21

Evaluating 200M positions per second and having a search tree of 20+ moves might not be enough to beat a World Champion if a computer searches in the wrong direction or if it's evaluating positions a human opponent would never consider (either by instinct and/or reasoning). Coding tactics is easy, coding strategy is hard.

Was Deep Blue human assisted when it beat Kasparov in 1997? Is there any evidence about that?

Can a commercial software chess game running on a "common machine" beat a chess master nowadays?

  • 1
    Crude comparison, but the #1 (public) supercomputer in 1997 was Sandia's ASCI Red with 1 TFLOP performance (Deep Blue had about 0.01 TFLOPs, but it was more purpose-built, so probably more efficient). A single NVIDIA GPU (P100) can deliver 5 TFLOPs in 2016. Consumer-grade GeForce hardware matches that; so $67 million of performance then is about $600 (divided by 5) today. – Nick T Dec 29 '16 at 22:50
22

Game Over: Kasparov and the Machine is a decent enough documentary on the subject. The coverage does kind of imply that Kasparov was grasping at straws abit. Like a typical World championship match there was a lot of behind the scenes tactics going on from both sides.

My feeling is that you only have to look at the ascent of Anand and other chess players who rely heavily on computers in their training and analysis to realise that, at the time of this match, there was a paradigm shift from human to computer as the dominant chess player.

Addendum:

There's a interesting tidbit in Nate Silver's “The Signal and the Noise: Why So Many Predictions Fail—But Some Don’t.” In it he describes an interview with one of the engineers who worked on Deep Blue:

Nevertheless, there were some bugs in Deep Blue’s inventory: not many, but a few. Toward the end of my interview with him, [Murray] Campbell somewhat mischievously referred to an incident that had occurred toward the end of the first game in their 1997 match with Kasparov. “A bug occurred in the game and it may have made Kasparov misunderstand the capabilities of Deep Blue,” Campbell told me. “He didn’t come up with the theory that the move it played was a bug.” The bug had arisen on the forty-fourth move of their first game against Kasparov; unable to select a move, the program had defaulted to a last-resort fail-safe in which it picked a play completely at random. The bug had been inconsequential, coming late in the game in a position that had already been lost; Campbell and team repaired it the next day. “We had seen it once before, in a test game played earlier in 1997, and thought that it was fixed,” he told me. “Unfortunately there was one case that we had missed.” In fact, the bug was anything but unfortunate for Deep Blue: it was likely what allowed the computer to beat Kasparov. In the popular recounting of Kasparov’s match against Deep Blue, it was the second game in which his problems originated—when he had made the almost unprecedented error of forfeiting a position that he could probably have drawn. But what had inspired Kasparov to commit this mistake? His anxiety over Deep Blue’s forty-fourth move in the first game—the move in which the computer had moved its rook for no apparent purpose. Kasparov had concluded that the counterintuitive play must be a sign of superior intelligence. He had never considered that it was simply a bug.

9

It isn't hard to find machines/programs rated 2200+ nowadays, so yes machines can beat masters.

Kasparov's comments versus Deep Blue are sour grapes. Kasparov once boasted that a machine could not beat him because he could create and a machine could not. He failed to realize that sufficiently deep tactics are indistinguishable from creation ;-)

  • 1
    To back up the comments a bit, Wikipedia has a nice summary of human vs. computer chess history. I believe this part sums it up nicely: In 2009 a chess engine running on slower hardware, a 528 MHz HTC Touch HD mobile phone, reached the grandmaster level. – Daniel B Dec 11 '12 at 11:42
5

Kasparov accused Deep Blue of cheating, but I don't believe there is any evidence about this.

Anything can happen on any day, so yes, a commercial software chess game running on a "common machine" can beat a chess master.

2

Kasparov did just (beginning of 2017) publish a new book on this topic. He comments on the whole IBM fair play issue in the book.

In the book Kasparov also reflects on why IBM placed Russian speaking security guards at his hotel and why a specific change was made to Deep Blue on the very same morning that the final game was to be played. And exactly a game play change related to the very opening that Kasparov had discussed with his team about the night before the final game.

Se also:
http://www.kasparov.com/deep-thinking-ai/

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.