Episode Details

What if the biggest mystery in AI is that it works at all? In this condensed recap of Into The Impossible, Brian Keating sits down with Terence Tao to unpack AI, machine learning, mathematics, quantum computing, and the philosophy of science in a shorter, time-saving version of the full conversation. Tao explains why patterns like Benford’s Law can reveal fraud, why induction has limits in the real world, and how high-dimensional geometry breaks intuition. He also explores cryptography, prime numbers, quantum computing, and the challenges of AI alignment, including hallucinations and unreliable textbook outputs. Along the way, he shares how neural networks are already finding new results in knot theory and why education must emphasize verification and critical thinking in an AI-driven future. Listen now to get the key ideas in minutes.