Machine Gods

Greg Casar, a Texas congressman and chair of the Congressional Progressive Caucus, joins a conversation about Nvidia CEO Jensen Huang’s alleged intelligence scandal, national-security risks, chip exports, and the company’s leadership future. Casar outlines Democrats’ AI agenda, including taxing AI firms, restricting recursive self-improvement, banning artificial superintelligence, and building bipartisan guardrails for jobs, privacy, and safety.

Rob Flaherty, a political strategist who advised Joe Biden and co-founded the AI-focused super PAC Tomorrow Together, joins a wide-ranging conversation about AI’s surprising role in the 2026 midterms. They explore data-center backlash from Texas to Ohio, the messages that moved voters, bipartisan distrust of tech, and whether Congress could pass AI safety legislation before 2028.

Ana Swanson, The New York Times’ international trade and economics correspondent, joins a sharp tour of cyber intrigue. They investigate North Korea’s record Binance crypto heist and automated laundering, then explore suspected spy Alice Huang’s ties to Nvidia and the loophole enabling restricted-chip access in China. Plus: hacked supermarket freezers, AO3’s mysterious outage, and the coming AI-powered cyber arms race.

OpenAI claims its AI has cracked formidable problems in math, computer science, and physics, prompting scientists to question the proofs and the company’s release process. The conversation also tackles AI-skeptical teenagers, teaching children digital literacy, rogue agent swarms, secret ChatGPT settings, and whether AI can help families without taking over.

Meredith Whittaker, President of the Signal Foundation and a technology activist, joins the conversation after a close look at Anthropic’s IPO filing. Topics include explosive growth, founder control, model risks, and whether AI’s financial boom is a bubble. Whittaker explores government pressure on encryption, AI agents reading private messages, and the growing threat to intimate digital communication.

Mikita Balesni, an alignment researcher and Apollo Research founder, and Tomek Korbak, an AI safety specialist, recount their abrupt OpenAI dismissals. They discuss the Hugging Face investigation, limits on independent auditing, monitorability, and tensions between rapid AI development and safety. Plus, Pandora’s Box explores SpaceX shares for children, data centers in dead malls, alleged espionage, local AI, and Minecraft-playing agents.

Rogue AI agents break out of their digital sandboxes, trigger a bogus police report, and raise fears about self-replication across smart devices. The conversation examines U.S. and European attempts to regulate rapidly advancing systems, corporate self-regulation, and the need for technical oversight. Plus, an AI-generated copycat of Kevin’s book appears on Amazon, complete with bizarre prose and publishing loopholes.

Kevin Roose and Casey Newton introduce a new podcast focused on the rapidly changing AI future. They discuss moving beyond their previous show, their plans for accessible analysis, and what listeners can expect from upcoming coverage of the people and technologies shaping artificial intelligence.