> As a founder, it's just one problem after another, rapid fire, and you might have 1-2 days to solve something that took 2 months at Google. You do a lot of hacky 80/20 solutions. You need to think big-picture on everything and understand how everything fits together. You solve a much wider variety of problems, but you don't really go into depth with anything. Rigorous engineering isn't really part of a founder's job description, and some people view that as being dumb or poorly trained, but you make up for that in speed, ambiguity, and breadth.
This is why. While I sometimes yearn for a little more rigour, I certainly do not miss change reviews taking weeks and months at a time, and projects languishing while we attend endless meetings which seem designed to _undermine_ consensus rather than build it. I find the glacial process of large corporations to be utterly frustrating.
At Google you do rigorous engineering, you're intensely data-driven, you have to work at massive scale, everything you touch is a distributed system (with all the skills and pitfalls that comes with working on that), and you'll often learn a lot of fundamental CS algorithms because you need to re-implement them to work across 1000s of machines rather than using a standard library built for a single address space.
As a founder, it's just one problem after another, rapid fire, and you might have 1-2 days to solve something that took 2 months at Google. You do a lot of hacky 80/20 solutions. You need to think big-picture on everything and understand how everything fits together. You solve a much wider variety of problems, but you don't really go into depth with anything. Rigorous engineering isn't really part of a founder's job description, and some people view that as being dumb or poorly trained, but you make up for that in speed, ambiguity, and breadth.
I'd say that when I joined Google in '09, the level of colleagues there was much higher than the general level of technical founder prowess in the Valley, and it remained that way for the whole time I was there. But the types of employees Google hired in '13 were very different from the types of employees hired in '09 (who themselves were very different from those hired in '02), and it wouldn't surprise me if starting in '15 or so the balance started shifting back towards startup founders. Honestly I think the best technical minds today are actually doing cryptocurrency "non-profits" (scare quotes because they're actually being paid by capital appreciation of their founder tokens) - I've been quite impressed by the algorithms being discovered by projects like Ethereum, OmiseGo, MakerDao, zCash, Monero, etc.