Keynote Interview: Peter Thiel / The Buttonwood Gathering

You've been on the road for a few months now pushing the messages of your book in terms of how the current dead hand of the state has been holding back innovation and innovation has been slowing down. Have you become clearer as to what in practice we would need to do to reshape the state to really encourage more innovation?

Well, it's like a very broad, difficult-to-answer question. I do think the regulatory structure matters tremendously. I'm always interested in the state with respect to technology. So I think there are all sorts of debates people have about marginal tax rates or all kinds of other areas of policy. But the area that I always think is very important is the amount of technological innovation that happens. Because I think that's what sets whether we're going to have a much better future or not. I think it gets driven by the rate of technological progress. And I think we've been living in a world for the last 40 years where it's really been sort of a tale of two worlds, where there's a world of bits, computers, the Internet, the information technology revolution, which has been very lightly regulated and has seen extremely rapid progress.

And then the world of atoms has become increasingly heavily regulated, and that's almost all other areas. So that would be everything from energy policy, where it's very hard to build a new, improved design on a nuclear reactor, for example. This would be a very difficult thing to do. Or drug development by the FDA, where it's costing more and more money to get drugs through the process. It costs something like $100,000 to start a software business. And people say up to a billion dollars to get a drug through the FDA. And so when you have four orders of magnitude higher costs, we get more video games and fewer life-changing medicines. And so I do think it's been this, I think we've had something of a—

—as close as you can get to a controlled experiment the last 40 years, where you've had tremendous innovation in bits, much less in atoms, and they've been much more heavily regulated. And I think that's something that needs to be rethought. There are many reasons they have been regulated. There are many reasons people are scared of technology. They're scared of innovation. We live in a society and a culture that's dominated, I think, by fear and hostility towards technology and science. And my view would be that a lot of these fears and hostilities are somewhat misplaced, but they are powerful drivers for what's happening.

But I suppose we're seeing with companies like Uber and Airbnb, this attempt of the world of bits to come out and shift the world of atoms a bit by challenging the regulation and betting that technology can be fleeter of foot than the regulators. I mean, do you think that is the way to change in some of these areas, and how far can it go?

Certainly, there is an aspect where you sort of don't ask for permission, you ask for forgiveness. And if you have a sort of consumer technology or consumer product that can get adopted very quickly, then it becomes very hard to reverse because if you have tens of thousands, hundreds of thousands of people using a product, then it's much harder for regulators to say this is a terrible thing, terrible idea. I don't know if that's a good template for something like drug development where you couldn't get hundreds of thousands of people to use drugs before the FDA approves them. And then when they work, the FDA would have a lot of pressure to approve the drugs. You never actually get to run that experiment.

And I mean, do you think that would—I mean, is there, is there therefore going to have to be a different strategy that we come up with to, to free up that process? I mean, if we want, if you want to go down the line of personalization of medicine, for example, it seems, it seems deeply political.

And so I, I don't know if there's a way to free it up independent of reform of the FDA or perhaps, uh, and perhaps I, I always wonder whether there's some point in which you have some sort of regulatory arbitrage that happens globally where, um, after all, the US is only 23% of world GDP, and yet the FDA somehow sets the drug policy for the entire world. So is there a point where, say, China says, 'We're not going to defer to the FDA, we will set our own policies, and maybe we test for safety, but we don't test for efficacy?' There's all sorts of different parameters one could come up with that would create a very different drug development policy. But that hasn't happened yet. But maybe there's some regulatory arbitrage in the next decade that could start happening.

But do you think, I mean, obviously with something like Uber, we've seen already the dramatic benefits that having a much more efficient system can produce? I mean, do you think that with medicine, we are missing out on a huge number of benefits because of the FDA, or is it just a bit slow?

Well, I think failure is often overdetermined, so there's a regulatory piece. Maybe the science is really hard, actually, and so we've hit some limits in our ability to do things. There are probably a lot of different reasons it's been hard. There's a history of failure where, if you've had little progress for 40 years, it discourages talented people from going into a field. So there's a certain hysteresis where success begets success and failure begets failure. So I think these things are very overdetermined, but one of the challenges for those of us who would push for a lighter regulatory framework is that the benefits are always counterfactual. Look at all the medicines that aren't being created because of these heavy regulations, and that's not that compelling somehow.

I mean, what do you think would happen if we did just abolish the FDA, for example?

