presentation is another test, which is, as a founder, if you're not good enough at your job as a CEO to get up and present to an institutional investor for 60 minutes and sell them on your thing, like we're easy.

Like this is the thing I think is misunderstood. VCs exist to give out money. Like, like we’re in the B—they love when somebody walks in and has a compelling pitch and we can give them a check.

Like that’s a successful day for us! In contrast, every other pitch you’re ever going to make is going to be to somebody who’s going to be much worse than us, right? Like customers, right, are going to be like, you know, no, I’m not going to give you any money.

Like is their default position, right? You know, engineers—you’re going to try to recruit engineers and they’ve got 20 other job offers, right? And so why is your pitch going to be so much better than the other 20 startups, right? Um, and so, again, I just—I view that as, like that’s the second big test, which is can you successfully execute a formal pitch?

Um, and then if that goes well, then generally you go into, um, you know, there—if it’s like super, super smoking hot, you could—you could move to term sheet very quickly after that. Usually it’s a week or two of additional work to kind of have some more discussions. Uh, various people follow up, you kind of get into the process of, you know, the—the firm—the venture firm really trying to make sure they understand what you do.

Um, you start to negotiate the term sheet. Um, and then at some point, you have the term sheet discussion, and how many checks do you write a year?

So we see, uh, 2,000 startups a year that are warm referrals, um, and we can say yes, uh, for at the institutional level for A's or B's, uh, we can say yes to about 20. Um, and we're a pretty high-scale firm, you know, eight general partners, um, and so we're—and we're up in, you know, large fund size. And so we—we operate at scale.

There are other top firms that will say yes three, four, or five times a year. Um, and so—and by the way, like on the one hand that’s very daunting, which is like, well, you know, like that’s 1%, right? It’s 1% of warm referrals basically.

Um, and so that seems very daunting. Um, I can just tell you on the other hand, um, it goes back to this concept of the tests—like it’s not that hard to get good at this part, like—and—and so I just—I think it’s worth the time to really understand and get good at how to get networked in and how to make that pitch because if you’re good at that, then, like I said, when it comes time to go sell your product to a customer, when it comes time to go hire somebody, those same skills, uh, reapply in the exact same way, and I think people—people should just—people should just—founders should just get good at that.

So, uh, there’s a pretty big, you know, gap between kind of an average engineer and a really great engineer, and—and there are ways to test that. Uh, how do you know an investor is actually good if you’re a founder and you’re raising money for the first time or, you know, you haven’t raised much money?

Yeah, so I think I have, uh, I think there are five key ways that a founder can know whether or an investor is good, um, and they are references, references, references, references, and references. What was the third one? The references? Yeah, references are really important.

Um, so, um, the good—the good news with the valley is super networked. Um, as a consequence of that, you know, we're people know. Um, so I’m a founder, and I, uh, you know, I’m going to seed fund B. What does it mean to do a reference?

Yeah, like do I just find people they’ve invested before? Or? Yeah, I mean, so there are sort of two different kinds of references. There are so-called front door references, which is you just—you ask me like, you know, who should I talk to? And I’ll give you a list of people who I hopefully know will say good things about me. Sometimes I’m wrong.

Um, then there’s the—those—the back door references, right? Which is—usually where the—the value is, which is people who have worked with the investor before—other entrepreneurs or people who have been at companies that that investor has funded. Or, by the way, for that matter, other investors, right? You can—you can reference—like the way a lot of people get reference checks is with the, you know, the angels, um, and so your angel investors, your seed investors, if you’re in YC, the YC partners, other YC founders—the internet, um, Twitter, um, uh, you know, these discussions take place, uh, you know, all over the place now.

So, um, so I think you just—you try to—it’s just like anything else when you’re evaluating people, is you just try to get to other people who have worked with them in the past. Again, I think that’s a general skill. I think that’s a good thing to know how to do because if you can reference check an investor, it also means you can reference check a candidate, yeah, right? Which is really critically important.

And I will just tell you, like in the valley—like in the valley, like there is a way to hire great people, and it involves references, references, references. Um, you just—everybody who’s good at hiring, what should one listen—let’s say I’m doing first reference check. What should I listen to that is like the double talk of, “This is not a good investor”? “This is not a good...”

