I could not disagree more with John Nunn. The man vs. machine
matches are far from over. Hydra is far from perfect and has been defeated
in correspondence chess match. However computers have many advantages over
humans. Firstly, they have complete opening books and endgame table bases.
Secondly, computers will never error in retrieval of this data and will
never make a short-term tactical oversight. Finally computers don’t get
tired or feel pressure. They will always play near or at their level of
strength. In reality, Hydra at its maximum strength has transformed these
matches into man vs. near perfection. As the machines get stronger so will
the top players.
I think these matches will have more meaning if the very best player
plays against it. I would like to see Anand take on Hydra. I would like to
see Kasparov play against Hydra, provided that Kasparov can check his
reputation
at the door. During the Kasparov vs. Deep Junior match, Kasparov was afraid
to lose another match to a computer. This fear prevented Kasparov from
fighting on in the final game, even though he had an advantage. Millions of
fans were watching on ESPN and the game and match ended in a draw. I think
we need to vary the experiment. Lets give the human player twice as much
time to solve the defenses of the computer. I was a little disappointed in
the final out come of the Kasimdzhanov vs. Accoona Al Toolbar man vs.
machine match. The game was very exciting; Kasimdzhanov went against the
conventional wisdom when playing against computers. Kasimdzhanov heroically
sacrificed a piece in order to repair his pawn structure and open lines of
attack against Al Toolbar’s king. However Kasimdzhanov found himself in time
difficulty and to force a draw by repetition. He felt he had a winning
position but against the computer he only had 10 seconds for each move, so
it was too dangerous for him to continue.
There are many ways to vary this experiment.
1)
Give the human player twice as much time
2)
Turn off the computer’s opening and endgame databases.
3)
Reduce
the computer’s opening databases
Eventually the computers
must learn how to play the opening without any opening book database.
Kasparov made this point at the last computer matches. With opening books
expanding well into the middle game, when does the computer actually start
showing its talent? Kasparov said that as computers get closer to
perfection, even one victory by man will signify mans dominance over the
machine. I have to agree with that statement.
Because of the one sided result of this match (Hydra winning 5.5-.5), many
people criticized the Hydra vs. Adams match, but I thought it was an
exciting match. I enjoyed the decisive games. I salute both Adams and Hydra
for the entertainment. Adams probably should have taken a serious look at
the Kasparov vs. Deep Blue matches. He should have employed more
anti-computer strategies It might be too early to say that Hydra
represents the next great leap in computer chess. It will be hard for Hydra
to captivate the world like Kasparov and Deep Blue did in 1997. However
Hydra vs. man matches will be very interesting and exciting matches. Bottom
line, the man vs. machine matches will be good for promoting the game of
chess. We also have many ways to vary these matches to make them more
exciting to the public. And if the computers get too strong, then they can
play a simul against the top players of the world.
|