Timestamps
(0:00:00) Intro
(0:00:28) Identifying talent on Tyler’s podcast
(0:03:48) Why are follow up questions overrated?
(0:04:21) Tyler’s preferred guest career stage
(0:05:16) Optimal frequency for recording podcasts
(0:05:50) Tyler’s podcast prep
(0:07:39) Importance of in-person episodes
(0:09:56) What would Tyler ask Paul McCartney
(0:10:16) When to read literature in translation
(0:13:31) Tyler on Dostoyevsky and Nietzsche
(0:15:39) Elena Ferrante and the importance of pseudonyms
(0:17:17) Misreading literature
(0:17:56) Will literature get better in the next 10 years?
(0:19:21) Watching complex film
(0:22:14) Enjoying art
(0:23:08) Jonathan Swift and Peter Thiel
(0:24:45) Crude comedy
(0:25:28) Can we trust elites in the internet age?
(0:26:44) Generational theories of politics
(0:27:57) Religious thinkers and Tyler’s implicit theology
(0:29:33) What did Tyler get from Plato at a young age?
(0:30:51) Why are Shakespeare, Proust, and Melville in a league of their own?
(0:31:32) Fernando Pessoa
(0:31:59) Implications of demand sloping down
(0:33:22) Innovation in governance structures
(0:36:14) Should business leaders study the greats?
(0:38:36) GDP growth from today’s AI models
(0:40:35) LessWrong and worries about AI
(0:41:46) Feminized societies in times of chaos
(0:43:17) Fernand Braudel
(0:43:52) Best argument against NGDP targeting
(0:47:06) AI business models
(0:48:19) What is a bubble?
(0:49:33) Where are there too few Emergent Ventures applications?
(0:50:09) Unsolved problems in economics
(0:51:27) Immigration policy
(0:52:56) If Tyler wasn’t an economist
(0:53:50) Would Tyler go to Burning Man?
(0:54:33) The MR Universe
(0:55:32) Tyler’s equanimity
(0:57:15) Tyrone
(0:59:27) Joyce Carol Oates and Susan Sontag
(1:04:45) Wasting time
(1:04:54) Bach’s productivity
(1:05:29) Competitors to MR
(1:07:51) New Jersey and The Sopranos
(1:11:56) Rapid fire - what Tyler learned from different people
Links
- New Yorker on Joyce Carol Oates
- Subscribe to Undertone on YouTube, Apple, Spotify, or Substack
- Share anonymous feedback on the podcast: https://forms.gle/w7LXMCXhdJ5Q2eB9A
Transcript
[00:00:12] Dan: Okay, my guest today is Tyler Cowen. He's a professor of economics at George Mason University, blogger at marginalrevolution.com, host of the podcast Conversations with Tyler, responder to email, and one of my intellectual heroes. Tyler, welcome.
[00:00:26] Tyler Cowen: Dan, thank you very much.
[00:00:28] Dan: Okay, so let's talk a little bit about your podcast. You've said before that the Straussian reading of Conversations with Tyler is that the entire conversation is really just a talent evaluation. How confidently do you feel you can make a talent assessment after you have an episode of with just one guest?
[00:00:42] Tyler: Well, virtually everyone who's on the podcast is already highly successful. In that sense, it's easy. You've decided just to pick the winners. Now, there's another question at stake, which is, how good will they be in a podcast? Someone can be mega successful, but completely boring on a podcast, that's a different kind of talent assessment. I would just ask the listeners to make their own judgment. I don't have really other reasons for having people on episodes, other than that I think they will be good. It's not like, "Oh, I had to have my best friend on or my cousin on." If there's a bad episode, I screwed up evaluating talent is the simple way to put it.
[00:01:24] Dan: I guess another way to phrase this is when you are evaluating talent, just for a hiring decision more explicitly, how different do those feel from your conversations on the podcast?
[00:01:33] Tyler: Oh, they're totally different. It's hard to have a very high hit rate on evaluating talent. These are typically markets where people are doing startups or they want to be important public intellectuals. They're a bit akin to power law or winner take all markets. If their success rate is, say, 2%, your ability to pick the winners, it just can't be that high. If you can build a portfolio, where say you're supporting 100 people, and 15 of them do very well, that's a very high hit rate. My guess would be, if you're a very good interviewer, and the people are operating in power law markets, you're doing quite well if you can manage 10 to 15%.
[00:02:20] Dan: Got it. I'm curious, when you have a guest in a return episode, how much marginal insight do you feel like you gain from meeting them a second time, other than maybe like you talked about their new book or another area of their work, but just like getting to know the person, what's the marginal improvement when you have a second episode?
[00:02:36] Tyler: It's quite slim. I don't know if you heard my episode with Paul Graham, but he said when he does interviews, typically he's making up his mind within seven or eight minutes, and often within two minutes. Now you can learn factual things about the person over many hours, years, decades, but your fundamental sense of them, it takes a very long time for it to change after, certainly after 20 minutes. As Paul noted, it can be less than 10 minutes.
[00:03:04] Dan: How do you choose what order to ask your questions in? You don't tend to start with softballs.
[00:03:09] Tyler: Depends if the person knows me or not or knows the podcast. Sometimes you want to signal that it's super serious. Other times, you just want to signal that it's friendly. When I had Vishy Anand on, he doesn't know anything about my worlds. What I did was I set up two chess boards in the room with important positions from earlier in his career, one of them was 30 years ago, and he saw the positions. Of course, he recognized them immediately. That put him at ease, and he thought, "Well, this will be at a high level, and this is someone who appreciates me." It depends on the guest what you signal at the very beginning.
[00:03:48] Dan: Yes. Why our follow-up questions overrated?
[00:03:52] Tyler: Often people just repeat the same thing they've said. Two-thirds of it is going through what they had said a moment ago, so why do too much of that? Sometimes true clarification is needed, or a person has said something you think is wrong, and you want to see if they can defend it, but most of the time, move on to something else. Do you have a follow up to that?
[00:04:15] Dan: [chuckles] I have another question. That's a response.
[00:04:18] Tyler: I could repeat my same answer, then we could show it's true.
[00:04:21] Dan: Yes, exactly, okay. What do you think the best career stage for someone to be at is to have them on as a guest? You can imagine you have people that are very early in their career, maybe like a Vitalik, you have Paul Graham, who's towards the end, you have other people who are operating in the middle of their career? Do you have a stage or any themes that you prefer to have somebody as far as how far they are along in their career?
[00:04:44] Tyler: I'm not sure it matters, and I'm not sure if Vitalik is well thought of as early in his career. I think of him as mid-career, even though he's super young because he did so much so early to his credit.
[00:04:55] Dan: Fair enough.
[00:04:56] Tyler: The person's basic temperament, are they willing to be open and somewhat controversial and engage doesn't change that much across the person's life. Maybe some people can be too old, or others too young. My intuition is mid-career is fast and there's a big wide long band where it's not changing so much.
[00:05:16] Dan: What's the optimal episode frequency? Of course, the more episodes you do, the more guests you get on, but then your less under prep for each guest.
[00:05:24] Tyler: Every day is optimal. Now, I can't manage that, given the rest of my life, but it is my aspiration to do what I hope is a good podcast every single day. In essence, you do a lot of prep in common, you would have people with common topics or common backgrounds. Doing one podcast, in fact, would help you prep for another one. It's always more frequent than whatever someone is doing.
[00:05:51] Dan: Say you're doing like two historians back to back. Today, what does the prep look like in terms of just like raw hours? Let's say that you haven't had a historian on yet, you've got a historian that's coming on next month, how many hours do you spend prepping for the conversation?
[00:06:07] Tyler: Well, first, I will try to pick someone in a field where I've already read a lot throughout my life. That's not always possible.
[00:06:11] Dan: Which is everywhere.
[00:06:13] Tyler: Well, not everywhere, but if it's the Byzantine Empire, there's someone with a new book out about the Byzantine Empire, I'm thinking of having them on, but it's an area where my background is pretty weak. I'm thinking, do I have enough time to prep for this person? In a sense, you need your whole life, but for someone like Foster, the Irish historian I had on, I spent four or five months reading Irish history pretty intensively. That was rewarding, but you can't do that for every guest. It depends who else you're having on, and you always prefer to have some guests that are relatively easy preps. For me, those are the economists, but it's not that I want to slack, it's that I want to put more total prep time in.