Oh, yeah, yeah, yeah. So it’s all in the little tells, right? So it’s all in the—like if somebody’s good, they’re like enthusiastic. They’re going to be beating the table like, “This person is great. You should take their money!” Like if you can get them on their board, you know, your board, you’re lucky. If you do, you get anything other than that, right? That’s like, “Uh-uh,” like trouble, right?

And it’s—it’s like the silences, like, you know, so what’s a person like in board meetings? And then there’s like the silence, and then there’s like the breath, like, you know, and it’s like, “Okay, great, trouble.” You know, here we—here we come. Um, or, you know, the other thing is that, you know, a lot of people don’t want to say bad things about people, even when nobody else is listening. And so, they’ll say these lukewarm—I mean my favorite is like, “Well, they’re very punctual, right? Sharp socks. They’re always on time, right? Very well dressed. Um, you know, always brought donuts,” um, yeah.

And so you just—you gotta look even for the back door with the front door references. That’s what you listen for, but—but the back door references, even still, that’s what you listen for. And even when it’s people who you know really well, they still may not actually tell you the full thing.

Um, and so you just—you just want to test really carefully. I just think it’s an incredibly useful skill. If you learn how to do it, you can basically test this out on investors, and then when you—when it does matter for investors, but then what it also really matters is when you’re hiring, especially executives later on.

So, uh, I want to step back a little bit. Uh, lots of founders—investors look up to you, and— and your ideas. Who do you look up to, and how do you make sure your ideas are actually, like, continue to be good?

Yeah, so, um, I’m a big fan of history. So there’s a sort of a classic cliché in the valley, which is like we don’t respect history very much. And a lot of people who visit the valley for the first time, that's their impression because they—I don’t—people may have had this.

I had this experience when I first came to the valley. It’s like, “Okay, I want to see Silicon Valley. I want to drive around and see Silicon Valley.” And it’s like, “Good luck with that!” Right? Like, okay, there’s a freeway, there’s a strip mall, there’s an office park, there’s a security guard who won’t let me into the office park. Like there’s no physical history here, right, really?

And then—and then in the conversations, I mean we're all about the future—like we're building the future—and so I think there's this natural tendency to assume that the past isn't very relevant. And I actually think—I actually criticize a little bit, you know, there’s sort of the Elon school of thought now, which is very strong in the valley, which is sort of think everything through from first principles, which has its huge strengths for sure.

Um, but it does kind of dismiss the idea that people who came before us had anything to teach us. Um, and my general view is the people who came before us, um, had a harder time doing what they did than we do because the world was in a more immature state. Like they didn’t have all the—they didn’t necessarily, you know, startups started before 50 years ago. They didn’t have venture capital, right? They didn’t have—there’s a lot of things startups before 20 years ago didn’t have. The internet—things were harder in the past, and so the people who were successful in the past, I think, were often better, uh, than we are because they had to be.

Uh, because it was harder. And so I like going in the—in the past, um, in the tech industry. Um, and I’ll just—I’ll recommend a few books. So I think the orig—the origin stories of companies—I think IBM. There’s a fantastic book about Thomas Watson, Sr. who built IBM by Kevin Man.

Um, and if you think that Steve Jobs is rough—was rough on people, um, he had nothing on Thomas Watson, Sr. There’s a—the book is—the book has transcripts of the executive staff meetings at IBM in the 1930s and 1940s, and, um, let’s just say he was an absolute terror and built an extraordinary company.

And so he’s a real role model. Um, not the terror part as much, be careful how this is all interpreted. Um, Bill Hewlett and Dave Packard, Michael Malone wrote a great book called “Bill and Dave,” talks about how HP got built over the course of 30 or 40 years—by Bill and Dave.

Um, and then I like going back further. So I like Edison. I like what was called the second industrial revolution a lot where, like, you got electricity and you got cars. So there’s a great Edison book called “The Wizard of Menlo Park.”

Um, it’s really good. There’s a great Ford book called “I Invented the Modern Age.”

Um, and if you go back and read the history of cars 100 years ago, Detroit 100 years ago was a lot more like Silicon Valley today than I think people even understand. In fact, Henry Ford, like Ford Motor Company, was not his first company, right?

So there’s—there’s all these very interesting backstories. And then I think you can go back even further. I like going back, um, and—and this is not to compare, you know, us to—to this, but just as inspiration and role models.

