There is something satisfying, and deeply weird, when an LLM reports that it has fixed 37 of 41 failing tests (a 90% improvement) and then asks what to do next, and correctly interprets "Onwards!"

Posters: patio11. patio11

2 likes recorded by the source corpus.

Topics: AI & LLMs

View Post: https://x.com/patio11/status/1962255139332313213