 Welcome to Free Thoughts. I'm Aaron Powell and I'm Paul Matzko, filling in for Trevor Burris. I am host of Libertarianism.org's newest podcast, Building Tomorrow. Joining us today is Matthew Feeney. He is director of the Cato Institute's new Project on Emerging Technology. Welcome back to Free Thoughts, Matthew. Thank you for having me. What is the project on emerging technology? Yeah, the project on emerging technologies is Cato's relatively new endeavor. I'm trying to count now. I think it began a couple of months ago, June or July. I should probably know that, but it's relatively new. I'm running it. It's a project of one at the moment. But the goal of the project is to highlight the difficult policy areas that are raised by what we're calling emerging technologies. Now, this is always a difficult thing to define, right? And of course, emerging tech is not just changing technologies, but new things arriving on the scene. And what I've done is to try and highlight a couple of issues where I think Cato has a unique capability to highlight interesting Libertarian policies associated with new tech. So some of the policy areas that we're focusing on include things like artificial intelligence, driverless cars, drones, data and privacy issues, and others. There are a lot of tech issues that have been around for a while, so I don't think net neutrality is going anywhere anytime soon, nor are the numerous antitrust issues associated with big tech companies. And we've certainly at Cato had people write about those issues before, but this new project is confining itself to five specific areas, but I'm sure that as the project grows and develops, the list of issues we'll be tackling will grow. How did you choose those five in particular? Yeah, so the five were areas where I thought Cato didn't have enough people writing about, and also areas where I think Libertarians have something new and interesting to contribute. So for example, I've first couple of years at Cato, I did write about the sharing economy. I also wrote a little bit about drones, body cameras, new tech issues, but my work with drones, for example, was just on law enforcement use of drones, specifically the concerns associated with drone surveillance. But I wasn't writing at all, really, on the commercial use of drones. So the exciting world of taco delivery drones and building inspection drones, and that's a whole different policy area, really, compared to drone surveillance. So that was an area where I thought we should really have someone who can direct a project that will commission work on those kind of issues. So artificial intelligence, right, is something that I think is very exciting, but poses difficult questions to Libertarians, and Libertarian commentary on that space has been not nearly, I think, as robust and as loud as it could be. So that's another reason why I picked that. But yeah, basically the five, I think, fulfill a criteria of being focused on new and emerging tech that Libertarians have something interesting to say about, and that Cato's in a good position to tackle. Can you give us an example of what you mean by Libertarians have something new and interesting to talk about? Because a lot of tech policy in the past has taken the form of its regulatory policy, and it's should this thing be regulated or not, typically. And then what form should it be regulated in? And that tends to break down along the standard lines. You have the people who are opposed to regulation. You have people who are generally pro-regulation. But what's uniquely, I guess, Libertarian in the way that you're approaching technology issues? Yeah, so I don't think that the way that we're approaching the project is much different to how a lot of us here in the building approach our other policy areas. And for me, it's to tackle the issues raised by this tech by embracing a presumption of freedom and trying to minimize coercion, right? So number one on the presumption of freedom that we should act in a way that allows for innovation and entrepreneurship and make sure that people working in this space are in a position where they're asking for forgiveness more often than they're asking for permission. And as far as minimizing coercion, this goes back to some of the work I discussed earlier when we're talking about data privacy and drones. We should be wary of some of the government use of the technology and making sure that exciting new technologies like drones can be used for really cool stuff like deliveries and other private applications while also trying to make sure that the scary aspects of it like surveillance are being put under lock and key as much as possible. Something like artificial intelligence might be another good example that we want to make sure that people working in this space are free to innovate and to explore new ideas, but we want to make sure that government use of it, especially when it comes to autonomous weapons and automated surveillance that we ensure that there are So emerging tech by its nature is still, you know, yet to come. It's already not yet. It's kind of here, but it's still in prototype or developmental form. So a lot of the potential benefits as well as potential risks are still in the future. So like as you're trying to decide what should be regulated and what shouldn't be regulated or in what ways it should or should not be regulated. Like what's your rule of thumb for trying to rule on decide on something that hasn't actually happened yet? Yeah, I suppose the libertarian response to this is comparatively straightforward, right? That we should proceed with caution