More than 90% of statups fail. Why do you expect better lower failure rate for *moonshots* than for startups? That's entirely unreasonable. It's very interest

More than 90% of statups fail. Why do you expect better lower failure rate for *moonshots* than for startups? That's entirely unreasonable. It's very interesting that you're focusing on failures rather than on graduated projects. Do you really expect 100% success rate for moonshots? It might actually be hard to estimate success rate as best ideas got absorbed into Google itself. E.g. Google Brain originated at X. Wikipedia has other examples: https://en.wikipedia.org/wiki/X_Development#Others https://en.wikipedia.org/wiki/X_Development#Others