Russ Roberts

Sam Altman on Start-ups, Venture Capital, and the Y Combinator

EconTalk Episode with Sam Altman
Hosted by Russ Roberts
PRINT
Continuing Conversation... Chr... Continuing Conversation...Sam ...

Sam Altman, president of startup accelerating firm Y Combinator, talks to EconTalk host Russ Roberts about Y Combinator's innovative strategy for discovering, funding, and coaching groundbreaking startups, what the company looks for in a potential startup, and Silicon Valley's attitude toward entrenched firms. The two also discuss Altman's thoughts on sectors of the economy that are ripe for innovation and how new firms are revolutionizing operations in these industries.

Size:30.1 MB
Right-click or Option-click, and select "Save Link/Target As MP3.

Readings and Links related to this podcast episode

Related Readings
HIDE READINGS
About this week's guest: About ideas and people mentioned in this podcast episode:

Highlights

Time
Podcast Episode Highlights
HIDE HIGHLIGHTS
0:33Intro. [Recording date: July 17, 2014.] Russ: Tell us a little bit about your background and how you got involved in the Y Combinator [YC]. We interviewed Paul Graham--is he the founder? Guest: Yes. Russ: Five years ago, which was about 4 years into the life of the Y Combinator. It's now about how old? Guest: 9 years. Russ: 9 years. So this is in some sense an update on the Y Combinator but it's also our chance to get to know you. So, tell us actually first what the Y Combinator does, and then how you got involved. Guest: So, Y Combinator is a firm that funds startups. We invest a small amount of capital and a lot of startups. We funded I think 716 so far. And then we try to help them. And our belief is having sort of this central organization like Y Combinator can make these startups way more likely to be successful. And in the time that we've been operating we've been fortunate enough to be part of some really successful companies. And I think helped them a lot. Really what we want to do is just make a lot more startups and thus a lot more innovation happen. Russ: You said 716. That's an enormous number. It's much more than a typical venture capital firm is going to be involved in. Guest: Way more. And probably in 5 years from now we'll do that many every year. Russ: And why is that a good idea? Guest: There are some network effects to this business. There is this great cultural value of Y Combinator of helping startups that are part of the community. And so by having this very powerful community, the start-ups can help each other. It also means other investors in big companies are less likely to maltreat any company that's part of the YC family. And there's some power there, and some benefits for the founders. And the more influential founders we have in the community, the more powerful the community becomes. Russ: Is there a physical location for the Y Combinator or is it a virtual thing? Guest: There is. We just moved into a new building, because we outgrew the old one. And we're in Mountain View. However, we don't let the startups work in our office space. That was kind of the bad, old kind of incubator. Russ: Right. That's an incubator model. The idea is they can share some oversight; they can share the copying machine. Guest: Yeah. It turns out that startups need to feel unique. Startups need to feel like they are kind of taking on the world and they are special and sort of this like chosen thing. So, the worst possible way to get that is to put everybody in one big bullpen. People feel like employees. People get distracted; they go out every night because it's always someone's birthday. The poach people from other teams. It feels like sort of like farce of startups. So we encourage our startups to work, usually, in their apartment, but certainly in their own office. Sometimes maybe share an office with one or two other companies, but not to feel like part of this incubation space. Russ: So, besides money, and being chosen--which confers a feeling of psychological benefit, I'm sure--what else does the organization provide for these firms that makes it unusual? Guest: Well, the main thing that we do is advise the startups. So, in the partners--there's 14 of us or something like that--we spend all of our time advising the startups. And we help them with a lot of what they need. And we see a lot of startups; we have a lot of data. And we can give them advice about their product or marketing or sales or further fundraising or legal issues or personal partnerships. Anything they need help with, we can help with. And that's what we do all day. Rather than have a number of people around the periphery that help occasionally, we just really get involved very intensively with these startups, meet with them whenever they need about any topic, email with them, 24-7. That's mostly what we do. The alumni network that you also get access to is very powerful--these 716 companies will try to do anything they can to try to help another company in the network.
4:33Russ: So, it's nice to have somebody to hold your hand, but of course the nature of a startup is, you are on your own. So, how do you balance that advising role with the need for them to be making their own decisions ultimately? Guest: We specifically are not hand-holders. And we tell them, in the first day, in the first dinner, we are not your boss. We will not hold your hand. If you want to fail, we will let you drive off that cliff. And luckily we fund a lot of startups. But we're very clear about our role. When you are an investor, you're along for the ride, but the founder runs the company. And some people freak out about this. There are plenty of founders that on paper would be very good, but their whole life is on this track and their whole life they've had someone telling them what to do. First they had professors that said, Turn this in. And then they had a boss at Google that said, Do this. And now they are off and they have to figure out themselves what to do. And some people stand up and rise to that occasion; and some people paralyze with fear. And some people just have no self-discipline. But it's very hard to predict who will do well in that kind of environment. Russ: Now, how do you discipline them when they call you 8 times a day and say, What do I do now, what do I do now, what do I do now? Guest: We happily tell them, like, You are running your own company. Why don't you like come to us with areas we can help you, but like no one but the founder is going to make these product decisions. You have to figure this out. Russ: Of the 716, how many have been phenomenally successful? Guest: Well I would argue at a minimum that anything over $100 million dollar valuation is phenomenally successful. There are more than 20 of those. Russ: What are some of the biggest successes? Guest: Oh--AirBnB, Stripe, Dropbox. Weebly, Optimizely, Zenefits, Teespring. Instacart. I could go on for a long time. One stat that I just heard that I love, is that since we sort of really got going, two years in, since 2007, half of the $10+ billion dollar technology companies that have been started have gone through YC. Russ: And they would be names like Dropbox, AirBnB; and then WhatsApp and then Uber did not. But a pretty good hit rate. Russ: Yeah, I'd say so. And the other 696--how many of them failed? Guest: Outright failure? I don't know. Maybe 100, something like that. Russ: That's all? Guest: Yeah. A lot more yet to fail. Russ: Oh, good. Oh, whew. Okay. Guest: But also a lot more yet to become--the question is like 10 years from now when the results are in, how many of the 716 that we've funded so far will have become these huge companies? So far there are like 3 that are over a billion dollar valuation. But I expect that when all the data are in, 10, 15 even, will be over that mark. Which I think proves that--if you look at the hit rate out of YC companies, hit rate over a billion dollars, versus start-ups in general, I think proves that there is some value to having this network and a community, and finding these early stage startups and helping them. If we're able to increase the success rate 10 or 100 times. Russ: It's hard to know, of course. Obviously some of them might have been successful without you. Guest: Well, some of them might be. But I think, if I had to pick the company that we have funded so far that was likely to be the biggest, I'd say AirBnB. And at the time we funded AirBnB, they were completely out of money. They were like living on credit card debt. They were like eating dollar store cereal. They had a lot left over because they couldn't afford other food. No investor would fund them. They were in the middle of what they call the Thousand Days of Darkness. Their product wasn't working--that was another problem. Kind of a big one. And we funded them when no one else would; and we also worked with them to figure out how to get growth and how to evolve their product. Russ: How old were they at that point? Guest: A year and a half, two years. Something like that. Russ: And what had happened in their company at that point. Guest: Nothing had worked. They first had this other startup. And then they were out of money. They couldn't afford rent. So they are like, Man, we really need to get some money quickly. And the idea of AirBnB was born out of the necessity of them paying their own rent. They looked at their assets; they had this air bed in this extra room, nothing else. They needed some money. So they figured out a way to find someone on the Internet. And then it evolved to people sharing their space for a conference; then it evolved from there and from there and from there. But that is a company that I am very confident would have died had it not been for YC. Russ: So, what do you think was--well, two questions. First, why did you guys think it was a good investment? What did you see or what did you hope that no one else did? Guest: It's sort of this, like piece of YC lore. But I will recount it as accurately as I can, given that it's 6 years ago and history gets somewhat re-written. They came in and pitched. Paul Graham, my predecessor, thought it was a terrible idea. Like shockingly terrible: why would anyone want this system? It's ridiculous. Russ: It is a horrible idea. Who wants to stay at--I don't like staying at my friend's house. Why would I want to stay at a stranger's house? Guest: Right. Horrible idea. And they had not yet invented of renting the whole apartment. It was really the idea of you were staying with a stranger. So they probably were on a pace not to get funded. But as we say at YC, we care much more about the founders than the idea. And the founders were clearly very good. And maybe that's one thing that was exceptional, which was at the end of the interview, they gave YC two boxes of cereal-- Russ: Oh, it's huge. Guest: they had made for the Presidential Campaign. This was the 2008 Presidential Campaign, Obama versus McCain. And the guys were graphic designers. They had made these ObamaO's and Cap'n McCain Crunch cereals. And they made these boxes of cereal. They printed out these things; they wrapped them around dollar-store cereal--that's why they had a lot of extra cereal. And they sold them on the Internet. Limited edition, 500 each, for $40 bucks. They made $40,000. Which they used to continue to operate their startup when nothing