The problem with commercialized research is more the attitudes of the founders than what they're building. They usually approach starting a startup as a solution in search of a problem. Sun (and Google) were exceptions in that even while working on the project within a university, they were building a product. We'd love to fund that kind of project, but they're rare.
... material science, sensors, robotics,
medical devices, life sciences, etc...
VCs whose firms would have looked at these
deals or invested in these sectors, are now
only interested in whether it runs on a
smart phone or tablet.
Paul is saying that founders aren't coming to him with the former sort of startups.
Facebook was born in a university dorm room, not a university lab (or department lounge). Facebook is a great product, and has spawned some interesting technology created to solve a massive scaling problem, but what cause the scaling problem to begin with was a product created with a good intuition for what people love, built with some existing, not academic, technologies.
