AI Risk Assessment
The conversation delves into the existential risks posed by AI and the importance of how these risks are assessed. Sayash argues that current probability estimates, particularly the concept of "p doom," may be too unreliable for effective policymaking. He emphasizes the need for a more structured approach to understanding risks, contrasting inductive and deductive methods with the subjective probabilities often used in the AI community.In this clip
From this podcast
Machine Learning Street Talk (MLST)
Sayash Kapoor - How seriously should we take AI X-risk? (ICML 1/13)
Related Questions