[00:06:58] Dan: I see. What do you think the optimal popularity is for the show? Why not put even more effort into marketing?
[00:07:06] Tyler: I don't know that marketing helps much for podcast, and the marginal listeners we would get I suspect would be lower quality ones than word-of-mouth listeners or people who know me. We market it a fair degree. We have swag, and we have a Twitter feed, and we send people emails. We don't buy ads on cable TV. I don't think this audience is so large anyway. Say you could pull in another thousand listeners by putting ads on TV, who cares? You're not doing it to have another thousand listeners.
[00:07:40] Dan: How important is an in-person episode? I suspect you're going to say very important, but especially if you were to aspire to do one episode per day, it seems like that'd be implausible. How do you determine whether or not it's worth going out of your way to make a conversation in person?
[00:07:55] Tyler: Well, before the pandemic, I thought it was incredibly important to have them be in person. Then, of course, all of a sudden, none of them were in person. We had feedback from listeners, and basically people said, "These are not worse," we asked a bunch of people, so I had to revise my view. The main reason often I want to do it in person is so I can meet the person, not to make the podcast better.
That said, there's some minority of guests, that's probably below or close to 10%, where you can put them at ease by being there with them and chatting a bit in advance, and those are better in person, but most of them aren't, but I've learned, and I was wrong at first.
[00:08:35] Dan: Interesting. How much better do you feel you've gotten at interviewing for the podcasts over time? Should we expect that the recent episodes of Conversations with Tyler are going to be a better you than the early ones?
[00:08:46] Tyler: No, I don't think so. There was a streak we had, one of them was Richard Prum, the ornithologist. There was the woman who does archaeology using satellite data. There was like five or six in a row, maybe a year and a half ago, that's like our best streak of episodes. Maybe it's just going to get a little worse. Katherine Rundell was great, Lazarus Lake, but no, I don't think I'm getting better. On average, I should expect to get a little worse. Too bad. Sorry.
[00:09:15] Dan: That's surprising to me. You have so much focus on practicing at what you do and it's self-improvement, but you don't feel like you're getting better at the podcast over time, why?
[00:09:25] Tyler: I just think there's an asymptote to a lot of processes. If I'm preparing for, say, a historian for four months, well, I could prepare for seven months. It would only be very slightly better, I think, and my basic skills, how well I speak or how well I think on my feet, I don't think they're getting better. There might be some areas where I'm in a better position to get desired guests, and that would make the podcast better. That's different from me getting better.
[00:09:56] Dan: If you're able to get Paul McCartney on, what's like the most important number one question you would want to ask him?
[00:10:01] Tyler: I would look at his early to mid solo career and dig deeply into b-sides, outtakes, different things he did in the studio and ask him about the details and see what he has to say, and not really talk about The Beatles very much at all.
[00:10:17] Dan: Switching gears a little bit off the podcast, one thing I noticed, it seems like you're often pretty critical of translated literature, but at the same time, you're a really big fan of it. You love Knausgaard, Ferrante, you talk about Houellebecq, and obviously much of Harold Bloom's Western Canon is all in non-English. At the margin, what's the right way for a reader to think about whether or not to read a translated work of fiction?
[00:10:37] Tyler: Well, the classics you should read anyway, but just realize you're probably getting something much worse. It's not true for every translation, but many things are just extremely conceptually in principle difficult to translate. It's one reason to learn a language as you get a better sense of what you're missing once you've seen how good or bad a translation can be, but you should never not read an important work because it's in translation. You mentioned Houellebecq. I actually read that in German first because-
[00:11:06] Dan: Oh, interesting.
[00:11:07] Tyler: -it was out in German before in English, and it's much better in German, I think. There's something about the seriousness of it, of sounding European, German with longer sentences in some ways being closer to French in that regard, and that worked. I don't read French and then I read it later in English and it felt a little more superficial, but I don't think that's the author's fault.
[00:11:30] Dan: What's your benchmark, though, if you don't read French? You could benchmark it against English. How do you know it's a good translation? Is it just the flow of how coherent it sounds or how do you benchmark it?
[00:11:43] Tyler: I don't think I always know. Typically, I can ask someone, and I would more or less trust what they had to say, but if it's a work with a very high reputation that I'm not enjoying at all, my first thought is to wonder if the translation isn't at fault because I think the market in classics is pretty efficient. That is everything considered a classic.
Pretty much all of it is quite high quality or very interesting. It could be the fault is not in the translation, but it's in me.
I've tried reading Count of Monte Cristo a few times, never enjoyed it. I don't think the fault's in the translation, because I've tried more than one. There's a new supposedly better translation. It's not philosophically deep in a way that might be hard to translate, so it's just like my defect.
[00:12:28] Dan: Interesting, because the other day, you also posted an addendum to your best fiction books of the year and you added Pedro Paramo the Spanish novel, and you said previously the translation, you didn't like it, this new one, you gave it high praise. Is this an example of that? You know it's a great work. You didn't like the previous translation and this one was just good enough? Or how did you know this one was good when you were reading it?
[00:12:51] Tyler: Well, I had read the original in Spanish, and that took me really a long time, even though it's only 120 pages, and the vocabulary is not hard, but what is going on, there's a paucity of information and you have to piece it together as a reader. That's hard in any language. If it's your third language, which Spanish is for me, it's harder yet.
The preexisting English translation just had zero humor in it and the Spanish had a lot of humor in it. This new one, I suspect, is as good as it can get. I bet if you asked the translator, he would know it's hard work to translate and that what he did wasn't perfect but was more or less optimized.
[00:13:32] Dan: On your year-end best books, another one that I thought actually surprised me was you put the new translation of Brothers Karamazov on there. I was surprised because you actually have talked-- you don't talk much about Dostoevsky at all, and then you actually even have a post where you respond to a reader who is like, "Why don't you talk about Dostoevsky?" You're negative towards him. What do you actually think of him? You were able to put this book on the best of the year? What is your view?
[00:13:57] Tyler: When I was in high school, Brothers Karamazov was my favorite novel of all time, and I just loved it. That was actually from the Constance Garnett translation. I don't think Dostoevsky is worse. I think it's that I have very different concerns and what he's obsessed with doesn't register with me really at all. These big, huge questions about God and death and evil and murder, Tolstoy is much more relatable in a way that he wasn't when I was younger.
I think that's a shift in me, not that I've gotten worse. Those questions to me, though, they just feel played out. Something like Thomas Mann, Magic Mountain, I think that I liked somewhat better when I was younger, but I don't think less of the work. It's just heavier stuff makes more sense in your 20s than in your 60s sometimes, and social nuance you'll find more interesting in your 60s for me. Nietzsche doesn't interest me anymore and he's clearly a brilliant writer, philosopher, but I pick it up and it's a kind of blank.
[00:15:03] Dan: I actually had a question on Nietzsche because you do list him. You have a list of authors where you should read all of their work and he's on it, but then you're usually pretty negative on him. In your conversation with Elijah Millgram, you're a scathing critique of him saying that he's just basically an anonymous Twitter poster. [laughs] Do you fit him in the same category as Dostoevsky or how do you think about Nietzsche?
[00:15:27] Tyler: He's still very important and was super insightful, but a lot of it's been absorbed. Again, I now have different concerns. Kierkegaard to me is more interesting, Nietzsche less so.
[00:15:39] Dan: Okay. There was an article claiming that Domenico Starnone, I'm not sure if I'm pronouncing his name right, but he's a male. The claim was that he is actually Elena Ferrante. Let's just pretend for a second that this is true. My question is do you think the Neapolitan novels would've had the same success under his name? Or how valuable is actually the pseudonym here?
[00:15:59] Tyler: First, as a side remark, my suspicion is that it's probably Starnone, but though I don't know that Starnone and his wife co-authored the Ferrante novels rather than just Starnone.
[00:16:11] Dan: I've seen that theory as well, yes.
[00:16:13] Tyler: His wife is a well-known translator, is very literary, seems to be extraordinarily intelligent. There's a lot in the Ferrante novels where I do truly feel only a woman could have written this. I don't know the actual story, but to address your question, if they came out as either co-authored or under a male name, yes, I think they would've done much, much less well in terms of number of copies sold. I don't know if that was their motive. There are significant downsides to fame. That may have been their motive.
