Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: ** Ethical Basics...** Indy and Indeep

Author: Rolf Tueschen

Date: 09:21:20 02/01/06

Go up one level in this thread


There is a main or key point behind all these aspects which isnt discussed
enough. Either because Kasparov created so much fog and Morovic explained the
double bind with all frankness.

What I called ethical basics is the perspective from human chess. If you are a
GM you must follow that commen consense. Only if your are below masterstrength,
say a good expert wit Elo 2000 or lower, then you have the freedom to play with
the so-called nasty mode and defend the honor of human beings against the - I
must clarify it - the operator side of computerchess. Programmers who mostly
have no strong insight into chess, do have the real relation in mind. They know
that their creations dont play chess. They know that it's all already unfair how
today machines can rely on all the normally forbidden tools in human chess. The
problem is the second row, the operators and eager freaks. They really believe -
like religious fanatics - that their computerprograms are really stronger than
human GM. Which is hyperbole and big nonsense. The solution for that delusion
was explained with the ethical basics of the typical GM who always plays his
very best chess (in a way every artist does his best) no matter against whom.
Against a lay a GM could otherwise sac two pieces and still win left-handedly.
Of course against a program he could win at will IF and ONLY if he would play
into the weaknesses of a machine. But no, he cant do it, he plays the opening
until move 25 with the machine as if the latter were a veritable GM with a
yearlong experience. In truth the machine has all the books and it looks what is
written down. In humanm chess this is called cheating. In computerchess this is
called "otherwise the machine wouldnt know that it's not a washing machine but a
chess player"! Again, the average operator believes in the veritable strength of
the machine as if it were a real GM, only much stronger and with Elo above
3000...

I said Kasparov made a lot of fog. What does it mean? Well, Kasparov was in
schizo mode. On the one hand he believed that he was the better chessplayer
which is true and false depending of the practical side. On the other hand he
made so much PR for Deep Blue and himself that he was obliged to make the
machine stronger than it really was. In the end Kasparov was confused by his own
weak preparation and a nasty aspect the other side brought into play on behalf
of IBM. They simply threw their science ethics into the bin and defined the
whole event as another form of war. But in that case Kasparov wasnt prepared at
all. And he would have had million of possibilities to make the machine look
like idiot. The way how he tried to make it he let look himself like the idiot.
That is the sad interpretation. But look, you shouldnt forget that this brought
him still a few hundreds of thousands of dollars... not bad for such a gamble.

The second ranks in the computerchess scene always thought that DB was now
really stronger than Kasparov, which is obviously crap. But it is also true
because Kasparov isnt the killer on the lose, he was always the well-prepared
academican who made something called autopsy, the same Kramnik later made with
him as the victim. As you know Kramnik to me is the best player ever from his
talent and concentration. He's so far above players like Leko or Topalov that he
gave Leko the advance and forcedly he had to win the last game to draw the match
and keep his title. If you know how difficult it is to beat Leko on command then
you know what Kramnik has done. Also if you consider that a draw is so natural
between the higher GM, in special for Leko. Kramnik is so strong that he can
draw Fritz to let the business have their PR and still take the money for
himself. The second row here in CC does really think that Kramnik is actually a
weaker GM just because he's so nice and let's his collegues also win a few
thousands whereas he has made millions. I digress.

Why I brought the ethical aspect into the debate is still a different. I am not
a GM, I have the freedom from science to make my own thoughts. And I say that
it's fundamentally false to take a program that is still in the making, that has
no endgame knowledge no nothing, at least for version 7, a beta, then take a GM,
make 5 pics in the cubby of the hotel. Below 1000 games there is no possibility
to declare a winner at all! Because with a 1,5:0,5 we have a confidence
intervall for Chile of +/- 370! It makes no sense. But the second ranks in CC
see the result as the proof of the superiority of Rybka. For me as a scientist
Fern did prove a single fact. That Chilenian women have black hair! I knew that
before, the best example is the new President of Chile...




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.