was working. And that is the kind of relentless resourcefulness that we like to see in founders that find a way when everything looks impossible to keep going. And they had done that. And, I was like, All right, this is our kind of team. And so we decided to fund them. And then during YC, they focused entirely on New York. They used to fly back and forth between New York and Silicon Valley every week. And they got it to work in New York. They evolved the model; they figured out whole apartment rental. And by the time YC was over--I think it was $7000 a month; maybe it was $7000 a week. But I believe they were doing $7000 a month in gross revenue. But it was growing on this beautifully exponential curve. And they raised some money and kept growing ever since. Russ: So I heard recently, I don't know if it's true, but it's probably roughly true, that New York City and Chicago have the same number of hotel rooms. And of course a lot more people want to visit New York--Chicago's a fine town, I went to school there, it's lovely; there are many fine things to see there. But many more people want to go to New York. And if they really do have the same number of hotel rooms, it's part of the reason that New York is so expensive. Overnight. It raises the question of why there aren't more there. We'll put that to the side. But that clearly is the opportunity for AirBnB to expand the supply of room. Guest: Yes. All of these situations where software platforms can expand supply, when there is latent demand--Uber is another great example. Russ: Right. When there's a snowstorm, rain, etc. Guest: All these people that--the cabs are too expensive, too inconvenient to give up their car. But now, like, I rarely drive any more. Because Uber is so great. Russ: You rarely drive your own car. Guest: Right. Russ: You get driven. Guest: It just unlocked, the software platform was able to unlock this asset that people had and that other people wanted. And that's been great.
12:34Russ: Yeah. We previously had an episode with Mike Munger, talking about the sharing economy, peer-to-peer economy and what it's potential is, and one of the issues we talked about because we're both interested in political economy and politics is the regulatory response from the push from people who have had an entrenched asset--their hotel or their cab or their medallion. And I'm curious: Do you think Uber is going to make it? Do you think AirBnB is going to make it, given that natural political response? Guest: I think they are both going to. Because I think consumer demand given enough time will overrule corrupt entrenched entities. However, I think that Silicon Valley is in general too disrespectful of trying to work with the rest of the world and with regulatory frameworks. Russ: It's not the personality. Guest: It's not. But what we need--right now there's this sense of like Silicon Valley trying to disrupt the rest of the world. Russ: Correct. Guest: So there's this like hatred among medallion owners or hotels or whatever. And I predict that if unchecked it will get a lot worse as we get into more out-there industries. What we need is we need a world where everyone sort of becomes more optimistic about the future and realizes that overregulating things is not the way that we fix all of our economic and social woes. But also that Silicon Valley is not sort of viewed as sort of like, Screw you all, we're the best and we're going to do this with our without you, thing. Russ: Right; we'll do it our own way. Guest: There is a balance here, and no one I think is on the right side of this balance. Russ: You think as I do, by the way--that great stuff is very hard to stop with regulation. Guest: The track record is basically nearly 100%. The great stuff, the stuff that people desperately want, will always win. Russ: Yeah. Even if harms the economic livelihood of a concentrated politically motivated group of people-- Guest: They can only hold out for a little while. The history on this is very strong. Russ: Yeah. That's what I said. Some of my listeners had a counterexample. We're going to come back to that maybe toward then end of this podcast. But that's exactly my thought as well. For whatever reason, the politics don't seem to work out for those entrenched--they can hold it off, they can slow it down, but they can't stop it. Guest: Although even if you look at how much they can slow it down, I would argue that the cycle time is compressing. Russ: Yes it is. Guest: And in a world of everyone having a mobile phone and social media and these things just spreading among people very quickly, that the amount of time these entrenched interests can hold something off is just condensing. Russ: But talking about Uber and AirBnB, when you are saying that about the compression, I was thinking about the recent story--I don't know if it's a true story--but the Uber driver who panicked when he was being chased by a taxi cab commission guy, drove his client all over town at high speeds to his dismay. Those kind of things are going to happen occasionally. They are prone, these kind of sharing things, to a very bad publicity event that could discourage folks. As a funder of AirBnB, did you guys talk about that? And how much did you or do you worry about it? Guest: There were a couple of events early on in AirBnB that were bad PR [Public Relations]. An those are going to happen. Any great platform is going to have terrible and wonderful uses. Usually a lot more good than bad. But the bad ones generate more attention. However, for all that people worry about this and for all that the press loves to say, 'Ah, this happened; this thing is over,' and sort of whip themselves up in this frenzy, it never actually kills the startup. It's a problem; people love to talk about it. And there's just this like schadenfreude of loving to hate on someone's perceived missteps or someone's problem. And so people love to talk about it. But then they go back and they book AirBnB for their next vacation. Even after swearing they never would again because this awful thing happened. Russ: Yeah, because it creeps them out or whatever. You think it's an age thing, by the way? Because I think younger people are much more prone to be comfortable with these products. Guest: I think it's age-correlated. But I think loving scandal in the press and loving others' misfortune is true for any age. Russ: No, that is. I'm talking about possible discomfort people would have in reaction to a bad PR event. I think older folks are more likely to say, 'I'm never going to do that again,' and younger people say, 'Well, stuff happens.' Guest: My general belief on this is whatever is considered the privacy and personal whatever-you-want to call it--whatever the privacy norms are, when you are say, 18, you are going to be comfortable with that for the rest of your life as you grow up. And then older people will slowly get more comfortable with these things-- Russ: But they're never going to be the same way as people who grow up that way. Guest: Right. And I think you can look at this like, first with IM [Instant Messenger] and then social networks and then, like, Phase 2 of social networks, and now things like AirBnB and stuff like that. And if you ask my little sister if she has a problem with AirBnB like it would be weird to her the idea that people might object at all, that staying in someone's house is a foreign concept. Russ: Yeah. It fascinates me. Obviously, the cultural interplay between technology and vice versa, how they work together--it fascinates me. Guest: There's this--what actually happens is technology shapes society. It's not the other way around. Russ: I agree. Guest: And so whatever is around as you are growing up, you are, like, 'This is great; I love it.' And that's my world. Russ: If it's good, yeah. The stuff that people love, they get used to, and then they internalize it and it drives the culture. Correct; I don't think it's the other way around at all. In fact, the more fabulous the product is, the more you want it, the easier it is for you to realign your way of thinking about the world--no big deal, it's fine. Guest: Right.
18:54Russ: So, you mention, which of course, I'm going to make a bizarre analogy; but when people ask me, 'What course should I take? I go to such-and-such university; what course should I take?' I always tell them, 'Don't take the course; take the teacher.' A great teacher--it doesn't matter what they teach; it's phenomenal. And you are echoing that in a different way. But similar. That a great founder is worth giving money to even if this particular product is going to have to change a lot. What else do you look for? What else matters? What do you care about? Guest: Since you mention the great founder, I'll tell something that I don't think we've ever said publicly before. We have something that we say after interviews sometimes, which is: Fund for the pivot. Which means that this idea is so terrible, but this founder is so good, we're going to give them money anyway; and we know this first idea is going to fail. But we want them in the YC community. That's how strongly we feel about it. So, the other things that we look for--I'll talk first about sort of founders and then I'll talk about ideas. On the founder front, we look for founders that are determined and smart. In that order. It is more important to be determined than smart. Success in startups is really a game of being willing to beat your head bloody against the wall for many years. And things like endurance at that occasion really matter. We look for founders that can communicate well. That turns out to be incredibly correlated with success. Russ: Talk about why. Guest: Founders eventually become the chief evangelist for the company. And if you can't communicate with people well, in the hundreds of people you have to talk to a week about your company--employees, press, customers, suppliers, everything--then you are really going to struggle. And so it turns out that if you can't, if you are not a good communicator, that's a big strike against you. We like teams that have known each other for a while and function together really well. We like founders that are particularly set on this idea. Founders that say, 'I want to start up, what should I do?' That's really bad. We want founders that are like, 'I am in love with this idea and a startup is the best way for me to get this done.' We like that a lot. Teams that have known each other a while and work well together and compliment each other's skills, we like. And then on the idea side, we don't want ideas that are whatever the current fashionable thing is. So by the time everyone is already starting something in some category, it's too late. Like, Facebook starts in 2004; everyone else starts a social network in 2006. The one of those that looks like the bad idea, Twitter, is the only one that goes on to be successful. Russ: Right. 