4 ms·
Could play against adversarial policies become part of the training alongside self-play? It's possible that would be a very slow, essentially "serial" process-
by twotwotwo 4y ago
Could play against adversarial policies become part of the training alongside self-play?
It's possible that would be a very slow, essentially "serial" process--adversaries don't really find a new trick until the AI can avoid the old, like with the pass attack in this paper. I'd also be curious whether training against adversaries makes the networks better or worse at general-purpose play or causes no change, but there's an argument that more-robust-but-slightly-worse-on-average is an overall improvement.
It also seems relevant to looking at weird behaviors of other models, like what we're seeing as people poke at LLMs more. Some of what we're seeing might simply be that they respond weirdly to some fraction of all prompts because they're not quite baked; also, we haven't defined their goals all that well yet. But automatic searches for problematic classes of input seem like a possible tool for keeping things more on the rails.
- janalsncm 4y agoAdversarial play is going to be huge moving forward. It’s not that exciting anymore to create a generic chess engine that can beat people. But an engine specifically tuned to exploit Magnus Carlsen’s weaknesses? I bet he would pay handsomely for that. Doubly so for his next opponent.
- hgsgm 4y agoWhy would an engine be tuned to beat Carlsen, when it would crush him on default settings?
- BlueTemplar 4y agoBe afraid of ChatGPT. Very afraid : https://www.youtube.com/watch?v=rSCNW1OCk_M https://www.youtube.com/watch?v=rSCNW1OCk_M
- janalsncm 4y agoIt’s like the difference between sparring with a teacher vs sparring with Tyson Fury. You don’t learn that much from being obliterated by Stockfish. Stockfish is too strong. An engine which is tuned to beat Carlsen would demonstrate his specific weaknesses, for example specific issues with an opening he plays. For his opponents, they already try to do this. “He plays this opening a lot” is a model of Carlsen, just limited to the first few moves of the game.
- redox99 4y agoI assume they mean instead of minimax (ie assume perfect from the opponent), to play around the flaws of the opponent. If you know the opponent won't spot a certain line, or will choose a certain sub optimal line, you would have an advantage by playing around that. Kind of like how you can play fools mate on a newbie.
- janalsncm 4y agoPrecisely. It’s called opponent modeling and it has only just recently (as in the last couple years) been possible in chess.
- rcme 4y agoIt seems like other players would pay for that, to learn his weaknesses.