Author: Robert Hyatt
Date: 20:25:47 05/06/00
Go up one level in this thread
On May 06, 2000 at 22:55:11, Hans Gerber wrote: >On May 06, 2000 at 20:57:44, Pete Galati wrote: > >>Specifically what "iron rules of scientifical methodology" are you referring to? > > >A little bit difficult to explain to laymen. Let me answer this way: If you >construct a machine you should be able to control the output of the machine. >Otherwise how (worst case) could _you_ understand if the machine is doing what >you wanted it to do, and how could you demonstrate to _others_ what the machine >is doing. How could you even discuss the results without having exact knowledge >of the 'nature' of the output? How could others repeat and understand your >experiment? To give just a few aspects of a complex problem. > What does this have to do with 'science'? Computer chess has been going on for 30+ years now. Search is pretty well understood. Parallel search less so. Evaluation is a bit of experimentalism... But _none_ of this has anything to do with being able to detect or prove the absence or presence of cheating. It simply can not be done. You give me an experimental set-up that you like. I'll explain how I will still cheat without your being able to detect it. > >> Because I don't really see any conflicts myself. If you are referring to those >>log files not being able to prove that IBM didn't cheat, I see no problem here, >>I only see that as the nature of the beast. > > >Only if the beast was constructed this way. Why so much confusion? If the moves >of the machine would be so strong to cause big trouble for one of the best >chessplayers, then why the need to make the output so confusing? Intention or >impotence? Unbelievable aspects in science ... what is confusing? I had no problem understanding their output at all. Many others had no problem. I think some _wanted_ there to be a deep plot or conspiracy. But the output was crystal clear from day 1. For those that wanted it to be clear... > > >> >>Labeling any of this science is a little bit hard for me to swallow, this was by >>no means a piece of cake to pull off, because there was enormous amounts of work >>involved, but I'd rather label it the greatest publicity stunt I ever saw, than >>science. > > >Good one. But it's not so difficult to explain why this was science. _Because_ >Hsu Feng and some of his aids are scientists. They are "forbidden" to violate >the rules of science. What if they did? Well, they would lose their reputation >as scientists. However they could still cooperate with David Copperfield ... > You say "violate". Can you give a clear example of what you mean? I don't see anything that fits this concept in the DB match, either number one or number two. > > > >> What I did find somewhat suspicious though was that Deeper Blue was >>never used again after that match (that I know of), so I'm unhappy with IBM >>about that. Did they create that beast only to play against Kasparov? A waste. >> >>Pete > > >If DEEP BLUE could have been regarded as the best chessplayer of the World >beyond any reasonable doubt, the destiny of the machine would have been >different. Beating Kasparov was _the_ goal. No one ever claimed anything else. Once that was done, they 'moved on' to other problems. I can't imagine a company continuing with computer chess after beating Kasparov. There is nothing else to win, a lot to lose, and it would cost a lot of money. They got their advertising landslide.. That was all they wanted. Hsu proved it could be done, IBM made it possible. Kasparov assisted. Basically, the end of the story.
This page took 0.04 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.