What this covers

Tyler Cowen discusses how COVID-19 has triggered the great reset he previously forecasted, examining where institutional stagnation persists and where genuine dynamism emerges. The conversation ranges across economics, politics, education, and existential risk, with Cowen arguing that the Great Stagnation remains real in education, healthcare delivery, and government, yet remains absent from technology and large business sectors. He positions the pandemic as a catalyst for biomedical innovation even as longer-term civilizational threats loom. A recurring tension threads through: contrasts between functioning institutions (the Federal Reserve, the Supreme Court) and dysfunctional ones (regulatory agencies, Congress), and between short-run productivity gains and durable growth shaped by liberal institutions over centuries.

The discussion turns substantive on existential risk, where Cowen departs from space optimist framings. He argues that cumulative probability over time makes a major war near-certain on sufficiently long horizons, especially as weapons of mass destruction become cheaper. The true danger, he contends, is not a coordinated totalitarian adversary but the asymmetry of destruction vs building: destruction is far cheaper than construction, making accidents, miscalculation, or rogue actors plausible doomsday triggers. He estimates over 70% odds that a nuclear weapon kills people within Dwarkesh's lifetime, though apocalyptic scenarios remain lower-probability. On near-term matters, Cowen argues online education supplements rather than replaces in-person learning, that mentorship and small-group friendship remain high-return life strategies, and that most talent development happens after age 18, not before.

Sharpest takeaway

Cowen argues the 'great reset' he forecast has arrived via COVID-19, and that while institutional decay and stagnation persist in education, government, and services, the crisis will spark a durable burst of biomedical dynamism even as long-run existential risks (especially war and cheap WMDs) remain civilization's ultimate threat.

  • Institutional failure (regulatory state, Congress) contrasts with functioning Fed and Supreme Court
  • The great stagnation persists in education, healthcare delivery, and government, not in big tech or big business
  • Over long enough time horizons a major war is near-certain, which caps humanity's durable future on Earth

The claims · ranked36 claims · weighted by value

This asset isn't compiled yet

You're seeing its claims, ranked. Compile it to build the argument threads, weight them, and check each claim against your library — the full view.

0.54

Mentors are enormously high-return because they teach only a few things but those things are critical: they give you a glimpse of what you can become, which you are oddly blind to in their absence even if very smart; the value comes mostly from what you see them do, not from advice you should literally follow.

causalcontestednovelty 3/4durability 4/4· Tyler Cowen

“They give you a glimpse of what you can be and you are oddly blind to that in the absence of those mentors, even if you're very, very smart.”

0.50

Being a space optimist lets you hold a very long-term horizon without obsessing over safety because humans could colonize galaxies for enormous multiplicative gains, but Cowen rejects this because the speed of light and travel difficulties are binding constraints, and because the framing devalues the world we actually live in via a Pascal's-wager-like trap.

causalcontestednovelty 3/4durability 3/4· Tyler Cowen

“If you're what I would call a space optimist, you can believe in a very long-term horizon without getting obsessed with safety. Because humans are the successors, the humans can colonize galaxies... but I'm not a space optimist.”

0.50

The likely civilizational threat is not a superior totalitarian enemy with better bombs but the fact that the cost of destruction is far lower than the cost of building, so a wacky general, an insane president, or a mistaken missile-detection reading is the more probable doomsday scenario.

causalcontestednovelty 3/4durability 3/4· Tyler Cowen

“the cost of destruction is so much lower than the cost of building, and it may not be some totalitarian enemy that does us in, but just some wacky general or an insane president or a mistake in measuring whether nuclear weapons are coming over.”

0.50

In any given year the chance of a major war is small, but letting the clock tick long enough makes a civilization-threatening war near-certain, especially as weapons of mass destruction become cheap (a nuclear weapon costing $50,000), so even if not every human dies, civilization may not persist.

forecastcontestednovelty 3/4durability 3/4· Tyler Cowen

“I think in any year, the chance of a war is quite small, but if you just let the clock tick for long enough, a major war will happen.”

0.50

Steven Pinker's data on declining violence is at least equally compatible with a ratchet hypothesis in which each major war is worse than the last, so they come less frequently but a future World War III would be far more catastrophic; Pinker's inability to answer the cheap-nuke question prompted a Bayesian update toward Cowen being right.

causalcontestednovelty 3/4durability 3/4· Tyler Cowen

“it's at least as equally compatible with a ratchet hypothesis. That major war is each one worse than the last, so they do come less frequently, right? World War II was a while ago, but World War III is going to be quite a doozy.”