when dealing with imaginary threats. So let's think of a good example, right? Maybe only because I work on it in my own research, right? But I think it's fair to say that in the coming decades that we will see more and more government use of unmanned aerial surveillance tools. I think that's a fair assumption. I also think it's fair to say that that technology will improve as much as it proliferates. And as I did, right, I wrote a paper saying, look, we should, in preparation for this world, we should have the following policies in place. What I'm very hesitant to do and not that it should never be done, right? But we should be hesitant, I think, to develop new rules because of a new thing coming onto the block. Drones, for example, raise interesting privacy concerns, but it's not clear that they're necessarily unique in the way that a lot of people think they are. So we don't like the fact that drones could be used by people to snoop on us in our bedrooms or to fly over our barbecues. And we don't like that police could use them to do surveillance. But we already have laws, peeping Tom laws. We have a tort system that can do with a lot of these complaints. And while the Supreme Court precedent on things like drone surveillance is not particularly not very satisfying, it is the case that states can and have gone above and beyond what the Supreme Court requires. So going forward, I think we should be hesitant to think of, well, we need a driverless car policy. We're going to write down, we need a drone policy. We should think about the kind of threats that come from these fields, but resist the temptation to write a lot of regulation and anticipation for the proliferation of the technology. But isn't that the problem that because these are emerging technologies, they're not technologies that we either as citizens or just ordinary people in our lives or as lawmakers or legislators or regulators, we don't have any experience with them. We haven't used them. We haven't seen like how they shake out. And so that notion of saying, well, we shouldn't just imagine threats. Isn't that what we're kind of forced to do? One of the things that distinguishes emerging technologies now from emerging technologies in the past is the pace at which they can become all pervasive. The pace at which they can spread. So either they're network technologies that just, you know, in a matter of years, suddenly everyone is on Facebook. Whereas, you know, the printing press took a lot longer to get books into everyone's hands. Don't we have to be anticipating threats because with a lot of this stuff, if we don't and we don't protect ourselves now, it might be too late? Well, too late for what, right? This is the question. I think history has enough examples of people exaggerating threats that we can learn from. So one of my favorite examples of this, right, is the British 1865 locomotive act, which required a vehicle that not pulled by an animal. So a steam powered locomotive. If it was on a road and towing something, it was legally required that you would have a man 60 yards ahead of it with a red flag, right? Because people were anticipating certain threats, right, that these new technologies are going to cause accidents. And so what we need is, it's obvious, right? We need a man running ahead of these things with a red flag to alert people that there's very dangerous things coming across. I don't know if that's the right kind of approach to dealing with emerging technology issues, right? We can anticipate that with the emergence of the locomotive that there will be occasional accidents and some people will get hurt. The early years of flight, for example, are just full of people killing themselves in these new flying machines. And you might, it sounds a little cold-hearted to say, but the price of innovation for something like that is that mistakes get made and people might get hurt. And that's difficult, especially in today's world where news travels so quickly that the moment that someone gets hit by a driverless car or a drone lands on someone's head, everyone's going to hear about it. And I think people are thirsty for news, for bad news, unfortunately. And that's something we're always going to be fighting against. So I actually would go on record right now saying I'm in favor of a law requiring that Elon Musk wave a red flag 60 feet in front of every driverless vehicle. Because he has more time on his hands in this case. So I hear you talking about essentially assumption of risk, that when it comes to tech, we have a long history of people overrating or exaggerating fears of the downsides of a tech and having a harder time imagining the beneficial applications. And so a light touch regulatory policy wedded with like a general cultural sense of, hey, if you want to experiment with this, as long as you limit the externalities that damage other people, go for it. I mean, is that kind of the attitude you bring to stuff like unmanned vehicles and the like? Yeah, I think that the barrier for government intervention in this space should be difficult to overcome, right? So a very high risk of death or serious injury is basically where I would say you can maybe argue for some kind of regulation. Again, we're sitting in the Cato Institute, right? I mean, our approach to regulation, this is a unique approach to emerging technology. I think libertarians across the board have light touch approach and I feel like you can have that approach while accepting that there are risks, right? And the problem, of course, though, is that with a lot of this stuff, an argument can be made that innovators