Hacker News Reader

My argument is that LLM generation and a correlated Lean verification are not sufficient conditions. Both are falliable.

So is peer review. False proofs have been published before and will again. Does that mean journals are publishing slop?