140 characters--who is going to...? Guest: [?] it's really different. Russ: Is that why it looks like a bad idea? Why did it look like a bad idea? Guest: People were like, tweeting, like 'I'm pooping.' It was so inane in the early days. Russ: Yeah, not like now. I'm kidding. I actually like Twitter. I learn a lot, and get a lot of intellectual stimulation from it. Guest: But what most people were doing in 2006 was starting social networks that looked a lot like Facebook. And that was wrong. So, we like ideas that are new, and that are almost unpopular. A friend of mine once drew this great Venn diagram, and one circle was things that look like a good idea, is a good idea; and the other circle was, looks like a bad idea. And you want the things that are at this intersection. So you want things that look bad but are in fact good. And that not a lot of people are doing. Russ: Marc Andreessen said the same. Different way, but said it similarly a few episodes back. Guest: It's really true. Russ: Because otherwise you won't make any money. The ones that look good that are good, everybody knows about and recognizes. And you're not going to make a lot of money. Guest: They always end up looking a little crazy. You also want something that is the appropriate size to start with. If AirBnB, since we've been talking about them, tried to start what they are doing today, they never would have gotten off the ground. Russ: Why? Too much to do at once. Too hard of a problem. You have to get a small beachhead somewhere. Russ: That scales. Guest: They started just in one city. They did things that didn't scale, actually; in that one city, they went door to door, they knocked on people's doors, they sent their own photographers-- Russ: They went door to door and did what? Guest: Would you like to list your apartment on AirBnB? And we'll make the listing for you, we'll take photographs of your place for free with a professional photographer--which turned out to be one of the co-founders. Russ: And how many people thought that was an attractive opportunity, at that point? Guest: Dozens. But that was enough. It was enough to get it going. And so they started with this very small subsection of the problem. Russ: They didn't say, 'We'll start with the East Coast and then move on.' Guest: Right. Stripe-- Russ: What is Stripe? Guest: They are a payments company for developers. I think for the first year and a half they only had like 4 customers or something. And they just worked relentlessly on this product to make it really good for these few customers. And then they were able to expand it. Another thing that we see among good founders is that they care so much about the product. Almost too much. They get obsessed with small details of the user experience, and the Stripe guys--[?] Russ: They read Steve Jobs's biography by Walter Isaacson so they learned that--the inside of the machine has got to be as pretty as the outside. Is that part of it? Guest: I wonder if they are trying to emulate Steve Jobs, or just if that's a shared trait among these sorts of people. Russ: Yeah. Who knows? Guest: I don't know which way it goes. Russ: [24:07] When they make their presentations, or better yet, when you make a decision, what amount of stuff do you expect? So, I used to run a business, plan completion program at Washington University in St. Louis's business school. I don't think I've told this--maybe I've told this story on EconTalk. I'm not sure. But one of the student teams had their business plan over the life of the company that they projected and their numbers; they made no profits. And I said, well that's, this is around, I don't know, somewhere in the late 1990s. And I said, this is a little bit discouraging--you're not making any profit. The student said, 'Well, all the best companies don't make any money at first.' So they viewed that as a selling point--they weren't making any money. And I explained that, 'Well, Amazon at this point had never made any money; and that really wasn't the idea.' Maybe it was to actually make money. So, I'm curious at what level of projection and detail--if after this interview is over I said to you, by the way, Sam, I've always had this great idea; and it grabbed your attention--is that level of--? Guest: Very little, honestly. Russ: Very little what? Detail? Guest: Yeah. So, on the specific question of profits, we actually like startups that are reinvesting the money they are making into making their company better. Russ: Oh, sure. Guest: They don't actually have profits when we see them. That's expected. The level of detail is quite low. The startups--we look at their demo, but we never read the business plan. Russ: You do not? Guest: Never, ever. Russ: Why? Guest: I've never written one in my life. At the stage that we are operating at, it's irrelevant. Like financial projections also we never look at. Russ: It's the wrong hoop to make them jump through. You'd rather make them bang their head against the wall for a few days. Guest: We would rather them spend the time working on their product, talking to users. What we care about is: Have you built a product? Have you spoken to users? Can we see that? Can we talk about where it may involve? The really good interviews end up being this back-and-forth brainstorming session about all the things they can do. The future directions the product can go in, what you can do to get users, what they can do to improve it. We care about stuff like that. But in terms of business plans or financial projections, nothing. Now, we get an application, which is like a text but not much; a video, a one-minute video of the team. Russ: Did you say one-minute? Guest: One minute video introducing the team. Russ: That short. It's 60 seconds. Guest: Very short. A demo if they have one. That's all that we have about whether we make our decision or not about whether we are going to interview the team. And if we interview them, we interview them for 10 minutes and we decide on the spot. So, it's pretty quick. Which means we get it wrong some of the time. The good news. Russ: Hold on. When you say you decide on the spot, there are many dimensions to deciding, right? So, are they coming at you with a particular ask in terms of money? Guest: With a standard funding offer. So our deal is always the same. It's $120,000 for 7% of the company. Russ: Oh, wow. Guest: And we don't really change that. Russ: Wow. Nobody else does anything remotely like that. Guest: No. But if you look at our time scale, like it would take 10 minutes to negotiate an offer. So at the speed that we have to operate, there's no time to negotiate. So we just say this is our offer, the same for all companies. And that's it. But, like, again, these are very much-- Russ: Does anybody say no? Guest: It's happened like once. I think. But not really. Russ: Did they say--well, 6. Guest: And then we don't say we don't negotiate; goodbye; and they say, 'Okay, we'll take it.' So, but like, you know, I probably shouldn't admit this--I'm on the Board of a public company: I can't read a balance sheet or income statement or anything like that. I have to have someone explain it to me, every Board meeting. Russ: Yeah, anyway. Guest: And so, I don't think any of the other partners can, either. Maybe one guy can. But that would all be wasted on us. Like, we'd much rather spend the time talking about the product than the strategy for the company. 'Cause that, we really understand. Russ: And hearing their reaction to those thoughts and how they respond to it. So, $120,000 is a very small sum of money. Guest: It is. Russ: So, that determines who walks in the door. And where they go after. What's next? Guest: Yeah. $120,000. And then the program lasts for about 3 months. And that's enough for them to survive on for 3 months. Russ: What do you mean, the program? Guest: Y Combinator. So we work in these batch--these cycles. We're in the middle of one right now. And during this 3-month cycle, they work very, very closely with us. They come in once a week to meet with the other companies and also to hear our dinner speaker. We had the Benchmark Partners last night. Mark Zuckerberg before that; Peter Thiel the week before that. But they come in. And we have other events. We help them get their company into the best shape possible. And at the end of the program, we have a day where they meet a lot of other later stage investors. Who, instead of writing a $120,000 check, write $3 million or $12 million dollar checks. Russ: And are you part of that? Guest: No. Russ: You don't invest in that stage. Guest: We do not. The historical reason was we just didn't have enough capital to do that. Now we do. But there would still be a conflict problem. If were invested in some companies at the end of our program and not others, it would really hurt the ones we didn't invest in. Russ: Sure. Guest: One of the other accelerators does this, and I think it basically is going to kill their program. Russ: We'll find out.
29:25Russ: So, you said you try not to invest in the copycat or the fad or trending thing. So I want to ask you about a few trending things. When I recently announced who I was interviewing, I said, coming up on EconTalk, Chris Blattman, who is a development economist at Columbia, Sam Altman of Y Combinator, and Daphne Koller of Coursera--MOOCs and artificial intelligence. And somebody Tweeted, 'OH, NO'--all caps--'Econtalk's getting into this, gonna start raving about how great technology and entrepreneurs are, like everything else on the Internet.' But I have to admit, I do kind of like technology and entrepreneurs. And one of the things that happens to me when I come out here in the summer--I live outside of Washington, D.C. and I come out every 6 or 7 weeks in the summer, and come to Stanford--I feel like I'm at the center of the universe. You know, Washington is--everyone in Washington, except for me-- Guest: Thinks they are-- Russ: Thinks they are in the center. And there are things they are in the center in. Obviously. But it's so placid there. And when I come to Stanford, the intellectual, the excitement about products and transforming concepts into reality, is palpable. And then I run into start-up people and venture capitalists. And they are so alive, compared to, say, a lobbyist in Washington, say, just to pick a random example. And there are certain things that just--again, it's almost palpable. You can almost feel them. So the thing is that I notice being here--which are already the next big thing, which at least they feel like they are. So, there's health. There's personalized health devices that are exploding in this area where people are trying to either customize medical care or provide diagnostic stuff through iPhone, through other devices, through simpler things. So, there's health stuff, through iPhone, through other devices, through simpler things. There's health. There's wearables--the excitement over the iWatch is growing despite the fact that the iWatch seems like a really bad idea--a Dick Tracy throwback device that everybody--that other people have done that don't seem to excite much excitement. Guest: Well, no one had gotten the tablet to work before the iPad. Russ: Correct. Or the--I was going to say the portable device. Palm did that, really, launched that, when the Newton didn't make it. Anyway, so, wearables, whether it's some better version of Google Glass, which doesn't seem to be moving along, but, the iWatch seems to be. We'll see. And then BitCoin. Guest: Yeah. Russ: Which is--well, those three things, first of all. Tell me if you think there's anything to the hype. And then, what am I missing? What are the other things that you are seeing dozens of, and you it's like, 'Enough, it's already the space is overfilled.' Guest: Well, instead of answering what I think is overfilled, I can talk about enterprise software or, like, news aggregators, or a whole bunch of other things. I'll try and answer what I think is really interesting that I think not a lot of people are paying attention to. Russ: Okay. Take a chance. That's valuable information. Just throw that out here. Guest: But first I'll comment on those three categories. Because I do actually believe there are interesting things happening in all 3 of those areas. But not what most people think they are. Russ: Okay. Guest: So, on health care. That is an area I think we're seeing great development in. After having been ignored for a long period of time. Most investors interestingly enough are still not paying a lot of attention. And probably in 2 years, when some of these health care companies get successful, there will be a true flood of investment into this space, and it will already be too late. Russ: So, some of that is a reaction to the disappointment of the genetic[?] space that didn't quite pan out yet, and people lost money. Guest: Yes. Often though these ideas are good but too early. So, biotech is probably a very good idea. But in the first biotech boom it was just too early. The costs were too high and the cycle time was too long. And I've really come to believe that those two things, low cost and low cycle time, are the most important things for startups in a given area to be successful. But now, in biotech specifically, which is an area where we've been active recently, the costs and the cycle time have come down quite dramatically. And so you are able to have a startup that for a few million dollars or in a year or something can get something really meaningful done. And that's a brand new thing. And that's hugely, like a gigantic megadeal--I think most people still fail to understand that. Also, there's all this data that's now available on health care. You can get your entire genome--the whole thing--for a thousand dollars, soon. You can get your like, your microbiome, the bacteria in your guts, sequenced. And all these other things. You can get your blood drawn and have like 800 biomarkers specific to you, and know exactly how they trend over time. So there's just so much data that's available. And we have these like devices-- Russ: I would add that it's real data. Unlike these claims that, 'Oh, we're going to figure out which drugs work the best'. And to me, epidemiology is a cesspool of intellectual failure. But we'll have more data to fail on. Guest: Right. Russ: But this is real data. This is stuff that is true about you. Which is different, say, from the characteristics of the whole population. Guest: Right. And so there's this huge amount of data and personalization. A company I just agreed to fund can take a piece of your tumor out, try hundreds of different drugs against that, in different combinations; [?] for you, for your particular tumor. And your body. Here's the combination that works the best. Rather than have a doctor like blindly try a few dozen things. Russ: Oh, let's try one. Yeah. Guest: So, this is like--these are huge developments. Russ: What technology is making that possible? Guest: Well, it's a bunch of things together? So, one is just-- Russ: So, like, why is that now available? Guest: Yeah. So, one is just like the software revolution. Which is that these are gigantic data sets that have been expensive to process. Two, is that the custom hardware has gotten much better. So like the new gene sequencers on that example are just unbelievable. And the robotics of the lab are unbelievable. Russ: Yeah, that's what I've heard about--is the robotics lab, it's ability to generate the test results. Which used to be a person with an eyedropper, is now-- Guest: So, I just went to the--I'm trying to decide if I should mention this conversation. Russ: Nah, you don't have to. Guest: I won't. But I just went to this one company, a company, a synthetic biology company that has a roboticized lab I could not believe. Where there's probably 20 different machines working in unison, a robotic arm on a track; and they automate this entire thing. And they can have the throughput now--you know, of thousands of humans in a lab with no mistakes, 24 hours a day. And that is a huge change. And then there are also these new--so that, but coming back to these concepts of cost and cycle time. Those have come down a great deal because of things like this. Russ: It sounds very useful. 'In 20 years, I'll be able to tell you the best drug for your tumor.' Guest: Right. Also this thing that was huge for the Internet, Amazon web services, has happened now for biotech. So this thing that was really important in the history of the web was this idea that you no longer had to set up your own servers and infrastructure and network equipment. But you could just throw it off to Amazon in the cloud and they would just host it and serve it for you. And you could just work on your product, and they would run it. And there are now these labs where you can say, 'Hey, do this physical life science experiment for me. Here are the parameters. Here's the experiment to run. Do it.' And you push a button on your computer, and just like Amazon, if you ask, they do it. And that's--so all these things have changed for health care and biotech. And that's important. Russ: Really big. Guest: So that's going to be really big. Guest: BitCoin. Russ: You skipped over wearables. Guest: Oh, I'll do wearables next. Russ: You want to make forecasts for the iWatch? Guest: I think it will do really well. There's been a lot of hype for wearables. They've all not been very good. But, that doesn't mean they are not going to be great some day. Before the iPhone, smart phones were pretty disappointing. Before the iPad, everyone else had failed with tablets. And I am confident that there is going to be a new computing platform on our bodies somehow. And that's going to be really important. Every time in history that there has been a new computing platform, it has been hugely important and a ton of value has been created. But they don't come along very often. Russ: What are the breakpoints--when you say a new platform? Do you mean going from mainframe to desktop? Guest: Supercomputer to desktop to laptop to phone to tablet, I would say. Russ: What's next? Something is next. We don't know what it is. Guest: Something is next. Well, I think two things are coming. And I think one is wearables, and I think the other is virtual reality. Within wearables, I include augmented reality. So in some ways I can say the future is a combination, the two future platforms are augmented reality and virtual reality. Russ: What are those? What do you mean by-- Guest: Like, Oculus Rift is virtual reality, where you put this thing and then you can go live in this computer world that you move around in. The world moves with you and it feels like you are immersed. Augmented reality is something like Google Glass. Hopefully a version much better in the future, or made by a company YC funded. Or even an iWatch-- [?] Russ: This will be where I'm at the baseball game and the hologram of the highlight pops up so I can watch some play again. Guest: Yeah. Or even if you have to hold your watch up to see something, but you are looking, without pulling something out of your pocket--you are looking there and seeing something else on your watch. That helps you, inform[?]--
39:16Russ: I'm somewhat excited about the iWatch. Of course, I don't know what it's going to do or what it will be. I have to confess, part of me just wants to be excited, because it's just fun when a new product comes out. But when I talk to people and say, what's the big deal, you just take it out--you take your phone out? Guest: They said this before the iPhone came, and then when the iPad came, they said, 'What's the big deal? It's just a big iPhone.' New platforms--the reason it's hard to conceptualize is that the things you are going to use those platforms for don't yet exist. And thus you can say I can already see the weather on my iPhone. Why do I need--field division? And you don't. That not what the really cool thing is going to be. The reason you are going to need some wearable computer is for some reason you haven't even thought about yet. It always makes it difficult to conceptualize what's going to do well. But the history of tech pundits saying we're going to need or not need some new technology is horrible. In both directions. Russ: Right. Is wearables going to be followed by embeddables? I assume we're going to be on to something there. Guest: You know, I'm thinking about getting an RFID [Radio-frequency identification] chip implanted in my hand. Russ: Me, too! No, I'm actually not. But what why? What is an RFID chip? Guest: It's a little device that can sort of emit a signal and identify itself. Russ: So I can find you? Guest: Well, it's more like I could start my car, get into my house, whatever, without having any devices. My phone is theoretically good enough for this. But there are all these reasons in practice why it doesn't quite work. Russ: It runs out of batteries. It's really annoying. Guest: That's one problem that's often an issue. Russ: Of course, that chip--what's going to be powering that chip? Guest: It's passively powered. Russ: What does that mean? Guest: Basically, as it moves through the flux of the receiving station it generates some charge, some power, and then can send it back. People have like--some crazy people have already done this. The current devices are pretty big. But I just met with someone who has produced this much, much smaller one. And I think that would be pretty cool. Russ: Part of your thumb, maybe? Guest: There's a number of places people can do it. Yeah. Finger pads are bad because those--the tactile stuff is so important there. Russ: [?] I agree. Guest: But I'd say I'm an extreme case here. Russ: Yeah, I think that's true. But that's just now. Guest: And who knows [?] will. In a way, though, like, I'm confident that for some definition of an embeddable, even if it's a little nano robot that gets injected in my bloodstream that does something, that doesn't really look like it's a computer, it's this biological thing, like I am very confident that at some point in my life, assuming I don't die in a car crash in the short term, that I will inject some sort of embeddable into my body. Russ: I think that's in the next three years, actually. I think I'll even live to see that, and I don't have the child within you. Guest: Wait out the car crash too long. Russ: Yeah. I think--I talked to a startup last summer that puts something the size of a grain of sand in a pill, so it's cheap. When people take their meds, phenomenal idea that implementing it is reasonable. It's not a fantasy. Guest: There are all these different ways that sort of the man-machine merge gets talked about. And there's the crazy, out-there stuff. But there are all these little versions like that there are already coming. Russ: They're coming. And you can't stop them. It gives some people the creeps. But it's coming. It's just a matter of time. Guest: So, on the wearables, I think it will be a big new computing platform. My sense is the risk-based stuff is much closer than the vision-based stuff, but that the future is likely to have a vision-based piece in it. Although these predictions are historically difficult to make. Russ: Yeah, I agree with that.
