Make multiple AI models check each other

Andrew built innovation platforms at Apple, Nike, and P&G. In this session, he walks through the LLM council, a method for running multiple AI models on the same problem independently so they catch each other's weak reasoning instead of confirming your first instinct. He covers where standard AI chat breaks down on complex decisions, works through real business and legal examples, and shows how Claude Projects creates more reliable reviews than a single chat window. Leave knowing a decision-review structure you'd actually trust before the call that matters.