Um, Florence, uh, you know, at the of the Medi—Leonardo da Vinci—I just got this great book called “Da Vinci’s Robots,” and it turns out Da Vinci, while he was doing everything else that he did in life, he was also designing robots.

And in his, you know, sketchbooks, there are fully-fledged designs for, like, everything mechanical, you know, all the Boston Dynamics he was like, we’re trying to invent all that stuff 500 years ago. Um, didn’t quite have the—you know, didn’t quite have the technology to be able to pull it off, couldn’t quite, you know, go to—couldn’t go to TI and get the microcontroller.

Um, so he had some issues, but, you know, he’s an inspiration. Um, there’s another great book called “The Lunar Men” about something called the Lunar Society, which was in England about 250 years ago, like James Watt invented the steam engine, and all these other guys who were doing this kind of thing back then all worked together. It’s kind of—they had kind of a way thing going back then, um, and so—it’s all basically—it's just, you know, what it is, you're trying to look for, okay?

The world is—you know, the points in time—the world is the way it is. Everybody takes it for granted, and then there’s a person or a set of—usually a set of people who think no, no, there’s a better and, you know, different and better way to do things.

Um, and then just all the practical challenges involved in causing that kind of change to happen and trying to build, you know, whether it's a new company or trying to build a new movement of some kind or a new kind of art. Um, and I think there’s all kinds of lessons to be drawn from all that.

And do you extract ideas from those people? Or, because there is—I think one of the problems of being in a—in the valley is there's just such an echo chamber of opinions. For every opinion, there is a counter opinion which is just as loudly and—and aggressively, uh, drummed. How do you—how do you separate what is actually good advice and what is not good advice as a founder?

Yeah, well, you just—you go back and, right, I guess what I draw from history is like it’s always hard. Um, and so you just—like Henry Ford, as an example, like there was nothing easy about building Ford Motor Company.

Um, or Thomas Edison—there was nothing easy about electricity. Like electricity—I mean, talk about like you would think that that would be a relatively easy thing to get people to want, um, and it turns out there was actually a lot of effort and work involved in getting people to think that was a good idea. Um, and so, um, uh, it just—it’s—it’s so—it goes back to Ben’s idea of the struggle.

It’s just there is so much work and effort involved. There are so many false starts; there are so many things that can go wrong. Um, and I think you draw on all the different lessons and theories from then and now, and you kind of view them as a mental toolkit to draw on.

But I guess I would say one of our beliefs that Ben also talks about in his book is there’s no silver bullet. Like there’s the temptation when you get kind of in the—it’s always like there’s gotta be a way out of this, and it's gotta be one thing. There’s gotta be some idea, or to your point, some piece of advice, right? Or some way to reframe—you know, if I could just reframe the pitch this way, or if I could just hire this person, or just get this customer, um, or something, you know, it's going to fix everything.

And we always like to say there are no silver bullets, there are only lead bullets. Um, and so, you just have to kind of do all the work. Um, and most of the work is just a grinding—it’s just, you know, grinding labor, and there’s just no substitute for it.

Um, and I think that, frankly, I think that comes out of the history as well. So, um, what do you think is, you know, we're kind of in the tail end of 2016. What’s in store in the next couple, two years, two and a half years for the startup ecosystem? Because there’s been a lot of talk and a lot of different opinions about both in verticals and where verticals are going in just the larger community and ecosystem, right?

Yeah, so I guess, you know, maybe two—two schools of thought. So, um, as you’re all probably well aware, um, experts, um, have been declaring an imminent bubble and crash continuously since 2004. Um, so we’re now in year 13. Um, next year will be year 14 of declarations of an imminent, uh, com crash redux all over again.

Um, they are—these—the people who predict this are getting increasingly frustrated that it hasn’t happened, um, and increasingly irate that we’re all still here. Um, uh, you know, um, maybe they're right, um, and we've just, you know, completely reinflated, you know, dot-com bubble 2.0, and we’re just—it’s just a—and now we’re just delaying until the inevitable crash—it’s possible. Um, you know, I mean, you seem skeptical and skeptical.

Well, that because the other—the other theory, right? So the other theory is we’ve—we’ve entered a phase of the industry where, you know, some things are just working now. Um, and in particular, right, the internet—the internet is working now, which if you were here 20 years ago, that was not necessarily a certain thing, and like the internet works; you know, mobile works; uh, the smartphone works.

