3 korrents from this piece
Every claim below was made in this piece, quoted word for word and numbered in the order the piece makes them, so you can read it there rather than take our word for it. The sentence above each quote is our reading of the claim, not their wording. Each quote was checked against a stored copy of the page at build time; where the two differ, the quote is the fact.
Our reading
AI agents will soon reach superhuman, AlphaZero-like performance across a wide range of domains.
Their wordswe should therefore expect AI agents to achieve AlphaZero-like, superhuman performance in a wide and expanding variety of domains.
Our reading
Human social cooperation is fundamentally norm-governed rather than purely consequentialist.
Their wordshuman sociality is not purely consequentialist but instead highly norm-governed.
Their wordsIn short, RLVR gives us no reason to expect the normative representations learned in pre-training will acquire motivational force over a given action, especially when post-training repeatedly selects trajectories for terminal task success.