It's probably a slightly too drastic binary, so I'm not sure the alternatives are incredibly heavy regulation or none at all. But certainly, certainly, if you, say, got rid of efficacy tests and simply limited it to safety tests, we would be developing drugs much more quickly. They'd be getting to market much faster. You might have some drugs that would be slightly less efficacious but much cheaper. So, you know, in a lot of areas, if you have a product that's 90% as good but costs one-tenth as much, people might want to do that. That's not an option in drugs where efficacy is always measured, efficacy relative to the best drug in the world. And so you can never develop drugs that are slightly worse but much cheaper today.

So there are a lot of things that would be very different. But you just don't see it. I mean, I wonder whether this is the case across the board where you're seeing too much regulation from your perspective. I mean, do you think any of these areas that remain dominated by regulation—I guess let's think about primary education, K-12 education is another area where I think a lot of investors have said, 'Well, we could do this better, but it's just such a regulatory barrier to success that it's not worth going into.' Do you see any hope in that area?

Well, I think there's always things you can do even in heavily regulated areas. So I think in certain areas like healthcare or education, which are fairly heavily regulated and also are very broken—so there are things you have to do, but you shouldn't underestimate the regulatory challenges. So if you have a simple solution, where we say the MOOCs were a relatively simple solution to improve university education, make it much cheaper, possibly somewhat better, that's very hard to do because the university system, sort of the self-regulating bureaucratic system, it's very hard to break into. So one has to think very hard about the kinds of business models one develops in these heavily regulated contexts.

We're always very bullish on contexts which are maybe heavily regulated and become just slightly less regulated. This was true, for example, of the space industry in the last decade where it became—there was an opening for private space companies to be built in the U.S. Something like SpaceX could be started in a way that would have been harder in the '80s or '90s.

As you look back at PayPal, which is where you first really came, got a successful business out there that made a lot of money—I mean, we've now had this revived interest in Bitcoin and digital currencies and so forth, which I think all run up against the problem of: will the state allow a private money system to exist? Do you think that PayPal could have evolved differently along those lines, or do you think that fundamentally that could have been the equivalent of Uber for finance, and is Bitcoin gonna do it this time around?

Let's see. So, I always think PayPal had these goals of creating a new currency. We failed at that. We just created a new payment system. I think Bitcoin has succeeded on the level of a new currency. The payment system is somewhat lacking, so it's actually very hard to use, and that's probably the big challenge that exists on the Bitcoin side. I would say—I would say that with respect to finance, we're generally in a more heavily regulated world than we were in '99–2000 when PayPal got started. And so it might actually be very difficult to start PayPal today. I'm not even sure you could build PayPal today because the regulations are tougher. And so a lot of this was in a somewhat gray zone in '99–2000.

Today, it might be much harder to get started, much bigger barriers to entry. At the same time, there's still a lot of people trying to do things in financial technology in one way or another. The challenge is that it's a sector where you're having increasingly heavy regulatory headwinds, and that's a big worry.

If you were to sort out the payment system around Bitcoin, do you think it could catch on in a way that would be seriously disruptive to existing fiat currencies?

Hard to say. The fiat currencies, if you look at a dollar bill, it says, 'This is legal tender for all debts, public and private.' And the stress you have to always put is on the word 'public.' And so as long as you have to have dollars to pay your taxes—if you don't have dollars to pay your taxes, people with guns will come after you and lock you up or do something bad to you. And so, you know, you can't use Bitcoin to pay taxes. You'll have to still convert it into dollars. And so on that level, it doesn't really threaten the supremacy of the dollar as a reserve currency, which is probably ultimately backed up by U.S. military power.

And what if a government somewhere, maybe the Swiss or someone, was to say, 'Let's embrace an algorithmic digital currency and put the weight of the state behind it'? Do you think that's what Bitcoin needs to take it to really catch on and be an effective currency?

Well, it becomes a threat. Let me put it this way: it becomes a threat to fiat money at a point where Bitcoin is encrypted in such a robust way that the tax authorities can't break the encryption, can't tell how much money you have, what transactions you're doing, where it all becomes untraceable. There was sort of a '90s cypher-anarchist vision of the future where encryption technology would make it very hard for the tax authorities to break in, and it was like this defensive technology where the state could not—would get diminishing returns on violence and extracting revenues or tax revenues from its subjects. Now, this is not the way the computer revolution has really gone in the last 17 years. So, in the '90s, we thought of IT...