[00:16:47] Dan: Should more authors do this, should they create up a pseudonym and try and restart their career?
[00:16:51] Tyler: It's hard to keep it secret. Indeed, in this particular case, it didn't remain a secret. Though I think most readers still don't have a sense of it. You have to do a lot of work, and it means you cannot do public appearances to promote your book, which is the main way people promote their book, not to mention podcasts. You're at a big disadvantage if you do not present yourself to the world. I don't think it will be that common.
[00:17:17] Dan: Yes. Say I'm like reading an author who many people can interpret as Straussian like Houellebecq, how worried should I actually be about misreading their work or is misreading their work kind of the point?
[00:17:30] Tyler: Maybe misreading is the point. I think with Houellebecq, there are multiple readings. When you're asking, "Well, how sympathetic is he toward Islam and Islam in France anyway?" You're not supposed to come away with a simple direct answer. It's all a misreading. None of it's a misreading. The multiple layers are important, and that's fine if you think it's simple, probably you've misread the book. Other than that, give your mind free play.
[00:17:56] Dan: The only take that I agree with pretty strongly, which is where you say that a lot of the recent literature is actually not so far behind maybe the 18th or 19th century. You've got Naskar, Bolaño, Sebald, a whole bunch of authors who've actually done really great work. Again, notably a lot in translation, though.
[00:18:13] Tyler: Yes.
[00:18:14] Dan: My question is do you expect it to keep up in the next decade? One of the confounders here I was thinking about is AI. I guess I won't give you any more perspective on that. I'm just curious for your overall view over the next 10 years.
[00:18:28] Tyler: I don't have a concrete prediction, but I don't see a reason why it should slow down, so I'm not pessimistic. Obviously, things come in waves and bulges and there's back and forth of Solenoid, which was translated this year, it's an incredible novel. Why think it has to stop? Now, maybe some longer time horizon people will read AIs.
There was just a Chinese science fiction story that won some award, and it was written by an AI.
I think short stories will come way before novels and even then a lot of people will prefer the human product and AI might help you write a great novel in different ways. Again, not a concrete prediction, but no reason to be pessimistic. Readers want it, humans can do it. What's in the way?
[00:19:18] Dan: That's great news. Let's get excited.
[00:19:20] Tyler: Yes.
[00:19:21] Dan: I asked Scott Sumner a similar question, but I really want to get your take on this. Say I'm just a casual movie-goer, and I go see the biggest hits every year, but I haven't really seen too many Hollywood classics and I'm not familiar with someone like Tchaikovsky, but I have the opportunity to go see Memoria in theathers, should I go and what should I do to prepare myself?
[00:19:40] Tyler: Oh, of course, you should go. To see it on a small screen is almost worthless. It's worse than reading literature and translation. Some movies you just have to see as they were made.
[00:19:52] Dan: Would I find it boring if I'm just a casual movie-goer, I haven't seen a lot of these "Art House" indie movies? How do I prepare myself to not be bored by it?
[00:20:01] Tyler: I don't know what you mean by the word casual. Not everyone likes any kind of movie, but I'll just say the first "deep foreign movies" I saw, I loved them immediately. I was at that point in time by definition just starting out. I don't know if you'll like it, but the returns to you trying it are very, very high. Again, you need to be ready to just walk out. Just leave.
What's the cost, but $14, $15, parking and driving, definitely do it. I don't think these things are hard to understand. It's whether or not you're interested, like am I interested in Nietzsche and Dostoevsky right now? Well, less. That's fine. You might be on that tier with Memoria, but there's going to be something maybe like Godzilla Minus One where, "Oh, wow, this is what I've been waiting for." For me, it's both. Godzilla was good too.
[00:20:55] Dan: Fair enough. Actually, I read you as like overall, though, somewhat negative on Hollywood. Maybe you like Godzilla, but it seems like you are generally more bullish on, again, foreign movies and lower-budget indie movies. What's just your overall sentiment on film for the next 10 years, similar question to with books?
[00:21:12] Tyler: I suspect it's at a turning point. Putting this year aside, Hollywood, to me, it seemed to have a four or five-year run that were just the worst four or five years in the mature history of Hollywood ever. That was enough to make me pessimistic. Then you had big 10 pull movies, a few of which were entertaining, but mostly a bad trend. This year, I went back and looked at my movies list, and it was incredibly good, and not just foreign movies like May December.
Last night, I saw Poor Things, two incredible movies. They're not mainstream Hollywood, but it's not South Korean or Iranian. It's broadly a Hollywood movie. My hope is we had this negative shock to theater going, couch potatoes, Netflix. Movie makers had to adjust. Now they've adjusted. I'm pretty optimistic looking forward. Though I admit there's only one year of a really good data point for cinema, but that one year is so strong. I'm optimistic.
[00:22:15] Dan: You indulge in a lot of different artistic mediums. My question to you is, how do you decide what to enjoy next? Say you've got some free time and you're trying to decide whether to read a novel, watch a movie, or maybe you're thinking between visiting a museum versus listening to music. Do you just do what you feel like or do you have some strategy just allocating your time to enjoy different types of art?
[00:22:35] Tyler: I think it's pretty impulsive and selfish. There's also a side constraint what is my wife interested in doing? She and I have broadly similar desires. I try to see all the major exhibits that come to town. I think I get to see most of the major movies I want to see. I'm trying to do it all. I wouldn't say I succeed. There's some areas I just don't touch upon that are probably very good, but something like dance I just don't follow. I know I wouldn't have the time, and that's a shame, but there's scarcity.
[00:23:09] Dan: What do Jonathan Swift and Peter Thiel have in common?
[00:23:12] Tyler: I wrote a paper about Jonathan Swift for a conference sponsored and chaired by Peter Thiel. That paper will be published, but Swift was pessimistic in some ways that Peter is. He was obsessed with questions of technology. He was doubtful as to how much moral improvement there really is in people. He thought that war was likely to recur perpetually or at least the risk of war. He was pretty skeptical about a lot of the academicians and intellectuals of his own time.
Swift even wrote about venture capital and innovation in Gulliver's Travels. He mocks the world of science, the flying island with all the innovations, and the scientists running around acting like kooks. It's also the only major world in Gulliver's Travels where there is no slavery.
To have an absurd science pursuing a lot of dead ends is better than the world of slavery. I'm not sure how much you could classify Peter's ideas in that same framework, but I don't think they're totally dissimilar the idea that there's something quite violent in human history. For Peter, it's Girard more than Swift and you just want continuing antidotes to that violence, I think you can find in both Peter and in Swift, and my paper makes these points that will at some point be out.
[00:24:36] Dan: When is it coming out?
[00:24:38] Tyler: I don't know. They're doing a book. Books take longer, I would guess more than a year, but less than two years.
[00:24:45] Dan: Got it. Should a classically liberal society draw the line on crude comedy? When, if ever, should a joke be off limits?
[00:24:51] Tyler: If by off limits you mean illegal, I don't think it should be illegal in most cases. Now there are extreme instances such as modern Germany, which I think has made certain jokes illegal for reasons related to Nazi history. I don't know if that's a good idea. I think I would grant there might be a few extreme exceptions of that kind, even there, I suspect it's not a good idea, but if you just mean the comic should be canceled, I don't know. I would prefer just not to patronize the person and not lead some online charge against them. If it's not funny and in poor taste, just ignore it.
[00:25:29] Dan: In your analysis of what we might call the new right, just broadly defined, it basically boiled down to how they have a huge mistrust of well-functioning elites. My take here is that the internet is maybe the most causal factor behind this like the Martin Gurri thesis. My question to you is do you think that we'll be able to trust elites in the internet age without censorship or is a free internet going to keep us trapped in this distrust?
[00:25:56] Tyler: I agree with you that Gurri is probably right and probably has the best explanation. Gurri himself is modestly optimistic. He thinks internet norms will somehow adapt to this over the next few decades. I don't at all dismiss that possibility. I would make the additional point, the whole internet is about to change because of AI, and whatever the internet gave us two years ago, it won't be what it's giving us five years from now. It's just going to be remixed in some very different way. Probably it will change that it's just making us cynical, but I don't know what that new change will be.
Imagine you wake up in the morning and you just say to your AI, what was on the internet last night that I might want to see, and it serves it to you. I think that will just be a very different experience, but I don't yet know how.
