I also think that the criticism that startups are not "big-thinking" enough is more appearance-based. She may be more impressed if startup pitches read like res

I also think that the criticism that startups are not "big-thinking" enough is more appearance-based. She may be more impressed if startup pitches read like research papers and less like, well, startups. However, there are a number of problems with conflating research problems with startup opportunities, one of the first being that the startup field is now a lot smaller than it was. She talks about Stanford and Harvard students as though they're the best of the best and while it's true that recent graduates may have great potential, what you really want for research-level efforts is domain experts. It usually takes years to become a domain export -- far longer than the typical beginning-phase startup. There are also a number of major problems with funding straight research efforts just from an expected-value point of view. Even Google was mostly working with known quantities (i.e., the exploratory research had already been done) when it started. If the result of a year-long startup effort is "turns out this is impossible" that's pretty bad. Basically, there's a sweet spot for startups in a fiscal and informational sense and I think it will always tend to sound a little less impressive given that context.