Revolution is trending towards a decentralized world that empowered individuals. Today, it's again perceived as trending towards centralized databases, centralized servers, a more transparent world, a world in which privacy is always perceived to be somewhat threatened. And that's one where, again, Bitcoin only threatens if you can't trace any of the things you're doing with it.

Now there's this discussion around—obviously, a lot of people are very worried about cybersecurity at the moment. And there is this sort of concern that maybe, in some quarters, the government is using this as the way to—some of these hacks and so forth—as a way to justify massive intrusion into people's lives. What do you feel about that threat at the moment or that concern?

You know, there's probably—I mean, I do think there are some—I do think our intuitions about cybersecurity are very, very poorly formed because we probably don't fully understand how much the data can be taken from us. We have sort of this Cartesian model where we think that we know what's going on, and if we're not saying something, the other person can't figure it out. And there are sort of all these ways on the internet where things can be figured out about you in really subtle ways that can be somewhat disconcerting. I think the way I like to frame the national security issue is always that technological innovation would mean more security with fewer privacy violations. This technology always involves doing more with less, so more security with fewer privacy violations. The, uh, the political debates we tend to have in our society are always anti-technological in character, and that they assume there can be no innovation, and so the debates are always, um, more with more versus less with less. This was, you know, there's a...

More government spending on more projects, spending versus austerity—it's never efficiency, which would be sort of a technological dimension. And I do think there has been sort of this runaway, very low-tech national security apparatus in the US that sort of hoovers up all the data in the world and has no clue what to do about it. The image I would present...

But is that better than having a super-effective state that really knows how to be Big Brother in a big way?

Well, there certainly are dystopian technologies we can envision, but I think the practical—because it's the Keystone Cops rather than Big Brother, you end up with sort of a very intrusive, low-tech state where you have incredibly intrusive airport security measures that probably don't do very much good. And these are sort of replicated on many different levels. So I think the intrusiveness on some level is a function of its ineffectiveness.

Now, you've talked about the state having this deadening effect on innovation. And one of the points that several people have made to me is that, you know, didn't the US government invent the internet in some ways? Didn't DARPA play a crucial role? Yeah, indeed. I mean, is it as black and white as maybe is often presented, that the state just deadens innovation, or is actually there a much more complex ecosystem there that we need to figure out how it works and how to replicate it?

Well, certainly in the 1950s and '60s, there was a certain part of state-driven innovation that happened. There was also much less regulation, though, so you would not be able to get the polio vaccine implemented. When they first tried the polio vaccine in the '50s, a whole bunch of people were given too much, and a number of people got killed, got infected with polio. If that happened today, put the kibosh on that for like 20 years or something. And so I do think the regulatory thing is much different from the '50s or '60s. But in terms of the state being a driver of innovation, there has been a decline. And it's worth thinking about that a little bit more. You could not do an Apollo space program today.

You could not do a Manhattan Project. If Einstein sent a letter to the White House, it would get lost in the mailroom. And so there is something very odd about the sort of... There's probably...

But do you think there is a potential Manhattan Project that we're sitting on and not doing because of the way modern states work?

You have to believe that complex coordination by the government is possible, and even most of the people on the left no longer believe that. So I would say the minimum condition... for the state to do things is that you have to believe that you can do it. And even the people on the left do not believe in complex coordination.

But we did do a Manhattan Project.

That was a different set of people in the '30s and '40s who believed in a form of central planning, coordinating hundreds of thousands of people on this one very specific determinate project. A state is not being run by scientists or engineers, it's being run by lawyers. Lawyers prefer process over substance. If the state is going to innovate, it has to make substantive determinations on what you would do, what you would prioritize. For example, in the area of clean energy, the substantive version would be you decide, OK, this is the energy we're going to build. And then there's this very complicated set of things you have to do in making the grid work. I personally believe there should be more nuclear power.

And that is a complex coordination problem. You have to figure out where to site the nuclear plants. You have to figure out where to put the waste. And so there is probably some state involvement in getting that coordination problem solved. That's not the way we think about what the state should do today. We think of it as a portfolio, financialized portfolio theory, where the state should invest in a hundred different projects, let a hundred flowers bloom, see what works, see what doesn't, and then you end up with white elephant projects like Solyndra, which go absolutely nowhere.

And do you think the great minds in Silicon Valley are capable of coming up with a solution to the failure of the state? Can Silicon Valley invent a better version or disrupt the existing model of the state?