and entrepreneurs might be hesitant to start doing a lot of this work and feel like they might get in trouble or they want to wait until there is a safe regulatory space. So Amazon, right, decided to test its delivery drones in England because they knew that the FAA had not cleared the delivery drone testing here. So I can understand why Amazon didn't say, yeah, we'll screw it, we'll do it anyway. People want to be, I think, if you want to be a respected private business, you don't want to get in trouble with the feds. I get that, but I think that's an unfortunate feature of FAA regulation, that the FAA should have an approach of, you know, you will better be careful because you will be in a position to ask forgiveness. But I still think that's a better position than people in the drone space asking for permission. But I mean, and going back to the question I asked before, with emerging technology and with the, to quote Donald Rumsfeld, the unknown unknowns in, you know, at play here, do we want people to be extra special careful in a lot of these areas? Because you even have situations where, so the story often is told. A lot of people, you know, this is the narrative is that all of a sudden a handful of people in Palo Alto, well, no one was watching, broke American democracy with social media, right? Or a situation where, you know, everyone's kind of out there innovating and then suddenly we have a rogue AI and we can't do much about it. Or, you know, like gene splicing, CRISPR, people making stuff in their garages and then we have a pandemic. Like that kind of threat of regulation or that asking for permission, does that help at least to mitigate against those kind of sudden catastrophes? Well, I think you're highlighting something interesting, namely that, well, first I'll say hindsight is always 2020, right? That it's easy to look back and be like, wow, if we had X regulation, why would never have happened? But it's easy for people to come up with scenarios. The difficult job is thinking of regulation that would hamper that scenario if I'm ever taking place while also not hurting innovation. So rampant AI. OK, so this is something anyone who's watched a science fiction film worries about. But what's the fix to that? We write a law saying no one shall build AI that will run amok on servers and take over. I mean, isolating a threat is not the same thing as coming up with a good regulation for that threat. And so social media companies ruined American democracy. So this is sometimes said by people. But what's the regulatory fix that would have stopped a lot of the bots and the trolls that got everyone concerned in the wake of the election? That's a much harder question, it seems to me. It's easy to get outraged and to get worried about possible threats, but coming up with solutions is much, much harder. And I think we should also keep in mind how likely the threat is. It would be a shame if developments in AI were seriously hampered because a couple of lawmakers watched too many science fiction films and got really, really worried about the terminator scenario. Well, how big of a problem is that? Specifically that this is an area where lawmakers, I mean, we put the Kato Institute, we often lament how little lawmakers seem to know about the subjects they plan to regulate. And in fact, we have named our auditorium the FA Hayek Auditorium who, you know, Hayek famously offered a theory for why it was lawmakers could never know enough about the stuff they wanted to regulate to regulate it well. But this seems to be an area where lawmakers are particularly ignorant. I mean, it's often cringe inducing to watch congressional testimony because these lawmakers have levels of understanding of the internet of networks of technology that is substantially worse than, you know, the typical middle schoolers. So is that, how do we deal with that kind of problem? That we've got a situation where lawmakers, there's this tech, you know, the urge is always to pass a law. Whenever there's a threat or potential threat, it's pass a law and they're doing that because they want to do it, they're also doing it because constituents, you know, demand pass a law. But that this is an area where almost like by definition, you can't know much about it. Yes, I defy anyone under the age of 30 to watch anything like Zuckerberg's testimony on the Hill and not have their head in their palms by the end of it. It is very worrying that many of the lawmakers on the Hill don't seem to know much about this. And that makes sense because a lot of the people who'd be qualified to be on staff in these offices to actually give advice and to explain to members of Congress how this stuff works could be paid much, much, much better almost doing anything else actually in the tech industry. And that's a serious worry. And there's also this worrying inclination among some lawmakers to urge technology companies to, and I quote, this isn't a phrase original to me, but to nerd harder, right? That whenever there's a problem like end to end encryption, people think, well, we don't like the fact that some terrorists can communicate using WhatsApp or Signal. But there must be a fix, you must, you know, how can you not fix this? And there's a frustration there. Where we're sitting, I think that we should maybe spend more time focusing on the benefits of this technology, not focusing on potential costs. So driverless cars will kill some people. They just will. And that's of course regrettable, but we should think about the lives that they could save. The vast majority of auto fatalities in the United States are