Jan LeikeAI alignment researcher, at DeepMind in 2017.On the timeline (1)Reinforcement learning from human preferences (June 12, 2017)