I don't know if it's a technological issue or more of a cultural or political set of questions. I certainly don't think that it's a question of different voting priorities or online voting. I mean, there's certainly all sorts of technological things like that that one can envision, but I believe most state action gets driven through these sort of quasi-permanent bureaucracies and agencies, and there is sort of a question how one would streamline or rejuvenate them, and that's the kind of debate one would have to have.

So let's take a couple of questions from the audience. Right at the front here. Please say who you are and keep the question short.

Yes, hi. I'm Carrie Sheffield. I blog at Forbes, and I noticed you're wearing pinstripes. In your book, you talk about pinstripe suits as being empty suits. Can you talk about...

It was about people pitching clean tech companies in 2005 to 2008. If they were wearing suits, you had to be skeptical and think they were just salesmen. So if I'm trying to promote my book in front of an audience in New York, very different context. So there are no timeless and eternal sartorial rules.

Very nice.

Steve Jobs got that wrong, didn't he? There can be personal brands or things, but not everybody should wear a black turtleneck sweater. That would be really silly.

Can you talk a bit about net neutrality? I know you've generally been against it, and it seems like with the new rules we had, or the outline that Wheeler proposed, it seems that it's coming down the pike. Is the answer to push back just bring litigation, or what do you see playing out here, and is there any sort of recourse that opponents can have?

I don't have a particularly strong view on the net neutrality issue. I don't think it's necessary for us to be doing this. I think the system we've had has worked well enough for a long time. That being said, there are aspects of the cable companies that are these fairly bad monopoly-like businesses that are sort of these static monopolies, trolls that collect tolls at the bridge, or they're tax collectors of one sort or another. And that is always the argument for, sort of, this antitrust argument for something like net neutrality. My skepticism is that I suspect, I think it's a necessary condition to have a bad monopoly for the government to do something, it's not a sufficient condition. You'd also want them to be able to replace it with something better, and I'm still not certain they will replace it with a saner pricing mechanism or anything like that.

When you bring up the antitrust question, and you got a lot of attention for your apparent enthusiasm for monopoly that you expressed in the book, which was very much that every entrepreneur ought to dream of creating a monopoly.

On the inside, you always want to aim for monopoly. If you're starting a company or if you're investing in a company or you're an early employee, you're much better off being in a monopoly like Google than in a Darwinian competitive business like a restaurant in Manhattan.

But from society's perspective, do you think there ought to be antitrust? For example, yeah, I mean, Microsoft have been making the case recently that if Google were treated like they were treated in the '90s, Google would be broken up now and so forth. Do you think those sort of interventions in tech make sense?

You know, there's a point where they do. I think you have to be worried about monopolies that are permanent and that become very static. I think in a dynamic, progressing world, they tend to get replaced by other ones as technology develops. The government historically, I think, has done a pretty bad job. It went after IBM in the late '70s, when the reality was that things were shifting from hardware to software. They went after Microsoft in the '90s, just as things were shifting from the desktop to the Internet. If the European governments were to go after Google today, I suspect that would be almost an indicator that somehow the Google monopoly had run its natural course. So they seem to time their monopoly.

Do you think Google might have run its natural course?

Well, they have a set of challenges in the shift to smartphones and whether they have as much pricing power on smartphones as they do on desktops. There are all sorts of ways that search can be broken down and aggregated in different ways. So it's not at all clear. So I think the government has historically gotten it quite wrong in technology. The really bad monopolies tend to be ones that are legally enforced by the governments. The Post Office is a bad monopoly, or urban slumlords who benefit from government zoning laws and are able to charge monopoly-like rents in cities like New York City. That's another example of a very bad monopoly.

Another question in the middle there.

Dov Heller.

You seem to think that the reason we can't get real research and innovation done through government now is a lack of confidence in government. So why is the answer not just to try to restore confidence in government?

Well, I'd say, I'm not saying that's a sufficient solution, but certainly it's necessary. You need to have people who have confidence in themselves, who believe they are capable of making determinations, and we don't have such people in the government. I'll go back to my Solyndra example. You had a cylindrically shaped solar panel versus a flat panel. High school geometry, it's 1 over pi as efficient as a flat thing. 2 pi r over 2 r, so it's 1 over pi as efficient as a flat panel. From high school geometry, you could tell this was a mathematically inferior design that was never going to work. You had a Nobel laureate, Stephen Chu, running the Energy Department. He had a Nobel Prize in physics.

