Conversations with Tyler 2024 Retrospective
What this covers
Tyler Cowen delivers a year-end accounting of where ideas come from, how institutions are changing, and what counts as success. The retrospective spans his year of conversations and moves across technology, economics, media, and the arts—touching on why the Great Stagnation has ended, what is hollowing out the economics profession, and why YouTube summaries may soon deliver more value than the books they critique. The through-line is Cowen's assessment that the discipline of economics is losing relevance even as it tightens its gatekeeping, that real intellectual progress now lives in engineered artifacts (generative AI, biomedical research) rather than published arguments, and that trying to measure success can obscure it.
The conversation moves between the diagnostic and the provocative. On economics specifically, Cowen argues that rising credentialing barriers and bureaucratization are driving talent away while the profession doubles down on failing strategies. He claims the Great Stagnation itself—once radical—is now accepted truth, ended by advances in AI and biomedicine, and that the semi-popular science book as a vessel for new thinking has flattened. On media and meaning, he contends that derivative works—LLM summaries, critical YouTube videos—increasingly eclipse their sources, that algorithms create vastly larger gaps between winners and losers than podcast downloads do, and that Engaging Without Consuming an original work may leave you better informed. On culture and constraint, he argues that Constraints Generate Drama, explaining why contemporary romance fails where Victorian literature and religious dating shows succeed. He also makes the counterintuitive case that Impact Without Credit—genuine influence that requires no metrics to prove it—is the truest form of success, and applies this principle to nonprofit work and institutional contributions alike.
In this year-end retrospective, Cowen argues that the most valuable ideas now live in things people build (LLMs, biomedicine) rather than in books, that traditional gatekeeping and bureaucratization are hollowing out economics, and that real impact is recognized without measurement.
- The Great Stagnation has likely ended due to AI and biomedical advances, while books have gone flat as a source of new ideas
- Rising entry barriers in economics are driving talent away and the profession is losing status
- Genuine success is self-evident and does not require metrics
AI is reshaping where real innovation happens, displacing traditional knowledge work and publishing.
- The core prediction of Average Is Over — that AI would be the next big thing and success would hinge on how well one can work with AI — has done very well, though whether AI will boost or lower income inequality remains undetermined.
“The predictions I made in my book Average Is Over, that AI would be the next big thing and it would all be about how well can you work with AI — that’s done very, very well. I think I deserve more credit for those predictions, in fact.”
- We are entering a future where the derivative product (YouTube criticism, LLM summaries) can attract far more attention and provide more value than the original work itself — e.g., classical recordings selling 300-400 copies while critical reviews reach hundreds of thousands.
“more and more, you can engage with a thing without consuming the thing. It might be the criticism on YouTube, or it might be that you ask the LLM to give you the summary, and you might get more value out of the derivative product than the product itself.”
- The Great Stagnation, considered radical when stated in 2011 but now widely accepted as having been true, ended a few years ago due to advances in AI and biomedical science.
“The Great Stagnation I think was seen as radical when I stated it in 2011. It’s now widely considered to have been true. I think it’s over. Not everyone thinks it’s over. I think with AI and advances in biomedical science, it ended a few years ago.”
- It is not a great time for new ideas in books — the semi-popular science/economics book trend that Jared Diamond started now seems over and flat — because the real new ideas are embedded in the things people build, like generative AI (with models like o1 pro beating humans at most things) and biomedicine.
“Those are the new ideas that matter, and the rest is often people just spinning their wheels. It’s not a great time for new ideas in books. The new ideas are all stuff — also in biomedicine — and it’s just been incredible, mind blowing.”
Art and narrative require intersubjective coherence and constraints to generate lasting canon and drama.
- Video games are unlikely to eclipse cinema as the great art medium because they are too context-specific to the individual player and fail to generate narrative and vision that is intersubjectively understandable enough to enter the canon, despite huge usage and revenue numbers.
“There’s something about video games — it’s too context specific to the person playing it and doing it. It doesn’t quite manage to generate narrative and vision in a way that is intersubjectively understandable, where it could become part of the canon.”
- Much contemporary romance fails dramatically because characters can sleep together too quickly or divorce too easily, eliminating tension; Victorian literature and the Modern Orthodox dating show Srugim succeed because their cultural constraints supply the necessary obstacles that generate drama.
“The problem with a lot of romance today is, where does the drama come from? They could just sleep with each other too quickly, or if someone needs to divorce, it just happens. Victorian literature had solutions to these problems, and Srugim... has solutions as well.”
- Music does not truly reach its peaks until Monteverdi and then Bach; Renaissance music, while including admirable composers like William Byrd and Palestrina, is limited because its means are too constrained, marking a genuine break in the level music attains.
“I think it’s not until Monteverdi that music truly starts to reach its peaks, that the means are just too limited... for me in music, there is a break with Monteverdi and then Bach, or I just think it attains much higher levels.”
- Christopher Nolan is not a good filmmaker — Interstellar has become nearly unwatchable because it is excessive, not properly cinematic, lacks actual drama for the most part, and is too parasitic on its own special effects, even if its craftsmanship and production design are strong.
“I don’t think Christopher Nolan’s a good movie maker. I know that’s heresy, but it’s all too excess, and it’s not really tinged with actual drama for the most part, only in some scenes.”
Professions that double down on gatekeeping and prior strategies instead of adapting lose status and talent.
- Economics as a profession is losing status and influence — the Biden administration relied much less on economists than Obama's did, and academic economists are unlikely to be highly influential under Trump — yet the profession is doubling down on prior strategies instead of adapting, which it will regret.
