Author: Robert Hyatt
Date: 15:25:22 08/11/99
Go up one level in this thread
On August 11, 1999 at 16:10:17, Ratko V Tomic wrote: > > In the computer chess world, Botwinnik was > > a complete fraud. > >Well, it's true he didn't have a full blown program, much >less a competitive one. He was probably too old to learn >enough to do it himself, and the folks he got to help him >weren't perhaps up to task and certainly didn't have an >adequate hardware and software (did you see the fortran >dumps in his book, it's a sad sight). The problem most of us have with his work is that it appears to be outright 'fakery' in places. IE Berliner took him to task for a paper he published in the ICCA Journal, where he gave analysis that was just outright faked. And Hans pretty well nailed him point by point for doing so. He always kept telling us how great it was going to be, yet it _never_ played a game of chess in 20+ years of promising. 20+ years is rediculous, of course. :) > >But his ideas on multilayered control, interactions >and job partition between layers, his field of play >construct, etc are deep and farsighted ideas, well >ahead of their time and the current levels of programming >techniques. When I first read his book on his Pioneer >project some years ago, it seemed like nonsense. But, >the more I accumulated knowledge and ideas on chess >programming, the better his vision looked. It's a kind >of work which grows on you, as you revisit it over years. > >Obviously, someone entangled deeply with the latest >game tree searching tricks, would not appreciate it, >thus I am not surprised at the general reaction he >received in the chess programming circles. Expecting >otherwise would be like hoping that a kid flipping >the burgers at McDonnalds would appreciate an advice >from a world class French chef. He would say that the >chef is a fraud in the fast food business. > >But the chess programming field evolves, too, and some >day they and the hardware and the software technology >will catch up with that kind of approach. Brute force >can go only so far, even in such a tiny domain (relative >to the real world) as chess playing. At least for the >domain of chess, his work has mapped out a concrete and >sound long term strategy how one might implement it.
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.