What this covers

Tyler Cowen and Russ Roberts discuss how organizations discover and develop talent, centering on Cowen's argument that the process resembles craft more than science. The conversation ranges across hiring practices, the economics of talent development, and the kinds of human qualities that predict long-term creative impact. Rather than credentials or measured intelligence, Cowen emphasizes that stamina—sustained, disciplined output over decades—and demonstrated drive emerge as far stronger predictors of transformative contribution. He contends that structured interviews suit routine hiring, where conscientiousness screens matter, but that identifying people for creative or senior roles requires unstructured conversation that can surface the ambition and determination that paper qualifications obscure.

The conversation moves through several contentious territories. Cowen argues that vast pools of undiscovered talent exist—particularly in India—because markets systematically under-supply investment in talent development; the externalities mean that employers who develop talent rarely capture the returns. He pushes back against the assumption that articulate, credentialed candidates are the best bets, suggesting instead that intellectual arbitrage—recognizing value in unlikely places and people—yields the highest gains. The speakers also examine how soft networks and belief-signaling shape trajectories, why deliberate practice must remain engaging rather than punishing to sustain, and whether the college degree requirement for ordinary work has become a costly barrier. Throughout, Cowen distinguishes functional meritocracy from moral claims about human worth, addressing a persistent source of friction in talent conversations.

Sharpest takeaway

Cowen argues that identifying transformative talent is more an art than a science, requiring unstructured interviews that probe for drive, stamina, and self-improvement practice rather than credentials or raw intelligence, and that vast pools of undiscovered talent exist that markets systematically under-supply.

  • For high-level creative positions, unstructured interviews capture signals that structured processes miss
  • Stamina and practice predict long-term success better than smarts, grit, or credentials
  • The geographic distribution of talent is uneven and shifting, leaving talent (e.g. in India) undiscovered

The argument · threads10 threads · 43 claims
0.85

Talent assessment should prioritize unstructured interpersonal signals over credentials or homogenized structured interviews for creative roles.

6 pointscentrality 5/5
  • For creative and high-level positions, the relevant data arrives in unstructured form whether you like it or not—e.g. a newspaper deciding on a columnist asks around among people who have interacted with the candidate—so the real question is not structured versus unstructured but how to interpret the unstructured signals you inevitably receive.

    “for a lot of creative or higher level positions, the actual interview process is unstructured, whether you like it or not. So, say the New York Times wants to hire a new columnist. They will ask around about that potential person... Those are in a sense unstructured interviews.”

  • For most ordinary jobs, you should rely on structured, homogenized interviews with centralized data collection and screen primarily for conscientiousness, which is a fundamentally different approach from the unstructured, idiosyncratic process appropriate for creative or high-level positions.

    “You need to rely then, much, much more on structured interviews, which are more or less homogenized... And also, in general for those candidates, conscientiousness tends to matter more than other qualities... most jobs are of a fairly ordinary sort, by definition. Right? Then look for conscientiousness, usually; have a structured interview process; and apply a lot of discipline.”

  • When assessing applicants in person, one should look for drive and determination rather than articulateness, because articulate people tend to succeed regardless, so a smart but inarticulate candidate is often more interesting as an underexploited opportunity.

    “When I speak to people, I don't look so much for articulateness. I look more for drive and determination. Someone who seems smart, but is not that articulate, often gets me more interested, right? Because, I know the articulate people often make their way in any case.”

  • Contrary to mainstream media skepticism, the actual evidence shows interviews do matter even for ordinary jobs and matter even more for higher-level ones; for example, simply asking an economist candidate two or three substantive questions about economics—rather than only asking them to describe their job market paper—is a highly underrated practice.

    “the actual evidence--if you get past some of the mainstream media pieces--the evidence is that interviews do matter... the simple notion that you sit down with someone and ask them some questions about economics is going to get you a long ways. It's not overrated. I would even say it's underrated.”