[00:26:44] Dan: That's fair. What do you make of Scott Alexander's theory that the political views swing on a pendulum between generations? I think the name of his post is called Right is the New Left, but the basic idea is you've got like the boomers in Gen X, some of their liberal sway he attributes to a rebellion against maybe more conservative parents. Then today the argument is that what you see with libertarians getting a little bit more interested in some of these either neo-reactionary, whatever you want to call it, like new right types of movements would be a reaction against their liberal parents. I'm just curious, how much weight do you put on a theory like that?
[00:27:21] Tyler: I would want to reread Scott's post, but I would say in general, I'm quite skeptical about generational theories of all kinds. I don't think there are usually discreet breaks. There can be with particular events like 9/11, the great financial crisis, COVID probably will be one, but they may or may not overlap with how we classify generations. I just think of the data. There's a lot of papers on generational effects and they don't impress me that much. On this, is there a clear statistical test to show that hypothesis is true? I haven't worked through that literature. My guess would be no, there's not such a test.
[00:27:58] Dan: You say the most important thinkers of the future will be religious, but you yourself are agnostic. Who specifically should be religious?
[00:28:05] Tyler: I don't know if people should be religious. I think on average you have a better life doesn't necessarily mean you should do it, but you mentioned Peter Thiel before. He's probably been the most influential public intellectual over the last five to 10 years. My other pick would be Elon Musk. Peter is the religious thinker of the two, and Elon is the secular thinker of the two. They used to work together.
The best columnist in the world right now, I would say, is Ross Douthat who's not only Catholic, but he's very religion and God-oriented in his writing. We already see this. A lot of people ask me about that statement. I'm amazed that to other people, it isn't trivially true. Now religious may be a minority. They just have this very rich source of ideas, inspiration, and motivation that a lot of secular people don't have.
I'm not religious, but I think in some ways I'm a religious thinker. The sense of personal mission is very strong in me and I think or at least hope it comes out in what I do. I'm not a total exception to the claim.
[00:29:11] Dan: What is your implicit theology?
[00:29:13] Tyler: A lot of it is American protestant with some Jewish elements mixed in. For someone who grew up in the northeast, basically the 1970s, that's extremely common and indeed garden variety ordinary. I like it, but it's nothing special about me. I'm a regional thinker like most people, and that's my region.
[00:29:34] Dan: What did you actually get out of reading Plato when you were very young? I think you said you read it when you were like 12, 13. Can young people really understand classic philosophy?
[00:29:42] Tyler: I strongly believe they can, maybe not all, but at least the ones who care. For one thing, you just get a sense of truth being conversational or dialogic and that there's not any single position that can answer all the complaints raised against it. That was one of the most important things I learned. You learn particular techniques of persuasion and argument, and you also learn why they fail. Most of the arguments in Plato's dialogues fail, and often they're pretty bad. It's an interesting question how much Plato knew that or designed them that way? I think he did, but people have different views on that.
Then just particular dialogue like the Euthyphro question. Well, is it good because God says it's good or does God say it's good because it's good? Whatever you think the answer might be. I'd never heard that question until I read Euthyphro, and I still use that as a mental device. You can just learn a lot of particulars and then a lot of the writing's beautiful. Symposium to me is and was very beautiful. Parmenides' dialogue, in a funny way, I found that beautiful. Yes, you can learn a lot at quite a young age.
[00:30:51] Dan: What makes Shakespeare, Proust, Moby-Dick, so much better than a Stendhal, Twain or Thomas Mann? Why is it that there's this tier up that's well above all other writers?
[00:31:02] Tyler: I never knew any of those people, of course, but I have to think they were just somehow more extraordinary as human beings, and they had that extra something they could draw upon. That's a hypothesis very hard to verify or refute. Maybe something about their time was especially rich as well. The times of Stendhal, that was obviously a rich, fertile time. There's something ultimately quite mysterious about it. I like how Harold Bloom puts his finger on that mysteriousness.
[00:31:32] Dan: Where has Fernando Pessoa most influenced you?
[00:31:35] Tyler: The style in which he wrote the book, the book of small things. It's very scattered, I think it's a more fruitful approach than what Nietzsche did. It's more positive. It's about how to find beauty in life in different ways. It's also about how to innovate. It's positive, but there's this melancholy to it as well. That to me is a very rewarding book. I think it should be much more prominent.
[00:31:59] Dan: How would the world be different if every person understood that demand slopes down?
[00:32:04] Tyler: I don't think it would be very different. I think most people do understand that. They may not articulate it as such. The problem arises when they interpret public events and policies through emotional lenses, mood affiliation, sometimes even evil motives. The people who do bad things, like did Hitler not understand that demand slopes down? I'm pretty sure he understood that. If the price of bullets went up, maybe he'd buy for the army fewer bullets and more of something else, but he was still extreme evil. Unfortunately, that act of education won't get us very far.
[00:32:41] Dan: Why is that concept, it seems like you valorize this concept a lot. Is it just that it's so important, but it's already well understood, so there's not much more we can do with it, or why do you often cite this as such an important insight?
[00:32:54] Tyler: I don't think it's well articulated. It's well understood in the sense that people live by it in the supermarket. When you're thinking through public policies to keep that opportunity cost, gains from trade, a few other central ideas foremost in your mind, and not let them be crowded out by mood affiliation, evil sentiments, what some people call tribalism. I don't like that word, but you know what I mean, to me it's very important. It's taking seriously what you already know that is difficult for people.
[00:33:23] Dan: Got it. Practically speaking, like what's the fundamental difference between a public and a private organization? At the end of the day, aren't these just groups of people that sit in a room to solve problems? Why is there so much fuss about whether work gets done with a private or public sector?
[00:33:38] Tyler: There are differences in many areas. There's some areas where you find no difference. In the data, a privately-owned water utility seems quite similar to a publicly-owned utility, noting that the private utility is still regulated heavily by the public sector. You do observe especially in young new firms, a lot more dynamism than you observe in old public sector bureaucracies. That said, if you look at public and private universities, an area I know pretty well, are the private ones, they can act much more quickly, which is good, but they're far less tolerant.
The problems that have surfaced lately with universities, overall, they're worse in the top-tier private schools than in the public ones, which, maybe, a priori is not what a lot of libertarians would've expected, but it's very clearly true. There's much less free speech, both de jure and de facto in the private places.
[00:34:34] Dan: On that topic of just like governance structure for an organization, do you think that this is a solved problem for business? Crypto in some ways is trying to innovate here with like the concept of a DAO, but like broadly speaking, the modern C-Corp seems like it's working pretty well. Would you expect this over the next 50 years to still be the standard for the way that corporations work? Maybe with some tweaks here and there, or do you see innovation coming in governance structures for businesses?
[00:35:00] Tyler: I suspect it will still be the standard. I'm not sure it works very well, so my core view is there's some minority, to be clear, of new firms that are super effective and large firms they're effective because of scale, but on the ground they're massively inefficient, and bureaucratized and can be worse than the public sector. I'm not sure we'll change that, but if there's a way to change it, it would be to carve out more dynamic spaces in large companies.
They're clearly doing well in terms of profitability. Again, that's mainly scale or sometimes because of regulation, but you deal with them, and you think like, "My goodness, this is horrible." Or like, "I could never work here, I would hate that." If anything changes corporate governance, I think it would be AI. I think what we will start doing is having companies, probably new and small ones, take all their data, shove it into the AI, just ask it, what should we change? [laughs] Then listen.
I don't think it's the established companies that will actually do that. They may play active doing it, and I don't know what all the suggestions will be, but I think that's going to matter quite a bit. It may not change the functional forms much, but we'll see.
[00:36:14] Dan: It's pretty striking to me like just how different John Rockefeller seems from Elon Musk from Lee Kuan Yew, all of these really charismatic successful leaders. If you were to lead an organization, how much effort should you put into studying the lives of these great business folks versus just doing?
[00:36:29] Tyler: Sam Altman says you should. Most of the people I know who are highly successful in CEO-type roles, they do it. I'm never sure what the marginal product of doing it is. Is it the doing it or is it the feeling that you've done it unless you have a certain level of confidence to just proceed with some decisions? I don't know, but look, the smartest people think you should do it, so I guess I'll side with them, but I'm not entirely convinced, I would say. You said Rockefeller, Musk, I was wondering which is the one you think is charismatic and which you think is not charismatic? That's going to depend, I guess I think they both are, or were.
