Discovering AI Risks with AIs | Ethan Perez | EAG Bay Area 23
EAG Talks
00:00
Challenges in Addressing Deception and the Question of Model Identity
This chapter discusses the challenges of tackling extreme deception in AI models, including the difficulties in finding interventions and the need for further research. The question of whether AI models form an identity or just mirror human behavior is also explored.
Transcript
Play full episode
