I thought maybe you could do that with human beings too. You could tell what huge masses of human beings would do, provided they didn't know what the predictions were so they couldn't distort their own behaviour, and provided you had a large enough number, and I felt that with the galactic empire you'd have a large enough number. I don't really believe it's going to work, but it made a good background for the stories, and I was always able to use my "psychohistory" to show how things became inevitable, economically or sociologically and so on. It made for interesting historical novels.....was much smarter--and vastly more humane--than Paul Krugman:
Not only are there not enough people, but actually their behaviour is far too complicated. They're not like individual molecules. Molecules have limited modes of behaviour and human beings are far less limited, so that human history is more chaotic. In fact, so chaotic that it probably can never be predicted, and in my later Foundation novels I dragged this in. But of course when I first started I didn't know anything about this new theory of chaos.
It is one of the few science fiction series that deals with social scientists—the “psychohistorians,” who use their understanding of the mathematics of society to save civilization as the Galactic Empire collapses. I loved Foundation, and in my early teens my secret fantasy was to become a psychohistorian. Unfortunately, there’s no such thing (yet). I was and am fascinated by history, but the craft of history is far better at the what and the when than the why, and I eventually wanted more. As for social sciences other than economics, I am interested in their subjects but cannot get excited about their methods—the power of economic models to show how plausible assumptions yield surprising conclusions, to distill clear insights from seemingly murky issues, has no counterpart yet in political science or sociology. Someday there will exist a unified social science of the kind that Asimov imagined, but for the time being economics is as close to psychohistory as you can get.