“The Biden administration relied much less on economists than the Obama administration did. Trump, we’ll see what happens, but I would bet against academic economists being highly influential there”
- Increasing bureaucratization and credentialing requirements in economics — needing a top undergraduate school, letters from highly credible people, and pre-docs — are the profession's most overrated feature and are driving away talented people who would otherwise have become economists.
“The thing that is most overrated is, just entry barriers in terms of bureaucratization are higher and higher. You have to have gone to a very good undergraduate school, have letters of recommendation from highly credible people, maybe have done a pre-doc, and that’s ruining it for people.”
Successful work and relationships demand deliberate effort, broad knowledge, and willingness to forgo credit.
- Generic guest pitches built around topical business or news books ('the impact of tech on America's youth') are immediately deleted, while pitches that demonstrate the person is an impressive polymath, has done something very strange, or have clearly done homework on the show will get a closer look.
“They should make a case that the person is an impressive polymath, even if they’ve done nothing and have no book. That, we would consider. Or the person has done something very strange, perhaps with their body.”
- Historians and scientists make the hardest interview preparation because, unlike economics where one already knows much history, a body of scientific work cannot be crammed for by simply reading more books, whereas guests whose work one already knows well (like Stiglitz) require little prep but teach you less.
“Science, you can’t prep for. History, you can just read more books... But a body of work in science, it’s like, “Oh, do I start reading a biology text or something?””
- As you get older you need to make real, deliberate efforts to make and keep new friends, but it can be done — exemplified by recent friendships with Rick Rubin and Noam Dworman.
“As you get older, you need to make real efforts to make and keep new friends, but you can do it. Rick Rubin or Noam Dworman, they’re pretty new friends for me.”
- There is no limit to what one can achieve if one is willing not to get much credit for it — the right attitude for an organization like Mercatus, exemplified by Hester Peirce's work on crypto leading to her SEC commissioner role being a personal rather than institutional triumph.
“Someone said on Twitter recently that — it was Joe Lonsdale — “There’s no limit to what you can achieve if you’re willing not to get much credit for it.” I think that should be our attitude.”
Individual lifespan, wealth, and media consumption follow natural limits that cannot be overcome by will alone.
- Settling Mars is not a new idea and will be tough, though worth rooting for.
“Settling Mars is not a new idea. Let’s root for it, but it’ll be tough, and it’s not new.”
- The right way to invest is to buy, hold, and diversify — doing very little, not selling, not rebalancing, buying more as you save — because you cannot beat the market and should not try.
“Buy, hold, and diversify. I do very little. I don’t sell things, don’t really rebalance; as I save more, buy more stuff; and that’s it... I don’t think I can beat the market, and I don’t try to.”
- By the logic of the winner's curse, if the show actively pursues and wins its audience's loyalty, it has effectively bid too much for them and is therefore overrating them — so the fanbase should be called The Overrateds.
“if winner’s curse holds, we’re pursuing them as an audience, and that’s winner’s curse for us. We’ve bid too much for their loyalty, so we must be overrating them, would be the simple economic argument.”
- Life expectancy will continue to increase slowly, but there is a natural upper limit to human lifespan of roughly 100-120 years imposed by the brain — you cannot repair the brain without replacing who the person is — so individuals cannot live forever, though most people who buckle their seat belts will likely reach their 90s.
“My guess is there’s some natural limit due to the brain at somewhere between 100 and 120. I don’t see how you repair the brain without replacing who is the person. I don’t think individuals can live forever.”
- In the mid-1980s the dollar was extraordinarily strong against the German mark — about 3.45 marks to the dollar, roughly double what purchasing power parity implied — which made it affordable to lease a car and travel widely in Europe.
“When I started in Germany, it was 3.45 marks to the dollar, which people now will forget, but that was an incredible exchange rate. Like 2X for the dollar, what it should have been by PPP.”
- Although nonprofits should create reports and metrics, genuine success is self-evident and recognized without measurement — as with Ralph Nader's impact or Peter Singer's success in increasing vegetarianism and pro-animal legislation despite not winning on most issues.
“every nonprofit does and should create reports and metrics, but ultimately my view is, if you’ve really succeeded, you know it without measuring it. It’s just obvious.”
- Comedies tend to be forgettable and cannot really be rewatched — you watch them once and then forget about them, and that is optimal.
“Comedies, I tend to forget, but I don’t mean that as a criticism of comedies. And you can’t really rewatch them. You just watch them once and then forget about it. That’s optimal.”
Large-scale systems—payments, transportation, policy—remain broken or inefficiently designed despite available solutions.
- The US payments system remains badly broken — sending a single international transfer to Ethiopia looked sketchy enough that his bank canceled his account — and while Stripe has done amazing work, the infrastructure still needs fixing.
“The first time I used it, my bank canceled my account. I had to go in person because it does look sketchy to be sending just a single transfer to Ethiopia all of a sudden... Payments in the United States is still screwed up.”
- DOGE is not a new idea — a version was tried in the first term of the Reagan administration (per David Stockman's The Triumph of Politics) and failed; the current attempt will likely try harder with the benefit of knowing past failures, but the concept itself is not novel.
“If you read David Stockman’s old book, The Triumph of Politics, they tried a version of that in the first term of the Reagan administration. It didn’t work. They’re trying it again.”
- Investment in urban bike lanes is generally overrated in the US because no one has produced a serious cost-benefit analysis accounting for all costs including land; DC bike lane usage is declining and lanes sit mostly empty, even though the approach works empirically in dense places like Manhattan, Amsterdam, and Copenhagen.
“As far as I can tell from the numbers I’ve seen, use of DC bike lanes has been declining. I suspect it’s not the future. It’s a relatively slow and late-19th-century mode of transport.”