  • Obscuring the direct visual connection between two people—as in Catholic confession or a therapist sitting behind a reclining patient—can lead to a more open conversation, which is why Zoom and online interviews are not merely worse versions of face-to-face but sometimes superior because people will open up more.

    “In classic Catholic confession, you don't see the priest at all... it is at least believed, based on centuries--millennia--of evidence, that this leads to a more open conversation because you don't see the other person... sometimes people will open up more.”

  • Zoom interactions are on average more egalitarian than in-person ones because the boss can't take the best seat, command attention, wave hands, or hog speaking time; turns become discrete, so one should exploit Zoom to be more egalitarian, while also recognizing it makes humor harder—meaning candidates should not be judged on their sense of humor online.

    “In person, the boss usually takes the best seat... On Zoom, it's much harder to do all of those. It's, on average, more egalitarian. There are discreet turns of when you stop and start talking. You can't manipulate the room in the same way.”

0.80

Daily deliberate practice habits and sustained determination matter more than raw intelligence or articulateness in identifying capable people.

5 pointscentrality 5/5
  • Intellectual breadth is currently undervalued and synthetic thinkers are important for starting projects, but breadth alone is insufficient—broad people who don't work hard at being really good at a particular thing don't stand out or succeed, which is why the book emphasizes how hard and how well a person practices.

    “I think breadth is currently undervalued, and synthetic thinkers are important for starting projects, but a lot of the questions Daniel and I propose are geared toward: How much does a person practice?”

  • A key interview question is asking what a person does to practice on an almost everyday basis—analogous to a concert pianist practicing scales or a basketball player practicing free throws—because a candidate with no answer at all is a worrying sign of no program for self-improvement.

    “What is it that you do to practice on an almost everyday basis, that is akin to how a classical concert pianist would practice scales? ... The person who has no answer at all, that tends to worry me, frankly.”

  • Intelligence is often overrated—especially by smart people themselves, though not necessarily by the world at large—so curiosity and determination matter more than raw smarts once someone is smart enough to do the thing.

    “People have to be smart enough to do something. But as we argue in the book, I think smarts are often overrated. At least they're overrated by other smart people. I wouldn't say they're overrated by the world as a whole.”

  • Effective practice ought to be fun, because if it is pure torture a person will eventually give it up; therefore one should assess how enthusiastic a candidate is about their own practice routine.

    “Practice ought to be fun, if you're going to excel. Because if it's just a torture, at some point, you're going to give it up. Right? So, you want to see, how enthusiastic is the person about their own practice routine?”

  • Stamina—sustained productive output over decades without letting up—is distinct from grit, which is gritting your teeth through tough times; in a knowledge economy the returns to stamina are extremely high, especially if started young, so for intellectual life one should generally prefer a person with stamina over one with grit, though grit may matter more in age-limited pursuits like bicycle racing.

    “I think in a knowledge economy, the returns to stamina are extremely high. Especially, if you can start fairly young... I think of grit is, kind of times are tough and you grit your teeth and you bear down... if you distinguish stamina and grit, I think you can make better decisions for the intellectual life. I think overall, I would prefer a person with stamina than a person with grit.”

0.75

Bureaucratic processes fail to identify transformative talent; lean applications and minimal monitoring outperform traditional gatekeeping.

4 pointscentrality 5/5
  • A talent-funding process can succeed by minimizing bureaucracy—one short application, no CVs or letters of recommendation, only one layer of 'no'—because it allows betting on ambitious people with new ideas without selecting only for those who excel at milestone-driven monitoring.

    “The goal is to find new talent and fund it in a very non-bureaucratic way. So, we have only one layer of no in the system. Our application is about a page. We don't ask for CVs or letters of recommendation; and we're willing to take chances on ambitious people with new ideas.”

  • The distribution of funding proposals is more bimodal than expected: most rejected proposals are perfectly competent but simply not transformative, so the decision is less about screening out incompetence than about identifying the rare proposals that could make a major difference.

    “the distribution of proposals is perhaps more bimodal than you might think. So, a lot of the nos seem perhaps perfectly fine, but they don't seem transformative.”

