LLMs have been outperforming humans at AI criticism for several years now. It's time that they become a lot more persuasive at AI research

→ View original post on X — @plinz