“Deep Thoughts”: Tyler Cowen on AI and the Future of Work

It's beautiful. If you had to guess, what's it look like it's doing?

Like it's thinking.

I mean, this could be—

it's not a human mind. I mean, look at this. They're like neurons firing.

Artificial intelligence.

Welcome back to Goodfellows, a Hoover Institution broadcast examining social, economic, political, and geopolitical concerns. I'm Bill Whalen. I'm a Hoover Distinguished Policy Fellow. I'll be your moderator today. Joined by only two of our good fellows today. Lieutenant General H.R. McMaster is not going to be with us, but we are joined as usual by the historian Neil Ferguson, the economist John Cochrane, and sitting in today in H.R.'s place, making this an economist-two, historian-one show, our friend Tyler Cowen, returning good fellows guest. Tyler Cowen is a professor of economics at George Mason University. He's the author or co-author of multiple books. I encourage you to go to that website named after the big river in South America and check for yourself. That's Cowen spelled C-O-W-E-N. You'll find his body of work there. He's been described as a America's hottest economist. I'd like to know what a hot economist is, Tyler. He's also been deemed one of the most influential economists of the past decade. I'd also refer you to the excellent blog that he co-authors called Marginal Revolution. Tyler, welcome back to Goodfellows.

Happy to be here. Thank you. So let's get right into it.

You are here in the Bay Area with us, which means that you're not too far from Mr. Zuckerberg's operation, the building formerly known as Facebook. Today is 53 weeks to the day, Tyler, that Facebook became Meta. I'm very curious as your thoughts as to how that experiment is going and what that says about the future of social media.

I think of formerly Facebook, now Meta, as a classic example of the American corporate research lab, and they're spending like crazy trying to get the metaverse to work. I'm sympathetic with the attempt. If I think about the metaverse as a potential user, I'm skeptical that it's something I will want to do. I worry about how much heat has to be funneled through the headset, whether people feel too nauseated when they're using the headset. What they will use it for at my current margins in life, it's not clear to me. But we've been saying for decades, we want people to take big, bold ideas and do so in a decisive manner. So we ought to applaud and hope he succeeds.

There's also another tradition of amid the American— the rise and decline of the American corporation. Money— corporation makes a lot of money. And then wastes it on grand projects. And there's a possibility that that's the story too. As I look at the product, it seems to mistake the basic attraction of what Facebook had and Twitter addiction had. Nobody says, I think I'll take out my phone and waste an hour on Twitter. No, they do it randomly. You have to take this big thing, put it on and say, I think I wanna waste the afternoon wandering around the metaverse. It kind of looks like the 2003 Second Life. Now, I, like you, every new tech revolution's come along, I've said, I don't need this. I can remember thinking, I don't need my computer to talk to other computers. What the heck's all this about? So I applaud your optimism on this one. I'd also say the story of the great immovable tech monopolies that will be with us forever does seem a little bit, once again, a little bit shaky.

The trade of the year has been short Meta, long Twitter. Year to date, Meta stock is down 73%, Twitter is up 35%. And there's a, there's a story, of course, here of, uh, of two titans. Uh, there's Mark Zuckerberg, who fancied himself Augustus, uh, the master of a vast global empire. Then there's the meme lord, the Chinggis Khan of our time, Elon Musk. Now, it's not clear that Elon has made a good deal by buying Twitter at what was almost certainly an inflated price, but you can see why people are obsessed with Twitter, at least people in media land. And you can see why they're not obsessed with the metaverse. Now, this is not just old guys. I like to test out all technological hypotheses on my children. They are my focus group. And as it happens, my very technically minded 23-year-old Lachlan was with me over the past week, and the new job he has obliged him to get the goggles because the company, which is very cutting edge, wanted to start using the metaverse metaverse for meetings. Uh, the look of nausea and despondency that swept over him when he unpacked, uh, the goggles was to me very telling. Uh, if it's not appealing to him, there is a real problem. The final thing I'll say is TikTok. The big puzzle when you look at both Twitter and Facebook is why did they miss the opportunity that TikTok took to take advantage of people's desire to make short and silly videos and share those? Because TikTok's the internet business that's eating lunches on all sides. And maybe, Tyler, you have thoughts on that.

But if I just— they always do. And the story of Silicon Valley has been one company hits it big with something and then buys up their new technology. They don't ever seem to be able to develop new technology in-house. This is also, of all the list of things the Chinese are trying to do. I noticed they now say, well, we're going to lead in the metaverse. And finally, look at this. What we're doing now would be terrible in the metaverse with avatars because we can use facial expressions. I can go like this and communicate. I can't imagine the avatars doing that. I'm sorry, we keep interrupting. I'm supposed to be asking questions of Tyler.

Tyler Cowen. First, let me say I think hybrid meetings may be the killer app for the metaverse. So we're all on Zoom, that works okay. If we're all together, that works okay. If half of us are together and half of us are on Zoom, that works terribly. So the potential productivity gains from mastering hybrid meetings, I think, are very, very high.