42:56Russ: How about BitCoin? Guest: Okay. I think the most interesting piece of BitCoin is this idea of the blockchain, this idea we can have the centralized network agree on what the truth is, and have an incentive system to make that work. Russ: The truth being, you've got the money and I don't. Guest: I gave it to you, and it's done. All that. But that's not just for the money. There's a startup in this current YC batch that is doing a distributed file storage system based on the concept of a blockchain. I've seen companies doing contracts, future revenue shares. Russ: So, explain--with the blockchain, the only thing that I understand about it is it's the thing that allows Bitcoin to do its magic. So give us some more background. Guest: Blockchain is basically this sort of shared ledger, where transactions--which could be like I give you money, I get money, or I have this file, I give you this file, I prove that I still have this file--these ledgers, these things get entered into this public ledger, and then they get verified. Russ: And then it's not-- Guest: And it happened and then it's not disputed. This is just sort of in the ink that's dried, it's done. And then we can look back forever at this shared ledger. And it works, just off of math and the Internet as a whole doing this. Which is very powerful. Russ: You are saying it has applications out of the payment systems. Guest: Yeah. So, BitCoin itself, specifically, like, is BitCoin going to win? Unfortunately I can only say the same thing that everyone else does, which has gotten tiring, which is probably not but if it does it'll be a big deal. I own like not that many BitCoins, and I have them as a hedge in case it does win. But I think it still is looking unlikely. The thing that you would need to see--so if you say that currencies can be used for speculation, illegal transactions, and legal transactions, you could then look at: What does the dollar mostly get used for? What does the RMB (renminbi) mostly get used for? What does BitCoin mostly get used for? And you could look at the dollar and you could say, definitely probably like hundred dollar bills are the illegal transaction currency of choice. But that is dwarfed by all of the legal transaction stuff that happens in dollars. And there is some speculation-- Russ: Absolutely. But not so much. Guest: But it's more of a, people are like, Russian dollars because they'll still be here, not because they are going to go way up in value. But they believe in sort of the authority of the government and all of its guns, and a dollar is worth a dollar. You can look at BitCoin and there's a huge amount of speculation. Most people that buy BitCoin today do it because they think they are going to get rich by holding their BitCoin, because they figured it out early. There is some illegal transaction value, but not actually that much. People buying like drugs, or worse. And then there is very little legitimate transaction. Russ: Correct. But there's some. Guest: Very little. Russ: Is it growing? Guest: That's the worst part, is that if you look at most of these estimated transaction volume things, which are notoriously hard to do, it at least does not appear to be undergoing runaway growth, as everyone has been predicting. And so I think the big challenge with BitCoin as a currency is, it still has not found what it's better at than other payment systems. And that's really important. Personally, I think the most promising green shoots are around international remittance. Sending money to my family in Mexico or whatever. And that seems to maybe be picking up real steam. But it's so early in the day the data's so noisy, it's hard to know for sure. So basically what I think about BitCoin is it's unbelievably interesting technology; it's one of these things that's lodged in my brain and won't go away; but I have been continually disappointed about the speed with which it has found real usage. Russ: But I'm sure you are also continuously cheered by the fact that it still exists. Despite that failure to take off. Guest: Yeah. Russ: Because I look at it, that's the way I look at it. It's amazing it exists. So it's all gravy for me. Guest: Well, it has this kind of great thing which is that there are all these people that are really incented [incentivized] for it to continue to exist, which is everyone that holds BitCoin. And all the VCs [Venture Capitalists] who invest a lot in BitCoin companies. One of the greatest things BitCoin has going for it is that pmarca made a bunch of BitCoin investments and that's what got all these tweet stories. Russ: That's Marc Andreessen [@pmarca], for those of you who don't follow him on Twitter, which is a very small group evidently, who don't. Guest: Actually, that was his old email address, so I always heard him and I'll always call him pmarca, way before Twitter. But there are like all these people who really want it to succeed. And hopefully that can give it the activation energy to get over this hump of the current state being ahead of the reality. But I think one of the things that I think, or I see it can do that's really important, is try to fund BitCoin companies that are actually enabling a real legitimate usage in transactions. Because that can make BitCoin survive. And BitCoin is good for the world, I believe.
48:00Russ: Any areas--I pick three that anybody would think about. Are there any areas you're thinking about that--and by the way, I want to ask you a preliminary question, which is: As you do something like this, running the Y Combinator, do you find yourself censoring yourself, saying, Oh, I can't talk about that? You just mentioned, should I say that thing about my company? So here I am asking you: What's the next big thing? Are you saying to yourself, Well, I can't say that? Guest: I'm definitely going to answer it. For better or for worse, I have like very little filter, and I just say what's on my mind. Sometimes that gets me in trouble in the press, and sometimes it doesn't. But I'm just going to tell you. I am very bad at self-censoring, is the short version of that. So, I think energy is one area I've been looking at very closely. All the clean tech stuff has basically failed. But it had all these weird problems--that companies were going after markets that were way too big. They never had an answer to how they were going to be economically competitive. It was just, the answer was like, we hope the government subsidies don't dry up. Russ: Yeah. Bad model. Guest: The teams were really broken. They raised way too much capital. So, energy is now a kind of wasteland that nobody's trying anything in. But I believe that-- Russ: Over in the physical world, we had an interview on fracking recently, where--it's unbelievable. That's been a very quiet revolution for people who don't live in Oklahoma. Guest: The innovation in fracking is remarkable. That is one of the great technology stories of the last 10 years. Russ: Yeah. It's kind of a quiet story. Guest: My sense is there's going to be a lot more coming; that it is an actual lower cost energy. And energy is the biggest market in the world. So if you can get that to work you are going to have a very, very large company. And the same things about cost and cycle time, I think have really come down. Russ: So, let's talk about that, because, so I was going to ask you this at the end. I said I would. So, when I had said, after a recent episode that I couldn't think of an historical example where a new technology came along but it was stopped by regulation, somebody emailed me or tweeted or commented: What about nuclear power? And I would say that's a different case. That's a case where most citizens--I mean, I could be wrong. Maybe that really is the case where the entrenched utility companies fought it, under the mask of saving people from a nuclear holocaust, it's a bootlegger and baptist example, for those listeners--that example speaks to. You've got people who altruistic versus people who are self-interested; they make this very powerful coalition. Maybe that's really what stops nuclear power in the United States. I don't think so. I think it's mainly an unease on most people's parts that comes naturally associating this with bad things. And so, you told me before we started this interview that you funded some nuclear power companies. Which shocks me. So, tell me what's going on in that space I don't know about. Guest: There are two nuclear energy companies in this batch. I believe that--the 20th century was clearly the carbon century. And I believe the 22nd century is going to be the atomic power century. I'm very convinced of that. It's just a question of how long it takes us. Russ: That's the next one. Not this one. Guest: Right. This one, the outcome is doubt. Russ: This is the Red Sox century--we've won 3 World Series in this century. That's the way I think about it. But that's a little bit narrow. So, go ahead. Guest: Incredible performance. Russ: We're going to eventually get to the nuclear. But what's going to happen in the meanwhile? Guest: So, I do believe that this is not a regulatory issue. I've actually spent a decent amount of time with the NRC [Nuclear Regulatory Commission]; I've been out there a number of times. And I think they want nuclear power to happen. They are mostly nuclear engineers. Nuclear is one of these rare things where the downside is so bad--it's not like some guy losing his taxi medallion. But it's like my city getting blown up. Russ: Exactly. So, you're going to make energy a little cheaper--I think I'll pass. Guest: I think I'll pass. And also, nuclear power is like way safer than coal. Like hundreds of thousands of times safer in terms of deaths. But when it fails, it fails spectacularly. And when you die of lung cancer you die very slowly. And humans very badly overrate these spectacular deaths. So I think the problem with nuclear is a public perception problem, not a regulatory problem first and foremost. Russ: I agree, but the regulatory response to that, if that changed the regulatory response would be different. Guest: Again, I know all of the physics, I know a lot about the engineering, and I know that it's totally safe. And I do not want a nuclear power plant on my block. I really don't. And it's totally irrational, but I understanding why people feel that way. Russ: You say it's totally safe, but there have been problems with nuclear power plants, right? So, are those problems exaggerated? Are they not problems? Guest: No. There have been three problems. But there's a lot of problems-- Russ: Three Mile Island, Chernobyl, and-- Guest: Fukushima. And some smaller ones. But those are the big ones. Russ: They weren't so bad-- Guest: How many people you think died in all of those? Russ: I know. Very, very few. But again, the specter-- Guest: It's the spectacle-- Russ: The specter, this ominous feeling that this 800 mile radius is going to be a nuclear wasteland with no life, with radiation death. Guest: It feels like we are doing this thing that is not a normal human thing. Burning coal, everyone's comfortable with. Russ: It's like burning wood. Guest: Changing the number--like transmuting elements, it's just a weird thing, how we feel about it. And also, the