[00:37:09] Dan: Yes. No. Fair enough. I said Rockefeller, Elon Musk, Lee Kuan Yew, I guess very, very different, but maybe charismatic to different people.
[00:37:17] Tyler: Exactly. I think some listener for any one of those people, there's a lot of listeners who wouldn't like them at all, or find them charismatic.
[00:37:27] Dan: Yes. Rockefeller is like a religious guy, totally abstains. Elon Musk is, I don't know what he is, but it's-- [laughs] I don't imagine that Rockefeller would have the same temperament, so yes. There's a quote about how people make steam engines, and it basically goes that you can make them without knowing thermodynamics, but the people who do know thermodynamics are going to make better steam engines. How much should business leaders actually understand economics? How much does it matter?
[00:37:55] Tyler: I think if you're a business leader, the return to knowing a few months' worth of economics is extremely high. You can even imagine packets designed around that. Past that, I think the return is quite low. You might need to know highly specific things about finance or your sector that would involve some additional learning of economics, but it's a learn-as-you-go sort of thing. The idea that a few years spent studying organizational theory, I'm pretty sure that's negative return. It's going to confuse you, get you looking for the wrong things. Again, to know basic economics, highly valuable, and the people in these top positions basically all do know that.
[00:38:37] Dan: Yes. Okay, let's assume, I've asked a few people this question, that all progress on AI model stops tomorrow. The doomers win, there's regulation, there's no more new models, but we still have GPT-4, Claude, Bard, the models that are out in existence today. Would you expect there to be an obvious causal boost in total factor productivity over the next decade, just due to the models that are out today?
[00:38:59] Tyler: Over the next 20 years. Decade, people are slow, and large corporations are bureaucratic. It could take 15 to 20 years. I'm also wondering in your thought experiment, are token costs allowed to fall or they just set it where they're at? If token costs can fall, that's de facto like having a huge amount of extra progress, even if it's just GPT-4, because there are ways you can layer current models that are quite expensive, but pretty more powerful than what's in your pocket. We would do those at lower token costs. Yes, I think there's a lot of productivity already embedded in what we've done.
[00:39:36] Dan: Got it. Why do you think there was, maybe there was, my question would be with Microsoft Excel, the release of that software seems like it would've driven a huge amount of total factor productivity growth. I actually don't know, but typically when I bring this up with people, the answer is, it didn't causally create some. I'm curious how you compare like GPT-4 to the leap that Excel made.
[00:39:59] Tyler: I think office software, I don't know if it's Excel or not. Probably it wouldn't be Excel, but in the time period, 1995 to 1998, give or take, office software and inventory management systems did lead to a high boost in productivity, so why did it end? I don't understand, but we got quite a bit out of it. We had some rates of productivity growth of about 3% for a few years. That was our mini escape from the great stagnation on an economy the size of US plus there's a global effect. It's worth a lot, but I agree it's a puzzle why it just seemed to end.
[00:40:35] Dan: Yes. To what degree is the internet actually amplifying worries about AI? What I'm thinking about here is LessWrong seems to have done a lot of work on helping people understand the risks, whether you agree with them or not, and it's been hugely influential with the PAUSE movement. If the internet were around during the Trinity test, do you think we would've had the same level of concern about nukes?
[00:40:59] Tyler: Oh, sure. If nuclear research had been open to the public in some manner, it probably would've been much worse, and the scientists themselves who knew what they were doing, by definition, they were extremely concerned. They just didn't have any internet. I think people are, by nature, we're rewards that may be optimal in many ways, but we're going to do this no matter what. We'll see who's right and wrong.
We're not going to let China rule us with Chinese AI. You might think that's a better world because we lower the chance of total doom by 2% and you do the hostile calculation, but the world just doesn't work that way, and blog posts are not going to make it so, and that, to me, is a more important point than whatever your P(doom) is.
[00:41:47] Dan: Yes. Sort of on that topic, one of my favorite takes you have is that, times right now, if they feel chaotic, they're actually not the unusual piece. The unusual piece was the relative period of peace we had after the war, up to 9/11. I am curious though, another observation you have is that society has become more feminized starting in maybe the 1970s. Do you think that a more feminized society is a positive or a negative in times of chaos?
[00:42:15] Tyler: Well, it's been a positive up until now because it's led to less war, but I think the data from the last three to five years show the trend of less war has been reversing only very recently, but obviously, that's bad. It's possible that the feminization trend is over, and what we're seeing now is an increase in the variation of feminization.
A lot of parts of the world will continue to feel much more feminized, but there's some sort of backlash where there's like men trying to act manly under different visions of what that means and even seeking out non-feminized world spaces. I think I admit, that worries me but it's a bit like AI. It's also inevitable if you're going to have that much feminization that quickly and just to try to yell stop is not so fruitful, so I would like to think there are more positive ways we could steer that, but I think that trend of a simple increase in feminization, that's probably over, even though it doesn't feel like it's over because of this variance effect.
[00:43:18] Dan: Yes. Interesting. Why should someone read all of Fernand Braudel's work?
[00:43:23] Tyler: I don't know if they should read all of it. The huge three-volume set on structures of material life. That's amazing. It's incredible. I got a lot out of those. The multi-volume set on the Mediterranean is excellent, like it as much as the other one. His later book on France was interesting but not really all that well-developed. There's other things he's written and I haven't read. You should read his major works. I'm not sure you should read all of it. I don't know.
[00:43:53] Dan: What is the best argument against Scott Sumner's proposal for NGDP futures targeting?
[00:43:57] Tyler: Well, there's a number of different parts of the proposal, but the first point I would make is we don't have an NGDP futures market, which Scott readily admits. You just can't do it. Now, Scott favors experimenting with subsidizing such a market, and I'm fine with that experiment. We've done that on a very, very small scale, and it hasn't taken off. My guess is those experiments won't work. Then there's the backup question. What about just targeting NGDP level in some manner?
Mostly, I agree with Scott, to be clear, but I think my pushback would be, can we ever make that a rule? Like Carl Schmidt, sovereign is he who decides the exception, I don't think you really can bind most parts of your government with rules. Certainly, not your central bank, not your treasury, not your foreign policy, so I favor NGDP is a kind of idea you inject into the mix to get central bankers more excited about doing the right thing and less fearful, but I don't think it can be a rule. I don't think there's some better rule per se. I just don't think it's an area where rules can make sense, and indeed, you saw in the pandemic, there's no rule that could have prepared us for what to do. Whatever you think we should have done, no rule specified in advance would've handled that well.
[00:45:15] Dan: Yes. I am curious, do you think that this is a bullish case for AI taking over humans then because, presumably, for AI to automate things, that's got to be, to some degree, like rule-based, but if we need like a human in the loop, would that suggest that like, there will still be quite a bit of human employment when AI takes off?
[00:45:36] Tyler: There's a bunch of different questions in there. The first is, how does AI relate to rules? I think one thing we learned from AI is the notion of a rule makes less sense than we think. Say, you ask GPT questions about monetary policy, about Ferrante, whatever, it'll keep on giving you answers. The answers might vary, and they're based on processes that are not transparent. Is that like GPT discretion, or do we want to say, our rule is always listen to ChatGPT?
It's like both a total rule and total discretion at the same time, and that dimensionality of how rule-like it is, it just seems to become moot. That's what I take is one big lesson of AI. Our rules make less sense than we thought, but then there's the question, well, how many people will it put out of work? I think 20 years from now, the number of people unfed staff can be lower than it is now, but not a tiny number. I'm struck by the fact that the research staff at Open AI itself, it was not too long ago, I was told nine people.
It's a remarkably small number for what might be one of the most important creations in all of human history. Some other things will become like that. I'm not sure how much and how many, but we're going to see other things like that are mid-journey. I think when it was doing peak innovation, total staff, I don't even mean research staff, just total employees was eight or nine, and maybe they had contractors. I get all the ambiguities, but still, that number should shock us.
[00:47:06] Dan: Yes, I've often wondered if, actually, that could be a business model where you just say like, pick your enterprise software or pick your existing business and your whole business model is like, we're just going to build this but with a 10th of the employees and use AI from the ground up. Presumably, for firms that already exist, the transition to downsize and leverage AI would be much harder than it would be for someone to start up and do it from the ground up, so that's been a pet theory of mine.