directly attributable to human error. So from that perspective, driverless cars that are better than human drivers, but not perfect will save thousands and thousands of lives a year. And once Congress eventually gets happy with the proliferation of driverless cars, we should expect that for the next couple of years there will be headlines of driverless cars killing people. And that's to be expected and it will be a big cultural shift. So emphasizing the benefits rather than the costs, I think, is worthwhile. But of course that's easy for me to say because I won't be the one sponsoring the bill that allows these things to run rampant and then who are they going to wag the finger at when the bad things do happen. But like I alluded to earlier, good news rarely makes headlines and it's also slow moving. It will take a long time for the benefits of driverless cars to be realized in the data, but the accidents and the deaths will be reported instantly. So what I hear from you, Matthew, is a sense that our cost accounting, our cost benefit accounting analysis is flawed. It's easy for us. It's kind of a seen versus the unseen situation. It's easier for us to imagine apocalyptic worst case scenarios and then to discount the possible benefits. So whether it's pharmaceutical regulation, something like the FDA has a notoriously stringent safety requirement that doesn't really account for the fact that not approving a life-saving drug costs thousands, even millions of lives. And that doesn't play a role. They just are asking whether or not the drug itself will harm lives. So in that sense we have a, you know, the ledger, the kind of accounting ledger is flawed when it comes to emerging technology. But I'm also interested in hearing you talk about ways in which regulators themselves by regulating too quickly can actually fulfill kind of self-fulfilling prophecy when it comes to kind of the downsides of that technology. So a good example of that would be what? I just want to make sure I understand the question. So I suppose you can imagine a situation where the FAA says, well, we haven't had as many drone accidents as other countries because we haven't let drones fly, which probably an accurate statement. But we need to keep in mind that while that's true and the FAA is tasked with safety, they need to make sure things are safe. We need to also keep into account what we're losing. I think when you ground drones, you incur a cost. Namely, you are not having as innovative and as exciting an economy as you could have. So yes, a federal safety agency can stand up and say bad things aren't happening because we're just not letting people experiment. But it's not a particularly useful thing to say, it seems to me. And it's also not helpful because no one who's rational is denying that emerging technologies will come at a price. We're just saying that in the long run, the benefits are way the cost. Given that, and given that bad regulation or overburden some regulation can not just slow down the pace of progress but can cost lives, can certainly reduce wealth, economic growth, when is it appropriate, and we've seen this happen a fair amount in the emerging tech space, when is it appropriate or is it ever appropriate to intentionally circumvent regulations? So we're at the part where Aaron asks me, when's it okay to break the law? I would like to point out that I think there are a lot of people who do this by accident. I don't know the number, but I imagine there are many people who got drones for Christmas or birthdays and flew them. Without adhering 100% to FAA regulation. I can say that with almost a certainty. The response from the FAA, I think, should not be to bring the hammer down. Now, when is it acceptable? The classic example being Uber, which Uber has arguably changed the world and frequently in a positive way. They've granted they have their problems as a company, but a lot of that came with them basically ignoring local regulations. Okay. In that case, I would argue that at least in some of the jurisdictions, Uber could have made the argument that well, we looked at the taxi regulations and we decided that we didn't fit the definition of taxi. So off we went. That's a much easier argument, it seems to me, than a drone operator saying that they're not an aircraft under FAA definitions. Uber, I think, was doing something very interesting, which was providing an obvious competition to an incumbent industry without being actually a very different thing. To customers behind the scenes, I think a lot of people found Uber and taxis to be very similar, but actually they're very different kind of businesses and it's a very different kind of technology. I take your point and of course, Uber's opponents would oftentimes portray Uber as a lawless invader. I think at least in some jurisdictions, Uber could make the argument that actually no, we just feel like we didn't fit into that regulatory definition and Uber does fit into this very, or at least when it began, fit into a very awkward regulatory gray area. So in a situation where you've taken a look at existing regulations and you think that you don't actually run afoul of any of them, I don't see why people shouldn't feel free to get into an area and innovate. Airbnb might be another example where you, okay, well, I took a look at local laws and I figured that I was in a hotel. It seems to be a reasonable thing for people to assume, but I wouldn't say this is without risk. I wouldn't advise anyone in a private company to deliberately break the law and to hope that you have good lawyers on hand. I don't know if that's the