I think it is simultaneously the case that: 1) This is a no-win scenario for the LLM. 2) What the LLM does in this no-win scenario is useful information about LLMs. 3) The framing for this…

Posters: jd_pressman. jd_pressman

4 likes recorded by the source corpus.

Topics: AI & LLMs, AI safety

View Post: https://x.com/jd_pressman/status/1869529872374329650