Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Answers (What means Super-GM and what is DJ?)

Author: Rolf Tueschen

Date: 10:14:02 02/16/03

Go up one level in this thread


On February 16, 2003 at 12:12:54, Robert Hyatt wrote:

>That seemed to be very good, and a move my program was not going to find, which
>only means I have more work to do.  However, I'm a bit concerned about making
>a great move here and an ugly move there.  The ugly move there will hurt more
>than the great move here, in the long run, if the goal is to play at the top of
>the heap (Super-GM chess).  At that level, one bad move is generally all it
>takes, while it might take several consecutive great moves just to hang on.

Of course it was no blunder, but I didn't ask for that reason. My reason was was
it a book move or not. Also, I remember a debate about CSTal when you explained
exactly what you mention here now, for the benefit of the overall performance
you can't allow a program to gamble with high risk. That wins 1x but loses 5x.
Something of that sense.

But I have a difficult question for you and it seems that exactly that one you
avoid to answer as if you feared the excommunication or a life ban to hell. Amir
and Kasparov are saying that DJ is now a super-GM. Kasparov says DJ is better
than DB2. Uri says that factor </=100x in speed is nothing if software made
progress. You say that you didn't know if the GM would destroy their future
income even if billions of dollars would wait.

I asked you a very sound question in the other message ad that goes like this:

Say GM are really determined and cooperate on the "weaknesses" of computer
programs, who are super-GM, would these GM find a way to kill a) these comps b)
the hype in the debate here, namely that DJ is a super-GM, or couldn't they.
I said, all that from you under the impression of your experiences with GM
thought processes in minor problems with Crafty. Or is that question one of
importance for the security of whole Nations? :)

And let me also confirm here, that I do NOT ask to get support or an alibi. Or
the motivation to put Amir into Formalin. I ask from a more science point of
view to have a sound argument against the actual Kasparov PR. For me the answers
are clear although I am no expert in CC. Because the games speak their own
language. But I ask for the benefit of many young readers here who could be
confused such business PR. I for one have NO interest whatsoever to dump Amir.
He was always IMO on the science side up to now. But after this 3-3 I can no
longer follow his logic. But perhaps this is also no myth for yourself because
you also had moments of comparable delusions after your Worldchampionship wins.
In short, is Amir a bit unrealistic or what is going on? Or is he just a
positivist claiming the Elo performance as the new Elo number? Which is wrong
after 6 games and a show event.

Rolf Tueschen



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.