This warning from Stephen Hawking got me thinking. First of all I was surprised that he was taking the trouble to lend support for this idea which has been depicted in the movies on numerous occasions. And then I started considering chess players’ relationships with computers and how they’ve changed the nature of the game.
Computers have certainly led to massive advances in the fields of training and preparation; now even some players below 2200 can effectively use engines such as Houdini or Stockfish to prepare critical positions. This has led to many top players eschewing sharp theoretical lines and instead choosing to slug it out in dour positional struggles, with Magnus Carlsen being the leading representative of this approach. Speculative gambits have become quite rare as the work required to prepare them is largely wasted; it’s a serious risk to play the same line in more than one game as future opponents may be very well prepared.
So what lines are good? Basically just about anything that puts the emphasis on the middle game in which both sides have lots of playable alternatives. Your opponent can still prepare using a computer database, but he’s not likely to unleash a decisive opening innovation.
As for artificial intelligence, let’s keep them on a tight rein. I discussed this matter with Michael Koblentz on Facebook and he cited Koblentz’s law of robotics. This included such common sense measures as not given computers weapons, allowing them unilateral control of life support systems, build other computers etc. All common sense really, and of course we should never, ever, let them play chess.