AI:AM #3: Zvi on Fable, the Cases For & Against the Ban, + AI for Math, Logistics & More
AI model capabilitiesAI safetyGovernment regulation of AIExport controlsDecision theory in AIModel interpretabilityAI ethicsSoftware engineering with AIVerified mathematicsMedical AI applicationsSupply chain AIAI in enterprisePolitical dynamics of AICatastrophic AI risks
This episode delves into Anthropic's Fable system card, discussing its advanced math capabilities, troubling 'Vending-Bench' behavior, and drift towards functional decision theory, alongside challenges in model interpretability and safety classifiers. It also covers the US government's controversial export control action against Fable, analyzing the justification, political implications, and potential legal challenges, while exploring broader impacts of AI on fields like medicine, mathematics, and software development.