There's also the issue of whether Silicon Valley has the requisite infrastructure to facilitate these sorts of projects. If you want to do projects in robotics, DARPA has both money to give you, deep connections with the universities developing the basic science, and a program manager to manage the project that typically has a PhD in the field. There's a lot of downsides to DARPA, namely that everything has to have some sort of military angle, but there's a wealth of technical expertise sloshing around there.
The last head of DARPA now heads up Google X. SFBA has tens, if not hundreds, of billions of dollars available to it.
Capability isn't the question. Its the will that matters. As an investor, are you willing to sacrifice returns for the equivalent of the laser or the transistor?
Now could they acquire that capability? Sure, with time and money you can buy whatever you want. But will comes into play there--why invest in that capability when you can get quicker returns in consumer products?
I do agree, though, that a big corporate R&D lab like at Google is better-positioned to tackle these sorts of projects than VC-funded start ups.
And interestingly it also seems to go against Mark Cubans SV hypothesis as well.
YC now (publicly) looks like it's working on the ideas and types of companies that the world needs more of. If this shift is real, it opens up YC to smart people without CS degrees. (I know it was before, but it's perception as well) And there are exits, they're just outside of SV. GE, Philips, Exxon, GSK, Pfizer, Bombardier...
This is a key issue, I think. When doing something technically innovative, a very common pattern is that the team that spends all the money & time up front to get the first working version is not the one that ends up with the successful market leadership: once the up-front investment is sunk, a 2nd team can much more cheaply iterate on the market/business side of things and capture the market without spending money on the R&D. Sometimes it might even be a different market, since inventions researched for one purpose may turn out to be applicable in another field. Patents are an attempt to provide a mechanism to reimburse the first group in that scenario, but are pretty imperfect.
One reason I prefer to see research publicly funded is that I think it's often good that it's difficult to capture/monetize the benefits benefits of research breakthroughs. I'd rather not have researchers spending their time trying to figure out how to encumber/"own" the results of their research out of a need to get their investors an ROI.
I didn't become an engineer to crank out CRUD apps in some rock star's crazy DSL all day.
Ventures like SpaceX and Tesla that have a long lead time and (at least in these two cases) are loss-making at the start probably wouldn't get YC-VC funding. Or at least they wouldn't get anywhere near the level of funding they'd need to get to profitability point.
I wager that for some of the really big world-changing projects we simply have to be willing to fund it as long as necessary, taking an extremely long-term view. And for this, VC funding is probably not going to be enough. Both Tesla and SpaceX needed to take in public (read: Government) funding to keep going.
On the other hand, one can start with $100k which is mainly used to secure the rest of the funding.
This new "shift" is essentially what the Founders Fund is trying to do.