Re: computer-go: Brains & Software

[Jeffrey Sorenson <jsorenson@xxxxxxxx>]

> Don Daily apparently said...

> I think there is a lot of "magic" in the way  we represent problems in
> our  brains AND on a computer.   I have often  wondered if Go or Chess
> was  poorly represented  on  a  computer by  our  natural tendency  to
> impose our own cognitive view upon the computer.

Jeff Sorenson replies:  As a programmer, I am sure this is the case.

> So perhaps  we are confined by our  own perceptions, not being able to
> see a much better  way to do things (from  the computers point of view
> or even our own?)
>
> Don

Jeff continues:  You have here identified a huge obstacle to the human
approaches to many kinds of problems.  I would propose that imagination is one
of our scarcest resources in any sort of problem solving.  Better analysis of
how our brains really work will eventually provide useful info on both how to
write problem solving programs, and point out some limitations of our (human)
approach.
One friend of mine has written a program that not only solves a variety of
calculus problems, but explicates its "reasoning" at each step.  Has anyone
else taken this approach with Go or know of anyone who has?   This would be at
one extreme of using the computer to model a human approach to Go, but there
are many many "less human" approaches that will be fruitful to investigate, I
am sure.

BTW, thanks to all contributors to this list, I am impressed with the
willingness of the top programmers to share info and ideas!  Happy CPU cycles
to all!