There are two chess videos among many on Youtube that have stuck in mind amid all the recently elevated discussions about the future of humans in the context of rapidly evolving AI capacities.
The first is a video of the American GM Yasser Seirawan taking a class through the famous Kasparov-Karpov "octopus knight" game in their 1985 world championship match. Seirawan is a fantastic chess teacher and presenter and brings alive the genius of the game. Kasparov makes a series of moves that eventually render Karpov's entire position helpless to the point of having almost no viable moves with most of the pieces still on the board. Karpov being Karpov continues to battle and Kasparov ends up finishing off the game with a series of tactical moves that most normal players wouldn't spot in a life-time. There are a number of occasions during the video that made me laugh out loud just even thinking that somebody could think of a move of that nature.
At the stage, chess computers were nowhere near being at a level where they could reliably beat someone like Kasparov. It was 1997 before Deep Blue defeated Kasparov in a match. It is often reported as a milestone in human/AI interactions, a marker of machines becoming more intelligent than humans. But the fact that Kasparov more than held his own over two matches and only very narrowly lost the last one makes it clear that the world was very different to now where computers are at a point of undisputed mastery in chess. Kasparov famously was disbelieving that Deep Blue could have made the decisive move in the last match, and claimed that a human must have intervened, on the basis that the move was more creative than he believed machines were capable of producing. In any case, at that point chess computers were incredibly impressive but by no means super-human, with Kasparov winning one and drawing four of the six games in the 1997 match, having won the 1996 match 4-2.
See below a video of the world number 2 Hikaru Nakamura talking about Stockfish, one of the main current chess programmes. Nakamura is easily in the top ten human beings ever to have played chess in terms of rating and talent. He is also a very charismatic broadcaster and speaker and drawings millions of viewings across his different channels. His description of the gulf between him and Stockfish gives a strong sense of how much has changed. There is basically no longer a hope a human being could beat a leading chess programme. He suggests some possibility that certain types of games might very occasionally end up in a draw but doesn't seem particularly hopeful of that.
And yet chess is flourishing. Audiences are growing and many more women are playing than was the case years ago. In countries such as India and China it has moved to being a key popular game in the broad culture and next to soccer must rate up there as one of the key truly global cultural exchanges. On a personal level, I don't emotionally feel anything has been lost in terms of the achievements of someone like Kasparov in producing a game like the above. Part of the experience of playing through a game like that is simply to imagine how someone could have produced those moves under human constraints, time pressure, and the pressure of the historic occasion. There is something oddly empty about working through the games of computer programmes rated well beyond humans playing one another. It is never clear how far to stretch chess analogies into other areas of life but it is certainly the case that the end of the world came to chess in terms of computers essentially solving the game, but it has come through the experience mostly fine (notwithstanding cheating scandals) and found ways to live alongside the AI.
No comments:
Post a Comment