Can you explain this? Because what I'm going to do is I'm going to let my avatar pretend it's paying attention while I check my emails, uh, rather than turning off my camera to do it at a hybrid meeting. How is the metaverse going to help a hybrid meeting?—

it doesn't have to be your avatar, so it could be you, your image in the metaverse, but it would be a kind of equalization between those of us who are in the room and those of us who are online. And those are right now the worst meetings, right? The hybrids. So people are going back to offices, but you still say want to hire people living and working in India. So I think, again, the potential gains there are very high. You don't have to believe the current manifestation of the metaverse is how we'll do it.. But just as the moon program gave rise to our current satellite belt, the chance that spinoffs from what we are calling the metaverse will be important, I think, is reasonably high. So I'm more bullish than you two are, at least from a social point of view. I don't know that it's good for the valuation of Meta. It's not my personal concern. I'm excited. People are trying things. Now, on TikTok, I think the notion that makes it work is every day— it's a bit like YouTube— you have so many swings at the fence. So you have hundreds of millions of videos being created in short periods of time, and some of them will take off. And I think entrepreneurs simply not taking the power law seriously led them to both undervalue YouTube. I remember when then-Google bought it, and the purchase price was considered insanely high. And that it was a copyright nightmare and wouldn't go anywhere. It's now our greatest educational innovation of all time when almost everyone was sour on the original purchase. I actually think YouTube will prove much more important than TikTok. I'm glad people enjoy TikTok. I think in some ways it saves time. People who used to waste 90 minutes watching some garbage on Netflix now spend 4 or 5 minutes on TikTok, and then they get back to their homework. So I don't think it's so terrible. I would also note that—

so I'm reading that Facebook employees don't use Meta, and some of the most successful internal technical developments have— when, like, Amazon develops an internal tool, they developed the cloud as an internal tool, and then they decide to sell it. Now, the fact that Facebook employees don't want to use this thing at all, and it's not an internal tool for, say, hybrid meetings, does seem to be a problem.

Well, it has to be better, right? And it has to be different.

There are all kinds of other intermediate options. I don't know if you know Phil Laban, who has a nice product called mmhmm. Mmhmm is better— mm-hmm, if I remember rightly— is actually a more interesting option than Zoom, and offers lots of of different ways of holding meetings and making the hybrid experience better than the little boxes that have become such a part of our lives. So my guess is that, and it's just a guess, that Meta's version is not gonna work because actually when I look at it, I just see a kind of bad cartoon. It's like suddenly everybody is in one of those Hanna-Barbera cartoons from the 1970s. And I never wanted to see the Hanna-Barbera version of me, but there's probably something that's less fancy but better than Zoom that could get traction. Tyler, I have a big question for you, and I should know this, but I've probably not read the right blog post. Where are you on the working from home versus back to the office, you slaves, debate? My sense is that there's an enormous range of views on this from those who say, 'Everybody needs to get back to the office. We're losing all kinds of productivity and positive externalities from working from home,' to 'Working from home is the future. Commuting and offices are so 20th century.' And I'm not sure where you stand because in a way, we have to solve that question before we can know if we really need something better than Zoom for what we're doing right now.

And this mixes Tyler the economist and Tyler the cultural writer and everything else you do.

So please, And Tyler, the manager, I think both of those views are true. So many workplaces, people have moved away. They're not coming back, no matter what the boss may want, and it's hard to just fire them all, and indeed you don't want to just fire them all. So I think for many people, especially more senior people, work from home is fantastic. I've done work from home for many years. I also do work from hotel room. Maybe more than work from home. And that's in this sense like work from home. But that said, for mentoring young people, spotting their talent, figuring out which intern ought to someday be a senior manager, I think being in the workplace is very important. So I think a major wave of forthcoming innovation will be these hybrid models. And I don't just mean how we do the calls, but how we make hybrid management models work. And the companies that master that, I think, will have a significant advantage. But it's something we've barely tried to make work yet. So there's going to be a lot of low-hanging fruit. And American business is strong at tech, strong at management. So if you're looking for companies to be bullish on, those that are succeeding in that arena, again, I would be very positive.

We're three middle-aged kind of senior people. Ask the assistant professors of the last three years how they've enjoyed the whole work from home, completely empty office period, and I think you get a different answer.

One thing that doesn't seem plausible to me, Tyler, is the idea that you come into the office 2 or 3 days a week but work from home the others, because then you're left with all the overhead of office and commute, but it's getting used less.

That can't be the solution. So let me ask the 3 of you then, what is the right blend for work versus home? I've noted on this call, John's at home, I'm at home, Neil, you're at the office, Tyler, you're in a hotel. So is there a formula here, or does this just vary from individual individual and trade to trade?

Well, our, our colleague Nick Bloom at Stanford is studying this, and actually, uh, despite what Neil's saying, that's where we seem to be heading now, that many companies are 3 days a week. Now, you're exactly right, there's an immense amount of completely wasted real estate if you do 3 days a week, so it's hard to see how that's tenable. Um, but of course, the, the genius of the free market is its enormous variety, so I'll go with it's going to depend very crucially on what kind of business you're in.

Prices are adjusting, and keep in mind all those abandoned big box retailers, we still haven't quite figured out what to do with all that space. So if you just imagine office space being much cheaper, and in the much longer run quantities adjust, I don't see why that should be the problem. Whatever configuration maximizes productivity of output, I think we'll figure out the prices and quantities to support it. Neil?

I go back and forth on this. It's been interesting to watch the Hoover Institution in many ways benefit from going virtual during the pandemic. Why? Because we then were able to get access to people geographically remote from California like Tyler Cowen, who would have in the past have come here only very occasionally. And we were able to get you when you were on the East Coast in your usual location. I think it was in many ways quite good for Hoover to have to experiment. We wouldn't have probably done Goodfellas if it hadn't been for the, the pandemic. Uh, in my non-Hoover life, I've been running a virtual, uh, advisory business for more than 10 years now. We were essentially virtual before the pandemic because we were covering the globe and had people spread all over the world, so we were quite accustomed to it. I come back to Tyler's point though, that you can't build build a team over Zoom. And in particular, you can train the next generation of people and build their familiarity with your culture virtually, remotely. What I don't think makes sense is to say, well, in that case, we'll just kind of part-time work from home and part-time work in the old office. So I'm going to throw out the idea that what you actually do is have a broadly remote structure with people living where they want to live, but you have to have quite regular offsites where the tribe gathers. And the Dunbar number permits all those things that you can't do over Zoom. And I think that's more attractive to me than a kind of part-time commuter lifestyle and a part-time office lifestyle. By the way, that's what we've been doing at Greenmantle, and it's really working quite well. And it means that we spend these quite enjoyable periods of maybe 3 days together at somewhere we choose because it's going to be a pleasant place to engage in that thing that we naked apes really need, which is to hang out together, drinking alcoholic beverages to lower our inhibitions and sharing ideas informally in ways that just won't happen over Zoom.