history of this was that nuclear power was developed as a byproduct of developing the atomic bomb. Russ: Not a good selling point. Guest: And so they've been--and the first nuclear power plants that most countries built, although advertised as energy-generating facilities, were really weapon development facilities. So they weren't that safe. And they weren't necessarily that cheap. Any new plants will be very safe; power will be much less, relatively inexpensive. And I believe that it will be very difficult to use to create materials for atomic weapons. But, like, a lot of people have talked about the tech slowdown. Like, why has innovation other than computers slowed down. Innovation in the bits world has been fantastic; innovation in the atoms world has been somewhat varied, depends on who you ask disappointing. And sort of what changed? And a lot of the people say the slowdown happened maybe in the early 1970s. We went to the moon in 1969, for the first time; and there was an incredible 20 years of science before that. Maybe 30 years. Really let's say from like the Bell Labs' heyday up through then. And what changed? And my view is that what it really was in this very big way the atomic bomb in 1945 that sort of changed the perception of technology. And it weirdly took another 25 years for the Cold War to like--like for a long time the United States was like, we have this great thing, other people don't. And then as the Cold War heated up, people were like, Wow, we unleashed this technology on the world and now we have to hide under our desks in our schools because it might come and get us. People got afraid of science. It was like the first time that something that could destroy on this huge scale. Russ: It's a Frankenstein story. Guest: On this huge scale. And I don't think we've ever recovered from that. There's still this fear of science Russ: And technology. Guest: And technology, that I think in many senses comes from that event.
55:40Russ: And so why is it, why would Y Combinator recently fund two nuclear power ventures? Guest: Well, there's the selfish version, which is we think they can be phenomenal investment returns. Russ: And why? Given that nobody likes it and everyone's afraid of it. Guest: [?] new sorts of nuclear. Without putting us in the companies, neither of them are traditional light water reactors. Brand new technology. Russ: Okay. Cool. Guest: But then the real reason is, I have studied a lot about what I think is sort of the best use of my time and money and what I think will help the world the most. And I really do believe that safe, cheap, clean energy is probably the most important thing an individual can do for the future of the world. Russ: Way up there. There are a few other things in the running, but it's certainly in the top 5, top 10. Guest: What would you put as 5 things above it? Russ: Health-- Guest: Related. Russ: Why? Talk about magic bullets and great stuff--we touched on some of them earlier. So, I interviewed Eric Topol a way back, and he suggested we are going to come to a world where we get a two-week advance notice you are getting a heart attack. That would be a big thing. That would be a really big thing. In the developed world. The undeveloped world, not so much. But eventually for everybody. Clean water would be an enormous thing. Guest: The thing that I find interesting about energy is a lot of these problems, the two that you just mentioned, dramatically cheaper energy, these problems reduce to an energy problem. If we can get energy to be one tenth the cost it is today, we can have all the clean water the world ever wants. Which is huge. Huge, huge. If we can stop destroying the environment I think that will have a huge long-term impact on all of our health. Russ: Maybe. I'm not as optimistic--I don't think that's as big a bang for the buck. But okay. Guest: A lot of the economic challenges we face, a lot of the wars that we fight--I think these do reduce to the cost and the availability of energy. Russ: Okay, I'll accept that point. And I derailed you. You started saying energy is the next big thing, and I steered you into the nuclear. Is nuclear where we are going? Guest: It's certainly a possibility. Russ: Are you excited about something else you want to mention? Guest: So when I say 'solar', I include anything that comes with one hop from like fusion in the sun. So anything that's like a photovoltaic or concentrating solar power, or wind. And even biofuels. Which is sort of--I count all of those things as solar power. Russ: Okay. So are they coming? Guest: Yeah. Russ: They're not doing so well. Guest: And so here's the question: Let's say that we believe that, in this weird way, like atomic energy includes anything we can do with a nuclear plant on earth; but also anything from the sun--because that's the same fundamental force. It's just a transport problem in one case. So, let's say we can agree that the 22nd century is likely to be powered by atomic energy, either terrestrially generated or sun generated. The thing I have no strong opinion yet is-- Russ: How we're going to get there. Guest: if it's going to be 80-20 one or the other. I think it will be 80-20 one way or the other, but I don't know which one is likely to win. There can only be one cheapest source of power, and historically whatever that has been, whether it's like the steam engine or an internal combustion engine or coal plants or whatever, has been dominant. But like which of those two is going to be cheaper? I think it would be very hard to call it at this point. Russ: So let me ask you a micro-energy problem. I would say another prize is waiting out there for somebody to figure out a way for batteries to last a lot longer, not just a little bit longer. Guest: I actually have a project going on that, but I can't share anything yet. Russ: Is that--but people are working on it, obviously. Guest: I will have something to announce there in the not-to-distant future. Russ: Okay. I look forward to that. Guest: But, I agree with you. The energy storage problem-- Russ: Big deal. Guest: So if you look at energy, the three kind of big areas I look at are energy generation, energy transmission, and energy storage. And I'm doing a little bit of work at least on all three of those areas. Russ: Cool.
59:48Russ: We're almost out of time. You alluded to what you called the man/machine interface--I don't know how you called it, but the merging of humans and technology. And let's close on that. We could spend another hour on it. It's an issue we've talked about a number of times here. The two things I'm interested in, sort of three things: One is, singularity. Is technology going to become self-replicating or take over? I'm not worried about that. Maybe I'm wrong. But I'm not worried about that. But I'll let you talk in a second about your views. The second is what's it going to do to the workforce. People are worried, we're humans, robots, smart machines, are going to be able to do everything. And the third area that seems interesting is there are all these people saying, as you have alluded to, that atomic--innovation in atoms--is kind of at a standstill. And yet, I look at the world around us and say we are living in extraordinary, innovative times and I'm not worried about stagnation or the world being an uninteresting or unproductive place. I think there's going to be really inexpensive things for almost everybody. There may be some inequality, some people have a little more access to it because they created it. Those are the three things. Guest: I am in the camp that technological innovation is actually going to that at a very impressive clip. But a lot of people think it don't. So, people like to talk about the theatrics of this man-machine merge, like there's going to be this moment in time where we are separate; and now we're together. And they call it the singularity and whatever, and they talk about AI [artificial intelligence]. Like I do actually believe AI is undervalued and people should pay more attention to it. But-- Russ: It's all about those cycle time problems. The early promise didn't quite pay off, and now people say forget about it. But they are making a mistake. It's important. Guest: But all of these things are, they happen gradually. So, maybe Step 1 is I can start keeping track of my thoughts in some software, and then I can ask that software, Hey, what was the answer to this thing? And I can have like an augmentation in my brain around my machine. And maybe it can source, you know, maybe it can search it across far more data than I can ever keep in my brain. Maybe it can search against all of Google's servers for us. There's this great science fiction book called The City and the Stars, which is like 50 or 60 years old but it's held up incredibly well--it was incredibly accurate in its predictions, I'll say. A lot of these ideas are finally coming true. Right? And if you look 50 or 60 years ago, they sound comically futuristic. And we're pretty far along down a lot of these paths. Computers are augmenting humans in big ways. Humans are augmenting computers in big ways. And that's already happening. So, like I think in some way we don't have the drama and the spectacle of computers implanted inside people or whatever. But when people look back in 100 years, I think we'll already be much further along this path than many people think. And I think if you look at what a human is capable of today compared to what they were capable of 30, 40, 50 years ago, you would say it's like qualitatively different at this point. And that, yes, this iPhone is technically not part of my body. Russ: It's pretty close. Guest: It never leaves it very far. Russ: It's pretty close. Guest: We have a pretty tight partnership. Russ: It's kind of already a wearable. Right? Because it's in one's hand, in my hand anyway, way more often than I would have thought. Guest: Right. The point I was trying to make though is we are already much further along this path, I think, than most people think. And even though we don't have the SciFi movie of things injected into our brain, cables going into our brain--which I hope we get someday, by the way. Like everyone's obsessed with biological and mortality and living forever. Which I am never that excited about. Because I think the risk of accidental death is still too high. What I would like to do is live forever in a computer. I would like to have some computer that could download my brain and thoughts, all of that, memories. That would seem like a lot less error prone and risk prone than just dying of cancer. But I would like to not do that either. So, yeah. I think like it's easy to look at the day-to-day and say we're not making enough progress. But if you look at year to year, decade to decade, it's really been wild. Russ: Yeah. I think we're on the cusp of an incredible transformation, actually. And people who say that we've thought of most of the good things, I think they're out of their minds. Guest: Well, here's the thing. It always feels like we're on the cusp. Because it's one of these exponential curves. So at any point on that curve when you look up it looks vertical. But if you can zoom out and look at how far we've already come--we've gotten over a lot of cusps already. Russ: Yeah, it's true. Guest: It's something that's really great.

