It’s that the artificial intelligence researchers were more... more flexible and experimental and if they weren’t successful on a particular problem they would move to another one, whereas the more professional people had to keep focusing on particular problems in electrical engineering or physics. And so the communities were different because the AI people were more like science fiction readers and writers and these people were more like business people who had to show a profit or had to get certain problems solved because they had to teach other students who would go out and become practitioners, so that you could say that, if you look at the world of research, there’s the kind of research where you’re exploring ideas that might take 20 or 30 years to mature and you can’t expect a company to pay for that because it will just be a minus on its balance sheet for as long as the stockholders… investors are looking at it.

Whereas, if you have an established profession then you can know how… what is the value of a particular advance and decide where to put your money and optimize things. So I think there’s a serious problem in the way governments and institutions have to allocate the long term basic res… when I say basic research we mean almost the opposite of basic. It means doing things where you’re not sure that they’re important yet. When they call it basic that’s an attempt to dignify it.