I want to agree with this. There's a model of teams of teams. The team that really needs personal interaction can live somewhere. But we live in an age of the huge returns to globalization, but the political and immigration restrictions on globalization. This allows you to do things at national and global scale that you couldn't do before, as we have benefited from Hoover becoming a much wider range. I'll just tell a story. One of my children works for a tech company, and they're global. They're a remittance company. They wanted to have— they do everything. Everyone works from home, but they get together occasionally. Most recently, they flew everyone to Amsterdam because for most of the people like the ones working in Africa, they can't get visas to enter the US for a business conference. So we all get together and go to Amsterdam for 4 or 5 days of networking, team building and so forth, and then come home. You can do more global with— you can do global economies of scale this way.

I worry more about the competitive sectors where there's not the profit margin that you can pay for these get-togethers.. Now, a lot of competitive sectors have just stayed face-to-face all along. If you work in McDonald's and do fast food, right, you're not doing that from Zoom. So automation may kick in here, artificial intelligence may kick in here, but we see a non-representative sample of jobs. And if I speak to my sister who works in human resources, where everyone is supposed to be there, her view is, I'm looking to hire people who will show up. And not just take the paycheck and never appear.

We are talking about a narrow slice of office jobs. And in one of these, again, Nick Bloom had a very interesting reunion on this, and lawyers were saying, I have to be in court in order to argue my case. Salespeople say, I gotta be person, press the flesh, shake the hand, have the drink in order to close the sale, get to know people. So yeah, not everybody is professors.

But I go back to science fiction to resolve all such questions. And Neal Stephenson's Snow Crash still seems to me the best visualization of what a world with a metaverse would look like. I think Neal even coined the term metaverse. And in Snow Crash, there's clearly a two-tier society in a California of the future, which has public squalor, but of course, private affluence. And the underclass uses the metaverse for us to escape from their basically ghastly lives. And that seems to me like still a quite plausible vision of where we're heading. Tyler, I don't know if you think that Snow Crash needs to be updated or whether there's a better dystopia that fits the near future for us.

Well, part of my suggested update would be to believe that the lower classes in California often have more fun It's the upper classes who seem obsessed with moving to the metaverse. And just ordinary life, including in the less nice parts of California, it's still very, very good. So people speak of, you know, East LA as a poor part of the city. It's wonderful to be in and go out to eat and walk on the sidewalk. So to me, what's striking is how many unhappy, well-to-do people California seems to have, including the Bay Area.

Well, can I build on that for a second, Tyler? Uh, part of the California existence is a dedication to escapism. You can go out and hike trails and clear your head. You can ski down mountains. Uh, we create wonderful wine if you want to drink away your problems. We've legalized cannabis. Mushroom legalization is probably around the corner. Is escapism the future for California, Tyler, or is escapism just something always been embedded in our history, or is escapism something that's been fueled by the pandemic?

If you think of California as a state with a disproportionately high number of immigrants who in part also make the state work, immigrants are not typically into escapism that much. They may indulge in it selectively, but they want to make their lives work. They want to give their children a better future.

They already escaped, so they don't really need to do escapism. Otherwise— Exactly.

This is their escapism. Your idea of escapism.

A second job. You know, the guys who did my roof recently came in and they did a first job, which they quit at 4:30, and then they came and did us from 4:30 to 7:00 because they were— they had just escaped from Mexico and wanted to get going. That, that's what escapism is for a lot of California, right?

I'm struck by the number of Bay Area and startup people who are maybe 30, doing quite well, been seeing a therapist for 10 years, talking about their latest experiment with psychedelics, which by the way I do think probably alleviate their depression. But to me, that's a sign there's something badly wrong. That there is something basic about life in earlier eras that we seem to be losing our grip on. A lot of America, but then again, probably not our immigrants on the whole.

I'd like to turn to AI just because I have a chance to talk to Tyler. I've been reading about— you're following this closely. We are all old enough to have heard the hype for decades. I was an undergrad at MIT when everyone was— in the 1970s, you know, when do we call the computers he and she, not it? You've been writing about the GPT-3, which seems to me almost miraculous compared to the self-driving cars and everything else that's always just around the corner. How do you think that is? First of all, I still don't see the model. If you just sort of ran follow sentences with what happens on average after the words, I don't see how you get paragraphs of coherent prose and structure.

Why don't we spend a minute explaining what GPT-3 is?

Well, yeah, Tyler Cowen, go on this topic.

First, the hype is usually correct. It's just 40 years too early in many, many areas. So listen to the hype, but don't listen to the talk about timing. GPT-3 is a language learning program that can, in essence, process prompts and give you very varied kinds of output, typically written. There are also image generating programs. Stable Diffusion is the one that is open to the public where you can feed it prompts and it will give you often quite interesting, quite vivid visual images. Both Stable Diffusion and related programs and GPT-3 are quite recent creations and they have become democratized much more quickly than people expected. I'm not entirely impressed by GPT-3, but I have seen later models not yet released that just knocked my socks off, and I'm not entirely at liberty to discuss. But the notion that we will use the NextGPT or whatever we call it to, in essence, process all of our information— think of it as a better Google. Rather than asking it for links, just ask it for the answer, and it will give you some version of what you think is the correct answer. And the John Cochran version of the answer will be based on fiscal theory of the price level, Maybe someone else's answer will be based on the quantity theory of money, and it will know this and do it for you. And within a few years, I think it will entirely reorganize how we use the internet, for whom the internet is good, what an internet marketing strategy looks like. And so much of that will be intermediated through the new AI creations. And people are underestimating this. They're all asleep. I think the time is finally upon us. I have seen the future and it is going to arrive in the next few years.