COMMENTS (17 to date)
Jeremey Arnold writes:

I liked your interview with Mr. Andressen. I LOVED your interview here. This is one of my favorite EconTalk's ever and I hope this one stands out on the end of the year list. Thank you.

Nivita Verma writes:

Enjoyed your interview style Russ and found this podcast w/Sam to be very informative - packed w/great advice for startups. Thanks!

Dave Ryan writes:

9 years, about the life of a single VC fund. 700 investments- quite a pool. Winners quoted are the same handful (Air BnB, Drop box, Heroku...) quoted.

All interesting chat, and the YC folk are very credible (Paul Graham's blogs in particular)...

But on a podcast called Econtalk shouldn't we be hearing some of the economics: what is the value (after dilution from follow on investments, not the headline "YC owns shares in companies worth billions) of investments YC has made (roughly 700 x $120k or $80m). Ask that question Russ, and you may find out if Sam really doesn't have a lot of filters in his responses.

This is an industry (VC, startups) where the returns distributions look like power curves and the data is full of survivorship bias, selective (very) reporting and use of multiple anecdotes in lieu of data.

But a sceptic might say, what has YC got to hide- if they are special people doing special things that's great, but if they are throwing mud at the wall and getting their fair share of luck perhaps that's an interesting econtalk discussion.

My suggestions: get a guest interviewer (Nassim Taleb) to ask more insightful questions of guests from the VC world (Dave McClure of 500 startups would be a good one!).Or crowd source questions ahead of meeting guests

Its still a great show with interesting guests, like Altman, but Russ... get your econ-mojo working!

john writes:

Overall a very interesting interview. Mr. Altman's comments on the biotech space were on the mark; particularly in regard to the the cycle time being compressed with automation and increased computing power. Hopefully this latest biotech boom is for real.

However, Mr. Altman has got to get rid of his annoying use of the word "like" every five seconds. He gives Moon Zappa a run for her money in the Valley Girl Speak competition.

Steve Sedio writes:

Really enjoyed this podcast.

If economics is the study of how to make the most out of a scare resource, then crowdsourcing and Y Combinator have found two (non-exclusive?) ways to increase the resource called innovation.

The success called Silicon Valley occurred when the number of tech workers exceeded critical mass, and innovation exploded. I expect Y Combinators providing access to alumni will reach that same type critical mass.

With the owners retaining more of the company (under Y Combinator than VC's), I see faster initial growth, more spin off growth, but fewer large companies (except for App / Software type companies). The one thing VC's bring to the party is good second phase management. There are a lot "retiring" Baby Boomer managers - maybe that could be another start-up, under Y Combinator.

I look forward to seeing how the Y Combinator companies do, and what other innovation crop up on growing innovation. I expect Y Combinator would coach those start-ups as well.

Michael Byrnes writes:

Very good podcast, and Y Combinator is a very interesting model.

When asked about why YC is so successful, Altman talked about network effects. But is there also just a "trial and error" effect? In the more traditional VC model, isn't there a lot more due diligence (a much bigger investment in trying to pick the eventual winner) followed by a much larger investment? YC turns that process upside down: due diligence is minimal but the investment is small.

As a result YC has put itself in a position to benefit from uncertainty and their lack of specific knowledge rather than be harmed by it (as might be the case in a model where they put more eggs in fewer baskets).

Would Taleb call the YC model "Antifragile"?

I also very much enjoyed Altman's insights on bitcoin and, especially, on nuclear.

Although I'm not sure I agree with him on conventional nuclear. His discussion of nuclear reminded me of the car vs airplane safety argument - because so many are killed in a plane crash, there are many people who won't fly but who drive (a more dangerous activity) every day.

That may apply to nuclear, but because a conventional nuclear disaster could be so devastating, I wonder if the relectance to go nuclear is reasonable. Just because Chernobyl is the worst disaster the world has had doesn't mean it is the worst possible disaster.

I wonder if any private enterprise operating a nuclear reactor would, in the event of a disaster, go bankrupt, with financial restitution completely out of the question.

I could be wrong - would love to hear an EconTalk episode on the economics of nuclear power.

That said, a clean nuclear technology would be an incredibly game changer. I think Altman is right that energy is central to virtually everything, since energy is so often a limiting constraint. From there a post-scarcity society really would seem possible.


Kevin writes:

Interesting discussion.

As someone in the cancer field I had to laugh at taking your tumor, growing it, testing chemo on it to find what one was the best for it. I am sure he is not stating that correctly (but maybe he is - guy is clearly smarter than me) because the number of agents we have discovered that will kill a tumor in a glass bottle is astronomical - only a few of those actually work in your body to kill cancer cells and its never clear until you get to animals and humans which agents will work based on prelim stuff.

But I hope they have found something new. I would love to be out of a job!

Brian writes:

Nuclear power is not even remotely an example of regulations killing innovation. Nuclear power is not viable on its own.

The two big users of nuclear power Japan and France are obscenely regulated and near bankrupt.

[url to Cato.org in comment revised with commenter's permission--Econlib Ed.]

John commented above on the idiosyncratic use of the word "like"

Mr. Altman has got to get rid of his annoying use of the word "like" every five seconds. He gives Moon Zappa a run for her money in the Valley Girl Speak competition.
I'm going to take a rare moment here to remark on how we at EconTalk handle some inside matters, with a specific focus on the word "like". I'd also like to give out some background on how the Highlights are typed up. Don't make me regret putting this out there!

