I want to invest in things where I can be most wrong, because that's where the upside is most asymmetric.
The areas where failure is most likely are also the areas where success would be most important.
I believe we can replace 80% of what doctors do with AI. That's not a put-down of doctors — that's the scale of the opportunity.
Venture capital should be about funding things that can't get funded any other way. The moment you only fund safe things, you've stopped doing venture capital.
People ask why I keep investing in clean energy after the failures. The answer is: the problem didn't go away. The technology just needs more time and more capital.
Failure is a feature of doing important things, not a bug. If you're not failing a lot, you're not trying hard enough.