Can the Market Regulate AI Better Than the Government? | With Peter Goettler, CATO Institute

Description

In this episode of the Regulating AI Podcast, host Sanjay Puri speaks with Peter Goettler, President and CEO of the Cato Institute — one of Washington DC's most influential public policy research organizations, founded on the principle that liberty is the highest political value. Peter spent 22 years in investment banking, including at Barclays, living through the Asian financial crisis, the dotcom bust, and the 2008 financial crisis from the front row. In 2008, he left Wall Street to build libertarian intellectual voices globally. He now leads an organization of over 200 policy scholars and a network of 20,000 teachers focused on restoring civic culture. Chapters: 0:00 - Financial Crisis Lessons 0:32 - Peter Goettler Introduction 1:20 - Cato’s Libertarian Philosophy 4:19 - Banking Regulation Debate 7:53 - Lessons for AI Regulation 9:06 - AI Risks and Hype 12:06 - Bias, Laws, and AI Safety 13:43 - Anthropic’s Mythos Model 15:32 - Limits of Regulation 18:26 - AI and Medical Breakthroughs 19:25 - Self-Regulation vs Guardrails 21:08 - Deepfakes and Media Literacy 23:03 - AI Speech and Free Speech 24:58 - AI, Government, and Defense 27:23 - AI, Freedom, and Inequality 29:33 - AI and Job Disruption 33:19 - UBI and Government Support 36:35 - Open Source AI Debate 38:51 - Federal vs State AI Rules 40:46 - Lightning Round on AI 42:19 - Outro Key Topics Discussed: The lessons from the 2008 financial crisis that should inform AI governance — including why more regulation is not always safer. Why the existential risks of AI are speculative — and why that does not mean AI carries no risk. The underappreciated upside: why AI bias can be corrected in a model almost instantly, unlike human bias which takes generations. Why Cato supports federal preemption of AI regulation over a patchwork of state laws — a departure from traditional federalist principles. The First Amendment and AI speech: why even deepfakes deserve careful, free speech-conscious treatment. Open source AI: who is right — and why looking at bot