> Though I guess the first year of a PhD often plays this role... In the U.S., the NSF Graduate Research Fellowships sort of do that for a bit longer (3 years)

> Though I guess the first year of a PhD often plays this role... In the U.S., the NSF Graduate Research Fellowships sort of do that for a bit longer (3 years). You do have to write a short proposal, but like with YC, there's not necessarily an expectation that you'll stick with the project that was in the proposal. It's just there to demonstrate that you can come up with a coherent project proposal. I actually did something completely unrelated to the original proposal, and they were fine with that. So it ends up being 3 years of guaranteed funding to do whatever you want. The funding is also tied to you, not through a faculty member, so you really can do whatever you want (e.g. if you change supervisors or even universities, the grant follows you). At the faculty level, there are also NSF CAREER awards, which are supposed to fund promising researchers, not specific projects. However at that point it's harder to dissociate the person and the research area: is it really possible to judge only the person unbiased by which area they work in and what they've been publishing? With the graduate fellowships it's a bit easier because, like with YC, applicants often have no real track record, so you're just judging college seniors' potential to do research, not judging their existing research.