He was not allowed to use ninth-grade high school geometry and say, "This is a dumb idea. I'm going to veto this. This will never work," because you're not allowed to use substance. It's all process. It is, no, they went through the proper processes. There were no bribes or kickbacks. They checked off all the boxes. And that's the way we think about it. So for the government to do this, you need to have far more substance, far less process. And that's a very different government from the sort we have today, and very different from the people who make it up. I think you'd have to have far fewer lawyers, far more engineers and scientists.

Those aren't people like that. One more question in the middle there. Yep, gentleman with the gray hat.

You say who you are. Yeah, sorry, Church Lewis. Going back historically, where government appears to have done useful things in the past, perhaps in the Great Depression, improvements in social policy that came out of government action during that period, the management of the financial crisis 2007, 2008 with quantitative easing, and an interesting juxtaposition in the Depression when Herbert Hoover, an engineer, pragmatist, et cetera, was elected to president and did not come up with adequate solutions for a national economic collapse and was replaced by a very statist kind of regime. What are your thoughts about those sort of examples from history where government appears to have done something constructive or useful for the society?

Wow, this is a whole crazy history perspective here. Look, I'm not a fan of FDR. I'm not a fan of the New Deal. I think a lot of it was very, very misguided. But... But the remarkable thing is that it actually worked and that by the late '40s, the US had a much larger GDP than it had say even in 1928, '29 before the start of the Great Depression. And so even though you had a government that had very high taxes, intervened in all sorts of ways, it was relatively light on the technological innovation, and you had tremendous amounts of innovation. So in the 1930s, just to name the '30s innovations, it was the beginning of the aerospace industry. You developed secondary oil recovery, plastics, household appliances like refrigerators, things like that became far more widespread.

There was tremendous amounts of innovation that happened, and when you have that, that gives you a powerful tailwind that can pay for even the most misguided of policies. In a world where there's much less innovation happening, in a world, Keynesian economic policy, sort of the idea that money grows on trees, it worked in the US of the early 20th century because we had this cornucopian technological progress and something like Keynesian stuff actually works in that world. It does not work that well in our world and this is why the current crisis has all the elements of going in reverse from the 1930s where there's always a sense there's no good ways to spend the money, it gets misallocated, and that's why I think the arc is this time around towards austerity, towards the idea that money is scarce because what it represents in the economy is also scarce, has become scarce, technological progress and growth.

It's interesting, we had a discussion earlier on the future of jobs, and there was this sense that technology may, you know, you've probably read The Second Machine Age. I mean, the technology is going to wipe out vast numbers of jobs, particularly artificial intelligence, machine learning. I disagree with all of this. You disagree with all of it? All of it, all of it.

Look, people have been making that argument about technology for 200 years. This was the argument the Luddites made in the mid-19th century. It's a problem we would like to have because if so many jobs were wiped out, you'd have such increases in productivity, it would free people up to far more productive things, which was the story of the 19th and early 20th centuries. I think these sorts of arguments gain a lot of traction because technology is a very convenient scapegoat in our society. It functions as a scapegoat. It's what gets blamed for everything that's wrong with our society. Middle-class jobs in the U.S. are not threatened by artificially intelligent robots. That's like science fiction, fantasy, might happen in 100 years, maybe never.

They are threatened by globalization, by people who look very much like us in China and India, who are getting paid much less. I'm not against globalization, but we should be very clear that that's the source of pressure on the middle class. It's not this scapegoat called technology.

And your former colleague, Elon Musk, he recently described artificial intelligence as the biggest potential existential threat to humanity. Do you share that view?

Well, it's very... I think our intuitions about it are very poorly formed. My sense is that it's still much further in the future than people think. It has been over-predicted for many decades now, and I suspect it's still quite far in the future. But it certainly is one where our intuitions are very poorly formed.

Well, our time is almost up. I just wanted to ask, if there's one reform to government in America that you could make that you think would have a big benefit, what would it be? Oh, there's so many different things.

But I think, you know, I would say massive reform of the FDA because I do think the biomedical area is incredibly important. It's one where I think so much more could be done.

And the particular thing would be to what? Allow more experimentation?

More experimentation, much... much less tight controls on drugs, especially in terminal cases and all sorts of contexts where I think we should be regulating drugs much more lightly.

Great.

Well, Peter Thiel, thank you very much for coming and sharing with us today.