The Constant: A History of Getting Things Wrong The Constant: A History of Getting Things Wrong

Your Hand in Mine (AI Yi Yi)

Sep 8, 2026 · 1h 15m

Summary

Host Mark Chrysler explores AI's history of failure, beginning with a story about an AI hallucinating a 1915 coroner scandal. He traces the origins of chatbots like Eliza and Perry, contrasting early optimism with Hubert Dreyfus’s 1964 critique that true AI is impossible due to the lack of human intuition. The episode examines how the first AI winter validated Dreyfus’s arguments, leading to the development of modern neural networks that bypass formal logic.

Topics discussed

Sponsor: Jack Archer Tech Pants Sponsor: Progressive Insurance The Coroner Episode and Historical Fabrications Gemini's Hallucinations and Apologies Introduction: The Mechanical Turk and AI Skepticism Sponsors: Toyota and Pfizer Host's Background in AI and Chatbots The Lambda Whistleblower and Eliza Weizenbaum's Eliza and the Turing Test Perry, Colby, and the Origins of the Turing Test Neuron Theory and Early Robot Turtles Logic Theorist and Early AI Predictions Sputnik, Cold War Funding, and Hubert Dreyfus Sponsors: Jira and Toyota The Dreyfus Affair and the First AI Winter Symbolic AI and the Limits of Logic Modern AI Triumphs and Dreyfus's Critique Vigellin, Newcomb, and the Return of Skepticism ANI vs AGI and the Problem of Truth Embodiment, Dualism, and Why AI Lacks Understanding Current AGI Predictions and the Risk of Another Winter Conclusion and Historical Context Sponsor: U.S. Bank
Listen ad-free on Castria