Friday 3 October 2008

Wise words of Solomonoff

I've been reading some of the work of Ray Solomonoff, having come across him in the context of Kolmogorov complexity/algorithmic information, and I like this (from The Discovery of Algorithmic Probability, Journal of Computer and System Sciences 55, 73-88 (1997)):
From Freud I got the idea of the unconscious mind: that
there were things going on in one's brain that one did not
have direct access to. Poincaré, made it clear that much of
his serious problem solving occurred in his subconscious,
and I felt this was very common in problem solving of all
kinds in the sciences and the arts.

This view was one important reason for my later rejection
of "expert systems" as a significant step toward artificial
intelligence. Expert systems were (at best) expressions of
peoples' conscious thought, which was, I felt, a very small
fraction of human problem solving activity.

Other implications: Memory is what you invent to explain
the things that you find in your head. Over the years, the
"facts" in this paper will be gradually revised as I reread my
research notes.

Explanations that people give for their own behavior are
not to be taken too seriously - including discussions in this
paper.

No comments: