Hacker News new | ask | show | jobs
by EGreg 1724 days ago
Sounds like how AlphaGo beat Stockfish or Rybka… massive precomputing vs alphabeta search in real time
1 comments
sillysaurusx 1724 days ago
Hmm. AlphaGo and AlphaZero were trained using self play. I don't think there was any precomputing, unless you count training as precomputation.
link
EGreg 1724 days ago
Precomputation on a huge cluster for a long time
link