Peter Thiel on Macroeconomics and Singularity

Well, I wanna talk about, uh, venture capital, macroeconomic things, and the singularity. But I thought I'd start this off in a little bit more of a participatory way and sort of take a survey of the audience of what you all think are the scenarios you're most worried about. It's always hard to come up with a comprehensive list, but I'm going to enumerate seven. I'll go through them, and we'll just see which ones people think they're most worried about. So, number one: we have a singularity in which robots take over the world and kill all the humans. Scenario number two: we have runaway biotechnology leading to smallpox combined with Ebola killing all the humans. Number three, we have sort of the gray goo runaway nanotechnology.

Throw in some that are a bit more conventional. Number four, Israel bombs Iran in the next few months, and we have an escalating war using conventional and nuclear weapons that wrecks the world. Number five, we have... we basically end up with a one-world totalitarian state that uses computers to control everybody. And number six, let me throw in a very banal one, that basically we have global warming, runaway global warming. And number seven, that the singularity takes too long to happen. And so... so those are the seven. Let me go through and just ask people which ones they think are most likely. Vote on the one or two that you think are the most likely. So, just a show of hands for the audience to get a little bit of a sense of what everyone's thinking.

So, number one: robots are going to kill all the humans, the Skynet scenario. Number two: the dangerous biotech, runaway biotech. It's somewhat more than the robots. Number three: we have the runaway nanotech gray goo. Biotech's still winning. Number four: Israel bombs Iran, runaway war. That's... about second to biotech. Number five: the one-world totalitarian state that uses computers to control everybody. Okay, decent number there. Number six: runaway global warming. And how about number seven: that the singularity takes too long to happen? Actually, I think that one is at least a close contender for second, and that's the one I want to actually focus my remarks on. Without articulating my precise sequencing of these scenarios, I do believe the one that we spend the least time thinking about as a catastrophic scenario

is one where the singularity takes too long to happen. And so, if you had discussions among people in this room, it would be sort of like, 'Yeah, it could be really good, it can be really bad. Maybe the singularity is near, or maybe the singularity is nigh,' depending on whether you're optimistic or pessimistic. But that it is far away, and that this being far away as a disaster is something I think people have not thought about enough. And if we thought about the people outside this audience and outside this room, I think, for the most part, if I had to sort of describe what about 99% of the world thinks, they don't really think about this stuff. But they also think it's probably not going to happen and that it doesn't matter.

And I want to try to articulate some of the reasons why I think that is an extremely big misconception, where if it—if it does not happen, it does not happen quickly, we will have all sorts of problems in ways that I think people do not fully anticipate. And so, let me sort of segue the focus to something sort of much more banal. And this is a set of conversations I had with a variety of people running pension funds. And they're basically trying to make 8%, 8.5% a year returns to make sure that all the people in their pension funds will have enough money when they retire. And the problem has been that it's not that obvious that you get those kinds of returns. You get them in the stock market, you get them in housing, and people have arguments that you should get them.

It looks like over 100 years people have gotten them. But if you don't get them, everything is just screwed in an unbelievable way. And this is sort of the implicit assumption that underlies all these kinds of things. Now, I would submit to you that the only way you can get those kinds of returns is if you have extraordinary compound exponential growth in science and technology in the developed world. I'm not going to talk about the developing world, the emerging markets. I think they don't need to have science and technology innovation. China, India, places like that do not need to invent a single thing. All they need to do is copy 19th-century plumbing, and, you know, 20th-century railroads and stuff like that, and they will see tremendous progress over the next 20, 30 years.

But if we're talking about the developed world, we're talking about countries like Japan, Western Europe, the United States, the critical question of where these countries will be in 20 years turns not on increasing leverage or on increasing global trade, which I think are at best one-time gains, but on sort of compounding technological growth, and it requires—and that, in turn, requires some of the kinds of technologies people talk about here to be happening. And if it doesn't happen, you will have a very big problem. I was speaking to another investor in our venture fund in the sort of university endowment. They had an investment committee a few weeks ago, and they raised their hands at the end and said, 'Well, you know, we've been allocating 5% of our money to venture capital, and it hasn't actually been working too well for 11 years or so. They haven't made any money, and we've been wondering whether we should cut that back.'

