If we take calculated risks where we know about the small probability of success (in the case of startups) and perhaps employ risk-mitigating measures and justify the whole enterprise with the problem to be solved or the possible payoff, we're still confident but not ignorant (and thus stupid).

I don't experience that people call me somehow stupid for what I do. Perhaps because I tell them I know about the dangers, how I manage them, that my approach is not necessarily the best approach for other people (invalidating their choices).