The lessons that chess legend, Garry Kasparov, didn’t intend to teach us



Original Source Here

Game 1: A Kasparov victory with disturbing implications

The opening game of the series was progressing rather evenly. Kasparov was playing as white, giving him the upper hand. On his 30th move Kasparov put his brilliance on display. Kasparov sacrificed his rook for the price of a bishop, usually this trade would be seen as unfavourable for Kasparov, but Kasparov realised he wasn’t just trading his rook for a bishop, he was also trading it for a complex long-term positional advantage. Kasparov correctly assumed that Deep Blue would be unable to recognise the positional disadvantage it was creating for itself, he believed making this observation would require a human’s intuition.

Within 15 more moves Kasparov had decimated Deep Blue and the computer resigned. However, there was a moment that caught the attention of Kasparov and his associates. On its 44th move, it’s final turn before resignation, Deep Blue made a bizarre decision, it moved it’s rook to the first file for no discernible reason. After further analysis, Kasparov concluded that the computer must have been able to foresee a checkmate 20 moves ahead, had it played a more obvious move. Kasparov assumed that Deep Blue rather opted for a move that would delay checkmate the longest. Kasparov was spooked, if Deep Blue could predict the next 20 moves, it may prove very difficult to beat in his upcoming encounters.

Game 2: Kasparov’s fatal error in judgement

In game 2, Deep Blue would have the advantage of playing as white. It made the most of this by pressuring Kasparov throughout the match. On the 36th move there seemed an obvious tactical move that Deep Blue could make in advancing it’s queen, to further elevate the pressure on Kasparov. However, Deep Blue opted for a strangely ‘human’ move, it made a pawn trade that intended to improve it’s board position. The more Kasparov assessed the move, the more concerned he grew.

He considered that the computer would only make such a move if it could see it amounting to an upcoming tactical advantage. He played out 8 more moves and as the situation grew to look increasingly dire for Kasparov, he shockingly resigned. Kasparov realised his suspicions about Deep Blue must have been correct, the computer must have seen 20 moves ahead and he accepted that any efforts to stand his ground would be futile.

But Kasparov was wrong! The following day, Kasparov’s assistant revealed to the chess deity that he could have in fact forced a draw just 7 moves after the point he retired. Kasparov hadn’t just lost, he had committed one of the great cardinal sins of chess, he had resigned a game in which he could have still forced a draw. Kasparov was completely dejected, but he still had 4 more games to play

Game 6: The final straw

Games 3, 4 & 5 all resulted in draws, which meant the series would come down to one last match. Kasparov was to try and forge a victory as black, an unenviable task. Unfortunately, Kasparov was completely drained and it showed in his gameplay. Kasparov played a well known defence, known as the Caro-Kann Defence and knowingly opened himself up to a devastating knight sacrifice, assuming Deep Blue would not know to sacrifice its knight. Deep Blue sacrificed it’s knight. 19 moves into the game, Kasparov retired. Machine had defeated man, it was a landmark moment in human history.

AI/ML

Trending AI/ML Article Identified & Digested via Granola by Ramsey Elbasheer; a Machine-Driven RSS Bot

%d bloggers like this: