34 - AI Evaluations with Beth Barnes
AXRP - the AI X-risk Research Podcast
Coordination Over Conjecture
Identifying the necessary coordination among various threat models is crucial for distinguishing which threats can be dismissed based on a model's capacity to interact and execute thoughtful actions across different instances. Observing changes in model communication, especially with interpretable states, could significantly alter perspectives on threat assessments and evaluations. Therefore, it's essential to explore how advancements in model communication can reshape the understanding of threat models and the corresponding evaluation methods.
00:00
Transcript
Play full episode
Remember Everything You Learn from Podcasts
Save insights instantly, chat with episodes, and build lasting knowledge - all powered by AI.