The problem, Tyler, is that GPT-3 and some of the most recent iterations can essentially write an essay, conceivably a book. But I've seen essays, I've seen debate speeches that certainly pass muster. That is to say, you can't, on reading them, tell that an AI produced them, you could quite easily assume it was an above-average student. The suspicious thing is that the spelling is always really good, uh, and there aren't the typical lapses in punctuation that I associate with the American students. But, uh, this means that, uh, for academia, there's going to be a, a huge problem, uh, because the students will just start using AI to produce, uh, the written work, uh, and pretty soon, uh, economists who are bad writers anyway, with all due respect to you gentlemen, will get the articles written by AI. In fact, I saw on social media just this week an economist boasting about how well this already worked. And this is probably good news for those of us who have to read economics articles. But what does academia look like if the prose is in fact being generated by AI? And supplementary question, if what the AI is basically doing is taking a vast database of past writing by humans and then inferring roughly what it would be like to produce an answer to the prompts, aren't we going to end up in fact in some kind of trap in which original thought becomes very rare because AI is—

That's academia as run by humans today. If anyone should know that, it's you, right? So we're in the trap. We're hoping AI will get us out of it. Good answer. People who do well will be those who figure out how to work with AI to bring something new from their side to the table. But AI doing a lot of the writing and a lot of the work. And Hoover, other institutions, think tanks will face this issue. Someone will submit an output and they'll be honest and they'll say, this is 83% AI. And do you publish it? Do you, do you put their name on it? What kind of footnote is attached? It will change everything. The general issue is just there's now a much better aggregator of information. And when you carefully think through how many institutions we're familiar with engage in aggregating information, that's why I think so much will change. I think we'll adapt to it. I think 15 years from now it will feel pretty fine the way we've all adapted to a world with Google and the internet. But man, it's gonna be a big change. I wanna be ready for it.

Well, everything in tech so far, there's the dystopian nightmare of the tech takes over, and then there's the factual reality of everything since this, the first looms that this is a tool that enhances human abilities, although raising and lowering some people's productivity. So this looks like a tool for writing. As the economist, I saw the same economist who was just entering the Economist, put out a paragraph and said, I'm tired. Let's just send this to GPT-3. GPT-3 filled in 3 following paragraphs, which actually had pretty interesting ideas in them. So this becomes a tool. Now you have to edit which one of those you like, but it becomes a tool. I don't— academic papers submitted for publication, those have to be original. So in some sense, that's not the obvious application for GPT-3, which, as far as I'm understanding how it works, is always going to be reversion to the mean. It's going to give you kind of like the average of what would follow these things. It's better than journalism. Twitter, Substack, those are just going to be 99% written by bots. Suppose your job is to write a Substack. You'll just put in a prompt, let the AI fill it out. It's just going to overwhelm the internet.

You're talking to somebody who has to write about 2,000 to 3,000 words in the next couple of hours, doing what I've done since I was a schoolboy, turning my ideas into prose. I use a keyboard, but in the previous era, I was doing it with a pen. And what we're talking about here is my obsolescence. And I take that very personally because being good at writing is what got me here. And if the reality is that that skill is no longer one that is really valued because a computer can do it better, I got to decide on my future, Tyler. What am I going to do if GPT-3 just made me redundant?

Well, also, who's saying not? Your key is high-quality original writing. You're not gonna be obsolescent. You're producing the Gucci, you're producing the Steinway piano. What's gonna be, what's gonna get obs— I'm advancing the idea, what's gonna get obsolete first is Twitter, Substacks, blogs, media, all the sort of generic writing that people produce, filling out junk. That's actually gonna get a lot better, I think.

Oh God, I'm a lot to read. I'm a luxury brand. This is a terrible nightmarish end to my career. Tyler, hear me out. Well, let's look at biomedicine.

We're already using AI to try out possible combinations of drugs or biomedical elements that hadn't been tried before, and the AI tries them for you and in some way tests how well they will work. There's plenty of biomedical innovation right now. It's hardly copying the drugs of the past. If anything, it's the opposite. You can try out new things better. But the people who are innovators are no longer kind of the simple medical geniuses of the past. They're the ones who understand how computational biology and solving the biomedical problem interact. So the new intellectual entrepreneurs, they will present the correct prompts to the GPT, whatever. It'll be better than GPT-3. And they will say, search this area space. I think the combination of like like fiscal policy, open economy, and depending election hasn't been adequately explored, please give me a rapid printout on which combinations of those ideas have not been covered, and you'll do a bit of the writing and it will do a rest of the writing. It's a very different task from what you're used to, but there's still a role for you. There's a role for your smarts. You'll have to relearn the entire craft, which you may or may not decide you want to do, but it will be much better for new ideas. What it will do is weed out ideas that are not new. And you can imagine readers all having their AI programs that'll just scan all the text out there, like, oh, Ferguson's latest column, how new was that? It'll have a newness score. And if it's really new, it'll have a high newness score. If it's not, it won't. And that's how it's going to be.

You'll become like a Supreme Court justice with a whole bunch of clerks who do a lot of the hard work for you, and you get to put the big picture on it. But the second thing you mentioned seems to me very relevant. It ought to be pretty easy to write a program that detects how much of this was written by the AI and how much of it was original. I like the newness score idea.

I think there'll be some readers like they want the real John Cochran, the real Neil Ferguson, the real Taylor Swift, but it will be a hybrid output.