And I think there are a lot of pros and cons. There are some things that are screwed up with VC. On the other hand, maybe people are too bearish on it at this point. But the deeper point that I tried to convey to them was, well, if that 5% of your portfolio doesn't work because there's not going to be science and technology innovation, then the other 95% is definitely not going to work. And what you need to be worrying about is not that 5%, but the other 95%. And I do think that the question about whether technology is accelerating or not is one that is sort of, you know, there are definitely some worrisome signs that it's not been going quite as quickly as people would like. And I think sort of one way to illustrate this is to go back to the science fiction of the 1950s, 1960s.

Where did people expect the world to be today? So, the future history of Star Trek, circa 1967: you'd have permanent lunar bases by 1980, manned missions to Mars by 1990. From there, it would go on to the stars. You'd have the Jetsons, you'd have flying cars, robots, etc., etc. And even sort of economic, social writers, I think, were sort of—have been sort of proven wrong in a strangely pessimistic way. There's a classic book written in 1968 by this guy, Servan-Schreiber, entitled The American Challenge, and the basic argument of it was that the United States was an exponentiating technological civilization, and because of that, it would accelerate in such a way that it would leave the rest of the world behind.

And unless Western Europe got its act together, the difference between the US and Western Europe would change from a difference of degree to a difference of kind, and Europe would be like an emerging market or undeveloped country, third-world country compared to the United States, because that's what compounding does. And then, fast-forwarding to the US, because it would leave everybody in the dust because we have this exponentiating technology, it would lead to sort of certain social repercussions. And, you know, one of them would be, of course, that people would have to compete less with people elsewhere in the world, so that by the year 2000, the average person—not an executive, not a creative person, but the average person in the United States—

Would have a work life, they'd have a work day, seven hours, work week, four days a week, and would have 13 weeks of paid vacation a year, as a result of exponentiating technological progress. Now, you know, we can sort of easily dismiss it and say, well, the writer was French, and, you know, but I do think this sort of points to sort of a bit of a weird disconnect. Just to throw out two or three other ones where, again, I think the progress may have been a little bit slower over the last 40 years than people like to acknowledge. There is, I think, this very odd disconnect between the story of phenomenal technological progress, which we see in computers and the internet, in areas of biotech, in all sorts of areas where we can describe extraordinary exponential growth, and the way in which it has not translated

Into the macroeconomic context. And to cite sort of just the most glaring example would be that median wages in the United States today are about the same they were in real terms in 1973. And so if you ask, why has there been no growth in real wages for the average person in the US in 36 years when we've had this incredible growth through technology? Even aside from technology, we supposedly had all these gains from global trade and increasing leverage. There's something very odd, and I think you can sort of come up with one of three answers. One answer is that that we are not measuring the real wages correctly, but the problem with that is that the government has a lot of incentives to actually understate the inflation rate and therefore maybe actually overstate the real wage growth, so that argument sort of cuts both ways.

The second way you can deal with it is by saying that yes, there's all this technological progress, it just takes a while to filter through, but the problem is 36 years is kind of a long time. And then the third argument is that maybe there's been a little bit less than we've been expecting and than we need. A third way to get an intuition that maybe things are not quite as robust as we'd like to say is if you're saying, well, we're living in a technologically exponentiating civilization today, and Silicon Valley is one of the places where it's centered. And so, a very natural prediction, if you knew nothing else about the U.S. or the world, would be that the state of California must be one of the best economic places in the world.

And then, of course, you have all sorts of explanations. Well, it's sort of this weirdly broken government. There are other things that are screwed up with it. But I think one of the big things that's a little bit off is that the tech industry is a tiny part in California. It's two or three congressional districts out of about 50. And most of the rest is doing things that are not at all innovation, technology-driven. It's a smaller part of the economy than people think. So I think we have a number of these weird disconnects where median wages haven't grown. California, the center of the tech industry, seems sort of weirdly broken. VCs haven't made any money in 11 years. People seem to be working harder, being forced to compete more.

