Episode Details

AI safety, AGI, and the limits of control collide in this tense conversation. In this condensed summary of the full Into The Impossible episode, Brian Keating sits down with Roman Yampolskiy and Emad Mostaque to debate frontier AI risks, open source models, job automation, regulation, and whether society can truly govern systems that may outpace human oversight. Compared with the full episode, this version cuts the conversation down to a quick listen while preserving the key arguments: why “P-doom” is hard to measure, how AI is already changing science and math, why narrow AI may deliver most of the benefits with less danger, and what ethical questions arise if advanced systems deserve moral consideration. If you want the core ideas on AI alignment, economic inequality, and global policy without the full runtime, this summary gets you there fast. Listen now to get the key ideas in minutes.