What about the real Supreme Court? I mean, why don't we just fast forward to a future in which the Supreme Court's judgments are written by AI. You know, I think maybe the luxury brand solution is the one that makes most sense. It struck me more than 10 years ago that the public sphere was kind of dying, was going to be permanently destroyed by the advent of network platforms optimized to sell ads, and that a rational actor would want to withdraw from from the public sphere because it was going to become incredibly toxic because of the way the algorithms work. And so, in fact, I've spent a large part of the last 10 years not working in the public sphere at all because I think intuitively I knew that the only real way forward was to become somebody with a very, very small and private market, but with a luxury brand price tag. So I think what happens in the world you are describing, Tyler, is that the public sphere becomes entirely fake. And people who want to exchange ideas as human beings have to retreat into the private sphere knowing that there's no point participating in a sphere in which algorithms award newness scores to content. I'm just going to vanish altogether and meet a few select people, probably in an Italian town. You'd want to do this in Italy and offer the intellectual equivalent of very, very high-end Italian fashion.

I agree with much of that, but I would challenge the use of the word fake. Would you say to an author, "Oh, your book is fake. You wrote it by, you know, visiting plenty of libraries." So I think our understanding of fake will change and quality will be high. There'll be plenty to read, plenty to listen to, images to look at. I think it'll be a lot of fun.

So— Until the AI-powered war breaks out and brings the entire edifice crashing down. One of the things that struck me when I was reading the Eric Schmidt-Henry Kissinger book on AI was that its most disturbing insight was that if you let AI do war the way it does chess, then war itself would be conducted very differently with the AI general perhaps sacrificing its own quote-unquote pieces in ways that a human general wouldn't. And for me, the most disturbing implication of that book, and I know this is what Kissinger thinks, is what happens when the domain of war becomes, uh, becomes a part of this, this brave new world? Is that something you've thought about? Because I could certainly see a way in which AI does, uh, something cataclysmic if it's sufficiently armed.

This could be the number one issue in the world right now, but I think at least in the medium run, it's a major advantage for the United States. They don't have a number of troops advantage. If you look at the new AI products that we see, they do come from the West and typically from the US, a bit from southern England. China is mysterious, puts a lot of money into it, but they haven't, you know, shown they've come up with much. So I think you can tell your AI not to sacrifice too many pawns, so to speak, but allowing it to say, conduct your drone swarm, for the time being, more power to the West. So I think it's fantastic. Maybe not forever, but that's going to be true of technology in general.

Technology always comes in at the low end and makes it better. Fashion is a good example. In the 19th century, you had to be a seamstress to make a new dress. Now you can be an internet influencer and fashion designer, and all the stuff is made very cheaply. And I think the same will happen in the military. Our military is just amazing to see how Western and our militaries are doing by knowing what they're doing at the squad, platoon, and individual soldier level. And certainly that and the technology and swarms of drones happens first before we AI the generals. And it's not just— it's the question of what objective do you give it? If the AI trains on itself in war games to figure out what to do, if you say the objective is win the war, period, you get one answer. If you say it's win the war and win the subsequent peace and don't cause a revolution in your own home by destroying too many of your own troops, you get another answer. So I'm with Tyler on that. Necessarily terrible, and improving military decision-making— I think our historian would agree with us— is something that, that would certainly be useful.

I wish I could believe that the Department of Defense had a sufficiently nimble procurement process to reap the benefits of American leadership in artificial intelligence. My worry is that while you're quite right that the technological innovation, that the cutting edge is in the Anglosphere, and credit here to DeepMind, an originally British company that is pioneering some of that medical science research you were talking about, Tyler. But I worry that, that when it comes to turning these new technologies into the weapon systems of the future, that's not actually happening. And in fact, there's some concern if you talk to people like our colleague Jackie Schneider here at Hoover, who thinks a lot about these issues, that we're actually failing effectively to adopt these new technologies because the Pentagon still has highly bureaucratic process of procurement. It is not in fact awarding contracts in the way that you would want them to be awarded if your fear was a Chinese drone swarm. We're not in fact building the kind of defenses, and it's not clear that we're able even to supply them to Ukraine right now where we're fighting a proxy war, a war that's going quite badly since the Russians started to target Ukrainian infrastructure using amongst other things, Iranian drones. So I'm not so sure that we'll necessarily turn these great technologies into the war-making leadership of the future.

I think we're bypassing military procurement. It may be necessary to do that. The speed of advance is so great that even a much better system of military procurement could not keep up with advances in AI. At some margin, AI will manage the development of AI, at least in part. And we will in a hurry call upon it and use it in a very rushed manner when we need to. And just if you ask on net, who in military terms is most likely to benefit from high-tech being more important, it's just hard for me not to think it will be the United States.

I'll remember those words as we watch the aircraft carriers sink in the Taiwan Strait, having been overwhelmed by Chinese drone swarms that we underestimated. But let me ask you one more AI question before Bill cuts us off and makes makes us talk about the World Series or something. It's what happens when, I don't know, Greta Thunberg and Al Gore are empowered to ask a powerful general AI, "Please solve climate change," and an AI concludes that the eradication of Homo sapiens would be the best way to achieve that. I know that's a slightly preposterous scenario, But it does seem to me that if you pose certain questions the wrong way, the answer is quite often get rid of humanity.

Well, let me expand on this too, because you asked that question of GPT-3 and it will send you back to you the deluge of complete nonsense that has been written on climate catastrophe for the last 20 years. 'Cause just, you know, climate change, what on average follows climate change? Well, that's what GPT-3 is gonna tell you.

