Some questions about great talent curation


Tony Kulesa recently wrote an interesting blog post explaining why Tyler Cowen is the best curator of talent in the world. It focuses on Tyler’s Emergent Ventures program which funds promising individuals early in their career.

Kulesa’s post raises some fascinating questions about great talent curation. Is it scalable? What kind of impact does it have? What does it reveals about our education system? Does it require deception? I address them here.

What does a Emergent Ventures winner win?

Emergent Ventures is not a prize, since it’s awarded to people at the beginning of their careers. According to Kulesa, it’s not really a grant either, since the amount of money awarded is relatively small. Kulesa also argues that it’s not a good credential, because most recruiters don’t (yet) know that Tyler Cowen is a great talent curator. So what precisely do you get when you become an Emergent Ventures winner?

Kulesa argues that Tyler uses EV to raise the winner’s ambition and give her a vision of what she could be. In this sense, it may benefit Tyler to have more people know about his talent curation abilities. The more the applicant believes that Tyler is great at picking talent, the greater her self-confidence and ambition will increase as a result of winning.

But is raising people’s ambition a free lunch? Applicants come in with some implicit prior about their own competence and potential (which will be very plastic because of their age). If an applicant wins, she finds out that Tyler thinks she is promising, and her confidence increases. But if she is rejected, she finds out that Tyler does not think she has much potential, and her confidence decreases. This may be net positive, because it is much more important to raise the ambitions of the very best people, but it does come at a cost. Under this model, Emergent Ventures works by transferring ambition from the simply great to the truly excellent. Average is over and Tyler killed it.

If this model is correct, one worry is that the young people who get picked out by this program are precisely the ones who didn’t need the help. The impact of philanthropy must be evaluated in comparison to the counterfactual in which it is not deployed. It is hard to imagine that a 17 year old building a fusion reactor in her garage or developing a new macroeconomic model for growth and sustainability isn’t already bound for greatness.

But we might be underestimating the extent to which mimetic forces will drive these kids towards low-volatility competitive tracks. Find a 30 year old who is on a mediocre career trajectory. If you ask him to tell you about the coolest thing he did before he turned 20, you often hear of some really interesting projects. A young writer who started a popular amatuer blog on ancient Rome in college is now churning out low value academic sludge. The coding prodigy who was a white hat hacker in high school is now nudging a few pixels here and there for Google.

Maybe the whole purpose of Emergent Ventures is to avoid this outcome. There is a crucial interval during which a person is old enough to show that he can do something great, but too young to have decided that he will. During this period, someone he really respects can drastically increase his ambition. This person can give to young adults a boon which Alex Dao claims VCs give to founders:

VCs don’t just give founders money, advice, and introductions. They give founders something powerful, and almost mystical: they bestow on founders a type of blessing. “You are the founder. You stand apart. Now go make the future real.”

Of course, this is a moot point if the main benefit of becoming an Emergent Ventures winner is the introduction to a talent-dense network - one which would not have existed without Tyler. The history of rapid progress is largely a catalog of legendary networks. Perhaps we will remember Emergent Ventures as one more item in this inventory which includes the Lunar Society of Birmingham and Building 1 at Bell Labs.

Is the signaling equilibrium for education shifting?

What does the success of Emergent Ventures say about the world? You can view EV as an indirect rebuttal to Bryan Caplan’s book The Case Against Education. Bryan’s claim that education is mostly about signaling relies on the premise that identifying talent is extremely hard. It is so hard that in the current equilibrium, employers do it by checking who is willing to burn four years doing mostly pointless cognitively demanding work. Depending on your perspective, you can either see this as an indictment of the current education system - it takes you four years to test what Tyler can learn in 1500 words? - or a vindication - if talent can be identified this efficiently, maybe higher education really is about education.

Emergent Ventures proves that with minimal signaling costs, it is possible to identify super right tail “teenager researching new biomaterials” talent. However, college may still be the best way to identify +1 standard deviation “someone who can just write me a goddamn report” talent. Tyler’s success at finding interesting contrarians matters little to most employers, who need conformity and are depending on colleges to test for it.

Conformity is harder to test for than weirdness, for the same reason that proving a negative is harder than proving a positive. All you need to do to check if someone is an interesting rebel is see if he can do something interesting and rebellious. And if he can’t do it within the 1500 word limit of the Emergent Ventures application, then he can’t do it at all.

Signalling conformity requires more work, because even the rebel can bide his time in a crew cut and stuffy suit for a while. You have to wait four years (and possibly more) to make sure you’ve filtered out most of the troublesome elements.

So whether programs like Emergent Ventures or the Thiel Fellowship will allow more and more young people to escape the academic gauntlet depends in part on whether signalling conformity will become more or less important in the future. Tyler thinks the former is more likely in Average is Over:

If you’re a young male hothead who just can’t follow orders, and you have your own ideas about how everything should be done, you’re probably going to have an ever-tougher time in the labor markets of the future. There won’t be much room for a “rebel without a cause” or, for that matter, a rebel with a cause …

Let’s draw up a simple list of some important characteristics in technologically advanced modern workplaces:

1. Exactness of execution becomes more important relative to an accumulated mass of brute force.

2. Consistent coordination over time is a significant advantage.

3. Morale is extremely important to motivate production and cooperation.

The new equilibrium for education may be one where most people have to receive more and more years of education in order to signal conformity and conscientiousness. However, the very best people, from whom innovation and creativity are expected, will be identified early, and they will be given enough funding and mentorship to get a head start on their careers. Once again, average is over.

Is talent curation scalable?

Kulesa writes:

Besides on Tyler’s blog, the winners are not promoted and there is zero resulting media attention. It remains obscure - and in fact, I suspect that if it becomes too well known Tyler will end the program - so it will never be a useful credential … Emergent Ventures appears to be designed against anyone looking for credentials, large amounts of cash, or status/attention. Instead, the program selects and rewards earnestness.

Is any sufficiently good talent curator bound to suffer from a version of Goodhart’s law? People notice how good he is at finding talent, turn his approval into a credential, and thus destroy his ability to exclusively attract earnest people with certain specific interests.

If this were the case, it would partly explain why an academic economist is the best curator of talent in the world, and not the corporations or venture capitalists spend billions on talent identification and recruitment.

In the beginning, YC was an obscure program where hackers who wanted to solve a problem could get some funding and advice. Now, it’s a well known credential you can use to later get a job at a tech company or venture firm. This newfound status hasn’t exactly helped YC filter out disingenuous or unserious people.

Talent curators face an entirely different set of incentives from most other kinds of traders. A successful hedge fund manager will usually want people to know that he has been beating the market. Status and success are self reinforcing. By being better known, he will have an easier time raising capital or getting the market to move in the direction he wants (notice what happens to a stock right after Warren Buffett gives his opinion on it). As a talent curator, however, you do not want the other fisherman to find your pond, nor the piranhas to bite on your line.

So then how can you scale your program?

  1. Dilution: You expand your program, knowing that the marginal admit will be lower quality than before. You simply want to take more shots on goal, and if that means shooting from an awkward angle with two defenders in front of you, so be it. This is the YC option.
  2. Inheritance: Your successful grantees create their own grant programs. They will have better access to new niches of talent, and they will (at least initially) be left alone by the overly credentialist applicants who keep knocking on your door. Think of Thiel Fellowship winner Vitalik Buterian funding the Ethereum Foundations’ grants (though maybe this lineage has progressed a generation or two beyond this point).
  3. Inspiration: Hope that by the time your program becomes too famous to remain a good filter, and you have to shut down shop, other curators will have noticed your methods and co-opted them. At some point, the curators who successfully replicate your methods can do the same. As Tyler writes, “Imagine a philanthropic world where, next year, you could give a million dollars to the Steven Pinker pop-up, to the Jhumpa Lahiri pop-up, to the Jordan Peterson intellectual venture fund, and so on. Three years later, you would have an entirely different choice, say intellectual venture funds from Ezra Klein, David Brooks, and Skip Gates”

Is Tyler hiding something?

While Kulesa’s piece overall is great and insightful, I found the section on how Tyler selects the best applicants to be lacking. Kulesa ascribes Tyler’s great taste for talent to his ability to “crack cultural codes,” one which he developed by studying different styles of art, music, cuisine, literature, and so on.

I want to contrast the kinds of advice Tyler gives when it comes to unearthing talent to the advice he gives for identifying good food.

When it comes to finding great meals, his advice is rational, straightforward, and actionable. He frames his analysis in terms of land, labor, and capital. Go where the raw ingredients are. Avoid locations with high fixed costs. Take advantage of cross subsidies. Consult with locals who are knowledgeable about the city’s market (for details, read Chapter 4 of his book An Economist Gets Lunch).

