“The field of RLVR (Reinforcement Learning from Verifiable Rewards) is expected to change so much in the next 18 months that it is not yet mature enough to be the subject of a book.”