0.50

On Earth, neither living space nor energy is a binding constraint on growth; human conflict is the one Cowen worries about most, with carbon emissions a short-run constraint he expects to be fixed.

factualcontestednovelty 3/4durability 3/4· Tyler Cowen

“It doesn't seem to me that living space is one and it doesn't seem to me that energy is one. Human conflict might be one, and that's indeed what I worry about the most.”

0.50

Common sense morality tends to be roughly workable because societies with anti-common-sense moralities (suicide at five, families with 20 or zero children, refusing to talk to grandparents) fail to reproduce or compete, so via a weak form of group selection and competitive pressure across families and clubs, surviving moralities stay near something workable — meaning a philosophy that totally contradicts common sense (as Peter Singer's sometimes does) should provoke worry.

causalcontestednovelty 3/4durability 3/4· Tyler Cowen

“societies that had uncommon sense morality like commit suicide when you're five... be unwilling to talk to your grandparents. That's the opposite of common sense morality. Those societies won't do well.”

0.50

Online education is a supplement, not a replacement: the right model is to replace the worst ~20% of one's education with chosen online material from a wonderful menu, while face-to-face remains the center of the educational experience because people fundamentally need to get together.

normativecontestednovelty 3/4durability 3/4· Tyler Cowen

“take the education you had and replace the worst 20% of it with things you choose online from a wonderful menu, is that a good deal for you? Obviously, in my view, the answer is yes.”

0.50

It is impossible to confidently identify the single biggest mistake in human history because we don't understand historical contingency well; even rerunning around World War I could produce a worse outcome, such as a large aggressive non-Nazified Germany developing atomic weapons first by 1942.

factualcontestednovelty 3/4durability 3/4· Tyler Cowen

“I'm just reluctant to make the comparison. I don't know is the answer. I don't think we understand historical contingency very well.”

0.49

Cowen's generic advice for young (or older) people is twofold: get at least one really good mentor, preferably two or three, and assemble a small group of good friends you love talking to and staying in touch with — 'small group theory and mentors.'

normativecontestednovelty 2/4durability 4/4· Tyler Cowen

“the first is get at least one really good mentor, preferably two or three. Second, get a small group of good friends that you love talking to, hanging out with... Small group theory and mentors, that's my generic advice.”

0.46

Contrary to common assumptions, after age 18 is the best time to develop talent because before 18 you are much less able; the exceptions are domains like elite sports where starting at 19 is too late, but for most actual jobs the post-18 window is optimal.

factualcontestednovelty 3/4durability 3/4· Tyler Cowen

“After 18 is the best time to do it, because before 18 you're much less able... but for most actual jobs, after 18 is the best time to develop.”

0.45

Cowen estimates over a 70% chance that a nuclear weapon is used to kill people within Dwarkesh's lifetime — most likely a terror attack or accident killing just a few or trading a city for a city — while the chance of an apocalyptic full nuclear war leaving only a few survivors remains low for quite a while.

forecastcontestednovelty 3/4durability 2/4· Tyler Cowen

“I would say the chance is over 70% that a nuke goes off in your lifetime.”

0.45

Today's chief innovators fighting complacency are big businesses to a considerable extent, which have far higher productivity than small-to-medium enterprises and remain extremely innovative (Alphabet's many products, Amazon's evolution from a bookseller to a cloud-computing leader), so the American productivity problem is not with big business.

factualcontestednovelty 3/4durability 2/4· Tyler Cowen

“if you want to look at who are the innovators right now who are fighting against complacency, it is big business to a considerable extent, and we should appreciate big business for that.”

0.45

The decline in startup formation is remarkably concentrated in retail (fewer mom-and-pop stores competing against the likes of Nordstrom), which is hardly a worrying development, making Cowen less pessimistic about the startup decline than when he wrote The Complacent Class.

factualcontestednovelty 3/4durability 2/4· Tyler Cowen

“If you look at the decline in startups, it's actually remarkably concentrated in retail. I don't think I knew that when I wrote The Complacent Class, so I'm less pessimistic about that fact than I was when I wrote Complacent Class.”

0.44

Coming out of the crisis the biomedical area will be incredibly dynamic, with the research activity into virology, immunology, vaccines, and pandemic-fighting going down in history as a phenomenal blast of innovation in a very important area.

forecastcontestednovelty 2/4durability 3/4· Tyler Cowen

“I think this will go down in history as a phenomenal blast of innovation in a very important area.”

