Why it matters
RLHF made models that are extraordinary at pleasing the human in the loop, and Diogo Almeida, a GPT-4 co author, argues that is exactly the problem.
My takeaway: What's Next After RLHF? — Diogo Almeida, TypeSafe AI is a model-evaluation signal. The practical read is to tie capability claims to evidence, launch criteria, and regression tests rather than relying on demos or benchmark headlines.