best approach because local law makers don't like that kind of confrontation for sure. I mean, I suppose there's a, some of that question comes down to one's own ethic, right? I mean, most people would imagine an ethical obligation to break the law when there is some kind of clear cost to life that comes from following the law. I mean, so, you know, civil disobedience writ large and no one, no one, well, some people did hold him responsible, but when Martin Luther King Jr. or another civil rights activist blocks the highway for a march on Selma or Birmingham or whatnot, right? Like, the idea is that laws are, it's okay to circumvent them when there's a clear ethical obligation to do so, that the law is less important than like ethical systems, but that gets complicated really quickly depending on. I will mention here though that Charles Murray, I think it's, I haven't read the book, but I think that in one of his most recent books, Charles Murray advocated for a law firm that specializes in protecting entrepreneurs like this to basically encourage people to go out into the wilderness. Adam Theer from Makedas who wrote an excellent book called Permissionless Innovation, he categorizes technologies as born free and born captive, that some are born captive into regulatory regimes and others are born free. They're truly new and innovative and regulators haven't caught up yet, but if you're born free as Adam might call them, I think you better be ready for certain fights and the Charles Murray's recommendation was, yeah, we should just basically have a law firm that specializes in helping entrepreneurs with these kind of fights. From the regulators point of view, I think they should perhaps just choose their fights more carefully and not scare away people, but that's not going to happen anytime soon. The costs that we've been talking about like deaths and injuries are, I think, easier to discuss, but the problem with a lot of technology or emerging technology discussions are you have these more difficult to pin down complaints about the impact on society and what's it doing to our children and isn't this making us more isolated, think about the citizenry, all that sort of stuff is... Thank you, Tipper Gore. Well, right, and it's interesting because this isn't a new kind of complaint, right, but nonetheless remains sticky. I wanted to briefly read out a quote I found from 1992. There was a Neil Postman, sorry, wrote a book called Technopoly, The Surrender of Culture to Technology, and he was on C-SPAN in 1992, and he'd previously complained about television, right? And he was on and he said, when I started to think about that issue, television, I realized that you don't get an accurate handle on what we Americans were all about by focusing on one medium that you had to see television as part of a kind of a system of techniques and technologies that are giving the shape to our culture. For instance, if one wants to think about what has happened to public life in America, one has to think, of course, first about television, but also about CDs and also about faxes and telephones and all the machinery that takes people out of public arenas and puts them fixed in their homes so that we have a kind of privatization of American life. This is a really interesting kind of complaint, but he goes on to describe a future that we're kind of in now where he says, when his people say with some considerable enthusiasm that in the future putting television, computers, and the telephone together, people will be able to shop at home, vote at home, express political preferences in many ways at home so that they never have to go out in the street at all and never have to meet their fellow citizens in any context because we've had this ensemble of technologies that keep us private away from citizens. And I hear complaints like this quite regularly. I mean, that's from 1992, but there is still this very persistent worry that emerging tech will make us bad citizens, make us isolated. AEI is exciting, but will our children say please and thank you to the robots? Will the robots become our friends or our sex partners? Isn't all this stuff making us kind of isolated? This isn't a new concern, frustrating and it's not going away. So we have been talking largely about policymaking, policymakers, regulators, people who are in the policy world. But how much of that is really just downstream of culture? Such that when we're dealing with these issues of emerging technology that where the real action is happening is in the culture, is in the cultural acceptance of it. And so to some extent focusing on strictly the policy is kind of missing where much of the influence is or will be. I certainly do think that it's important to communicate to the public about this because like you mentioned, some of these policy concerns are downstream from the public. And in preparation for the podcast, I was finding articles from 1859 editorials in the New York Times complaining about the telegraph and a 1913 New York Times article complaining about the telephone and how it's in bad manners. All this stuff isn't new, but I think when we're sitting in a think tank, we should be ready to communicate with the public in addition to regulators and lawmakers. If we have a optimistic forward thinking public, then you hope that that will translate somehow to lawmakers. But lawmakers are made up of human beings and the public are human beings and they have a pessimism bias. And I think though when you focus again on benefits that maybe more parents would be happy if driverless cars could take their kids to baseball