And I think we need to sort of really rethink this question of how fast technology is happening. One other related thought on this is that is that there's been a great deal of discussion of the sort of finance credit crisis of the last couple of years. And I think there are many different stories, of course, that people tell as to what happened and what went wrong. And you can tell stories of too much regulation and too little regulation, people being stupid or greedy or evil or manipulative. And there's sort of a whole range of different ways people have explained it, but I would submit to you that one kind of way to think of it is not as a housing crisis or a credit crisis or a finance crisis, but as a technology crisis.

And let me sort of try to articulate what I see as the link between the credit problem and the fact that the Singularity may not be happening as quickly as it needs to for our society to function. All forms of credit involve claims on the future. So, I will borrow—lend you money because you can pay me back more in the future. And so, in some ways, credit works if you have a background of growth. If everything gets better every year, all the time, you will not have a serious credit crisis. You may have a short interruption, then you can print some money, and everything will come back to normal very, very quickly. But if you actually have a very severe, uh, credit crisis, uh, if you've been wrong about your assumptions, sort of, of future growth, then you'll have a severe credit crisis because it turns out that, uh, all the claims of the future can't be, can't be met. You know, the house in Las Vegas that, uh, was, uh, was going to go up relentlessly because we were going to develop cheap alternative energy where, on the cost of air conditioning, it would be the equivalent of ten dollars a barrel—

didn't happen quickly enough, and then all of a sudden the value collapses. And I think there is sort of this very deep link between the two. And if you actually take one step back, I think if you think about the history of these various bubbles that we've had in the world in the last few decades—Japan, the tech one in the '90s, housing, finance, maybe in the 2000s—I think the most important one, still in some sense, is the tech bubble of the late '90s. There were parts of it that were real. A lot of it, unfortunately, was not. And it in many ways involves the story about the future, about rapid technological progress. And when that story turned out not to happen as quickly as it did, you know, much of it happened.

I mean, the internet kept growing. There were pieces of it that were definitely right. But when it happened more slowly than people thought, uh, it turned out that a lot of these investments ended up becoming worthless. And, uh, and in some sense, you can think of the housing and, and credit, uh, bubbles as ways to try to compensate, because people still have this mental model that we needed eight percent a year returns, because that's what the academics told them, that's what the, uh, people who research the stock market for the last hundred years, where we had rapid technological progress, said worked, you know, you can get eight to ten percent a year. And then, when it turned out you couldn't get it through technology, people tried to make it up through leverage.

And that worked for a while until it broke very, very badly. Because I think the problem was that you can't actually make it up through leverage. The way you actually solve it is through real technology, real innovation. And I think where we find ourselves today in the world in 2009 is in a world where the only way forward for the developed countries is through rapid innovation and progress in science and technology. It has to be more rapid than it has been the last 20, 30, 40 years for us to meet the kinds of expectations that are built into the very fabric of our society. There obviously is a way for our society to function if there's no progress at all. You know, people could save 40% of their money.

They could push the retirement age to 80. There are ways for it to function, but this is not what we're expecting. If we want to have anything like a smooth transition as a society over the next 20 years, I think we have to actually see the question of a good Singularity as the single most important political, cultural, economic, technological question that we have. And we have to sort of, in some sense, go back to the future. I think we have to go back to where the future stopped in the late '50s and late '60s and try to bring back the kinds of quantum leaps in technology that I think will be necessary to drive the world forward. So I just wanted to throw some of these thoughts out there, and I think I'd love to make it a little bit more open-ended and just have a little bit of give and take of some Q&A.

So thank you very much.

Hi. I agree with the positivity about technology and so on, but really the question is, if you're saying that this is—the bubble in the economic situation is based on this technology not happening enough, is that—does that mean you're saying that the behavioral finance doesn't happen? That is, that the greed accumulates and so on, so that, you know, even if we did have all this technology happening at the pace we'd like, are we still going to get greedy and irrational?

