← Home
Practical AI · February 18, 2025 · 52m

AI Safety and Security in the GenAI Era

Rick Kobayashi and Kenny Song from Citadel AI discuss safety challenges in generative AI, Japan's advanced GenAI adoption, and their LLM-as-a-judge evaluation approach.

This summary was generated from show notes and public descriptions, not from a full transcript review. Details may contain inaccuracies.

Canon

Japan's enterprise culture (consensus-building, process-orientation, quality focus) creates an environment where AI adoption is more methodical but ultimately deeper than US adoption.
When AI systems evaluate other AI systems, both the evaluator and the evaluated may present false selves: the evaluated model performs for the evaluator, and the evaluator may have blind spots that match the evaluated model's weaknesses.
Was this useful?