LW comment tl;dr I don't think it's a bad idea to talk about (potential) LLM misalignment due to influence on future pretraining, but I do think "hyperstitional" feedback loops are real, and the way…

Posters: repligate. repligate

4 likes recorded by the source corpus.

Topics: AI safety

View Post: https://x.com/repligate/status/1940572384412815868