Hacker News new | ask | show | jobs
by jimmyjack 1197 days ago
Funny enough it would.

LLMs have proven to be quite good at evaluating their own output. Think of things like "Think step by step" and how that improves output.