Self-driving cars promise to make roads safer, but measuring safety is far more complicated than simply counting crashes. A vehicle may perform well statistically while still creating serious risks in rare but consequential situations—like blocking an emergency vehicle.

Join this episode of InsideAnalysis to explore the gap between safety statistics and real-world edge cases. Legendary Carnegie Mellon visionary Philip Koopman, who helped develop safety standards for self-driving cars, will be joined by AI architecture expert Mark Madsen of Third Nature to examine what it really means to build and evaluate safe autonomous vehicles.