practice. And it would be better for people if their elderly parents have appliances in homes that can monitor if they've fallen down or if they have had a medical emergency. It would be good if we were able to travel more safely to have our homes know more about us. It would be nice to come home and to have the home set at the right temperature and playing the right kind of music. Making sure that people realize the benefits of a lot of this stuff is I certainly think part of the mission. My only audience is not lawmakers, that's for sure. All of that, the home that knows a lot about you, all these things that can predict stuff about you, keep track of things about you. There's a lot of data there. There's a lot of data gathering. It depends on devices that can surveil us in one way or another. And we as libertarians, we as Cato Institute scholars, we spend a lot of time talking about the problems of government having access to data and government surveillance programs. But are we concerned? Should we be concerned about the level of pervasive private surveillance that that rosy future you just sketched out demands? I think we should be worried. You can listen and read a lot of Cato material on the concerns that we have about government access to data. And I certainly don't want to sound blasé about that. So my primary worry is the government, mostly because as creepy as a lot of this might be when it comes to Amazon and Google, Amazon and Google can't arrest me or put me in a cage. I think that is a big difference. People might be a little creeped out by the shopping algorithms. They might be a little freaked out by the fact that these companies do know a lot about us. But I want the heavy lifting there to be on government access to that data. You buy a lot of these appliances. There's a certain degree of you assume that there will be collecting information about you. But I'm not as worried about Amazon as I am the government for the reasons I just outlined. And I don't think Amazon has an interest in creeping out as customers too much. Should we be worried though about companies like Amazon gathering all this data, centralizing all this data and then that data suddenly becoming either through the passage of legislation or through subpoenas or warrants or through government hacking accessible to the government. Yeah, there's a degree of trust you have in these big companies. They need to do a good job at being custodians of data. I don't want to speak to the... I don't know a lot about Amazon's actual security just using them as an example. They have a very strong profit seeking incentive to make sure that their customers privacy is not violated. There's not much though that they can do when the government comes to them with a valid court order. They are put in a tough spot there. And again, that's why I think that's where we should have the focus. But we shouldn't be in any doubt that a lot of these companies have a huge amount of information on us. And I think it was my colleague Julian who once said that if Google was a state, it would be a pretty powerful police state given the amount of information it has. My apologies to Julian if I'm butchering your quote. But the point being that they do gather a huge amount of information on us and people, even like me, do incur a cost when you use proton mail instead of Gmail or you use DuckDuckGo instead of Google for web searches. And that cost is that Google now knows a little less about you and can't provide you with the degree of service that most people have. But that's fine by me. There's still choice. Google's not a monopoly when it comes to this sort of stuff. And people value their privacy subjectively. And maybe I value it as slightly higher than the average person. But I have no problem with people using Google products to make their lives better. I do worry about government access to that data to conduct investigation. It feels like forever ago now, but it's only a few years ago folks were there was buzz about Mark Zuckerberg running for president. It's that blend of a major tech company with the power of the state while it's unlikely now. It's not outside the realm of possibility, even if it's not as literal as the head of one being the head of the other. To go back to something you mentioned before, Matthew, you teased a bit about how in Great Britain, I think it was regulatory policy towards unmanned aerial vehicles was more favorable. So it pushed Amazon to conduct tests overseas. So to broaden that out, how would you say on the net international regulatory landscape, how it compares the United States? Where is the U.S. rank when it comes to relative freedom and regulation of emerging technology? I think it's difficult to say for the following reason. But saying technology policy is a bit like saying economic policy. It's a huge range of things. So let's think of the plus side first. So the United States is still a global leader when it comes to tech innovation. This country is home to some of the best known largest and most interesting tech companies. Global data recently produced a list of the 25 most valuable tech companies in the world. 