←back to thread

Grandmaster-level chess without search

(github.com)
365 points lawrenceyan | 4 comments | 17 Oct 24 19:13 UTC | HN request time: 0.656s | source
1. BoardsOfCanada ◴[18 Oct 24 05:38 UTC] No.41876628[source]▶
It's somewhat telling that they chose Stockfish as the oracle and not AlphaZero.
replies(2): >>41877566 #>>41878266 #
2. Davidzheng ◴[18 Oct 24 09:00 UTC] No.41877566[source]▶
It's well established that SF today is stronger. Btw this is pretty bullish case for open source recreations of RL advancements to be catching up with ~2 year lag. Hopefully alphaproof is also reproduced in that timeframe
3. epolanski ◴[18 Oct 24 11:08 UTC] No.41878266[source]▶
Stockfish is stronger than AlphaZero or any other chess engine from quite some time.

Another interesting tidbit, both engines share (or at least, shared) the same lead developer.

replies(1): >>41880543 #
4. vlovich123 ◴[18 Oct 24 15:51 UTC] No.41880543[source]▶
To be fair though, deep mind hasn’t been working to improve alpha zero for chess playing. Their focus is a bit difference than the dedicated volunteers working on stockfish (+ it’s been a few years)