What this covers
Nassim Nicholas Taleb in conversation with Tyler Cowen ranges across history, politics, economics, and philosophy to argue that genuine legitimacy and robustness depend on exposure to real downside risk. The discussion touches geopolitics (why some governments appear stable yet prove fragile), expertise (why theory-only economists differ fundamentally from practitioners), social media's role in restoring natural human discourse, religious doctrine, and the education system. Throughout, Taleb builds from a single principle: systems and people insulated from consequences are not protected but weakened, and no one should hold power or public influence without staking something meaningful on their claims.
The conversation moves through surprising terrain. Taleb argues that volatile, frequently-changing governments like Italy are more robust than artificially stable ones like Syria or Saudi Arabia, because the latter sit on undetected fragility. He draws on Lebanese history and genetics to contest narratives of Arab identity, traces how Sufism was dismantled by foreign funding, and proposes that religion as law versus religion as faith creates a category error when comparing Islam, Judaism, and Christianity. On expertise, he argues the gap between feedback-based practitioners—pilots, dentists, dancers—and macroeconomists is unbridgeable. He defends Twitter and social media as Lindy-proof, discusses emergent behavior in self-driving cars, and applies inverse role models as a learning method. He also explores why theosis, the Orthodox concept of self-sacrifice, or why the Christian Trinity's insistence on Christ's genuine humanity and divinity, both hinge on the logic of real risk-taking and genuine skin in the game.
Taleb argues that risk-taking ('skin in the game') is the foundation of honor, legitimacy, and robustness across history, religion, and society, and that systems and people without exposure to downside (stable governments, theory-only experts, the over-protected wealthy) are fragile in ways conventional analysis misses.
- Stability that is propped up (Saudi Arabia, Syria) is more fragile than volatile systems (Italy, post-Soviet states) that have proven they can survive shocks
- Theory-only experts like macroeconomists lack the feedback that makes practitioners genuinely expert
- No one should rise in society or be a public intellectual without taking risk for others
Artificial stability and concentrated control breed fragility; volatility and diversity confer resilience.
- Post-Soviet states (Armenia, Georgia, Russia, Romania) all took severe hardship during a brutal transition phase yet improved through it, which proves their ability to survive future crises, whereas a country like France, never having been tested, may not be able to manage a comparable disruption.
“All of them — Armenia, Georgia, Russia, Romania — all of them, in fact, improved through that phase. Currently, it makes me believe that they could take another economic crisis, any crisis, and survive.”
- Countries with too much stability, especially when propped up, become weaker and more vulnerable to a catastrophic black swan; predicting country stability from past stability is backward, because the same-government-for-40-years countries (Syria, Saudi Arabia) proved fragile while the volatile, frequently-changing-government Italy remained standing — like companies with artificially compressed volatility that are sitting on dynamite.
“There is metrics in this town of predicting stability of countries based on past stability, which to me is absurd”
- A person with a very stable salary appears antifragile by day but is actually fragile, because after 30 years of employment and being laid off at 65 they can't do anything, whereas someone with a little volatile income is more adaptive and resilient to disruption.
“if, for example, someone has a very stable salary, he or she is employed as an antifragile by day. But then you’re laid off, at 65, after 30 years of employment, and basically can’t do anything, whereas someone who has a little volatile income is more adaptive.”
- Self-driving cars are a good idea only if someone in the car remains responsible and suable for accidents; the deeper risk is emergent — like flocks of birds, a system of self-driving cars interacting may behave differently from the sum of individual cars, and a single computer controlling all cars would work but would let someone take over the world.
“they don’t know how self-driving cars will react to other self-driving cars. The sum, it’s like flocks of birds. It behaves differently from the sum of birds, like markets behave differently from individuals.”
- Blockchain, like the centuries-old practice of letters of credit and the Phoenician trust-based trading model, is robust as long as transactions are not concentrated on one single model; concentration will eventually reveal its flaws (which exist but are not yet all known), at which point someone will introduce a counter-trick with its own different flaws — so resilience comes from diversity of mechanisms, not any single perfect one.
“So long as you don’t have a lot of transactions on one single model. Then you figure out if it doesn’t work, that someone will bring something with a counter — a trick that will have other flaws.”
Power and legitimacy require visible risk-taking and personal stake in outcomes.
- Trump's appeal in the primary debate, against opponents in suits, came from his exhibited risk-taking and visible 'scar' — having reportedly lost over a billion dollars of his own money — which acted like a sign of devastation from war that elevates a person above mere 'pencil-pushers' by demonstrating he is real and has skin in the game.
“is there anything more human than showing a scar, to say, ‘At least I’m real, and you guys are just like pencil-pushers,’”
- One should never let anyone rise in society unless he or she is taking risks for the sake of others, and one should never be a public intellectual unless every statement entails risk-taking; inequality is acceptable only if the unequal person keeps taking risks, but a frozen upper class that uses government to prop itself up is illegitimate.
“you should never have someone rise in society if he or she is not taking risks for the sake of others, period. That’s one rule.”
- Theosis is the Orthodox idea that humans, accepting their imperfection, can rise above their condition through duties, self-sacrifice, and risk-taking — an equal-opportunity openness to anyone — meaning life's purpose is to sacrifice and do something meaningful rather than to accumulate money or pursue pleasure.
“theosis is a way for us humans to rise above our condition as human, and it’s given to us openness, this equal opportunity for anyone.”
- As people get rich they become controlled by external preferences rather than their own, and sophistication (like a three-star Michelin meal that costs $200 and is anxious work to eat) is actually a burden even from a pure pleasure standpoint; people are happier in small quarters and would rather eat good bread with someone than dine in a fancy restaurant — so reclaiming authentic preferences is itself a form of skin in the game.
“as people get rich, they get controlled by the preferences, they’re controlled by the outside.”
Religion and law are distinct categories; Christianity separated sacred from temporal while Islam and Judaism remain unified law systems.
- Islam and Judaism are fundamentally bodies of law (the word 'din' means law) with no distinction between holy and profane, whereas Christianity is uniquely a secular religion that introduced separation of church and state — because the Romans already had sophisticated law, Christianity (render unto Caesar) left law to the state and confined itself to the private sphere; therefore comparing religions naively is a category error, and some law-systems like Salafism are political systems rather than religions.
“Judaism and Islam are not religions like Christianity is a religion, the exact opposite.”
- The Christian Trinity, which seems absurd to non-Christians, is a necessary doctrine because Christ had to be sufficiently man to have genuine skin in the game (to truly suffer and take risk) and sufficiently God to be God — like a tightrope walker without a parachute, the risk must be real for it to count, which is why the Trinity doctrine kept reasserting itself.
“If I go to the circus and I have a fellow walking on a tightrope with a parachute, I ask for my money back.”
- The religious schisms of the Levant (Maronites, Nestorians, Orthodox) were driven not by theology but by ethnicity and language, mapping onto Aramaic speakers versus Greek speakers, analogous to how English Protestant versus Irish Catholic division was really ethnic; if the English had been Catholic, the Irish would have been Protestant.
“If you see the religion schisms, they map to Aramaic speakers versus Greek speakers.”
- A peaceful branch of Islam (Sufism) that even permitted drinking, exemplified by places like Tripoli in Lebanon, was quietly destroyed by Saudi Arabian funding, which shows how a religion can be destroyed without anyone noticing simply by indoctrinating two generations.
“The problem, that it was destroyed by Saudi Arabian funding. What can we learn from them is how your religion can be destroyed without anybody noticing.”
Local genetic and economic realities, not ideology or religion, determine group identity and behavior.
- Despite their Syriac (Eastern Aramaic) liturgical language suggesting an Eastern migrant origin, the Maronites are genetically as local as possible and are descendants of the Phoenicians, as are the Shiites of Lebanon; only the religion migrated, not the population.
“it turned out that, in fact, it’s only the religion. The DNA is as local as it can be. The Maronites are the descendants of the Phoenicians.”
- Lebanon's Arab identity is a 150-year-old distortion: until the 1860s Lebanese Christians wrote in garshouni, not Arabic, and the Syrian Protestant College's Arabic Bible translation began an Arabization that pulled the region's identity away from its 1,500-year Greco-Roman/eastern-Mediterranean heritage; this Arabization ultimately had no effect on the actual local population, which is genetically ancient and unchanged for ~3,700 years.
“In 1860, the Christians of Lebanon did not write in Arabic. They wrote in a character called garshounè.”
- Phoenician city-states (and small states like Singapore) are robust against conquest because conquest yields nothing — there is no valuable land to take — so the rational arrangement is for the would-be conqueror to accept a small percentage (a tribute/patronage/vassal model) rather than 100 percent of nothing; integration historically came through commerce, not military conquest.
“Listen, what do you want? You kill us all, you get nothing. Land is not interesting. What are you going to get? We’ll give you 5 percent. What do you want, 5 percent of something or 100 percent of nothing?”