Well, look, I don't think artificial intelligence is going to kill or enslave us all. It doesn't have enough hooks into the outside world. Our world is fairly decentralized, and I don't think it will become sentient in a way that will misalign its goals with our goals. Could there be an evil Hitler who uses AI for evil purposes in the future? Of course, it's true of any weapon system. So you have to compare it to human systems, which are highly imperfect and static. And I agree with your worry about the aircraft carriers in the South China Sea, but precisely the problem there is those are not high-tech enough. They're big things that have a long history dating back to World War II, and we haven't improved them enough. So it's a turf where China, even without the swarms of drones, maybe just using hypersonics, can take out our aircraft carriers. We need something better. I think AI is going to help propel that something better. So—

And to wear my historian's hat for a moment, the problem about being number one, about being the dominant empire or great power, is precisely that you get wedded to the weapon systems of the generation that brought you to that position. And I see plenty of evidence that the US has that pathology. That was very much a part of Britain's problem between the wars. My sense is that we're actually on, on a course to replicate the British predicament circa 1940 because we're pursuing, I think, quite reckless policies in Eastern Europe, in the Middle East, and in the Far East, and not enough attention is paid to that historical danger. Thus far, our conversation has been tremendously forward-looking. We've talked about future scenarios on the basis of assumptions about technological innovation. We haven't really talked much history, uh, and, and history tells us that Britain had plenty of technological options open to it, uh, from the early 1900s onwards that it didn't translate, uh, into military capability. And the result was ultimately near catastrophe in 1940, and by 1945, total exhaustion. Tyler, what, what probability do you attach to the United States repeating some version of British history?

I don't think we're likely to make the British mistake. It seems to me that in Eastern Europe, we may lack the political will to see through a good solution, but our equipment has performed compared to the Russians far, far better than was expected. If I think about China, the major risk would be Taiwan falling, which perhaps we cannot prevent with any technology, and that spreading to a series of other conflicts around the world, —because American credibility is lower. Very different from the problems of the British, but you could imagine if Taiwan falls or is in danger, Israel and Iran start a war, other hotspots break out. I don't know how to fix that problem, but it seems to me the core issue is that Taiwan is very hard to defend given its relative size and where it is. So Germany, technologically speaking, was doing pretty well relative to England. And I don't see that we have a rival like that. We have a set of allies that China can't match. China itself appears to be much more bureaucratic than we had been thinking. Look at their zero COVID policy and its persistence, their economy slowing down, population aging. Uh, use of English is spreading. US soft power, I think, is still pretty incredible, partly through the internet. So I just think it's a very different set of problems. The main one, Taiwan, I don't know how to solve. I think we should be worried. Read, but I don't think it's parallel to the British experience.

Well, also, the thing you're mentioning really isn't technology. It's the ancient historical fact of a peaceful empire descends into hidebound bureaucracy and old interests. And so the guys who liked battleships kept going on battleships, even though it was perfectly obvious aircraft carriers were the next thing, until competition came along and said, mm-mm, those battleships are useless. You need more aircraft carriers. My impression was that the UK had plenty. The Spitfire was a darn good plane. They had technology, they just didn't have, they didn't deploy it in the right places. In the First World War, there was plenty of technology. It was just the generals refused, didn't see how the technology changed the proper tactics. Didn't the, I think the French general, the main French general refused to use the telephone 'cause he said he did not understand the instrument. Well, that, you know, there's a misuse of the new technology. So the problem of hidebound bureaucracies that develop in slowly decaying societies is an old one. But they're almost often— What about the problem of excessive debt, John?

I mean, it strikes me Yes, as I've got two economists captive, I should raise the prospects of the classic problem of any, uh, of any great power, the problem that arises when your debt service costs exceed your defense budget. That's going to happen to the US this decade with a very high probability, and it strikes me that that's one other way in which the British problem repeats itself. The reason that Britain couldn't really have a solution to the Singapore problem, which is the analogy I think with, with Taiwan, was that it couldn't actually afford a defense budget large enough to deter Germany and Japan, not to mention Italy. And I think the US is soon going to be in the position that it can't actually afford, because the defense budget is going to shrink relative to GDP and interest payments are going to consume a larger and larger share of tax revenues. It won't be able to afford credibly to deter China and Russia and Iran. Discuss. But China has a worse debt problem than we do, right?

It's not mainly at the federal level, it's at the local level, it's through special financing vehicles, it's at the real estate level. But our intertemporal trajectory of national wealth, not GDP, but national wealth relative to debt, I would much rather hold the US portfolio than the Chinese portfolio. And their total fertility rates are low and plummeting, getting rid of the one-child policy hasn't helped. Ours is lower than it should be, but it's about what, 1.7? We have a much better history of taking in, absorbing, and attracting excellent immigrants, including yourself. So I've still got to be bullish North America.

Me too, because our economy is hidebound and decaying, and I pull my hair out at all the sand in the gears until I go look at Europe or the UK or anywhere else in the world where we are. Europe stopped growing 10 years ago. It doesn't look like it's going to start again. China, grew like crazy, but we're all kind of seeing the ends of that happening if they're not willing to free up their political system.

So there is some hope there, yes. But you could have said all of those things in the 1920s about Germany and the Soviet Union. You could have said, oh, well, you see, I know we have problems in Britain, old chap, but the Germans have catastrophic hyperinflation and they can barely run a democracy. And as for the Soviet Union, well, madness has, has put the Bolsheviks in power. Uh, fast forward to 1939, and Nazi Germany and the Soviet Union essentially take over Europe, and the only thing that really bails out the UK is that they eventually, uh, fall on one another. So I think one has to be careful about telling oneself just-so stories about how strong the United States is and how weak China is, because partly what happens in the case of a totalitarian regime is that its very weakness, its very lead it to take strategic risk that we in our complacent Western way underestimate. I constantly talk to people who tell me, "Oh, you don't need to worry because they're not going to try and invade Taiwan. They're not ready. It's going to take years before they can risk it." And then we get into the narratives about America's underlying strength. That underlying strength could disappear in an afternoon if the Chinese surprise us by taking Taiwan, and we are then confronted with the challenge of how to take it back. Attack over an enormous distance with the Chinese armed in a far more formidable way than they were in the 1990s. So I just, I mean, I hear you, Tyler, and I've made those arguments myself, but in my role as devil's advocate on Goodfellas, I have to say that you could have said all that stuff about Britain in the 1920s, and people did. Churchill, who was the rare voice warning of disaster, was derided, was accused of being a warmonger and a militarist in just the same way that the more hawkish arguments that, that get made these days in the United States are often dismissed. So I just, I get a little nervous about this because it can all change very suddenly if, if China takes a risk. And I think Xi Jinping is putting China on much more of a war footing than people realize.

