RLHF predates the acronym: a 1993 robot shaping paper already described human trainers

andrewprock · x · 2026-10-11

andrewprock cites the 1993 paper "Robot Shaping: Developing Autonomous Agents through Learning" to argue RLHF-style training existed decades before the acronym: "In principle, the trainer could be a human being observing the agent's interaction with the environment."

Related event: Debate Erupts Over Who Really Invented RLHF(2 posts)→

Original post →