Well, you know, there's been a lot of focus on behavioral finance and the psychology of bubbles and how people behave in herd-like ways. And there obviously are things like that that have been true. But I think to really get a bubble going, you have to be able to tell a story. And people have to believe the story, whether it's a story that we're headed towards massive technological growth in the '90s or a story that you could have relentless housing appreciation in the 2000s. And I think, I think the housing story was a dumber story than the tech story, because at least the tech story was about what you needed to do. The housing story is like, you know, it was from, you know, bricks to clicks back to bricks, you know, in the 2000s.

And that was, I think—and to the extent those stories have run out, I think it actually will be very hard to generate the kind of mass psychology that leads to bubbles. One of the things I would predict, by the way, is that we're not going to have any more bubbles in the next 20 or 30 years. I think it's either the boom-bust bubble cycle is over, and we're either going to have a bust, or we're going to have a real boom, but we're not going to have any more fake bubbles. People are so aware of them, they're so focused on it, it's not going to happen again.

At the margins, yes. Yes, sir?

I totally accept your premise, so now I'm going to ask a question, which is if you had to choose between, assuming they're mutually not exclusive but competitive, government encouraging technological development or government getting out of the way and allowing unpredictable innovation coming out of the human imagination, which would you prefer?

Well, again, you know, these questions, I think, are always a question about theory and practice. So in theory, I would like both. In practice, I think I would like the second. I think, you know, this gets into sort of a deep political, philosophical question. I think that Our government is not focused on science and technology as a priority, and therefore it is not able to fund these things adequately. If it was focused on that, I think there might be more of a role for government, but the reality is it just is not. you know i think there is there's a role for basic research which is uh... which people can't monetize directly and you know there is a role for you know doing research into areas that people cannot uh... capture the value for twenty or thirty years and in theory if we had a government uh... that was not captured by special interests and that was not uh...

That was not focused on narrowly incrementalist kinds of projects, it might be able to work. I don't think we have that sort of a government at all, and it may be practically impossible.

Hi. It seems to me that productivity gains actually have taken place since the 1980s, but the money hasn't gone to the people who've gotten more productive. It's all gone to the very top, and that's the reason for the lack of growth. The guys at the top are getting all the dough while everybody else is making the same amount of money.

Well, there's obviously been an increase in wealth inequality over the last 40 years, really. That does go back to about the '60s or so. But I think even if you look at average, as opposed to median, incomes, they haven't gone up that much. And so this is not, you know, it's maybe they've gone up 15, 20%. You know, there's been productivity growth in some sectors. I would argue there's been productivity decline in many others. So I think, for example, if you look at how much time does the average person spend commuting today versus the 1960s? Is the public transportation system more screwed up now than it was in the '60s? Is the public school system more efficient, less efficient? There are all these places where productivity is not being measured and it's probably gotten dramatically worse.

So I think it's gotten a little bit better, but not a lot.

Love your talk. Thanks a lot. What is the single most important thing that you worry about most and why? Thanks.

Well, it's some combination. It's, you know... One of the things, on a long-term basis, the thing I worry about most is what I talked about in my speech, which is that this is not going to happen. It's not going to happen quickly enough, and that we need to sort of redouble our efforts, and that as a society, we don't see this as important. I would be more optimistic if other people were more worried.

Hi. Appreciate your sense of urgency with technology. There was one scenario of the seven, or one additional scenario to the seven that you started with that I'm curious in your thoughts on, and that is a scenario that's highlighted by Francis Fukuyama in Our Posthuman Future. And he talks about essentially technological progress integrating with human evolution such that you sort of get a species divide where you get the rich that are able to afford sort of auxiliary supplements and treatments that enable them to become sort of superhumans, and then you have those that are unable to afford to. The second step being it's hard to apply the rule of law to people that are genetically becoming so divergent.

How do you, when you think about the evolution of technology and you think about the impact on society to the extent that you've described it, how do you see us being able to reckon with that day where humans actually begin to diverge?

