I used to wrongly think you need a PhD to be able to win a Turing. Diffe proved me wrong. He serves as an inspiration to anyone who mentally feels inferior to PhDs when it comes to making significant contributions to the field of Computer Science
I originally wanted to do a PhD because of the feeling smart aspect; but actuall, after I had started it (actually during the middle) I had a realisation that the people doing PhDs weren't necessarily smart; but rather they were either highly passionate about what they were doing or really enjoyed doing research.
As it happens I have probably created more outside of doing my PhD than in it; and whenever I was advising people on whether to do a PhD or not, would strongly suggest that they didn't. (The ones that ignored that advice and pressed on were exactly the type of people who could keep up the slog of doing it.)
Lots of people don't make it to the end and the going gets really tough towards the end, particularly when you are writing up and especially problematic if you have a young family. There's only so much time you can burn the candle at both ends.
So don't worry about not having a PhD; in fact, I'd argue that some of my undergraduate friends went on to much bigger and better things than I primarily because they got into the industry much earlier and therefore worked their way up.
Plus creating a startup now is so much easier than it used to be - if you see a business need that's not being satisfied, go for it. Most startups success is entirely down to chance and not educational background in any case.
You absolutely shouldn't feel "mentally inferior" to folks with PhDs. Doing academic research, which probably involves getting a PhD, is almost certainly the best way to make a significant contribution to Computer Science. These two facts aren't really in conflict, I don't think.
