After nearly making a terrible mistake not buying See's, we've made this mistake many times. We are apparently slow learners. These opportunity costs don't show up on financial statements, but have cost us many billions.

Our success has come from the lack of oversight we've provided, and our success will continue to be from a lack of oversight. But if you're going to provide minimal oversight, you have to buy carefully. It's a different model from GE's. GE's works - it's just very different from ours.

If you have competence, you know the edge. It wouldnt be a competence if you didnt know where the boundaries lie. Asking whether youve passed the boundary is a question that almost answers itself.

I knew a guy who had $5 million and owned his house free and clear. But he wanted to make a bit more money to support his spending, so at the peak of the internet bubble he was selling puts on internet stocks. He lost all of his money and his house and now works in a restaurant. It's not a smart thing for the country to legalize gambling [in the stock market] and make it very accessible.

The cash register did more for human morality than the Congregational Church. It was a really powerful phenomenon to make an economic system work better, just as, in reverse, a system that can be easily defrauded ruins a civilization. A system that's very hard to defraud, like a cash register, helped the economic performance of a civilization by reducing vice, but very few people within economics talk about it in those terms.

The idea of excessive diversification is madness.

The hedge fund known as "Long Term Capital Management" collapsed last fall through overconfidence in its highly leveraged methods, despite I.Q.'s of its principals that must have averaged 160. Smart people aren't exempt from professional disasters from overconfidence. Often, they just run aground in the more difficult voyages they choose, relying on their self-appraisals that they have superior talents and methods.