Re: Some Refactoring Notes



Brenden T. wrote:

Having played go and read up on the theoretical aspects of its AI, I'd agree that it's at least not as well understood as chess, if not actually harder.

The new version of GnuGo (3.5.something) is being tested on the server where I play (kgs.kiseido.com) and it's darn good. They've made several nice improvements to the AI. It's merciless against low level players and will give even higher level players a tough time. One good player (8 kyu) was playing against it and said "sweet" and "this computer [program] rocks" in the chat channel in response to it's moves.

It's just happy coincidence that GnuGo was on my mind when I saw the activity on this list pick up.

I absolutely agree that Go AI is likely harder to code. Nothing in go is clean. Even "opening book" stuff. The reason is that openings are a localized thing, But can get influenced by stuff even on the opposite side of the board... It's a very interesting game. (As a note, I am about 18k and therefore have much to learn)



[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]