0.44

The great stagnation thesis is unchanged because so much of stagnation is in education, healthcare delivery, and non-tech services plus government, which still stagnate, whereas tech companies and big business do not — and the pathetic American pandemic response illustrates low government productivity.

factualcontestednovelty 2/4durability 3/4· Tyler Cowen

“So much of the great stagnation is about education and healthcare and services, non-tech, non-internet services, and those are still stagnating.”

0.44

What a university provides that online videos cannot is interaction with different personality types, learning how hierarchies work, meeting mentors who inspire and help secure jobs, and socialization — so the best ~10% of face-to-face classes are irreplaceably valuable.

factualcontestednovelty 2/4durability 3/4· Tyler Cowen

“you interact with different personality types. You learn how hierarchies work. You meet mentors who inspire you, you meet mentors who help get you jobs and you socialize and you figure out how a lot of things work.”

0.44

It is likely (well over 50% probability) that biomedicine five to ten years out will be much better because of this pandemic, leaving us better prepared for the next one — the silver lining in the cloud, even though it doesn't mean the pandemic was worth it.

forecastcontestednovelty 2/4durability 3/4· Tyler Cowen

“it is not just plausible, but likely to believe that biomedicine five to 10 years from now will be really much, much better because we had this pandemic.”

0.44

Growth is cyclical in the short run with periods of slower and faster growth, but because culture and institutions are persistent, maintaining reasonably liberal institutions allows steadily advancing economic growth over the long run of centuries, reconciling a cyclical view of history with the case for long-term growth.

causalcontestednovelty 2/4durability 3/4· Tyler Cowen

“if you keep reasonably liberal institutions, I think over quite a long run, let's say centuries, you can have steadily advancing economic growth.”

0.44

Economic growth over time raises all boats even when the gains are unequally distributed — an insight lost sight of because of impatience, since you can go a long while without seeing it happen, but it is nonetheless true.

causalcontestednovelty 2/4durability 3/4· Tyler Cowen

“Economic growth over time, even when the gains are unequally distributed, over time, it does raise all boats. That's an insight we've lost sight of because of our own impatience.”

0.44

We do not actually know what nuclear leaders would do if they saw missiles incoming; the available evidence indicates they would retaliate, but Cowen is genuinely not sure, given that mutually assured destruction's logic is psychologically fraught.

factualcontestednovelty 2/4durability 3/4· Tyler Cowen

“I think we don't know what any of these leaders would do if they saw the missiles coming. I think the evidence we have indicates they would retaliate, but I'm genuinely not sure.”

0.44

During an extreme recovery from a weird event, productivity measures (like other economic statistics) don't mean much because it is impossible to distinguish output gains from people working more versus working smarter.

factualestablishednovelty 2/4durability 2/4· Tyler Cowen

“when you have an extreme recovery from a weird event, for one thing, productivity measures don't mean anything, like a lot of other economic statistics. Like how much of the output gains, people working more versus working smarter. I don't think we'll ever know.”

0.41

Temperament is relatively stable for most people from roughly 18 through 70, differing only at the very young and very old extremes or in cases of so-called mental illness; interests and skills can change but core temperament largely does not.

factualcontestednovelty 2/4durability 3/4· Tyler Cowen

“from like 18 through 70, I think temperament is fairly stable for most people that I've known.”

0.40

The Complacent Class predicted that racial issues and segregation would be a central part of the great reset because American Blacks are often among the most vulnerable members of society, so trouble would flare there early as a warning signal — a prediction Cowen views as having held up.

factualcontestednovelty 2/4durability 2/4· Tyler Cowen

“segregation will be a central part of it because American Blacks are so often among the most vulnerable members of our society and trouble will flare there early as a warning signal.”

0.40

The naive thesis that China will become more democratic as it gets wealthier has no evidence; China has become less democratic, and the much wealthier potential Chinese median voter does not even seem to want democracy, while autocracy has proven a far better competitor than was assumed 20 years ago.

factualcontestednovelty 2/4durability 2/4· Tyler Cowen

“the naive thesis that as they get wealthier, they'll be more democratic, there's no evidence for that. They've become less democratic. The much wealthier Chinese potential median voter doesn't even seem to want democracy.”

