AI Safety Cannot be Solved at the Model Level - Anthropic's Latest Fiasco - The Wrong Approach | Artificial Intelligence Masterclass | Podwise