Author: Robert Hyatt
Date: 15:49:55 05/10/00
Go up one level in this thread
On May 10, 2000 at 16:20:23, Hans Gerber wrote: >On May 09, 2000 at 22:06:39, Robert Hyatt wrote: > >>Sure. Post any position you want here, which no computer can solve. I'll >>post "crafty's" solution quickly, proving beyond a shadow of a doubt that it >>can make the right move. I can type quickly, and I know VI inside-out. :) >> >> >> >>we did. we decided the problem was not solvable. For every measure a cheater >>tries, you have to have a counter-measure. And for every counter-measure you >>produce, he will produce a counter-counter-measure. >> >>It would be an endless battle. If I thought someone would cheat, I would >>simply elect to not play them. > >You fall into a typical trap for the expert. Of course it's enough to secure a >first control. You don't have to make safe computerchess like a paranoid. Of >course a criminal or a psychiatric case could destroy all normal social >behavior. No need to worry. If a scientist believes in his honest attempts of >control it is enough. But to deny the necessity of control is forbidden. > > >> >>the output comes from _inside_ the machine. I simply 'doctor' it before it >>comes out... IE it likes e4. I tell it to play f4. It plays f4 and the log >>shows analysis for f4, produced by the machine... >> >> >> > >Wouldn't you agree that a machine that had to be doctored by you is still not >the best chessplayer of the world? I mean the general "you" as a programmer and >expert. If you could program a machine to play chess on its own so that the >machine can decide when to switch to f4 depending the opponent and such, then we >have made some progress and nobody would speak of cheating anymore ... I would agree that if I helped it, it would not be 'best'. But we are on two different wavelengths here. I claim it is impossible to prove whether I helped the machine or not. Which means your question is not important since it has no possible answer. Is it possible to prove DB cheated? No. Is it possible to prove it didn't? No. What is left? Trust... Is it possible to do something to DB to make it more trustworthy? I don't see how...
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.