[00:47:32] Tyler: I think that would be a big trend. I very much agree with that, and it will shock people. It gets back to your earlier question, how might corporate governance change? Even if the formal structures don't change, if a whole tier of firms is way, way smaller, it will feel very different.
[00:47:48] Dan: Yes.
[00:47:50] Tyler: Then what you rope into the people who are not doing the jobs that now don't exist, that will be its own change, so there's going to be like a lot more gardeners everywhere, I don't know, carpenters, people greeting you at the restaurant, I think there'll be full employment, but you're going to see a lot of very weird jobs, like you do that, like you come over and you massage dogs' paws for rich people or something, I don't know, it's going to be weird.
[00:48:17] Dan: Yes. [laughs] One of Scott Summer's takes, I've heard you say you agree with, and this is one of my favorites, is that he doesn't believe that the 2008 housing, asset runup, was actually a bubble, and he doesn't believe Bitcoin was a bubble either, the price runup in both of those. My question to you is like, how long of a time horizon do we need to wait to assess whether or not something is a bubble? If two lips come back, we're not going to say the Dutch were just too early, it wasn't actually a bubble. Well, what's the time horizon we need to make that assessment?
[00:48:48] Tyler: I don't think it's time per se, but if you look at the whole context, I think you usually can say, so I mostly agree with Scott, but if you're saying, well, some Las Vegas suburbs, homes there were indeed a bubble. That's probably true, and it's past the point where, oh, if 100 years from now, they're worth a lot because of desalination in AI, and you want to say, oh, it was never a bubble. Oh, come on.
Most real estate has definitely come back, but there's some markets where I think you can say it was a bubble, quite a few crypto assets that most people haven't even heard of where you, like maybe it wasn't even a bubble, it might have just been a fraud, but it had bubbly-like aspects in early stages and you can render judgment, I think, in many cases.
[00:49:33] Dan: In what area do you think there are too few emergent ventures applications?
[00:49:38] Tyler: You mean geographic area or area of work?
[00:49:40] Dan: Area of work, yes.
[00:49:42] Tyler: Well, really good people working with ideas, and I think especially in small countries, it's so much easier as an individual to have an impact in a smallish country. I lived in New Zealand for a while, so I saw that I lived it, then in a large country like Brazil or India. People working with ideas in small countries, to me, still seems radically under provided. I would support more of it if the proposals were good.
[00:50:10] Dan: Do you think we'll ever have a coherent theory of what happened to the economy in 1972 even if we just poured more people into the economics profession to work on it? Is that a problem that can be solved?
[00:50:19] Tyler: I don't think we'll know much more than what we know now. Our opinions will change a lot based on what the future looks like. Say, if AI is a big deal as I think it will be, we'll then be more inclined to say we had exhausted the previous general-purpose technology. There won't really be new evidence about '72, '73, but it will feel like that's a critical idea maybe in a way that's even a little unfair. I think that's what will happen, but we won't actually get new evidence, not much.
[00:50:51] Dan: What about this other big mystery? Will we ever have a coherent theory of how differences of culture interact with traditional economic mechanisms? Basically, what is culture's impact on traditional economics?
[00:51:03] Tyler: You've deployed these two words, culture and theory. They're not literal direct opposites, but once they're both in play, I know they're not really going to fit together. That's in a sense what culture is. It's some of what's left over after theory is gone. Theory is demand curve slope downwards, and culture is something else. We're not going to make that much progress. We will catalog it much, much better, however.
[00:51:28] Dan: Who is more right about immigration? Brian Caplan or Garrett Jones?
[00:51:31] Tyler: People dispute what Garrett thinks and says, to be clear. I know Garrett well, and I feel I know what he thinks. I would say Garrett is clearly correct. Garrett, in my view, is pro-immigration, but he wants to be very careful about who is let in. Now, that said, I think Garrett should be more pro-immigration than he is. I view myself as more sympathetic to taking in what typically would be called low-skill immigrants. I'm not sure they are also low-skilled than Garrett might be.
I think Brian's open-borders idea is crazy and totally unworkable. The actual effect of it would be to elect someone fascist-like in countries that would try it. You wouldn't even get open borders. You would have a severe backlash, and open borders for a wealthy country just wouldn't prove workable. I would very much like to increase immigration levels. It just has to be politically sustainable.
The skilled people, they need nannies and people to clean their homes and wash their cars, whatever. Skilled immigration, you also want "unskilled immigration". Again, these unskilled people, they can do so many things I can't. They're awesome. Try to put together your own ping pong table, can't do it. Have over a "unskilled immigrant" like that. Awesome. Thank you.
[00:52:56] Dan: Yes. If you had to restart your career as an academic but you couldn't do economics, what would you do?
[00:53:03] Tyler: I have to be an academic. It's a bad time now to enter academia. It's shrinking, there's a lot less freedom of speech. Maybe I would consider being a legal academic or philosophy, but it would be grim. I feel I got very lucky, my area I ended up in when I entered it was really a golden age.
[00:53:23] Dan: If you weren't restricted to being an academic, what would you do?
[00:53:28] Tyler: You could ask what do I do now. A lot of what I do now, you could call publishing. I guess I would be a publisher. It's not a hypothetical, I am a publisher, and I work for a publisher, which is Bloomberg Opinion. In this sense, the question's already answered, right?
[00:53:43] Dan: It is, yes. Detached from GMU and do the same thing.
[00:53:48] Tyler: Yes, that would be it, and I could.
[00:53:52] Dan: How should one decide which cultural codes are actually worth cracking? For example, would you ever go to Burning Man?
[00:53:58] Tyler: I've been told at Burning Man, there's no internet, which for me would be a work problem and just a fun problem. I've been told there's a lot of drugs which doesn't interest me. I guess the cost would be high. I think it would be very interesting, but probably I wouldn't go. I think there's infinite number of cultural codes to crack. You just have to choose what you're interested in and/or what overlaps with work you need to get done. There'll always be more, you don't have to worry about running out. I wouldn't quite say it's all equally interesting, but all of it is so interesting. You don't even have to worry about that either.
[00:54:33] Dan: Okay. I really like this quote. There's an interview on The Browser between Yuri Bram and Applied Divinity Studies. One of the quotes is just that the marginal revolution extended universe is incredible. I can't get over how many writers and thinkers I know seem to have got their start by a link from Tyler Alex, the MR Universe. It doesn't have a name like the rationalist or the effect of altruist. My question to you is, should it or is that by design?
[00:54:59] Tyler: It's not by design, but I don't feel I want it to have a name. It may be better operating under the radar. Also, the original vision for Emergent Ventures was to operate under the radar. There's zero advertising, but now, it is, in fact, pretty well known that over time might make it worse, probably will make it worse. You want to keep a lot of things under the radar. That's hard for people because they want to earn income from it or fame or some kind of rent or some reputational rent. The world is always trying to convert you into that existence, and then you have to keep on pushing back.
[00:55:33] Dan: Fair enough. A question I wanted to ask you that I find really interesting is that you seem to maybe be one of the least likely people to get into like an emotional confrontation with someone. I think this was really on display. For example, in your conversation with Amia Srinivasan, your post on how you practice at what you do though has no mention of working on your temperament. Is your temperament something that you work at?
[00:55:55] Tyler: I think temperament is pretty genetic, I don't know the literature on that. That's what I observed, and when you see young babies, toddlers, it seems that how they're going to be, that those parts of it are pretty built in. There are advantages and disadvantages to detachment. I think I have both those advantages and disadvantages, and that's just my hand of cards, and I'm going to play it, so to speak. I don't sit around trying to be more detached. I think as you get older, you do become more reasonable operating from almost any basic initial temperament. That's happened to me as well.
[00:56:33] Dan: If I'm someone who says, "Man, I get too hotheaded on Twitter and I say things I shouldn't," is this just baked into the cake of my genetic predisposition or is it something that you think you can work at?
[00:56:43] Tyler: Something that operational, you can work at. I know people, they'll take Twitter off their phones. I don't know how effective that is, but it seems if demand slopes downward, it should be at least somewhat effective. Then something that specific and operational, just don't be on Twitter at all. A lot of people have left, so that, you can manage. I don't know that you can change yourself so much, but particular things like, should I go visit Uncle Joe and get mad every Thanksgiving because he supports Trump? That, you can fix for sure. I'm not sure you can, again, change how you engage with people more generally.
