Tony Kulesa writes that Tyler Cowen is “the best curator of talent in the world” and describes four components of Cowen’s approach that he thinks makes Cowen’s success possible. The first is the most interesting:
Distribution: Tyler promotes the opportunity in such a way that the talent level of the application pool is extraordinarily high and the people who apply are uniquely earnest.
I think there’s a larger theory of talent curation here, which helps shed more light on how Cowen has done what he’s done.
In my view, there’s two fundamental and overlapping ways to curate talent. The first is the way we usually think about curating talent. Everyone knows Harvard University is the best school in the world, so everyone applies to Harvard University. There’s some implicit curation—don’t waste the $75 application fee if you have no chance of getting in—but so long as you have some chance it’s not a bad bet. So Harvard has a good brand that attracts everyone they would potentially want to enroll and many (20x) more.
The problem is that Harvard’s image and brand have no way to discern between the types of good students: the ones simply looking for a high-paying gig and the ones that are going to invent nuclear fusion. So to the extent Harvard cares about such differences at all, they have to find the right students through filtration. We can call this the Filtration Model of talent acquisition: Cast a wide net, and use internal mechanisms to find the talent. The success of their filtration depends on Harvard’s admissions department being big and good at figuring out which students are the ones they want.
This is where Kulesa’s notion of distribution comes into play. Cowen, intentionally and unintentionally, has taken a different approach than Harvard. Let’s call it the Lamplight Model. As opposed to the Filtration Model, the Lamplight Model seeks to do most of the filtration externally: Cast a narrow net, but one that catches the people you want.
Cowen’s intellectual presence is representative of this approach. Marginal Revolution rewards careful readers, ones who can read between the lines of posts and notice recurring themes. Cowen doesn’t engage in typically viral or trendy topics, and when he does the angle is often diametric to the popular discussion. Linked posts are often obscure, especially for a blog ostensibly about economics. But they begin making sense when you think about the world like Cowen does: How does this new piece of information change your model of the world? What are the underlying incentives? Who gains status and who loses it?
So Marginal Revolution pushes away many potential readers but cultivates a deep loyaly from readers who see and think more like Cowen. He can convert some of those readers—usually young and ambitious ones—into EV recipients.
The structure of Emergent Ventures itself depends on the Lamplight Model: Cowen reviews every application himself, so receiving 40,000 applications a year (as Harvard does) simply isn’t feasible. He needs to use the Lamplight Model to do a lot of his filtering for him.
The Lamplight Model shows up in the data as higher acceptance rates. Kulesa writes:
It isn’t just a matter of more elite selection. In fact, Emergent Ventures has a higher acceptance rate than elite colleges. In May 2020, Tyler reported in an interview with Tim Ferriss that the award rate is ~10%. For comparison, the 2021 acceptance rates of Harvard, Princeton, and Yale were 5%, 6%, and 7%. It also isn’t a wider pool. At that time, he had only ~800 total applications since 2018.
Obviously, the 10% acceptance rate suggests that there is substantial filtering going on too—it’s always going to be a bit of both—but more of the work is happening externally.
You can see a similar system working for the Thiel Fellowship, as described in a post on Strange Loop Canon. Thiel’s program immediately ticked off much of respectable society and thus drew little interest from conformists and status-seeking students. Instead, it attracted weirdos, the very people Thiel wanted to attract. It repelled the wrong people and attracted the right people. There’s a number of familiar names even in the first cohort, including Laura Deming and Nick Cammarata. Vitalik Buterin was in the fourth cohort, in 2014.
YCombinator seems to have had a similar effect in their first cohort. Per Paul Graham:
That first batch included reddit, Justin Kan and Emmett Shear, who went on to found Twitch, Aaron Swartz, who had already helped write the RSS spec and would a few years later become a martyr for open access, and Sam Altman, who would later become the second president of YC.
Graham seems to identify the same effect I describe here:
I don't think it was entirely luck that the first batch was so good. You had to be pretty bold to sign up for a weird thing like the Summer Founders Program instead of a summer job at a legit place like Microsoft or Goldman Sachs.
Again, we see the power of not just attracting the right people but also repelling the wrong people.
This leads to interesting conclusions about publicity: Word of mouth is very good, as it will generally mean that the desired group is attracting more people similar to themselves. But large-scale publicity, features in mass media outlets and the like could be actively harmful to the program.
All of this suggests a difficult problem. Cowen’s reach seems to have been growing recently. This is good for the world, as Cowen is an important intellectual. It may even be good for Emergent Ventures initially, as its reach and network expand to an optimum level of word-of-mouth transmission. But if Emergent Ventures becomes a status symbol or meritocratic badge for non-weirdos, it could break the secret mechanisms of EV, just as it might imperil the Thiel Fellowship if they keep minting billionaires faster than any VC could dream of. I don’t have a strong opinion on YCombinator, but the common opinion seems to be that its fallen victim to this problem, now just a next career step for FAANG employees.
There are a few obvious ways to prevent this. Cowen could become less legible to non-weirdos. Or he could become more offensive to them. Both of these have obvious downsides as well, though. He could also hire an excellent admissions department and change the curation model. So it will be interesting to see whether EV is able to keep its track record and if other similar organizations that are beginning to pop up can emulate its success.
With the announcement of MothMinds, we might be entering a golden age of micro-granting programs. We’ll be able to see how these programs compare: the ones that are formal, with many people filtering applicants, versus the ones that are one person’s side project, where the program’s success depends on their Lamplight Effect. And someday, we will get to see who comes out of which.