Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Shredder improvements

Author: Russell Reagan

Date: 15:09:38 05/17/03

Go up one level in this thread


On May 17, 2003 at 14:16:23, Peter Stayne wrote:

>Are we getting any closer to the point where, given the same
>hardware, it's almost impossible to make an engine stronger? Or do we still have
>a long way to go?

I'm no expert, but I think we haven't even crossed the half-way point yet. Alpha
beta based programs have been seriously studied for decades, and there is still
probably room for improvement and new ideas even in that area. If you consider
the fact that other ideas haven't had the amount of research that alpha-beta has
had (and it's not even close), then it seems to me that we have only explored
one hall of a giant mansion, and we haven't thoroughly exhausted the exploration
of that one hall yet.

There are so many areas of AI that are rarely (if ever) used in computer chess,
and a lot of them are relatively new. Things like neural networks, genetic
algorithms, and the like certainly haven't been studied as much as alpha-beta.
In 50 years, think of what will have been discovered in those areas, and those
are just some of the more popular new things in AI. There are hundreds or
thousands of ideas (or maybe more) that are very cutting edge and are only being
explored by a handful of people.

Think about it. Maybe 99% of the ideas in AI aren't used in chess. As machines
get faster, and more and more areas of AI become better researched, well...the
possibilities are endless I think (at least for a few thousand years).



This page took 0 seconds to execute

Last modified: Thu, 15 Apr 21 08:11:13 -0700

Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.