[00:57:16] Dan: Yes. Should more people create their own version of Tyrone?
[00:57:19] Tyler: I hope not. Should more people blog at all? There, I would say yes. I would like to see more blogs and less Twitter, less TikTok, but clearly, it's gone the other way and for reasons that are probably permanent until AI changes things again.
[00:57:36] Dan: I think what I'm getting at here is maybe more at the conceptual level. The way that I view Tyrone anyways is a framework for thinking through points of view from someone else's perspective.
[00:57:45] Tyler: Oh, yes. That, much. much more of, and you should have hundreds of Tyrones in your head of different kinds.
[00:57:52] Dan: Do you have others that are not Tyrone that don't come out?
[00:57:54] Tyler: Sure. I even had a phrase for this once, I called it Phantom Tyler Cowen or Phantom fill in the blank. I said to someone, there should be Phantom Tyler Cowen sitting on your shoulder, not as the only phantom by any means, responding to what you're thinking, saying, writing, and you test it against the phantoms. I'm a big believer in that.
[00:58:14] Dan: Interesting. You use phantoms when you're thinking through something. Actually, who are your phantoms? Who do you look up to and get advice from today?
[00:58:22] Tyler: It depends what I'm writing on, but in economics, it would basically be the famous economist. I don't think there'd be any big surprise in there. I don't write fiction. If I'm doing a podcast, I'm not really thinking as I'm going along about that. You might have a thought ex-post, but I don't really think like, "Oh, Joe Rogan wouldn't have liked that there," or I don't know. The podcast is just done and you're already busy with the next one.
[00:58:48] Dan: Another thing I find interesting about Tyrone is that you said that he noticeably has less posts recently. One of the reasons you gave for this before is that you spent less time on him because the world has become a more bizarre place. I would actually think that Tyrone is much more valuable when the world has become a more bizarre place, like what gives there?
[00:59:08] Tyler: I think Tyrone works to the extent he does because he's shocking or was shocking. Now, it's much harder to shock people given what we accept in the normal course of ordinary affairs. Tyrone just sounds like another person writing, it doesn't mean Tyrone is never effective, but I do think it lowers the value of Tyrone.
[00:59:28] Dan: I'm going to read to you a passage. There was a profile of Joyce Carol Oates in The New Yorker about a month ago, and I think he linked to it one day. I found this really interesting. Here's a quote about Joyce, to waste time made her feel slithering centerless. She wrote in her journal, a 500-pound jellyfish unable to get to this desk. Oates was friends with Susan Sontag who had a busy social life. After the two spent time together in New York City, Oates told her, "In some respects, I'm appalled by the way you seem to be squandering your energy."
Then she reminded Sontag that the pages you perfect day after day will be the means by which you define your deeper and more permanent self. Now, I found it interesting, Sontag also wrote in her diary that Joyce Carol Oates writes all the time, she can meditate while writing. She says she has no feelings. There's a dispute here. Who do you think is right?
[01:00:13] Tyler: Well, they're both right for what they were born with. Joyce Carol Oates, don't know her, but I strongly suspect she could not have been to Sontag. I also strongly suspect Sontag will endure much more than Joyce Carol Oates. I don't know that enduring is the standard. Robin Hanson and I have this chat sometimes. He thinks you want to do things that will last a long time. I don't really see the value in that.
I think you want to do things that somehow generate some energy now, and you're not going to know it's going to last. Probably none of it will last. Very few things last, and don't worry about lasting. It's hard enough to generate energy or response now. Now, Oates has written a large number of novels, but the best things of Sontag, I don't think we'll read them 100 years from now, but I think we'll read them 50 years from now. That's pretty impressive.
[01:00:59] Dan: You think it's just her output that was more valuable than Joyce? Not sort of the way she chose to spend her time?
[01:01:04] Tyler: Her peaks?
[01:01:04] Dan: Yes, her peaks.
[01:01:05] Tyler: She may have had lower bottoms. The Chairman Mouse stuff, I'm really not crazy about. Oates, by being so obsessed as worker bee, I think avoids the flaws of Maoism and all those overreactions. There's some benefits to the Oates method that are sometimes under-discussed. You can't sit around for a long time and talk yourself into something really stupid because you need to get the next novel out.
[01:01:31] Dan: What's the right amount someone should invest in their social life? Because really, the argument here seems also over like how much time someone wants to spend being a New York socialite versus really committing this to work. What's your view on that, or does that just vary by person?
[01:01:45] Tyler: The Sontag, when you obviously have no internet, I strongly suspect, and I've read the Benjamin Moser biography of her, which is very good. I suspect you have to do a lot of that in New York or somewhere like that to be connected to ideas at all. I would be surprised if she had overinvested in social life. It's more of a cutting question now. There's just a lot more dimensions along which you can try to optimize. Like, "Oh, I'm going to stay at home, but I'll be in these great WhatsApp groups." I don't know. I guess it just depends. I've tried to have my work and social life overlap as much as possible. I like that, but it's not for everyone.
[01:02:21] Dan: That's a good point. I guess socializing could have been a totally different thing back before the internet.
[01:02:26] Tyler: It was the way you got in touch with ideas. Like you would hear things on the radio and, again, then especially, New York was the place to be in a way it isn't now. Sontag was well-connected, well-located, had everything going for her. Of course, she built it that way. Paris was another place you could be. There were a few others, but my goodness, like most of this country, you just didn't have a chance of being an important thing. It was very hard.
[01:02:55] Dan: To what degree do you think it's still important to be in somewhere central like that? Obviously less, but are there still benefits to being in a San Francisco, New York?
[01:03:03] Tyler: Depends on what you do. Certainly, if you're doing startups, there's a labor market you need to worry about. There's a reason why so many of those are concentrated. Inspiration, levels of ambition, I think it's very important to be surrounded by other ambitious people, that makes most of Europe much worse. It's still important. If you're an academic in Ann Arbor, which has always had a very good school, but you are much better off in relative terms than you would've been before the internet. Absolutely.
[01:03:32] Dan: Do you think that Europe's problems in general, just with lagging innovation lately, are more cultural or more political? In other words, do you think that electing new officials into the EU could get this thing turned around, or do you think it's the culture in Europe is really just turned away from ambition?
[01:03:51] Tyler: I think it's the culture, but I'm more positive on that culture than many people are. All the criticisms, growth is too slow, too bureaucratic, overregulated, that's all true. There is some cultural capital still embedded deep in Europe that they can have public conversations about things and arrive at non-crazy answers. That's very deeply embedded there. That's part of the culture too. It may even be related to some of the reasons why they overregulate, and Europe's pretty robust. I wrote a column recently about France. France has become an underrated nation, oddly enough. I'm not a doomster on Europe, but obviously, they face serious challenges.
[01:04:30] Dan: If you didn't need to sleep and had a marginal seven hours per day over everybody else, where would you spend the time?
[01:04:36] Tyler: Oh, seven hours. I can't really get anywhere in that seven hours. I don't think I would do anything very differently with it than what I do now. I would just have more of it.
[01:04:45] Dan: Do you ever waste time? What would that even mean to you?
[01:04:48] Tyler: Isn't it all waste time? I don't know, what's the waste?
[01:04:52] Dan: Fair enough. Fair enough. Do you think that Bach would've been even more productive with the internet, or would it be a distraction?
[01:04:59] Tyler: I don't see how he could have been more productive. It seems the thing he could have used was better-functioning pens, and that would've made him more productive. Even if you think, well, what if Bach could've had an AI and he talks into the AI and it writes out all the musical parts for him? We don't know the Bach production function. He did so much. I somewhat suspect it was the writing it out by hand that generated the musical ideas. It was actually pretty well set to optimize the bar.
[01:05:29] Dan: Why there are no real competitors to marginal revolution? No one out there just said, "Hey, I'm going to go blog five times a day and put assorted links out there." Correct me if I'm wrong, but I haven't seen a single competitor truly emerge.
[01:05:41] Tyler: I don't know of one. It's been over 20 years. Well, it's hard to do. That's a stupid kind of answer. The downward-sloping demand curve point reemerges. You need to put a lot really, in a way, your whole life into it. It's not that writing takes that long, but you have to live the kind of life where you're always discovering new content, and most people just don't want to do that, or they're not good at it. They can't do it. Some mix of the above.
I guess that's the binding constraint. More generally, I'm struck by how many of the people who are still around doing well. They're not bloggers anymore, but Matt Yglesias, Ezra Klein, Megan McArdle, they all came from that early burst. There was something in the air then. I don't know what it was. There've been some follow-ups. Scott Alexander comes more recently, but that's now even like 10 years ago, I think. These things are weird. You see similar patterns in music, in the arts, bursts, and then dry spells. We don't understand them.
[01:06:42] Dan: Why have you kept blogging?
[01:06:44] Tyler: I still learn from it all the time. Gives me access to a lot of smart and interesting people. It helps me get other things I do like the podcast. I have no plan to quit anytime soon. Not at all. I'll probably do it for as long as I can.
[01:06:58] Dan: Do you think that you just value those things more than the other people who quit, or why did you have the say in power and everyone else did not?
[01:07:06] Tyler: I don't find it strenuous or stressful to have a deadline of every day. That's part of it. That may get to my equanimity, our detachment that you mentioned earlier. I can read and absorb material much faster than other people. Even if comparable like IQ or education, that's an important underlying factor. If I could read, say, good, read 5 to 10 times faster than a lot of comparable people, that's just a huge edge.
I know a lot of people don't believe I do it. Fine, think what you want. I don't want you to believe me. I don't want competitors. There's always something. Then the notion that I could go almost anywhere in the world except North Korea and meet up with interesting people who would engage with me. That's a super high value, and I don't want to give that up.
[01:07:52] Dan: Okay. Some questions loosely about New Jersey. I think you mentioned before that you're a fan of The Sopranos and I believe, did you grow up around James Gandolfini? I don't know if you had a relationship with him.
[01:08:02] Tyler: He and I worked at the same Valley Fair in Hillsdale, New Jersey at the same point in time. I was in the produce department and he was pushing carts. Now, I didn't know him. My sister told me this, I almost certainly saw him around. I didn't think, "Oh, hey, that's James Gandolfini of The Sopranos," because we were 15, 16 years old. He was there and I was there the same supermarket you could call it. It was like a totally inferior, immature Costco, to be fair.
[01:08:35] Dan: Okay.
[01:08:36] Tyler: The Sopranos, what they show in the episodes, I recognize an awful lot of these places. Satriale is where they meet, especially early on, the butcher shop on Kearny Avenue. I was born in Kearny in the Kearny Hospital, which is only a few blocks from that butcher shop. That whole street view, I know well. The views driving at the 95 overpass and all that. It's like, "Yes, yes. I know that." The sporting goods store on Route 17, very familiar to me.
[01:09:06] Dan: Oh, fascinating. Here's the question. Why is Tony such a horrible talent spotter? He gave Christopher so many chances. His best friends are constantly betraying him. What is Tony Soprano's fundamental flaw with spotting talent?
[01:09:19] Tyler: Well, there's negative adverse selection into all of those roles and indeed into Tony's role. You're selecting for who is brutal, who will be at least superficially loyal to you, which is not a good way to select. You end up with fools, cretins, unreliable people who have been trained in enough brutality. They can also do you a lot of harm. Even if they don't kill you, they may go out and kill someone else in a way where you get whacked in return and you're liable for them. That's what happens. Selection so often matters more than incentives.
[01:09:51] Dan: Interesting. One of your learnings that you mentioned from the produce department and actually being from rural Michigan, I've sensed this to some degree as well is that there's many people who are highly intelligent and capable but don't quite have the conceptual frameworks to put them on a road to success. My question to you is, what's the real problem here? In what ways would society have to change for you to have the most capable coworkers that you had in the produce department, have a shot at getting a CEO-type job when they grow up?
[01:10:19] Tyler: Some of the people I worked with I thought were quite smart and certainly not lazy and charismatic, but there was some conceptual issue. Could they realize on their own they needed to go out and acquire all of this conceptual equipment, and they didn't have that. That's really quite important. It's related also to your Sopranos question. Maybe we should value that more, and also, in some way, teach it more. It sounds like a weird thing to teach. Maybe we teach it a lot of indirect ways by showing people role models. Maybe there would be effective ways to teach it more explicitly.
I would look into that. In a sense, that's what I try to do, teach people. There's this life you can live where you're an infovore. It's not for most people, but it might be for you, or maybe at the margin, you want a little bit more of it, and here's what it looks like. It's not that bad. I'm trying to do that in a way.
[01:11:14] Dan: Say, a talented 14-year-old who has broad intellectual ambitions comes to you for advice, under what circumstances are you going to recommend that they spend some time doing some sort of manual labor? Is it always, or are there special circumstances where you say, "Yes, you should go work construction for two years, or the produce market"?
[01:11:30] Tyler: It depends what country they're from, but if they're from the United States and a middle-class or upper family, I would recommend that. If you're from a country where you're just trying to get away from that, probably different. Why not, most of all? It used to be standard that people would work as teenagers. Teen employment rates have plummeted. There's way too much homework, and we've replaced actual work with extracurriculars, which I think is a terrible trade-off.
[01:11:58] Dan: Getting close to wrapping up here. I have just a quick rapid-fire question for you. I'm going to name a few people and I want you to just say the number one thing that you've learned from them. All right. We'll start with Thomas Schelling.
[01:12:09] Tyler: Game theory came alive for me when I read Thomas Schelling, which is before I met him. There's a separate question, what did I learn from knowing Thomas Schelling? He had the ability of always having an appropriate anecdote or good or insightful story ready no matter what the topic was. Even at quite an old age, he could do that better than anyone else could in the room. I guess I think I've never learned that from him like I should have. I failed to learn that. I'm not really good at that the way he was. I would say the story there is what I didn't learn.
[01:12:44] Dan: What about Jonathan Swift?
[01:12:46] Tyler: I think Oliver's Travels is one of the all-time great books. You can keep on rereading it. It always has additional richness and human motivation. He has a very deep understanding of how people respond to situations, and what strangeness is like, and how easily we can turn bigoted or intolerant. Those would be some things I learned from Swift.
[01:13:08] Dan: The Gioia brothers.
[01:13:09] Tyler: You mean in painting?
[01:13:11] Dan: I mean Ted Gioia from The Honest Broker, and Dana Gioia from--
[01:13:16] Tyler: Oh, Gioia. Sure. I only met Ted once. I've learned a lot from his writings on music and from the podcast with him, but I haven't had really much direct contact at all. Dana, I've spent a lot of time with. American music of the sort like Samuel Barber, I understand much better because of Dana. A lot of opera, I understand much better, a good deal of poetry theater. I would say the American Arts to the first half of the 20th century. A lot about the American West. Quite a bit about how the NEA, National Endowment for the Arts, works, which he was head of for a number of years, and how to negotiate bureaucracies. Some amount about marketing. I've learned a lot of different things from Dana.
[01:14:04] Dan: Camille Paglia.
[01:14:06] Tyler: I only met her once. Her book, Sexual Persona, really quite inspired me, and it made me realize I should write for broader audiences, and that one could take what is underlying scholarly material and make it something people might want to read. Now, in the particulars I learned from her, my whole understanding of Edmund Spenser comes from her, and I think that's basically correct. She changed my mind about quite a few things in the Renaissance. Androgyny and Shakespeare, I got quite a bit of that from her. The Pre-Raphaelites, I understand better because of her. There's a number of other details I could relate but I've learned a lot from her.
[01:14:44] Dan: Shruti.
[01:14:46] Tyler: Oh, it's an enormous amount about India and talent search in India from Shruti. The Indian Constitution, most of all, how the Indian policy world works, how to think about different regions of India, what's the best Indian food. Really quite a bit about Veena playing and Carnatic music. A little bit about Bollywood, which she knows incredibly well. It just doesn't interest me that much. One of my failings is I haven't learned enough from Shruti about Bollywood. Maybe that will be remedied, but the movies for me are too long. I do think they're good. It's just very time-intensive. That would be my Shruti answer.
[01:15:20] Dan: Tyler, it's been wonderful talking to you today. Thank you so much for your time.
[01:15:25] Tyler: Same here. A real pleasure.
Tyler Cowen