- Civil war in Lebanon today is unlikely despite recurring fears because everyone is heavily invested in real estate and therefore has skin in the game and no interest in waging war, unlike 1975 when factions had nothing to lose; groups with nothing to lose are the ones who can plunge into civil war.
“they don’t realize that everyone is up to here in real estate. Nobody has any interest in waging war.”
Practice-grounded expertise with real feedback beats theoretical knowledge lacking accountability.
- There are two classes of experts: those whose work has feedback (belly dancers, dentists, carpenters, pilots, car-radio thieves) who are genuinely expert, and those like macroeconomists and many policymakers who, lacking real feedback, may not know anything about anything — so equating a macroeconomist with a pilot is a category error.
“There are two classes of experts. Belly dancers are experts at belly dancing... Dentists are experts at dentistry. I’m not sure macroeconomists know anything about anything.”
- A productive method is to identify inverse mentors and inverse role models — people who are doing something wrong or whom you don't want to resemble — and then counter-imitate them; for example, observing a kind colleague who failed because he always got caught in one set of details taught the lesson that you cannot get into more than one set of details.
“I know how to find inverse mentors. People — you know they’re doing something wrong, and you figure out what makes them do something wrong.”
- Linguists are not fluent in the languages they categorize, which produces wrong categorizations; their metrics fail to recognize that Levantine is close to Aramaic, whereas using a statistical method or actual practice reveals the connection — reflecting a broader divide between people who come from practice and people who come from theory, with no third category having both.
“linguists are not fluent in the languages they’re categorizing, and it makes a huge difference.”
Modern technological disruption and external forces can overwhelm previously predictable local equilibria.
- The 1975 Lebanese Civil War may have been predictable as an event, but not on the scale it reached, because nobody understood the destructive effect of modern weapons (prior conflicts were local and died down quickly) and the influx of armed Palestinians/PLO after 1973 disrupted the internal balance from the outside, triggering the Christian reaction.
“It might have been predictable, but not on that scale for several reasons. The first one is that nobody understood the effect of modern weapons because previous conflicts were more local, and more confined.”
- Social media is Lindy and good because it restores the natural human condition in which people are simultaneously recipients and purveyors of news (like the Athenian barbershop, fish market, or funeral) — the postwar nuclear-family arrangement of passively receiving TV news without transmitting was a historical anomaly; consequently an author no longer needs the New York Times to succeed, as Taleb's book hit #12 on the bestseller list with no US reviews, relying on Twitter and Facebook.
“it opened number 12 on best-seller list. It tells you that you don’t need the New York Times to exist as an author, that Twitter is sufficient”
- Judging a dictator like Assad requires looking at both sides of the scissors — in a civil war between two groups, the least-bad faction looks good by comparison, so one-sided naive condemnation ('this guy is an asshole') is analytically incomplete; one must compare relative options, which is why many Iraqis might prefer Saddam back and Taleb hates Assad's family yet hates the jihadis more.
“You got to look at both sides of the scissors. That when you have civil war you have two groups fighting, so you take the least asshole becomes someone good in your eyes, but you’re only analyzing one portion.”
Self-directed learning and authentic personal preference matter more than credential acquisition.
- The best educational advice is to get a degree but become an autodidact: read voraciously by yourself, including books that will be useful 20 years from now rather than the latest bestseller, and aim for the lowest possible passing grades, because you only remember what you read on your own, not what was assigned; avoid studying subjects that get revised like history and geography, since math is probably the only school subject worth picking up.
“Get a degree from school, but become an autodidact. Don’t waste time trying to get an A because you’re not going to remember all that shit.”
- A formative lesson from Dostoyevsky's autobiographical mock-execution scene in The Idiot is that the resolve to savor every minute of life, formed under the intensity of imminent death, could not be sustained afterward once survival was assured — the man forgot the episode and reverted, illustrating the fleeting nature of preferences formed in the moment.
“The lesson was he no longer did that. It was about the preferences of the moment. He couldn’t carry on later.”
- The best dark chocolate is not the fancy, labeled artisanal products made in Brooklyn.
“the best way to find and consume is not those labels, the fancy things made in Brooklyn.”
- Cryptocurrencies are valuable primarily because they force the Federal Reserve to recognize it does not have a monopoly over people's lives, and central banks are genuinely scared of this challenge.
“one thing for sure that’s working with cryptocurrencies is that it’s forcing those fucks here in Federal Reserve to understand that they don’t have a monopoly over our lives.”