0.39

The argument for prioritizing growth requires that the time horizon for successful growth be neither too long nor too short; if too long, you become completely risk-averse because you don't want the world to end, ending up caring only about safety rather than growth.

causalspeaker onlynovelty 3/4durability 3/4· Tyler Cowen

“if the time horizon for successful growth is too long, you just become completely risk-averse because you just don't want the world to end. Things can run on for so long and you end up not caring much about growth at all. You just care about safety.”

0.36

The Fed and Supreme Court are functioning well as institutions regardless of whether one agrees with each decision, while the rest of American government is rotting, reflecting a willingness to 'just let things happen' that has caught up with the country.

factualcontestednovelty 2/4durability 2/4· Tyler Cowen

“The Fed and Supreme Court in my view have been quite good, whether you agree with every decision or not, but just as institutions they're fully up and running and the rest is rotting.”

0.35

Society growing more feminine is on balance good because it brings safety — the chance an American dies in combat has been very low for decades — and feminization is inevitable once you have anything approaching equal rights, so the right move is to build a dynamic highly feminized society rather than the myth that reversing feminization would restore high growth.

causalspeaker onlynovelty 3/4durability 2/4· Tyler Cowen

“that's a significant upside from feminization and we should be thankful for it every day.”

0.35

Women are more likely to change later in life and more radically than men, partly because of the formative experience of childbearing and because women are more disadvantaged by society, taking longer to break through barriers, whereas a smart nerdy white guy at 18 will likely get his chance early.

causalspeaker onlynovelty 3/4durability 2/4· Tyler Cowen

“Especially men. Women are more likely to change.”

0.35

If the crisis runs on for two or more years, the plausible (not tail) outcomes are psychological scarring, unemployment, collapse of international cooperation, and ongoing collapse of migration and immigration — and the belief that Trump's removal would undo these is wrong.

forecastcontestednovelty 2/4durability 1/4· Tyler Cowen

“just the psychological scarring, the unemployment, the collapse of international cooperation, ongoing collapse of migration and immigration. I worry about all those things, and I think those quite possibly are plausible outcomes. They're not like tail scenarios.”

0.35

The 'great reset' predicted in The Complacent Class has arrived because America can still borrow at low rates but has failed to show the flexibility to adopt mask-wearing or a serious test-and-trace regime, with the regulatory state, executive branch, and Congress all failing.

factualcontestednovelty 2/4durability 1/4· Tyler Cowen

“We can still borrow at low rates of interest, but we haven't shown the flexibility to adopt mask-wearing or set up a serious enough test and trace regime, and the American regulatory state and executive branch, and also Congress, it's failed.”

0.34

Cowen aims for each of his books to be 98% fresh material with no overlap among his blog, Bloomberg columns, and books, because readers and listeners are owed that the different parts of his output stand independently.

normativespeaker onlynovelty 2/4durability 3/4· Tyler Cowen

“When the book appears it will be 98% fresh material, as I try for all of my books to be.”

0.33

Despite absolute wealth in Parkersburg being far higher today than when its beautiful 1890-1910 homes were built, its architecture declined, suggesting the cause is not merely economic decline but a loss of cultural self-confidence and sense of centrality and importance.

causalspeaker onlynovelty 3/4durability 2/4· Tyler Cowen

“absolute wealth in Parkersburg is much higher today than when they built these wonderful homes in 1910, 1890... Why don't people give a damn? So there's some loss of cultural self-confidence, notion of centrality, ethos importance.”

0.30

Cowen estimates, semi-jokingly but semi-seriously, that humanity has another good 700 or 800 years left before civilization may not persist, primarily because of weapons of mass destruction.

forecastspeaker onlynovelty 3/4durability 2/4· Tyler Cowen

“I thought we had another good 700 or 800 years left in us.”

0.26

Learning returns follow a curve of high increasing returns from around age 14 through the late twenties, after which they decline somewhat but remain fairly high, eventually approaching an asymptote — so although one learns more between 15-25 than later, this reflects diminishing marginal returns rather than knowledge failing to compound.

factualspeaker onlynovelty 2/4durability 2/4· Tyler Cowen

“I think I started high returns at around age 14 and up through my late 20s, my learning returns were just very, very high per year. Since that time they've been somewhat lower, but I think still pretty high.”

0.18

Cowen does not see why Tesla should be worth more than General Motors, Ford, and several other auto companies combined, but defers to the market's judgment as a signal of dynamism.

factualcontestednovelty 1/4durability 0/4· Tyler Cowen

“I guess I ought to side with the market there though I don't quite see why it should be worth more than General Motors, Ford, and a few other auto companies put together.”