Hacker News new | ask | show | jobs
by applicative 30 days ago
LLM output is a kind of dreaming but with the whole of past human text output as dream material. It turns out to be useful if you can direct the hallucination