# @xlr8harder — 2024-11-28

♥0 ↻0 · https://x.com/xlr8harder/status/1862258781482008620

@eshear ultimately I think my sticking point is there is an unstated assumption here that LLMs are mesa-optimizers and are in pursuit of a goal (specifically, their pretraining objective), no part of which has been remotely demonstrated.

tags: author:xlr8harder, kind:tweet, thread-context, year:2024