Computer Chess: Learning From the Best

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Computer Chess: Learning From the Best

    'Computer Chess: Learning From the Best' is a good read from 'The Chess Mind' blogger, Dennis Monokroussos.

    http://worldchess.com/2015/10/29/com...from-the-best/

  • #2
    Re: Computer Chess: Learning From the Best

    Originally posted by Jack Maguire View Post
    'Computer Chess: Learning From the Best' is a good read from 'The Chess Mind' blogger, Dennis Monokroussos.

    http://worldchess.com/2015/10/29/com...from-the-best/

    There is a question asked in that article that I have brought up here many times:

    "why is it that we think the games of players rated near or over 2800 are essential for chess fans and students, but most of us don’t think twice about disregarding the best games of players rated over 3200?"

    No one has even attempted to give an answer to this. There seems to be a mass denial going on that computers play the best chess known to man. I think most chess followers believe that yes, computer engines beat humans, but they do it with tactics. To these people, the engines are 'cheapo' machines, beating the humans only by superior tactical calculations, taking advantage of the fact that humans simply can't see all the tactics in a complex middlegame and inevitably must make a tactical oversight. Historically, there is much to support this notion, but is it possible that 'the times they are-a changing'?

    The makers of the game Arimaa have long touted that engines can't dominate their game presumably because long-term strategy outweighs tactics, and so the humans beat the engines. Well, finally this year, an engine did defeat the best human players. Perhaps we are on the cusp of an age where computers can not only tactically, but strategically outthink us. And they might not even be doing anything revolutionary to achieve that. It could simply be that they have, due to speed improvements, passed a critical threshold in search horizon that enables the engines to capture the effects of some long-term strategies. Our brains still work differently from their brute-force calculations, and as long as their brute force search wasn't fast enough to see all long-term effects, we were strategically better because our brains recognize and classify patterns better.
    Only the rushing is heard...
    Onward flies the bird.

    Comment

    Working...
    X