  • Despite minimal monitoring, a non-bureaucratic grant program has a better-than-average track record on malfeasance, because the bureaucratic status quo—foundations with large staffs giving academics money for books that never get written—loses money just as readily.

    “I see so many cases under the status quo, where foundations with huge staffs, say, give academics money to write books. And, the books are never written, and the money disappears. So, this is hardly something new. I actually think, our record so far is much better average.”

  • Deliberately not advertising a free-money program and keeping it relatively secret acts as a valuable selection filter, because having to hear about it and sensing you ought to apply is itself a meaningful upfront screen—so wider advertising would actually weaken the quality of applicants.

    “we don't advertise. It's free money... you have to hear about it. And, if you've even heard about it and have a sense you ought to apply, that's already, up front, I think a great selection filter.”

0.61

Multiple strong signals and high standards are needed to avoid early affection bias and ensure candidates can truly move ventures forward.

4 pointscentrality 4/5
  • To avoid the trap of stopping evaluation once you find someone you like, you should hold candidates to a higher standard by requiring multiple signals that they are strong and ambitious enough to push a venture over the top—so 'liking isn't enough' even when a candidate is clearly smart and credentialed.

    “You want to receive a bunch of different signals, that they are strong enough and ambitious enough to push the thing over the top. So, once you get into the mindset of liking isn't enough, you're not just trying to fill a position”

  • For receptive individuals placed in a peer group of similarly talented others, an award and presentational charisma matter greatly because, as venture capitalists report, the money is often not the main thing—what unlocks a person is seeing that someone else believes in them.

    “venture capitalists will tell you the same. They say in a lot of cases, the money is not the main thing. It's the person, all of a sudden, seeing someone else believes in them. And, now they can get this thing done.”

  • Remarkably few interviewees lie about their goals—people who merely want to publish papers and get tenure usually say so, because they lack the mental vocabulary to concoct a convincing alternative—and those who do fabricate collapse quickly when pressed on details.

    “I find it remarkable, first of all, how few people lie... they don't have the mental vocabulary or frameworks to even concoct a good lie... if you think they're lying, just push them on the details and see how well they know the thing they're claiming they want to do. And, typically people who are making it up will collapse pretty quickly.”

  • Good interview questions are designed to induce variance in answers rather than to elicit a correct answer, which means many answers will be bad; the interviewer cares not about the answer's correctness but about how the candidate defends it, so feeling your answer is bad may actually be a sign it is good.

    “The point is to induce variance in the answers, which will mean a lot of the answers are bad. So, if you feel your answer's bad, that's actually a sign your answer might be good, because the question is a variance-inducing one”

0.61

Credentials are expensive informational shortcuts; greatest gains come from intellectual arbitrage recognizing overlooked talent in unexpected places.

4 pointscentrality 4/5
  • Requiring college degrees for many ordinary jobs makes no sense, acts as a barrier to minorities and to women who had children at the wrong time or left school to raise families, and is one of the worst things American society does—a path like Cowen's father's success without a degree is no longer possible today.

    “There are so many just ordinary jobs where you have to have a college degree, and it makes no sense... it's a barrier to minorities, that can be a barrier to women who had children earlier... And, it's one of the worst things we do in American society.”

  • Credentials are a useful but expensive informational shortcut—the top Harvard Law graduate will do well anyway and will cost a lot to hire—so the greatest gains come from intellectual arbitrage: looking for talent in other kinds of people and other countries where their value is not yet recognized.

    “if you want to do some kind of, what I would call intellectual arbitrage, you need to look other places. Because the person who comes out top of their class at Harvard Law, they're doing well anyway... you will make a lot of your greatest gains looking for other kinds of people. And including from other countries.”

  • The geographic distribution of talent is far from even but moves around over time—early 20th-century Hungarian high schools excelled at math and science, the Florentine Renaissance and Germanic classical music had their moments—and right now the place with the most undiscovered talent is India, which is not a permanent or intrinsic state of affairs.

