Google ran an experiment: "We are surrounded by stodgy, calcified software firms. Can we build a company that isn't like that and is also successful?"

For a long time, they succeeded. But I think that experiment has run to a conclusion and they've discovered for themselves why their competition became stodgy and calcified (it wasn't because that's what they wanted to be; it's because the larger you get, the more diverse work you're doing, the more you're rewarded for being reliable and efficient, not scrappy and disruptive).