Author: Mark Young
Date: 14:50:53 01/29/98
Go up one level in this thread
On January 29, 1998 at 02:59:12, Bruce Moreland wrote: > >On January 29, 1998 at 01:46:38, Mark Young wrote: > >>You're right to a point with today's computers, but chess itself is all >>tactics. Would not a 32 man tablebase, or a simple full width chess >>program that could see to mate from move one, play perfect chess? It >>would not care if a bishop is bad or a knight is good, only that a >>position is won, drawn or lost. The term positional play is a human >>concept to explain what we can not see tactically. >> >>While its true that the type of program you mentioned in your example is >>written from a "positional" point of view, that is only due to the >>limitations of both today's computers and the human mind, not due to the >>true nature of chess. You may have written a strong chess program, but >>you're still wrong on this point. > >This is true of any game of this type (checkers, chess, go, etc.). Of >course chess is solvable, you mention two different ways of doing it. >This doesn't make the game less interesting for humans and it shouldn't >make it less interesting for computers. > >bruce The original point which was being discussed had nothing to do with how interesting chess is, however. My point was that chess isn't a great way to gauge "intelligence" in computers (or in humans for that matter). Intelligence would imply a way to adapt to unpredictable situations. This simply isn't the case with chess. Chess is finite and fairly predictable while a true test of intellignce would require the ability to react to what is mostly unpredictable and nearly infinit. A good example are the fledgling attempts at AI. Computers being able take input and modify their own instructions based on the outcome. The more complicated the task that the AI is able to handle correctly the more intelligent the computer is. By definition this is the only real way to gauge intelligence. Unfortunatly I don't know of any good I.Q. tests for computers yet, but I think it might be something that would be an iteresting future test that all new computers would be rated with. Just think, when you go to buy a new computer instead of rating computers by processor speed, video speed and amount of hard drive and ram only, you might be more concerned with what its I.Q. is.
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.