Imagining a Profitable Thiel Fellowship
How to create a profitable genius factory. On the best education service incentive structures.
[Note: I wrote this in Apr 2023 but never published it, so some numbers in here might be outdated]
The Thiel Fellowship succeeded in creating an extremely high status credential/club membership. And it succeeded in helping create about $220 billion in value.
While it did create an alternative to college for a select few, it didn’t create a sustainable alternative to a broken education system.
But you could make it an incredibly profitable model. By simply taking 2% or so of each fellow’s (future) companies.
Multiples
In total, the fellowship granted $22.8 million in $100k fellowships. It helped create about 220 billion in venture returns. Let’s assume another 400k in total every year in operating costs of the fellowship. Since its inception in 2010, that’s $5.2 million. So the total cost might be around $28 million.
A 2% stake in the $220 billion outcome would be worth $4.4 billion. This represents a return multiple of approximately 157x from the initial $28 million investment.
In comparison, top venture capital firms like Sequoia have an average multiple of 4-8x. The Thiel Fellowship for equity model could generate returns that are approximately 19.6 to 39.2 times greater than those generated by the best VC funds.
Even if we assume a higher price (equity payment) discouraged many from applying to the fellowship, and it returned only $110 billion, that's still a return of $2.2 billion, i.e., a return multiple of approximately 78.6x from the initial $28 million investment.
However, the fellowship also served as a deal flow pipeline for Thiel and Founders Fund, thus allowing them to capture a respectable part of the mentioned $4.4B return.
Status signal dilution
Now you might think: But this will reduce the value of the signal? If fellows paid equity for a Thiel credential it becomes less valuable. That is true, but it’s not the true limitation of the “Thiel Fellowship for equity”-model. Stanford and Harvard students pay exorbitant amounts (avg of $200-350k depending on financial aid) for every semester and the status signal is high regardless. (The credential probably retains a high status value despite the high cost exactly because the price is opaque for others).
The true limitation is scaling up the batch sizes. This is what happened to YC. I couldn’t find precise “reputation metrics” on this but a lot of founders and VCs will confirm that YC isn’t as attractive anymore because of the ubiquity of the signal.
The best way to scale up the batch size is to increase the perceived value of the deal. Stanford’s high status is not only due to its exclusivity but also due to the program quality. Stanford students learn from the best of the best and live in an environment that resembles a resort.
Potential Value proposition
Since we’re now talking about a venture capital/school model we need to re-evaluate our value proposition. The service needs to be extremely attractive which allows us to scale up the batches. We want this to be a model that can revolutionize the education system.
We also need to make sure joining the program has a high success rate in making its users into geniuses that build massive companies. Otherwise, this won't be an attractive offering. How do we do that? By offering the world's best education. The company that’s running this program will be an institution that is more incentivized to do so than any other existing educational institution.
It’s trivial to deliver something significantly better than the current education system. Currently, students spend a big portion of their lives following a mandated curriculum cramming facts for tests which they then forget about.
The best education is 1:1 tutoring combined with project based, curiosity driven learning. Most of the people we call geniuses received 1:1 (aristocratic) tutoring from an early age. Famous examples include Darwin, John Stuart Mill, Hannah Arendt, Einstein, Jon van Neumann, Bill Gates, and Bertrand Russel1. Bloom’s 2-sigma problem is often quoted as proof of this effect.
The earlier this program starts, the better for our program’s success rate. However, that also increases the operating cost and our return cycles significantly.
Other objections
Incentives Ethics
Some might still argue that this model is “profiting from young entrepreneurs ideas and efforts” (describing it as something negative). Yes. That is the definition of an efficient education system. In the current system the tutors and school administrators have no incentive to innovate and improve the education of the students. They don’t care if the students hate school.
This is the same reason non-profit’s or NGO’s incentive structures are inefficient. The user of the service doesn’t pay for the product directly. A third party pays for it. This is also why I would not trust news sources that are donation or government funded, as they do not have an incentive to grow their customer base to the widest possible audience and deliver the best product.
Even private schools are misaligned in some ways. They get paid by their customers for a credential that demonstrates that they were able to pass some test, and personality traits such as conscientiousness and conformity. The incentive of a conventional college is to persuade employers that you are a conscientious and conformity seeking worker.
The Thiel Fellowship for equity model skips this step completely. Its only incentive for its students is to generate massive amounts of value for society.
Talent Pool limitation
Another possible objection is that the talent pool for this kind of program is too limited. This assumes that building an incredibly impressive company is something magical that you can’t learn or that requires an incredibly high IQ. That is not true. Building a company is not significantly more difficult than working as senior ranking executive or programmer.
Indentured servitude
The association of a program like this with indentured servitude appears to be the biggest obstacle to making the long-term strategy version of this a mainstream educational model.
Indentured servitude is a historical labor system where workers (mostly immigrants) were bound to work for a specific period of time in exchange for a passage, room, or some other service.
There are a number of important differences with this historical phenomenon: There is no upfront cost in the form of physical labor. Our version is voluntary, fairly compensated, and the value exchange happens asynchronously, in the form of a work artifact (a valuable company) that is created by the student after we provide our educational and incubation service.
The association is likely less strong in Europe where a version of this model is already practiced as apprenticeships. In Switzerland, for example, 70% of teens are apprentices, giving the country a steady stream of trained, experienced workers (including areas such as accounting, healthcare, aerospace, and software engineering).
Others who wrote about this
In modern times, the first notable mention of the concept of Income Share Agreements (ISAs) was by Nobel-prize winning economist Milton Friedman in his 1955 essay The Role of Government in Education. Last year The New Yorker published a piece “Is Selling Shares in Yourself the Way of the Future?” about two brothers that sold shares in their personal holding company2. The Residency provides, housing, food, and cash for 3 months in exchange for 2% equity in resident's startups3. Sam Altman also tweeted about this last year4.
Sam Lessin (co-founder and GP of Slow Ventures) also talked about the concept in an interview with Brian Requarth.
Questions I still have
How important is it to have an initial status signal / association with some high status individual such as Peter Thiel or Elon Musk?
How can you effectively sell the program to 10 or 16-year-olds? The earlier you start with the program the more you have to persuade the parents.
What legal edge cases could arise? If a student disputes the contract later, as contracts with minors are often voidable. However, contracts deemed beneficial, such as those involving apprenticeships or services, may still be enforceable if their terms are fair and reasonable.
How profitable was the fellowship for Thiel and founders fund? If anyone knows how much Thiel put into Luminar, Figma, DoNotPay, Scale AI, Oyo Rooms, etc. that would be interesting.
Bett Hoel wrote a great post about aristocratic tutoring and follow up defending some of the objections of his article. Henrik Karlsson also wrote about the childhoods of exceptional people.