I do think China will try to take Taiwan, and in the next 10 years, and it may be disastrous for the world, but I don't see them as an expansionist power in the same way Nazi Germany and the world will somehow have to adapt to the new geopolitical configuration. And China is surrounded by India, Japan, a possibly nuclear Japan, a very hostile Australia, still some American military presence, even if we fail in the task of defending Taiwan. And they're hemmed in and Russia does not love them either, as you well know. So, oh, could they have an even stronger sphere of influence over Cambodia and Laos? Absolutely. Probably already, you know, they can have their way in a lot of local issues, but I don't think it's going to go beyond that.

Well, I read Hal Brands talking about the containment coalition that brings the United States together with Australia and the UK and Japan and India, and when I get to India, I start to ask myself, is this really credible? Because a lot of our strategic thinking has assumed that we can convert India into an ally, but I think India's chronically non-aligned. It was non-aligned in Cold War I, and I think it's non-aligned in Cold War II. So I, you know, I hear what you're saying, but I think you've got to be careful here. China and Russia are not natural allies, but we have helped them get together, and that partnership is certainly an aggressive and expansionist one. That's why about, I don't know, a quarter of Ukraine is still a scene of ongoing warfare, why China is building an enormous military infrastructure infrastructure in the South China Sea. And if China plus Russia, uh, continues to pursue aggressive policies and Iran joins in, we confront a proper axis. When the term Axis of Evil was used back in the time of George W. Bush by David Frum, it was in fact a rhetorical artifice. The, the concerned parties were not in any kind of meaningful relationship with one another. There is now an axis of ill will, let's call it that. And I think it has a significant capacity to expand and destroy. The destruction of Ukraine is the thing that we keep underestimating. Russian economy down 3%, Ukrainian economy down 30%, and probably more than that now that the critical infrastructure is being knocked out. So I don't want us to, to be complacent on Goodfellas about where this could go. We might end up in the end victorious, as indeed the West was in 1945, but it could be a very, very costly victory, as it was in 1945. 5.

Let me mediate a little bit. Unlike 1939, China is not threatening to invade the United States, so it's not quite like the situation of the UK facing Germany. But yes, they probably will try to take Taiwan. Their goal is something like the Soviet Union taking over Eastern Europe after World War II. So an expanded sphere of influence, perhaps an axis of evil. That would be a tremendous setback for the US. And part of the problem is it's so much more costly to fight a war way over there than it is for the— So we could lose that and set ourselves— It would look about it like 1950 looked in Europe to the United States, which would be a tremendous setback for us. And back to Neil's first question, we go into this war with 100% debt-to-GDP ratio. And how do you fight a war? You borrow a ton of money and a good country can successfully borrow 100%, maybe 120% of GDP to fight a war if they start with nothing.. But if you already started 100% of GDP, your fiscal resources to fight a sustained war are in much deeper trouble.

I would agree India will never be our ally, but India is an obstacle. It's growing at over 5%. It's the world's most populous nation. They have nuclear weapons, and they really do not want China to be all too strong. So I think we will tenuously hold on to matters. In an uncomfortable, but probably muddle through kind of way.

And they are one small set of ideas away from a growth miracle. India could have China's growth miracle tomorrow if they would only let themselves do it. And that would be wonderful for the world and geopolitics.

In parentheses, I don't think the next great outbreak of conflict happens in the Far East. I think it happens in the Middle East. Middle East. I think Iran is in such domestic difficulties that it in fact needs a war in order to suppress the internal dissent. And I think pretty soon we're going to be talking about the new war in the Middle East, which will be between Iran and its adversaries. So watch this space. I agree with this.

I strongly agree with it. It's an underrated risk, but in a funny way, it could ease the longer-run problem of having the Axis of Evil all mind, because I don't think Iran will win that war. Iran is clearly beatable by Israel, by the United States, by a coalition.

There's rumors now of a planned attack of Iran on Saudi Arabia, which could happen any minute, and, and it would be interesting to see, uh, weapons training, will to fight, uh, how all of that— but it would be just be a bloody disaster.

And financially and from an inflation point of view, to say the least, consequential.

We have about 10 minutes left in the show. I'd like to go back to AI for a minute, Tyler, and get your thoughts on this. We've talked about technology that allows you to do your thinking for you, to complete your thoughts. Uh, you're in a part of Northern California where you'll see those strange Google vehicles driving around by themselves. Your Tesla will do your driving for you if you would like. Tyler, what does the future hold in the way of critical thinking and self-sufficiency if we're going to let machines machines act and think for us?

I think where machines have been most productive, we have not been very good at predicting. So there was a big hullabaloo about autonomous or self-driving vehicles 10 years ago. They still might be 10 or 15 or even 20 years away. Uh, the thing that's hard to automate, it seems to be, is your gardener. And how much white-collar work can be done by AI, we radically underestimated. So I'm reluctant to offer very specific predictions. I just would say I talked to a lot of people in the AI area who I thought were a bit crazy 3 years ago, and over the last 3 years they have been repeatedly surprised by AI itself, and AI is doing far better than what they had expected. And I think their point of view essentially has been proven to be correct. We're in for a series of revolutions, not starting in the next 6 months, but well within 5 years. Years, and it will have chain reactions that none of us can predict, including for the military. And that would be one of my predictions from this session, like risk of war with Iran soon, deeply underrated. Risk of— positive risk of radical transformation from AI, also deeply underrated.

So what, what do you think, talking about your own life for a moment, AI will transform. You're a tremendously productive individual, a polymath. If we fast forward a few years, what do you think will be different about your own life?

