I also liked Sam’s point about how it’s easier for startups than big companies to work on non-consensus ideas. At a big company, you need manager 1 and manager 2 and manager 3 etc. to approve a project. Whereas at a startup you need investor 1 or investor 2 or investor 3 etc to fund you. At a big company, a single no kills you. At a startup, a single yes saves you.

(Not sure the point is entirely true - like, Sundar Pichai is not approving every small project at Google - but there is a very different structure to ‘and’ problems vs ‘or’ problems.)