Um, uh, you know, God willing, the new Apple headphones will work. Um, you know, some stuff is just starting to work. Um, and so you’ve got these marvelous—you’ve got this, you know, giant install base now—people all over the world—on the internet. You’ve got a, you know, this massive growing consumer acceptance of e-commerce. You’ve got these giant new social platforms.

You’ve got these, you know, all these new—these new playbooks for how to build these companies. Um, uh, you got the rise, you know, developer adoption of enterprise technology is like a whole thing that’s working now. Um, and so these things are working at very big scale, and each of these things that—the big things that work are all platforms to build on top of, right?

They’re—they’re kind of opportunities for more things to then get built that, right, they’re—they’re kind of opportunities for more things to build that right on top. And so I think a lot of those things are working really well. So, which isn’t to say that everything’s working right? And so I think the—the valley and startups will always be characterized by, you know, there will always be what's—what's called a power law curve where you’ll always have some big successes, and then you’ll have a long tail of things that don’t work, and you’ll have a bunch of stuff in the middle.

And so you always get this kind of distribution of success. Um, but you have enough things that are working at scale that are serious. Um, and so, you know, I think the things that are working are working really, really well. What—what do you think in terms of—or do you believe in this kind of thematic of these three or four markets or these three or four spaces in the next couple of years are really going to be most likely the places where some of these outlying returns are going to sit, or do you—are you more agnostic?

Yeah, we’re totally schizophrenic on that point. Um, so, um, anybody in the field can’t resist talking about this all the time, and by the way, there is substance to it, right? It’s like there—there was a point at which, you know, there was a point between, you know, when the iPhone came out to probably 2012, was like there’s this point where all the mobile killer apps that were going to be on the home screen of the iPhone were going to get built—and, by the way, got built.

Um, there was a point where, you know, early—earlier, there was a point when the PC came out that all the PC killer apps are going to get built. So there are these things that happen, these new platforms get established, and then—and then there’s this period of time where things get built on top. There’s a bunch of candidates for that today that are very exciting.

So broadly, I give a few examples: AI, machine learning, deep learning is one cluster of things. Um, very, um, you know, tons of activity happening there. Intersection of, uh, biology, healthcare, um, and computer science—a big cluster of things happening there. Transportation, right? Autonomous ground vehicles, air vehicles—actually now, sea vehicles of different kinds, uh, different sizes, shapes, descriptions, uses, um, are really starting to click.

Um, and you know, we could probably name another half dozen, so, you know, and we—part of our job is to track those fields and try to—try to figure out the best new startups. How do you guys actually track those fields?

Oh, just—it’s just a constant—it’s just kind of the action of the firm, the flow of the firm. So it’s just constantly trying to review, trying to, you know, trying to map the landscape, trying to understand all the companies, um, trying to meet all— all the founders.

Um, so that’s the one hand. I would just say, on the other hand—on the other hand, um, I would say that I don’t think any VC really does that good a job predicting these things in the sense of you—first of all, it’s so—the—the big successes are so dependent on the specifics. They're contingent on the specifics, and so it’s the right founder with the right idea with the right team with the right breakthrough at the right time and the right market with the right, you know, pricing model.

It’s all these things have to come together, and you just—you never—my experience is you just never know when that company is going to crystallize. Um, and it often happens five or ten years after you think it should. Um, and then it can often be the case that you get these companies that are breakthrough companies in fields that people thought were dead.

So we just—an example, we have this company, Tanium, people have written about now, which is a—it’s a security company. Um, and, um, it’s in a field where I think a lot of VCs thought that the opportunity to build big franchises was over. And like this company is is growing like unbelievably fast—very, but very specific, uh, set of people who really understand the problem and, you know, became available at a certain point to go after it.

They’re going after it, you know? VMware, um, is another great example—virtualization. VMware popularized virtualization. Virtualization was a 30-year-old mainframe technology; it goes back to the 1960s. Um, and Diane Greene and Mingle Rosen Bloom, uh, out of Stanford, um, Mendel out of Stanford and Diane out of Sun, started VMware in 1999 at a time when everybody knew that like enterprise software was on its way out.

And it turned into one of the big breakthroughs. Google—I always like to say Google was like the 35th search engine, right? It was like 34 other search engines had come along. Like we all knew for a fact that number one search didn’t work, like it would give you crappy results, and number two, there was no way to make money on it. Like those two things we knew for sure.