When it comes to evaluating talent, (Kulesa’s compilation of) Tyler’s advice is vague, confusing, and indirect. Does one simply start reading Harold Bloom’s Western Canon and hope to come out the other end as a great talent curator1?

Two qualifiers. First, Tyler has not yet published his book on identifying talent, so it’s no surprise that we don’t yet have his book quality analysis on the topic. It’s entirely likely that his upcoming book will have a few chapters of, “Let’s roll up our sleeves and analyze the factors of production here.”

Second, we shouldn’t just dismiss the idea that maybe Tyler really did learn how to identify talent from studying the humanities. The people who call themselves practitioners instead of theorists will be skeptical of this claim. But if they really took praxis seriously, they would learn from Tyler’s success and read the Iliad or book a trip to Senegal.

Here’s a conspiratorial view which I don’t really believe but do find amusing. Tyler benefits from an increase in the number of knowledgeable foodies who raise the demand for good meals. But great talent curators are his competition, and overly credentialist applicants who can optimize for any specific heuristics he articulates are his adversary. So he keeps true methods secret while spreading vague and misleading advice which has the effect of tripping up the credentialist applicants worrying too much about passing the art test or something.

Tyler claims that he wants to keep Emergent Venture low profile to preserve the quality of the applicant pool. But he hasn’t exactly been shy about advertising the program on popular platforms like Tim Ferriss’s podcast.

What is Tyler up to? The Straussian explanation is that he provides the wrong pronunciation of Shiboleth on public platforms. He communicates the actual qualities he is interested in to previous winners who are integrated into talent dense social scenes. Here, they tell the exceptionally talented exactly what Tyler is looking for.

I don’t actually believe that Tyler is trying to mislead applicants or other recruiters. My guess is that his advice accurately describes how he identifies talent. But would there be any greater act of trolling than sending people who want to find talent on a pilgrimage around the world looking for under-appreciated art, cuisine, and books while you snag up all the talent for yourself?

Why aren’t more people curating talent?

In Average is Over, Tyler writes:

In today’s global economy here is what is scarce … Quality labor with unique skills. Here is what is not scarce these days: … Money in the bank or held in government securities, which you can think of as simple capital

The implication is clear: if talent, not capital, is the bottleneck to growth, we should use our excess capital to empower the world’s underleveraged talent.

The question is, why aren’t more people trying to curate talent, whether for philanthropic reasons or simply for their own benefit?

The first and most obvious reason is that identifying talent often doesn’t personally benefit a scout. He may get a collector’s satisfaction and a philanthropist’s pleasure, but not necessarily a hedge fund manager’s returns. Talent curators, just like innovators, often can’t internalize most of the gains from their activities, so it’s no surprise that both are undersupplied2. In some sense, better talent curation actually is a type of innovation.

An intellectual does not personally benefit from the ideas which his protegee ends up developing3. Similarly, a venture capitalist has nothing to gain from funding someone early in their career who clearly has potential but doesn’t yet have a promising business idea. Neither does he profit from funding someone working in areas like basic science or the humanities which can’t easily be spun off into startups.

We need new financial and reputational instruments which help scouts like Tyler capture part of the massive value they create by boosting a career. What these instruments will look like remains to be seen. Maybe our culture will give more status to the early scouts who first identified successful people. Perhaps innovations on the blockchain will allow people to sell a fraction of their future net worth or status. As more and more such tools are created, great talent curators will be able to do to global brainpower what venture capital did to startups starting in the 1970s.

Thanks to Applied Divinity Studies and Tyler Cowen for comments.

  1. If you must learn to crack cultural codes before you can identify talent well, then should we replace hiring managers with comparative religion PhDs? Come to think of it, most college admissions officers probably do have a graduate degree in some kind of humanities which involves cracking cultural codes. Does Tyler think this is optimal or efficient? ↩︎

  2. One notable exception is sports, where there is an entire infrastructure set up to scout, evaluate, sponsor, and train promising athletes. This is because colleges can internalize the value of the players they recruit and train (completely so, in fact, since the players are not paid). ↩︎

  3. To the extent that intellectuals simply care about spreading their ideas (narrator: they don’t), on the margins, they should spend more time helping and promoting budding thinkers. I suspect, for example, that Thomas Sowell’s ideas would be better off had he spent the time it took to write his 30th book on identifying and mentoring young writers like Coleman Hughes. Sowell himself is the product of similar efforts by Milton Freedman. ↩︎