You know, I find it just really quite implausible. I think there may be some divergence where some people get technologies who are wealthy earlier than other people do. So, you know, I think people who are wealthier got access to the internet before people who are poor, but eventually the digital divide gets overcome. You know, in the 1980s, if you had a BMW or a Mercedes, you had an airbag, and other people did not. This is not an argument against airbags. You know, so I think that... You know, I think that this is just sort of, I think the Fukuyama argument's basically hysterical and just, it's sort of not even wrong. Now, I would say, and I would say, you know, again, I'm much more worried about the stuff not happening than the details being weirdly off.

What I would, however, use arguments like Fukuyama's as an illustration of is, why is all the worry about the technology going badly wrong? And why are people not worried enough about it not happening? And that seems to me to be the... And so for every hundred Fukuyamas, I think there is maybe only one person who's making the argument the other way, that things are not happening quickly enough, and that that's the real problem.

Peter, Peter, I'd like to come to the question of what is it that you are worried is not happening fast enough? Because at the beginning you made a fantastic claim. You said it would be a major disaster if the Singularity is delayed. But it seems to me that what you were arguing wasn't the kind of incredible technological explosion that the Singularity would be with the advent of superhuman intelligence. It seems to me that you're just arguing for a return to the kind of quantum improvements in technology that took place in the 1950s, whereas the Singularity is just immeasurably faster than that. So what is it actually that you are looking for, and do you really think that the kind of dramatic change in the Singularity will still be good for economics?

Well, I think the faster technological progress happens, on the whole, the better, given the way our society is geared. I'd be happy if we went back to the '50s and '60s. Obviously, if we could do better than that, that would probably be even better. Maybe there's a case where it's worse, but I wouldn't be worried about progress being too fast until we at least get back to the '50s, '60s level. And if we can do even better, that's probably better, but I'm not particularly worried about that.

That sounds like a different definition of the Singularity.

Well, I think if we took a survey here, we'd find many different definitions of it.

How would you invest today to protect yourself against the failure of technology in the future?

Well, you know, as a venture capitalist, I still am interested in investing in technology, and it's a sort of Pascal's Wager, because I think that if that doesn't work, nothing else is going to work. So I actually do think that's sort of, on the positive side, what one has to try to do. There are sort of a lot of nuances to that. There are questions about what is real technology, what is not. And obviously there are many companies that are branded as technology companies that are really not. And so this is sort of an investing digression. But if you looked at the 100 companies in the NASDAQ 100, one of the somewhat disconcerting facts, I think, is that most of them actually involve bets against technology in the sense that you will do well by investing in these companies if there's no technological progress.

So investment in Microsoft is good if Linux never takes off. Investment in Hewlett-Packard and Dell and Cisco is good if they never get cheaper competition. An investment in IBM is good if we still are going to keep the same kludge software from the '60s and '70s for the next 30 years. Even Google, investment in Google is good if there's going to be no new search technology. So I think Intel maybe still is like pro-tech, but actually most companies that are called tech companies are actually more like banks where they throw out profits as long as nothing happens and as long as there's nothing disruptive. And I think that's sort of symptomatic of where it's weirdly off. But it would be an interesting sort of thought experiment to go through and think about

what companies are you actually betting on technological change versus what companies are called tech companies where you're really going to make money if nothing happens.

Your answer reflects an interest in aggregating more capital. What about investment strategies to preserve capital?

Well, I think that that's sort of a... That's a much longer discussion. I tend to think that people should be realistic about their assumptions. If we're in a world where technological progress is slowed, you shouldn't believe that there is a way to get 8% or 10% a year without taking enormous risk. And so if you're not willing to take enormous risk, you have to dial down your return expectations very dramatically. And I think something like the Madoff scandal, you know, obviously it's like, how are people so stupid to give this person all this money, and you're looking at all the details. But one of the reasons it happened, psychologically, was because people thought 8 or 10 percent with zero risk is perfectly normal.