First, I am grateful that our commenters generally refrain from making side-remarks about our guest speakers concerning their speaking styles. Griping about someone's speaking style or verbal mannerisms when that person is being recorded live--be it about a guest's foreign accent or U.S. dialect, or silences or stammering or overused words used as filler to extemporize while the guest formulates how best to bespeak more cohesive thoughts--is at best ineffective in changing the guest speaker's style, and at worst discourteous and humiliating for our guests. If you have a gripe or suggestion on speech idiosyncrasies or speech impediments, please send it via email to mail@econtalk.org or webmaster@econlib.org so that it may be relayed to our guests privately rather than publicly. The EconTalk comment section is intended for content discussion. There is no need to make our guests feel self-conscious and apologetic for their speaking styles.

But, as it happens, this matter of the many and pervasive uses of the word "like" is an interesting topic I've recently been reconsidering.

I type up the Highlights--which some of you think of as transcripts--almost in real time. I slightly slow down the playback, but for the most part I type while I listen. Maybe 120+ words per minute. [Thanks Mom! I hated you for making me take that 2-week summer course in touch-typing but it did help typing college papers.]

Toward the start of each episode, it's usually easiest for me to type up verbatim various idiosyncratic uses of utterances or phrases, such as "um," "you know," "like," "so," "well," etc. As an episode progresses, I get to know a person's speaking style, so I try to accommodate when I can, leaving out obvious idiosyncrasies. Sometimes I'm typing so fast that it's too hard for me to back up and remove them; but other times, when I'm listening ahead and typing from memory, I can accommodate and leave out known idiosyncrasies in the Highlights. Idiosyncratic words have nothing to do with content; and the focus is on content.

Sometimes guests listen to themselves afterwards and are horrified. An occasional guest has asked after the fact to remove an idiosyncratic word-usage from the Highlights, and I've done that. The Highlights, exact as they may appear, are not a verbatim transcript. There's no reason to include hemming and hawing unless it is critical to conveying the content. We appreciate our many guests' willingness to let us type it up as we are able.

But sometimes that hemming and hawing does convey content.

And that leads me to the word "like". This word is a special case.

In a large portion of the United States, starting on the West Coast but all across the country since the late 1970s or even earlier, this word in combination with the verb "to be" means and is understood to mean "said". As in, "She was like, 'So, you bought it?' And I'm like, 'Yeah. I bought it. It was great.'"

The word "like" also often indicates a kind of iffiness or inexactitude about what one is saying. It's shorthand for someone's indicating he doesn't know or remember precisely what the data or exact words are, while focusing on something else as an agreed-on conversational topic. Again, using the word "like" in this way goes back not only to the 1970s or earlier, but was used this way in speech all across the United States at that time or soon after.

So, here's the deal. When I'm typing the Highlights, my focus is mainly on conveying the content and meaning. The word "like" is one of the words I do my best with. In my estimation, it's often most informative and accurate for me to type it as is in the Highlights than to try to rephrase it or leave it out. The word "like" is not merely just a place-holder, such as "um". It may or may not be. Better often for me to include it than not, lest I leave out a content-oriented usage.

EconTalk is not the grammar police. Sometimes, even beyond content or grammar, word uses convey evolving meaning or even just regionalisms or styles of speech used dialectically that matter or reflect ways of speech that are happening at the time the person is speaking. EconTalk may even be recording language in use that is helpful to understanding language as it evolves, in the future.

Or in the case of "like", EconTalk may be recording matters that may have started as dialect and been scoffed at or spoofed, but that have been in widespread use for 50 years. If Valley Girls started using language that caught on and is now in widespread use every day all over the country and indeed around the world for those who speak or listen to American English, then they were onto something useful and productive. It's not just a momentary fad any more. Spoken English doesn't match prescriptive written English grammar. That's just reality.

I suggest reading great authors like Mark Twain for a long U.S. history of listening and writing in the actual way people speak. But that aside, using the word "like" is the way people speak. It does not behoove EconTalk to get all up on some high horse and tell guests to not use the word the way they do when they speak. There's a certain cultural disconnect involved; but, like, I'm like, "Like it or not, using 'like' is likely well understood by listeners and readers alike."

Greg G writes:

I never realized that moderating and transcript writing decisions on the use of the word "like" were nearly that complicated but you have done a great job of showing that they are Lauren.

Like it or not (not in this case) the overuse of the word does sometimes subtly change the meanings of the sentences it is used in. There are tradeoffs between being kind to the guest and changing the meanings of their statements.

Everything always comes back to tradeoffs in economics.

Mort Dubois writes:

Lauren - thank you for that comment, which was much more interesting than the episode, and for all of the work you do. Let me add my sincere admiration and thanks for everyone who works behind the scenes to bring out the best, most informative podcast on the internet.

Mort

Wyn Lydecker writes:

Wonderful, fascinating interview. This interview has finally given me the clearest understanding I have ever gotten of how Y Combinator works and invests. Thank you.

Martin writes:

I'm curious about the guest's (and Russ's) thoughts on nuclear power which came up in the discussion. On the regulatory issue, as Brian points out, Sam is incorrect. But also on the risk & cost benefit part I have a disagreement. The issue, to me, is not that the physics/engineering of the plants is not safe during the operational stage of their life-cycle (my interpretation of Sam's statements). I agree that the risk of a catastrophic meltdown like Chernobyl is close enough to 0 to ignore. The issue is there is no permanent retirement plan for nuclear power plants and their waste. So while reactors may be fine to operate in the scope of, say, the Sam, Russ and my lifetime the waste is unsafe in the scope of millions of years - and the downside of nuclear waste proliferation is greater than a meltdown because they can be used to create weapons.

So let's say the 22nd century is the 'nuclear century' as Sam proposes - what happens in the 23rd and 24th century when there's an exponential increase in the amount of world-wide nuclear waste over that time?

Think of the risk of, for example, a government no longer being able to maintain a plant due to political instability (according to world-nuclear.org China has approximately 26 reactors under construction). Is Sam considering also the risk that the waste from these facilities becomes available to people who are willing to use them as weapons?

Rather than saying - well, nuclear is bad but it's better than coal (or the inverse) - why not agree that both pose a risk of catastrophic consequences to humans which is above 0 and therefore the precautionary principle ought to be applied?

Arnim Sauerbier writes:

Martin, I expect the guest is referring to new reactor types, such as integrated fast reactors (IFRs) and deep-burn modular helium reactors, which can use current nuclear waste as an energy source, producing much shorter-lived isotopes as waste.

Such technologies can solve the nuclear waste problem, if there is either enough political will to support them, or good sense among regulators to get out of the way of their production.

B. Reynolds writes:

This was an extremely interesting episode. I loved it!

It had the feel of the kinds of things you'd read in Wired magazine back in its early days -- when there was excitement in the air with all the changes and you could hardly wait for the next issue to arrive.

Despite the fact that much change is still happening, I've not found anything that quite taps in to potential of what's next and what it could mean for our lives. There's lots reporting on the latest gadgets, but nothing that really looks at the big picture and what shifts could be around the corner.

This episode, the one on Bitcoin, and the interview of Andreessen filled that niche for me!


Ron Crossland writes:

Enjoyable show. Terrific concepts, interesting methods YC uses.

Would have enjoyed hearing about Altman's 99 not-so-big success stories. Were I a newcomer to Econtalk, I might hear that small business is good, but all the examples discussed are about making it big. What about small businesses that remain smallish by design or function?

Tom Burnett writes:

When Altman said, "The point I was trying to make though is we are already much further along this path, I think, than most people think. And even though we don't have the SciFi movie of things injected into our brain, cables going into our brain--which I hope we get someday, by the way. Like everyone's obsessed with biological and mortality and living forever. Which I am never that excited about. Because I think the risk of accidental death is still too high. What I would like to do is live forever in a computer. I would like to have some computer that could download my brain and thoughts, all of that, memories. That would seem like a lot less error prone and risk prone than just dying of cancer. But I would like to not do that either," I thought of Fukuyama's book, Our Post-human Future and Jonathan Swift's struldbruggs in Gulliver's Travels.

Immortality on a hard disk is not for me.

[italics fixed--Econlib Ed.]

Comments for this podcast episode have been closed
Return to top