Hacker News new | ask | show | jobs
by rsoto2 17 days ago
They have no preferences, LLMs spit out what's in the training data or what the filters tell it to do.