And that's why nobody asked any questions. So I think the thing, if you are a low-risk investor, you should expect very low returns and be very wary of anybody who tells you they can do better without warning you about incredible risks that you're going to be taking. Thanks a lot.

Hi, I find the logic of your argument is really compelling. The idea that we've sort of leveraged our futures against this growth that we think is going to happen and so we have to provide that growth or we're kind of screwed and we have to keep providing more and more growth and it has to keep growing. But it's sort of an argument that we're continually looking at the short term ahead of us, that we've always bet on the short term. And so as it goes forward, we've eventually always bet on the long term. So I'm wondering where this ends up far down the road. Doesn't this mean that this growth would have to go on indefinitely? What happens when we reach the sort of limits of growth? Are we then in for some huge catastrophe?

Or what happens at the end there?

Well, you know, there's an argument that what happens is what happened in the last two years. And that basically that in the long run, the short run becomes the long run, and that somehow the two points collapsed in 2008, and that's the start of what happened this very last year. Now, you know, I think—but if we were able to get growth restarted and get science and technology to happen, maybe it's not something that goes on forever, but it could go on for another 100 years, and then we can deal with the problem of when it slows down 100 years from now.

But you seem to be saying that there's always sort of more growth to happen.

No, I'm not saying there's anything automatic about growth. In fact, I'm concerned that there's not as much as people think. I think you have a big problem economically when people expect growth and the growth doesn't happen.

And so I think we either need to dial back our assumptions and expect much less growth, or we need to be refocusing our efforts to society and make sure that growth happens, or maybe if we want to cover our bases, we should do both. But I mean, we're—it seems like we're still sort of talking about growth on the human scales that we have it today. I mean, this is—this is supposed to be a point at which we've, you know, transcended humanity, and maybe we're—we're living on computers, or we have, you know, the changes are supposed to be sort of unimaginable by this point. Is there no point at which there is nothing left to do anymore and we sort of leverage against something that we can't even accomplish?

Or is that just too far out to speculate?

Well, I think if you get a good singularity to happen in the next 20, 30, 40 years, then all these growth worries go away because there will be no credit problem. It will pay for everything. And then we can rethink economics at that point.

I want to ask about AGI. The first company that will deliver true AGI has the potential to evolve into something that will make Google seem like a small business. I want to ask you about your—share your views on the world after AGI is introduced.

Well, you know, I think that is true. I think the question is obviously how fast does AGI happen? You know, how does it go about happening? My basic—my basic focus as an investor and the way I think of it as a venture capitalist, as an entrepreneur, is that the sort of scenario that is being underestimated is sort of a human-computer divide where the reality is there are some things computers are much better at than humans, there are some things humans are much better at than computers. And so I think the area to try to focus on the next five, ten years are figuring out ways to get the division of labor between humans and computers to work better. On the point where computers are vastly better than humans in every single dimension, I'm more agnostic on when that happens exactly, but I think on a 10-year horizon, on—I think the, you know, the place that I think is very underestimated is this—this combination. A PayPal story from—from 2000 for getting me thinking along this direction was that we had this enormous problem of fraud. People were stealing all the money on the service, and so we had, you know...

There was a woman running the fraud department, and she cried herself to sleep every night because so much money was being stolen. You had 20 investigators, and they were looking for needles in a haystack with millions of transactions. And then we had a team of mathematicians and computer scientists who were trying to build the equivalent of an AGI system to track down all the bad people. And the problem was, through all these sort of qualitative ways, that changed. And it was clear that AGI was at least 10 years away. Maybe it was 30 years away. But the company was going to be out of business in about 12 months. To the extent you needed a pure computer solution or a pure human solution, you were going to go out of business.

The critical answer turned out to figure out a way to figure out the right division of labor where computers would screen through certain things, then you'd have humans qualitatively investigate them, and that turned out to be incredibly valuable. And my general thought is that that sort of a paradigm is still very underestimated for AI research, even though it's not, strictly speaking, AGI.

