Game Over: Kasparov and the Machine

In the May 1997, Gary Kasparov, the reigning Chess World Champion and by the opinion of many, the greatest chess player ever played Deep Blue, an IMB Supercomputer. At its best scenes, the film is an entertaining look at the never ending competition of human intellect against artificial. The greatest player on Earth does not like and does not know how to lose, and his account of the match and its result is quite bitter. He can't believe that the computer program, the combination of 0s and 1s may appear to think like a human. It was sad and nostalgic for me to see Gary like that. I remember him back in 1985, 22 years old World Champion after his victorious match with Anatoly Karpov. In his (and former mine) country millions of people that knew nothing or next to nothing about ancient game of chess (All I know that the first move e2 – e4 will not bring me any problems, at least for a little while) passionately wished him to win. Gary was not just a brilliant chess genius, a wonder-boy – he was also a symbol of hope, of changes not only in the chess politics but in the life of the whole country that was ready for changes.

Not sure why everyone is saying that this is biased. There was lots of screen time for IBM. Their reasoning for concealing everything was pretty weak in my opinion. Legal mumbo jumbo about private vs public matches, absurd excuses about concealing code and game logs. Not on the up and up in the least. The movie itself was a fascinating discussion about the nature of corporate PR and the possible taking advantage of a tightly controlled publicity campaign. They made it seem like everything was out in the open, but if you look a layer below, the signs of opaqueness are, at the least, concerning. Engineers being allowed to tinker in between games, putting the computer in a locked room, refusing to display code or log files, immediately abandoning a rematch and destroying the computer. The IBM engineer in the movie even seemed to be gaslighting and turning the screw deeper on film. He relished in the controversy and mind games. If that is the man behind deep blue, I would be very concerned. He seemed like a sociopath. But that's just my two cents. If they played it fair, they certainly didn't behave like they did after the fact. Let alone the suspicious mistakes mixed with brilliant moves in game 1 and 2. The movie is fascinating and less biased than people give credit for. The turk B Roll is a bit gratuitous, but overall, IBM got to march out their weak excuses for hiding and controlling everything. The upside for them was way too high to not make all the circumstantial evidence baseless. Its a must see if you are a chess fan.

Vikram Jayanti's documentary on the 1997 clash between IBM super-computer Deep Blue and Russian chess grand master Garry Kasparov is frustrating and fascinating in equal measure.

Kasparov's insistence that IBM cheated (by using one or more chess masters to influence the computer during the match) seems perfectly possible, but the computer manufacturers refusal to allow either a rematch or analysis of their data makes it impossible to substantiate the claims. Sadly that doesn't stop Jayanti, his subjective approach is unaffected by the apparent stalemate.

Fortunately the film has more to offer. The actual match becomes hugely dramatic and exciting in Jayanti's hands. Kasparov himself is an extremely enigmatic and passionate figure, the footage of him in the press conferences that took place after each of the six games is dynamite. In the second, he accuses IBM of cheating. In the last, after nine days of play, he appears on stage looking physically and mentally destroyed, the applause that greets him (and the boos for IBM) would seem to indicate a general feeling of suspicion of IBM's sportsmanship and honesty.

Too subjective to be a 'great' documentary, it is still a fascinating insight into a game and community that would seem to offer much potential for future study.

I'm not sure who this movie is targeting. There are interesting tidbits concerning the history of the challenge to make a chess machine. These might intrigue both chess fans and non-fans alike, though much more could be made of this, as the history is richer than even this film implies. More could also be made of the history between Kasparov and his arch-rival Karpov (two almost perfectly matched players, though you'd never guess from this movie). More could be made about the connection between chess champions and paranoia, or between chess and politics in the USSR (a connection which makes one understand better why chess players are so paranoid).

Instead, the makers of this film push the silly idea that IBM's Deep Blue beat Kasparov in '97 because of human intervention (ie, IBM cheated). The film bases this on one piece of evidence: Kasparov believes his loss in game two of the match was the result of a move that no computer would ever make. This is made all the sillier because a typical home chess program (Fritz 7) makes the very same move as Deep Blue after only a moment's thought. The film also claims that IBM never released the logs of Deep Blue's analysis after the game (just go to IBM's historical site concerning this match, and you will see this is not accurate).

Are documentaries getting lazier with their facts, or am I just finally wising up after years of taking them at their word?

I saw the documentary, and saw the actual games in 1997.

Kasparov could surely be a fine actor, since the guy is very expressive and charismatic.

Whenever he felt good and winning, you could see it. And when he was losing and crumbling, you could see it too. Was way obvious. Like the documentary say, Kasparov would be the worst poker player in the world.

Now, did IBM cheat?

Who knows. Anything is possible

Arguments against it: Kasparov could've taken for granted Deep Blue's playing antics as the one of a normal computer, and since IBM had a grandmaster chess player advising the programmers, it's not so wild to conceive that they managed to program Deep Blue to be able to spot traps like the one Kasparov set up that wouldv'e nailed any other computer. And he's a paranoid (coming from the U.R.S.S. no one can blame him), since chess is also psychologic warfare, IBM surely did it's best to psyche out Kasparov and play his paranoid assumptions.

Arguments for it: Kasparov won fist game easy, but lost second when Deep Blue didn't take a bait a compute would've taken. Maybe losing the first game was intended to lower Kasparov's defenses so he would try a play like that, and there Deep Blue would surprise him, psyche him out and steamroll. But that's a human strategy, not a machine's (the computer only knows the game in front of it, doesn't know there are six games total, so it would NEVER sacrifice one to try to surprise Kasparov in the next one).

Also IBM,s attitude, while could be attributed to psyching out Kasparov (fueling his paranoia), looked totally like a cheater's conduct. Also when they won (no rematch, no further research, dismantling of Deep Blue) doesn't look like honest behavior (first truly artificial intelligence? Who would NOT follow through with research?), but like a cheater who won and now has to skip town before he's discovered.

And, the final nail: Why shouldn't IBM cheat? To IBM, it's nothing but a marketing stunt, nothing else. The whole point was not to beat Kasparov or improve artificial intelligence (or they would've continued the work on Deep Blue, published the groundbreaking work, patented programming code, etc), but only to improve stock value and reposition themselves on the market. So why not cheat if necessary? Like a company would be above that (Enron, etc.) or anything for that matter to increase profit.

But since there's no way to verify what Deep Blue did (thanks to IBM, like e-voting paperless machines, "trust us"), we'll never know...