Quote:
Originally Posted by CyberSEO
A machine has won a chess party against Kasparov in 20th century. So what? How does it relate to the AI development program?
|
I've just skimmed the article but I think the general idea is that the number of possible moves in a Go game is significantly greater than that of a chess game. You cannot simply brute force by simulating the next X moves. This is more about finding short-cuts to solutions in an impossible vast search space (too big for brute forcing).
Despite what DV and the BBC say, I doubt this is really anything "new." Heuristic searches, genetic algorithms, hill climbers etc have been around for decades. Refined, perhaps.