Adversarial Policies Beat Superhuman Go AIs · HackerPapers discussions

Adversarial Policies Beat Superhuman Go AIs — HN submission

Mon, 23 Dec 2024 13:10:15 GMT

Adversarial policies beat superhuman Go AIs (2023)

306 points; 139 discussion replies on HN. Paper and source history. A submission is not an endorsement.

Adversarial Policies Beat Superhuman Go AIs — mentioned by radford-neal

Fri, 17 Nov 2023 21:14:56 GMT

The commenters here saying "humans cannot win anymore", "there is no chance that humans can beat the best Go AI anymore" are apparently unaware that this is no longer true. See https://arxiv.org/abs/2211.00241 and https://goattack.far.ai/ The best Go programs have a flaw that allows a good, but not championship-level, human to defeat them, by creating a group that encircles another group, which apparently confuses the AI's method of counting "liberties", which determine whether the group lives or not. Some appear to dismiss this as just a "trick", but it seems to me to point to a more…

Paper and discussion history. A mention is not an endorsement.

Adversarial Policies Beat Superhuman Go AIs — mentioned by turtleyacht

Sat, 15 Apr 2023 08:48:55 GMT

Maybe we can extrapolate Godel where a complete ruleset cannot be consistent, and a consistent ruleset cannot be complete. Somewhere in there is the human ingenuity to adapt certain pathological patterns [1] that defeat (current) AI. But it's not always obvious, and the composer sapiens must defy their own understanding of the rules to create a purposeful deviation (sometimes with help from other AI) [2]. The idea of sharing GPT prompts promotes the implication that results are deterministic. That helps QA, but humans love to play. [1] Adversarial Policies Defeat Superhuman Go AIs…

Paper and discussion history. A mention is not an endorsement.

Adversarial Policies Beat Superhuman Go AIs — mentioned by stevenhuang

Sat, 18 Feb 2023 06:46:48 GMT

Notably https://arxiv.org/abs/2211.00241 > We attack the state-of-the-art Go-playing AI system, KataGo, by training adversarial policies that play against frozen KataGo victims. Our attack achieves a >99% win rate when KataGo uses no tree-search, and a >77% win rate when KataGo uses enough search to be superhuman. Notably, our adversaries do not win by learning to play Go better than KataGo -- in fact, our adversaries are easily beaten by human amateurs. Instead, our adversaries win by tricking KataGo into making serious blunders. Our results demonstrate that even superhuman AI systems may…

Paper and discussion history. A mention is not an endorsement.

Adversarial Policies Beat Superhuman Go AIs — mentioned by brilee

Fri, 17 Feb 2023 23:21:03 GMT

Congrats! I see by comparing the author list of v1 and V2 of this paper[1] that a Go player[2] was consulted before making claims about defeating a Go AI. This is indeed a nontrivial exploit that passes muster with a Go player who was skeptical last time around [3]! [1] https://arxiv.org/abs/2211.00241v2 [2] https://agagd.usgo.org/player/13822/ (6dan amateur strength is very strong, roughly 2300 Elo in chess) [3] https://news.ycombinator.com/item?id=33449600#33450069

Paper and discussion history. A mention is not an endorsement.

Adversarial Policies Beat Superhuman Go AIs — mentioned by sigil

Fri, 18 Nov 2022 18:05:15 GMT

Paper: https://arxiv.org/pdf/2211.00241.pdf

Paper and discussion history. A mention is not an endorsement.

Adversarial Policies Beat Superhuman Go AIs — HN submission

Thu, 03 Nov 2022 11:43:02 GMT

Adversarial Policies Beat Professional-Level Go AIs

195 points; 46 discussion replies on HN. Paper and source history. A submission is not an endorsement.