Understanding AI Manipulation
Dylan delves into the complexities of measuring and preventing manipulation in AI systems, highlighting the challenges in defining and detecting manipulation. Kanjun emphasizes the difficulty in discerning intrinsic interests from manipulated actions, shedding light on the intricate nature of manipulation in AI.In this clip
From this podcast
Generally Intelligent
Episode 10: Dylan Hadfield-Menell, UC Berkeley/MIT, on the value alignment problem in AI
Related Questions