Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Along tired arguement finally dead. Comps are GM strength

Author: blass uri

Date: 19:20:43 05/15/00

Go up one level in this thread


On May 15, 2000 at 19:22:43, Hans Gerber wrote:

>On May 15, 2000 at 16:20:10, pete wrote:
>
>
>>
>>It seems the humans are able to learn how to play a computer .
>>
>
>You can bet!
>
>The bad side of all these events, challenges and "matches" is the fact that a
>brand new machine (program) is presented the humans couldn't train on yet. It
>was the same with Fritz, with Genius, with Rebel and not to forget with Deep
>Blue.
>
>If you then limit the number of the games down to three or maximum six, you can
>understand that a serious preparation is not possible.

This is not the same.

Deeper blue was clearly something that humans could not prepare against but it
is not the case with Fritz,Rebel...

People can buy a program that is similiar to the last version.

In the Israeli league upgrades were not allowed so humans could prepare against
the exact Fritz that they played(except opening book).

The fact that humans can buy program when programs cannot buy humans is a big
advantage for the humans.

>
>The main fallacy of all such computer vs human games is the absence of the
>human's best talents, the one you mentioned.
>
>The knowledge of only a few weaknesses, and there are many in reality, would
>enable the humans to show that computers play worse than IMs.


I do not think that we need to do it now.
We will need to do it when computers will be better but not now.

 In other words,
>there is no IM so far with such weaknesses. Point is humans can communicate.
>That humans often do not compete successfully against computers, this is due to
>the unwillingness of really good players to make use of anti-computer chess.
>They are too proud for that. Otherwise the results would be much better.

I agree that humans have the ability to cause computers to play at 2400 level or
even 2300 level but it does not change the fact that humans do not use this
ability in most of their games so some program are practically in the 2500-2600
level.

Fritz may be practically only in the 2400-2500 level not because it is weaker
but because of the fact that humans know it more than other programs.

Uri



This page took 0.01 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.