Author: Art Basham
Date: 09:00:13 12/02/02
Go up one level in this thread
Thanks...Very interesting! Maybe the problem is.. every time a chess engine wins a game, the programer thinks it is a "brilliant win"...etc...:=) Also, maybe "planning" was a poor word to use here... I rather think of it as "seeing" into the future, and then, making it happen...:-) Maybe someday the computers will do just that! Who knows! Art ++++++++++++++++++++++++++++++++ On December 02, 2002 at 11:24:53, Bob Durrett wrote: >On December 02, 2002 at 10:28:27, Art Basham wrote: > ><snip> > >>I know that most chess computers know *how* to get two Rooks and the Queen on >>the same file for an attack -- (I have seen it happen). ..but do they know how >>to "sieze the moment" or sieze the opportunity in a game to do this..? >>that is the question I ask... >> >> To do this I think requires "imagination" >>or the ability to "make a Plan" right in the middle of a game... >>Thanks again! >>Art > >><snip> > >Perhaps most "hard core" chess programmers [like Bob Hyatt] will insist that >chess engines do not need the ability to plan at all. "Planning" is a process >which takes place in the human mind. But chess engines do not have a human >mind. What works best in the human mind may not work well in a chess engine. > >There is also an issue as to whether or not "beautiful human chess" is, or >should be, the same as "beautiful chess engine chess." > >Typically, [as I see it] human chess wins brilliancy prizes when the game >contains stunning hard-to-find moves and plans which are hard for a human to >find. > >But something hard for a human to find may be easy for a chess engine and visa >versa. Often, test positions are presented here with the message: "How long >does it take for your engine to find the correct move?" Occasionally, one of >those test positions presents exceptional difficulties for all or most of the >current chess engines. If one engine finds the correct move in a short time, >all the programmers should stand up and salute and award a "brilliant move" >prize to the good engine. > >Suppose someone were to find a game which contained MANY such positions, and >only one engine was able to quickly solve all those positions? Then maybe that >game should be awarded the prize: "brilliant chess engine game." > >Why? because the exceptional moves would satisfy the criteria: (1) hard for a >chess engine to find, (2) stunning [in the eyes of chess programmers]. > >Chess programmers are truly a wierd lot. They have a different concept of chess >beauty. : ) > >Bob D.
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.