I'm finding (not that this is a unique observation at all, though) that there are two kinds of founders: engineers who are accidental entrepreneurs, and entrepreneurs who are founders. If you need to think hard to find even a sector to found a startup in, you're probably doing it wrong. Not that you can't be successful, and I wish the best of luck to anyone, but it seems like the giants we all know come from founders who started coding to solve some problem, and let the company grow from there on its own, instead of the other way around.