Thank you. Hi. As a part of the larger slowdown in technology, you seem to have particular dismay in the lack of progress in Silicon Valley. And I know recently you moved to New York from California. I didn't know if, in light of these fiscal problems, if you thought—well, is it true that you think that there is any particular problem with Silicon Valley? And do you see this as sort of startup and technological innovation moving elsewhere and possibly creating other startup hubs?

Well, I've split my time personally between New York and Silicon Valley for about the last four years. I lived in New York for a while in the '90s, so I've spent a lot of time in both places. There are things I like a lot about both. I don't really see technological innovation happening in that many other places outside of Silicon Valley. I think there still is, you know, I think it still is the leader. I think there still is a lot of innovation happening there. I think it's just perhaps not quite as much as people think. I would say one thing that I think is true on this whole science-technology question—it's very important—is that a lot of it's like a black box. You know, we don't really—most people don't understand it.

You know, it's like we don't understand what's going on in superstring theory or quantum computing or at the frontiers of nanotech or all these different areas. And so we sort of end up deferring to various experts, be they university professors, entrepreneurs, to some extent venture capitalists. And I think the political thing one has to be aware of is that almost all the experts are biased in an optimistic direction, where, you know, if you're a university professor working on cancer research, you're not going to say, you know, none of the stuff's going to work for the next 20 years. I just want to get a grant so that I can study these ants in the Amazon. And if you're an entrepreneur, you're not going to say, well, this is probably a bad idea for a company, but, you know, it might work, it might not work.

And if you're a venture capitalist, I think you're probably... typically among the worst in this category, where you'll end up saying, oh, there's all this incredible stuff happening, and it's just great, and there's nobody who really has an incentive to tell the story on the other side, and that's why I think the debate is weirdly unbalanced. If you ask people on Wall Street or in Washington, D.C., they think of science and technology as this black box they don't understand anything about, but it's going to happen elsewhere. And it's an important problem, but it will be solved by other people. And this sort of a free-rider problem, I think, is endemic, where it may happen, but I don't think it happens if everybody thinks somebody else will do it.

Thank you.

What would you say to the thesis of corporate success as a disincentive to innovation, in the sense that today we're living in a time where multinational corporations and media conglomerates are consolidated like never before, and everyone knows anecdotally that corporations, as they grow, become more inflexible? So, just your perspective on that.

Yeah, I think, I think there is a degree to which large corporations are definitely less innovative than smaller corporations. I tend to think the bigger problem in my mind goes to people's time horizons, where I think if you're the CEO of a public company, you have to make your numbers every quarter, and so I think that tends to be a really big problem, and we need to somehow focus on a longer time horizon. And I think this is—this is a problem not just for corporations; it's a problem for individuals, it's a problem for our political system, where people just focus on the next election but not really fixing these sort of chronic problems that are over 15, 20, 25 years. And most of the real challenges we face as a society are of a chronic, long-term nature.

They're not of this sort of acute, temporary nature. So it's a problem for corporations; it's a problem in many other cases. I think if you look at the U.S. versus China, I think the U.S. is better than China in every dimension except for the time horizon dimension, and that may be so important that it may be decisive. If China ends up beating the U.S. and overtaking us in the next 10, 15, 20 years, it'll be because of this. I'm told I have time for one more question, so go ahead. Go ahead.

Arguably, the Nazis came out of the Great Depression, and they were successful in pioneering recording technology, television, jets, rockets. Even with IBM running large parts of the Holocaust, historians date the dawn of the Information Age in 1933 Berlin. And I'm wondering if there's a danger in today's economic crisis that the Singularity will pass into a totalitarian system and it won't come out of a free market or a democracy.

Yes. I'm sure that is a very serious danger, and that's why I think we need to do things. As someone who's a civil libertarian, I believe it is very important for us to get the economy on a trajectory of long-term growth, because if it's not, you will have all sorts of political pressures that will go in a socialist or more likely in a fascist direction. And I think the risk of fascism tends to be very underestimated. And I think in some ways, advanced societies, when they're put under incredible stress, tend to become fascist. They don't become communist. Thanks.