I think markets are often not thinking on a long-time horizon, I think that our government structurally is doing even less so. When we have a government where we have people who are up for election at most once every six years for a U.S. senator, that's a time horizon that is much shorter than in a market that a company is looking at 10, 15, 20 years which is a time horizon over which a stock price is typically valued.

As you craft a plan to expand to adjacent markets, don't disrupt: Avoid competition as much as possible.

I think we have a bubble in the US in government bonds, because of the quantitative easing and the negative real interest rates, and to some extent, that increases asset values across the board, including in startups.

I do tend to think that things that have incredibly long time horizons often do involve market failures.

You'll attract the employees you need if you can explain why your mission is compelling: not why it's important in general, but why you're doing something important that no one else is going to get done.

If you invest in Microsoft or Oracle, or a number of other companies for that matter, you're fundamentally making a bet that there's going to be no innovation. So an investment in Microsoft is a bet that the operating system is going to stay the same, it won't be replaced by Linux, Google Docs, or a mobile platform like iOS or Android.

Technology is probably the single biggest driver of productivity gains for the developed countries. For example, I think it's much more important than free trade.