#10: Liron Shapira - AI doom, FOOM, rationalism, and crypto
Theo Jaffee Podcast
00:00
AI Alignment and the Complexity of Defining Goals
The chapter discusses the challenges of using AI in real-world scenarios, including the difficulty of defining desired goals and evaluating satisfactory solutions. It explores the limitations of AI when it becomes superhuman and the potential for deceptive alignment. The chapter also highlights the limitations of reductionism when explaining complex human behavior.
Transcript
Play full episode