15 are in North America, 7 in the Asia Pacific, only 3 are in Europe. And that, I think, is not an accident. Europe is, as you alluded, is slightly, I would say, ahead of the United States when it comes to drone policy. But they slapped Google with a huge, I think it was $5 billion fine on antitrust. So it depends on the technology you're talking about. They're certainly ahead when it comes to, I would say, drone policy. But when you're leveling billion fines worth billions of dollars on Google, right? It's not a great look. And so examine the policy, the technology of Pacific policy. I wouldn't want to go to a big generalization. I would say, though, that there's probably a reason that the United States is still today a massive hub and funder innovator when it comes to technology. Does competition work in that area? So do you see, is there evidence that countries look over at other countries that have better tech policy and so are getting better, bigger country companies, more innovative products and say, well, it's probably good for me to loosen things up a bit, too? I don't know. I'd have to look at data. I think the problem is for a lot of these countries is that a lot of the Silicon Valley is still a massive talent suck for a lot of these countries. That's a gut assumption. I'd have to look at data on that. Competition, of course, is an interesting point when you're talking about big companies like Google, Apple, Amazon and Facebook because a lot of those companies are big enough that they can buy interesting smaller companies. So what would be a good example? YouTube, Instagram, WhatsApp, these are all companies that were bought by much bigger companies. And that's not necessarily a bad thing. And it's not necessarily something that we should complain about. But for the foreseeable future, I imagine that Amazon, Google, Facebook and Apple are going to be on the lookout for interesting new companies to buy. One, because they view them as competition down the road, but two, they also feel that they can do interesting things with those companies. And that's not a bad thing necessarily. If you are building something that competes with Amazon and you're presented with a life-changing amount of money, there will be some people who say, no thanks. I'll keep plugging away at what I'm doing. I believe it's the case. I'm not a historian when it comes to Facebook, but I believe Facebook faced a buyout option at a certain point. Didn't someone want to buy Facebook? I could be making that up. But my point is that there are very large successful companies today that said no to buyout. Netflix is a famous example. Blockbuster had the offer on the table for some miniscule fraction of what Netflix's value that. And keep in mind that this competition question is something we're going to hear more of as long as Trump is the president, because there's a perceived anti-conservative bias in Silicon Valley that people think is actually affecting the product. So I think it's fair to perceive that most people who work in these big tech companies are probably to the left of the average American. I think that's fair to say that I'm not convinced that that personal bias among employees has had a direct impact on the product. And you're in this weird situation where self-professed conservatives are now saying, well, they're too big and we should talk about antitrust. When we're thinking about the big four, Google, Amazon, Facebook and Apple, I'm not convinced that these companies are monopolies in the true sense. And I think it would be a mistake to bring antitrust action against them. So the example that comes to my mind of international kind of regulatory competition from TechCrunch Disrupt out in San Francisco, a number of panels hit on the idea that when full self-driving cars, level five, no steering wheel, when that gets rolled out, it'll be rolled out in China before it gets rolled out in the rest of the world. And that will be because according to a number of speakers, the central government in China has just established by fiat we are going to be open to autonomous vehicle technology. And actually by like the dollar value of investment in China just over the past year has matched, in AV technology has matched the rest of the world combined. So you're seeing kind of that they're shifting to a place because in China, the central party can cut through local and state level competition. What that brings to mind for me though is a question for you, Matthew, about how emerging tech should be regulated by local and state authorities versus federal authorities. Like the question of federalism and emerging tech policy, how do you approach that as someone analyzing emerging tech? I'm very interested in a lot of the local regulations that handle industries like ride sharing and other things you see in the sharing economy. But when it comes to a lot of the technologies we've discussed that are very powerful federal regulators, the FAA, the FCC with bioengineering and all that, the FDA. So I am in a position where I am mostly focused on federal regulations, but I'm certainly keeping an eye on what's happening at the local level. And as we discussed earlier, state and local governments can take it upon themselves to address some of the concerns we've discussed, especially when it comes to drone surveillance was an example I used. And there are state and local governments that have been comparatively welcoming to the sharing economy that they have decided no, we're going to be a home of innovation and entrepreneurship and that's what we want. But I think it's fair to say that for some of the big issues we've been discussing today, driverless cars and drones and things like that, ultimately it's probably going to have to take some federal leadership to get the kind of regulatory playing field we want implemented. Thanks for listening. Free Thoughts is produced by Tess Terrible. If you enjoyed today's show, please rate and review us on iTunes. And if you'd like to learn more about libertarianism, find us on the web at www.libertarianism.org.