AI Safety Fundamentals Model Evaluation for Extreme Risks
16 snips
May 13, 2023 The podcast highlights the significance of model evaluation in addressing extreme risks posed by AI systems. It discusses the importance of evaluating dangerous capabilities and assessing the propensity of models to cause harm. The chapters explore different aspects of model evaluation, including alignment evaluations and evaluating agency in AI systems. The podcast also discusses the limitations and hazards of model evaluation, risks related to conducting dangerous capability evaluations and sharing materials, and the importance of effective evaluations in AI safety and governance.
Chapters
Transcript
Episode notes
1 2 3 4 5 6 7 8
Introduction
00:00 • 3min
Model Evaluation for Extreme Risks in AI
02:45 • 15min
Model Evaluation for Extreme Risks in AI
17:40 • 22min
Challenges in Alignment Evaluation and Evaluating Agency in AI Systems
39:39 • 3min
Limitations and Hazards of Model Evaluation for Extreme Risks
42:25 • 3min
Risks related to conducting dangerous capability evaluations and sharing materials
45:28 • 4min
Superficial Improvements to Model Safety and Hazards during Evaluation
49:34 • 2min
Model Evaluation for Extreme Risks in AI Safety and Governance
51:31 • 5min