How—how do you know as a founder that you're not too early? Because, you know, we talk to a lot of founders, and—and how do you know that you're, you know, 35th rather than maybe first or second, that your timing is right?

So my experience is the great founders almost always feel like they’re—you almost always feel like you’re too late, and you’re almost always too early. Um, and you almost always feel like you’re too late. The reason is because for you, it’s become obvious. Like for all of you—You know, everybody listening to this, like you’ve got some idea in your head, and like as far as you’re concerned, the world should already work this way, which is why you’re pursuing it.

And so, to you, it’s a little bit inexplicable as to why it hasn’t happened yet, and there must be 80 other people going after the same idea, and it must be just about to happen, and I must be too late. Which is, by the way, which is actually how I felt with that—that’s Gabe. I was like we’re too late.

Um, in retrospect, we weren’t. Um, the reality is you’re almost always too early. Um, uh, we almost never see a qualified founder fail because they’re too late to market. It’s—it’s almost always because they’re too early to market, and I think, by the way, and I don’t say that critically; I think we make—when we screw up investments, I think that’s often the reason as well.

I think we’re in the exact same boat as the founder in that degree, which is we believe it’s going to happen, therefore we invest, and it just turns out the way the history gets written is the world just wasn’t ready yet.

So, I always like to point out like the—the iPad was like this breakthrough product in 2009, and you know Apple, oh my God, they invent all these things. Well, there was this thing called the Newton, um, 20 years earlier, uh, in 1989. If you haven’t heard of it, look it up on Wikipedia. It was the iPad 20 years earlier, and it was like virtually the same thing.

Um, and the world just wasn’t ready yet. The technologies weren’t in place—you didn’t have mobile broadband, you didn’t have the high-resolution screens, you didn’t have the battery technology—and you just—they brought the thing to market, and it crashed and burned, and it convinced people for 20 years that tablet computing would never work. And then they did it again in—in the iPad, and it worked.

Um, and so I—I think that in a lot of ways, that’s the biggest risk, the biggest conceptual risk we all deal with is just that. I think that’s the permanent curse of the entrepreneur.

Yeah, I remember you—you gave a talk at YC one time, and you said, uh, you know, if you—the one way to figure out if you’re not too late is if what you’re working on was being hotly covered three, four years ago. You’re probably right on time because now the infrastructure and consumer behavior has now caught up to what was, you know, exciting in 2012, 2013.

Um, last question, you know, through whatever fate or whatever—you lose everything, and at the end of the day today, you’re 22 years old again. Um, what do you start? What's—what's the company you would start?

So, if I’m 22 again, um, I think what I do is I go try to find the best, um, uh, the—the hottest company—the company in the valley that’s growing the fastest, um, and that has a really good culture and a really good foundation, um, for training.

Um, and so an example of that would be an Airbnb. Um, but there, you know, there are many others—even by the way, still Facebook. Um, I would go try to—I would go to plant myself at a company where I can really learn.

Um, fast forward now, you’re 25, you’ve got—you’ve got your three years at Airbnb in. What do you start now?

Well, so, I—I spend that time. So, one is I work my butt off to establish a good reputation so that when people reference check me, right? I reference check well. Um, but the other is I—I spend that time thinking, and I really critically I spend that time building my network and the people who I know that I want to work with, you know, down the road.

Um, and then—and then I either, yeah, I either come up with an idea or I don’t, but I—I would focus much—I think it’s very hard to—I'm deliberately avoiding your question because I think it’s very hard to—to—to answer in the abstract.

Let me answer it a different way: If I were 18 and I were going to college, um, I would do—I would really—I would do computer science again in a heartbeat. Um, um, and then I would either focus on, uh, cryptocurrency, uh, distributed systems, and the broad, uh, domain of cryptocurrency, um, number one. Number two, AI, machine learning, deep learning; or number three, um, U intersection of biology, uh, biology and CS.

So genomics and synthetic biology, I would personally—I would do those, one of those three areas. Um, and the hard part for me would be picking between them because I think they’re all—they’re all—to me, it just seems like all three of those are going to have transformative work happen in the next 20 years is going to change a lot of how the world works.

Awesome! Thank you for your time.

Good, thanks. [Applause]

Everybody, all right, next up is a quick bathroom break. If you lost your car keys, head to the registration to find them. Um, and we'll see you back here at 4:30.