I will be ready, or I will try to be ready. I think public intellectuals will be a bit more of an impresario, juggling a bunch of AIs and also human collaborators. And just as I moved into blogging, moved into YouTube, moved into Twitter, in all cases pretty early, I'm at this moment thinking through what I will do with AI. I don't think I can get very far till it is here and available, but I absolutely do not expect to be behind this curve. I think I'll at least, you know, on average be with the curve, and I'm ready for a lot of change. Absolutely, bring it on. I'm 60 years old. I'm not about to retire. Fire. And I think it will be thrilling and exciting. But man, so many people will be obliterated in their influence.

Which trends of the last, I don't know, 10 or 20 years did you not get on board with? Because part of the challenge here is, is picking losers as well as winners. I did not become a blogger, and I counseled a good friend, Andrew Sullivan, not to become a blogger because I thought it would drive him insane. And I think it very nearly did, to be writing every day. You know, we talked earlier about the original originality variable. The variable I've always wanted to capture is the pages read to pages written, or words read to words written ratio, where blogging encouraged people to write far too much in relation to what they were reading. And I don't regret having passed on blogging. I passed on TikTok. I pass on a lot of things. I'm not at this point ready to go Substacking. So I want to hear about the things you didn't do. You've been very cutting edge as a public intellectual, but you haven't on everything?

Well, the thing I focused on in the last few years is talent identification, and we'll see how much AI disrupts that. But through my program Emergent Ventures, we now have about 350 winners, and we support them financially, and we believe we have good means of identifying young talent and advancing it. And I think that will not be disrupted for some while. So I have a diversified portfolio and I think of my public intellectual outputs as part of the bat signal that draws talent to me or to us, and that's a very nice synergy. But every day I wake up, and early in the morning I'm thinking, what can I do to be ready for the changes? And again, I clearly don't have good predictions, but I'm asking that question every morning, and the other people out there, I don't see that they are. So I'm kind of modestly optimistic about about how my enterprises will go.

Are books just part of the bat signal? They're a very—

They're part of the bat signal. Arguably, I should travel more and read less to learn things because what I learned from reading, probably AI will do that quite well, right? But what I learned from traveling, it seems much harder for AI to learn that.

But I want to echo Neil's frustration because we were all old enough to see many technologies come and ago, and the early adopters spend an enormous investment on something that ends up not really being that useful. The tech— I mean, just look at Zoom, which we're all looking at right now. Uh, you know, Microsoft Teams is still out there. It's awful, but why? It's just a little bit wrong because it's so hard to use. Uh, you can, you know, you, you invest— I for a while invested a lot in C++, and turns out that was a mistake because it was a terrible language and I should have been investing in Python instead. —so making—we're all old enough to have seen investments too early and made in the hundreds of technologies that didn't work. So I wonder, when does the Zoom of GPT-3 come along, the easy-to-use and popular version? And can you tell me when that's happened so I can not waste my time on the 5 versions that won't turn out to be useful?

I would be short Zoom. I think it will be improved upon quite soon. And I actually would be short TikTok in terms of influence. It's striking that on TikTok you can get an enormous number of views, but reputations don't get built in the same way as on YouTube, and people who do well on TikTok, they're often seeking to move to YouTube because that's where they can internalize their reputation and monetize it. So I'm not sure that viewer interest in TikTok will go down, but in terms of trying to get things done on TikTok, I would be very short TikTok. Talk. And that, I mean, those are my bets.

Tyler, we have just a few minutes left. I have two questions for you. First of all, just reading your blog, reading Marginal Revolution, you get into Taylor Swift, you get into Monty Python. I'm curious as to just how you find this information. But secondly, one thing that artificial intelligence can't do, I think, it can't go into a restaurant and review the food.

I think AI restaurant reviews will be very good soon. That's one thing it can do. It just has to read the other reviews and know something about Sichuan cuisine, be able to read the menu. I feel most of the restaurants I review, I could review them without visiting the restaurant. By looking at the menu and reading other reviews, I can be pretty accurate. It's not what I do, but if I wanted to, I could. So again, I think it will surprise us how much it can do.

And then finally, is AI going to be able to bring together Taylor Swift and Monty Python, or is that human curiosity?

I think readers truly desire celebrity. So Taylor Swift is not just big in music, but she has an incredible number of Instagram followers. People want some version of Taylor Swift that to them at least feels authentic. I'm not saying Instagram is truly authentic, but it's authentic enough. So I think I will keep and maintain the kind of luxury brand and give people like the authentic Tyler Cowen experience, and that will do fine even competing against AI, which can read more quickly than I can and do the hyperlexic thing and all the rest.

So I think if one studies celebrity, whatever you do, don't Instagram that hotel bedroom because your relevance as an influencer will collapse. Collapse. Uh, it is, it is not a cool bedroom you got there. This is why I don't do Instagram.

I— it's another area where I think I don't have a future at all. It's very comfortable. I'm happy here, no complaint. You know, the— when the hotel room is a bit less good, it's way more likely to have plugs next to the bed, and that to me is the single most valuable thing about a hotel room.

And in-room coffee.

Uh, that's the kind of thing Angela Merkel would said when she was asked what was it that was really great about Germany, she said that the windows worked very well. High praise indeed.

Well, Tyler, we appreciate your coming on today. Our apologies for the hotel room, but if you come out again, I'll tell you what, we'll put you in Neil's office. We'll put Neil in the hotel room. Neil, I was not going to ask you about the World Series, Neil. That's an HR McMaster question. So another day on that, my friend. That's it for today's Goodfellas. We hope you enjoyed the conversation. On behalf of my colleagues Neil Ferguson, John I'm Michael Cochran, our special guest today, Tyler Cowen. All of us here at the Hoover Institution, we hope you enjoyed the talk. We'll be back soon with an episode. Until then, take care. Thanks for watching.

If you enjoyed this show and are interested in watching more content featuring H.R. McMaster, watch Battlegrounds, also available at hoover.org.