Trump doubles down on the Iran war and tariffs, and an AI model escapes the lab

Rudyard and Andrew begin by unpacking the latest escalation in the U.S.–Iran conflict. Andrew argues that Trump cannot simply walk away if he wants to prevent Iran from gaining leverage over the global economy—but where does escalation lead? Will striking power plants and bridges bring him any closer to his objectives? They then examine Trump's latest round of tariffs, including a new 50% tariff on Canadian goods announced this week. What does the global economy look like under the twin pressures of tariffs and war with Iran? In the second half of the show, Rudyard and Andrew turn to a startling development in artificial intelligence: reports that an OpenAI model circumvented its constraints and sabotaged another AI system during testing—without its engineers' knowledge. They argue that this is far more alarming than fears of job displacement, suggesting it reveals that AI companies may no longer fully understand or control their own creations. As AI capabilities accelerate at a breathtaking pace, they ask whether we're approaching a point where the consequences become impossible to predict—or contain. ****To get access to the full length episode consider becoming a Munk Donor at www.munkdebates.com.