    “the geographic distribution of talent is far from even, but it moves around... right now, I think the place with the most undiscovered talent by far is India. That's why we have Emergent Ventures, India. I don't think that's a permanent state of affairs.”

  • Soft networks—people who know of you or follow your work but aren't close friends—become increasingly important as the talent world grows, because the best strategy is to build your soft network so that talent searches for you, a decentralized Hayekian solution that inverts the usual search-for-talent framing.

    “the more you rely on your soft networks to bring you talent. So, the book says, yes, you're looking for talent, but the best way to succeed at that problem as Daniel and I stress, is to have talent looking for you... That's a decentralized Hayekian solution.”

0.57

Meritocratic selection is functional necessity distinct from moral claims; excellence, stamina, and high standards deserve cultural reaffirmation.

6 pointscentrality 3/5
  • Everyone believes in meritocracy at some level even if they rhetorically reject it—virtually no one wants a non-meritocratic surgeon for their open-heart surgery—and supporting a talented person does not commit you to claiming they are a morally better human being, distinguishing functional meritocracy from moral meritocracy.

    “everyone believes in meritocracy at some level. The question is, what kind of rhetorical cloaking do they put around what they do? So, if you're having open heart surgery, there are very few people who do not believe in meritocracy.”

  • The social value of staying power and stamina is extremely high, even though it is not necessarily morally better for every individual—reflecting the gap between functional and moral meritocracy—so the culture could use more people unabashedly standing for excellence, hard work, and determination.

    “The social value of staying power and stamina is extremely high. So, is it morally better for every single individual? I don't think it is. This gets to the difference between moral meritocracy and meritocracy. But, I think in our culture, we could use for some number of additional people, taking a stand for excellence and hard work”

  • There is something brutal about the book's approach to rejection: instead of an impersonal reason like lacking a Harvard degree, the implied message is 'we thought you as an individual weren't good enough for this job,' a more painful basis for rejection that may not always be made transparent in society.

    “We're saying, 'We didn't hire you because, we thought you as an individual, weren't good enough for this job.' And, that is a more brutal reason for rejection. And, maybe it can't always be made transparent in society.”

  • Given Germany's intelligence levels and cultural sophistication, its underwhelming output of significant public intellectuals in the Anglo world reflects ideas that are too verbose and complicated and that aren't honed in rough-and-tumble back-and-forth—a cultural complacency partly visible in slow email norms that would be unacceptable in Silicon Valley.

    “how many significant public intellectuals in the Anglo world are German? To me, it's a very much an underwhelming performance. I think they're too verbose, and their ideas are too complicated. And, they're not really out there in the rough and tumble, back and forth, honing them”

  • Speed of response to emails is a significant variable in assessing people: for startup founders and potential public intellectuals it is a strongly positive signal of prioritization, clean inboxes, and eagerness to act, whereas for a lab-bound chemist or cancer researcher fast responsiveness could well be a negative signal.

    “If people respond to your emails quickly, it tells you something about them... For people doing startups or potential public intellectuals, that's a strongly positive signal. If you're looking to support or hire, say a chemist or cancer researcher who will be spending a lot of time working in a lab... it could well be a negative signal.”

  • Early exposure can change a young person's lifetime trajectory ('slope') by letting them rub shoulders with the best in their field, which convinces them that becoming such a person is achievable—a non-one-off gain that funding trips for young people can produce.

    “You can change their slope, especially early in their lives. I'm a big believer in funding trips for young people, just so they can, in some way, interact with the best in their field... you see that a person can become that... it's changing an entire lifetime trajectory.”

0.51

Talent means creative spark producing meaningful ideas, distinct from technical competence or ability to execute routine tasks well.

2 pointscentrality 4/5
  • Talent, in the sense the book addresses, means people with a creative spark who can produce ideas that make a difference—as distinct from the ability to perform a rote task 10,000 times well, which is valuable but a different quality entirely.

    “People with a creative spark who can produce ideas that will make a difference. So, we're talking about a particular kind of talent. Not the ability to perform a rote task 10,000 times in a row and do it well.”

  • The ability to write a novel that does well signals that a person understands people, communication, and organizational culture, which is why an unusual bet on a two-time novelist for corporate-culture work can be justified—the goal being to find non-obvious bets rather than slam dunks.

    “If you can write a novel that does well, it shows you understand something about people. You understand something about communication. You must have a certain kind of breadth... and that's a sign you understand culture of organizations.”

0.39

Selection inherently involves discernment and discrimination; understanding oneself—one's own talent, practice, networks—is as important as assessing others.

5 pointscentrality 2/5
  • Instead of asking the standard contrarian question 'What do you believe that no one else does?', a more useful values-based screening question is 'What's the consensus view that you agree with?'—because it forces candidates to locate themselves within the space of existing ideas.

    “Instead of asking, 'What's the thing you believe that no one else believes?' It says, 'What's the consensus view that you agree with?' So, it's saying, 'Place yourself in the space of things that exist.' I find that very useful.”

  • Though framed as searching for others' talent, the book is equally about understanding your own talent—asking 'Who am I, what can I do?' and even 'What are my open browser tabs?'—and its principles extend to making friends and finding a romantic partner.

    “It's about how to understand your own talent. So, the framing is search, searching for others. But, how about searching for yourself? Like, 'Who am I, what can I do?'”

  • Selection inherently involves discrimination in the older sense of showing discernment or good taste—everyone does it daily, choosing only a few restaurants over the many they pass over—so any process where something of value is at stake produces more nos than yeses, making such discernment the essence of a finite life.

    “Any process where there's something of value, there are more nos than yeses, right?”

  • Emergent Ventures was the first supporter of Recidiviz, a nonprofit that gathers data on which prisoners can be released early without reoffending, and that data proved especially important during COVID when prisons sought to release people to make conditions manageable.

    “there's a group called Recidiviz. We were their first supporter... what they do is gather data on which prisoners can be released early, without committing crime again. This turned out to be very important during COVID times”

  • Emergent Ventures' very first grant went to American Ukrainian economist Tymofiy Mylovanov for popular economics writing in Ukrainian, and he later became Ukraine's Economy Minister under Zelensky—an early validation of the program's picks.

    “Our very first grant was to a American Ukrainian economist called Timothy Malave, who wanted to do some popular writing about economics in Ukrainian. And, he ended up becoming the Economy Minister of Ukraine under Zelensky.”

0.33

Founder skills like speaking, audience connection, and responsiveness are teachable through deliberate practice like comedy or podcasting.

1 pointcentrality 3/5
  • Many seemingly intangible founder skills are in fact practicable—addressing a team, connecting with an audience, being quicker and more responsive to questions—and people deliberately train these through stand-up comedy, taping themselves, and going on podcasts.

    “I know people who go to stand up comedy routines to just practice connecting with an audience. I know people who tape themselves. I know people who go on podcasts to learn the art of being quicker or be more responsive when questions are asked.”

0.24

Talent investment is under-supplied due to market failure; elevating talent produces gains that accrue elsewhere, reducing private incentive to develop it.

2 pointscentrality 2/5
  • Investing in talent is under-supplied because of a market failure: if you elevate a talented person, you don't internalize the gains—they or another employer reap them—so absent altruism or some other reason, there is too little private incentive to develop hidden talent.

    “There is a potential market failure. So, if you spot a talented person and help them, you may just elevate them so they go away and work for someone else... you need either altruism or some other reason to want to help them.”

  • Good interviewing is a form of conversation that, in an Adam Smithian sense, requires putting yourself in others' shoes and having sympathy for their perspectives, which demands understanding them—sensitizing you to be genuinely inquisitive about the people in your life, not just your employees.

    “It's an Adam Smithian book in this regard, influenced by your own work on Smith, about putting yourself into the shoes of others, right? Having more sympathy or empathy for their perspectives. But, to do that, you need to understand them somewhat.”