Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Two of the Deep Blue moves protested by GM Kasparov

Author: Rolf Tueschen

Date: 15:05:00 07/22/02

Go up one level in this thread


On July 22, 2002 at 13:35:40, Matthew Hull wrote:

>
>>>If it hadn't been, we would have needed to search for _other_ reasons why DB2
>>>might have chosen Qb6.  But, we don't need to now, because we _know_ why it
>>>played that move.
>>
>>Objection as before. You can't compare bananas with asparagus!
>
>A better analogy here would be Red apples to green apples.

Please, how could you interfere in my private tutoring of Bob (Dr. Hyatt)? I
pick up my students in their usual neighbourhood. This wasn't against computers.

>Computers are well
>known for making "computer" moves, which is of course a pejorative.  In general,
>this is due to the fact that computers cannot (in general) search to the depth
>where GM strategy is operating.

Why is it pejorative? It's against those who show too much hybris for their
progs and lack of respect for GM. For a scientist it's trivial but for normal
people it's very difficult to understand why proofs with commercial machines
today can't prove the play of DB2 of 1997. You can't simply reduce the question
on speed. In principal it's not possible to prove something after 5 years and
with different machines. But this is too difficult too to understand. This is
the bad situation. Bob is intentiously misleading the public against his own
better knowledge as a scientist. He knows what is possible and what is not -
methodologically! In 1997 he presented game results of the sort of 10:0 without
a single score record. Without a single move! This is unsound and unallowed. I
could give more examples.



>
>This is slowly changing however.  I think this is what shocked Kasparov.  Here
>was a machine whose speed could take it to the next level in some positions.
>When the machine did not make the expected "computer move", K became suspicious.
> But there need not have been suspicion, because we know that compterchess will
>"get there" eventually.



To a certain degree you are quite right. But do not dream of such a devolopment
in a complete sense! For decades to come human chessplayers will have certain
advantages after some training. Computers have a principal blindness from a
depth on. This will infect their chess. The moment human GM know such holes they
will try to exploit the weaknesses of the machines. To prevent exactly this
programmers and other CC workers dive into secrecy and gambling. But super GM
can't even be disturbed by such tricks. Do not believe in such myst that
Kasparov was shocked! It was more the other way round as we've found out. He was
perhaps too little prepared. But then this was the DB2 side who didn't publish a
single game of the preparation time. It's well known in chess that you can't
conjure wins at will if you have absolutely no knowledge of the idendity of the
opponent. Look, in human chess players have been known for years when the enter
the tournament circle. Perhaps you also heard of the event that the best GM
often lose or draw in simuls against total amateurs. This is nothing spectacular
in show events. So basically Kasparov wasn't beaten fair and straight in 1997.
It was a show event and IBM simply psyched out Kasparov. It wasn't for the World
title! That Americans tend to think in World title dimensions is regarded as
very amusing in other regions of the World. For me the show didn't prove
anything. But if USA thinks that IBM has won the Fredkin prize, then it's fine.
It soesn't mean much, because under serious parameters the show wasn't a fair
match.

> The fact that DB demonstrated this to some degree
>(axb5/Qb6) is being comfirmed by long searches of lesser machines.

This is what Bob is saying too. But as I showed there is no way of confirming
DB2 by today's machines.


>
>To try to define the debate down to whether the DB machine could have found the
>move, versus another program is just silly.  One day all programs will find the
>GM moves as a matter of course.  The fact that DB was getting there sooner than
>the others is only logical, considering it's wonderful speed!

This is a strange kind of logic. This is totally alien for me. You simply can't
constitute reality with describing possibilities. THis is wishful thinking.

>
>Wouldn't you expect that to be the case?  It's just a logical progression.

No. There is no progress in logic. Logic is over two thousand years old. What
you mean is that we can logically expect something to happen. But for the moment
we want to know if DB2 was able to find axb and to deny Qb6. In combination with
the deconstruction of the machine it wasn't proven by no means. The DB2 team had
the duty to show evidence. A print weeks later is no proof since such a print
could be doctored within minutes as Bob explained!

Thanks for your interesting argumentation.

Rolf Tueschen

>
>Regards,



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.