Re: My top 3 SF writers of all time

:: the issue is, what can T.H. do "in her head", and there's no
:: particular reason she should be as good at that as an AI who
:: specializes in attempting to understand chess deeply, even if they
:: both had tons and tons of computer power.

: "Gene Ward Smith" <genewardsmith@xxxxxxxxx>
: I didn't say as good as KingFisher, but what about Bobby Fischer?

I don't think there's any reason to suppose she would be as good as Bobby
Fischer, either. I've probably got as much computational horsepower as
he does in my haid (and if I don't plenty of other people do), yet I'm
a total incompetent at chess. And billiards, and many other things,
but stick to chess.

Again, the issue is what she can do "in her head", not whether she could
run a program and quote the moves it makes; I'm quite sure she could
do that. And even use googleish workings to fake a certain plausible
veneer of expertise.

And naict iirc the events in the stories don't conflict with that model.

: That, and the fact that KingFisher, a Turing-level AI, specializing in
: chess, doesn't seem to be the world champion. They both should beat
: the pants off of mere humans.

KF should be able to beat most humans "in his head". Do we have textev
that he doesn't?. TH should be able to beat quite a few humans by running
standard software packages. But heuristic and AI-like approaches to chess
aren't as fruitful as all that. Or haven't been, afaik. And isn't it
still true that computers that beat the top humans still have specialized
hardware assist? Or has stock hardware gotten fast enough?

It is entirely possible I'm just mistaking the state of computer
chess. But while this issue raised a blip on my implausibility radar,
it didn't sound klaxons or anything.


Wayne Throop throopw@xxxxxxxxx http://sheol.org/throopw
.

Relevant Pages

  • Re: Understanding descriptions
    ... humans, is that they have'nt processed board state class descriptions, ... The reason why chess programs are so simple is that they are very ... and play chess not at all like humans. ...
    (comp.ai.philosophy)
  • Re: Understanding descriptions
    ... chess playing programs had search trees only two deep (Newell et al. ... and play chess not at all like humans. ... and there is no pressure to make Chess programs that play using ...
    (comp.ai.philosophy)
  • Re: the Mathematics of Chess
    ... to Stravinsky or some other atonal composers, such as Scriabin or Shostakovich, ... to analyze in terms of harmony, because he continuously breaks the rules. ... Although certain aspects of chess also can defy understanding as far as ... but with humans this is not always the case. ...
    (sci.math)
  • Re: Hydra crushes Pono in 29 moves
    ... >> open file) looked deeply, ... Humans and machines _are_ different when it ... Chess, of course, entails quite a ... Pitting a human against a machine in OTB play ...
    (rec.games.chess.analysis)
  • Re: Understanding descriptions
    ... humans, is that they have'nt processed board state class descriptions, ... but can be applied only to chess and few ... chess playing programs had search trees only two deep, our modern chess programs completely rely on deep search trees and play chess not at all like humans. ...
    (comp.ai.philosophy)
Loading