Those who will not face improvements because they are changes, will face changes that are not improvements.

The possibility that stock value in aggregate can become irrationally high is contrary to the hard-form "efficient market" theory that many of you once learned as gospel from your mistaken professors of yore. Your mistaken professors were too much influenced by "rational man" models of human behavior from economics and too little by "foolish man" models from psychology and real-world experience.

To finish first you have to first finish. Don't get in a position where you go back to go. What's interesting is that some guy whose grandfather was a lawyer and a judge-hurriedly going to Harvard Law with a wave of veterans-I was willing to go into so many different businesses. I was constantly going right into the other fellow's business and doing better than the other fellow did. The reason it was possible? Self-education- developing mental discipline, big ideas that really work.

For society, the Internet is wonderful, but for capitalists, it will be a net negative. It will increase efficiency, but lots of things increase efficiency without increasing profits. It is way more likely to make American businesses less profitable than more profitable. This is perfectly obvious, but very little understood.

See's candy company was the first high-quality business we ever bought.

I don't invest in what I don't understand. And I don't want to understand Facebook.

There are exceptional loyalties and there are old fashion ideas about how you get loyalties, and after all the auditorium is full of people who have co-owned shares with the managers for many decades, and in many cases they co-invested when everyone was young and obscure. Also when you come back to a place like that you are celebrating old loyalties, and of course the basic idea behind so much of Berkshire is the old fashioned idea that the best way to get loyalty is to deserve loyalty.