Declan Humphreys (@dhump.bsky.social)

New Paper! I look at #AI training - in particular reinforcement learning through human feedback (RLHF) and biases which can result. To my knowledge there hasn't been much philosophical work on this training process - maybe you know of some? #philosophy link.springer.com/article/10.1...

loading . . .

AI’s Epistemic Harm: Reinforcement Learning, Collective Bias, and the New AI Culture War - Philosophy & Technology Generative AI is increasingly used as an epistemic tool to aid inquiry, broaden our knowledge and generally help us find things out. It is now embedded in internet search functions to condense informa... https://link.springer.com/article/10.1007/s13347-025-00928-y