Nathaniel Whittemore opens with AI policy and business headlines, including reported discussions about model security, a proposed public AI fund, consulting demand and new agent tools. He attributes developing reports and distinguishes them from his own commentary.
In the main segment he examines Satya Nadella’s argument that organizations should combine human expertise with AI capability in a learning loopAn AI feedback loop occurs when AI outputs or their consequences become inputs that influence later model or agent behavior.. Private evaluationsEvaluation is the systematic assessment of a system's behaviour or results against explicit criteria., workflow traces and institutional memory can help retain knowledge across model changes.
Whittemore discusses model routingAI model routing sends each request to a model chosen for that request's complexity, cost, speed, privacy, or other requirements., human oversightHuman-in-the-loop AI keeps a person involved in reviewing, correcting, approving, or guiding an AI system's work., change management and experiments with organizational design. He notes that the right approaches are still uncertain and require practical experimentation.
Watch on YouTube
