The episode begins with Anthropic researcher Jacob Coxon’s resignation and viral warning that AI could destroy humanity, then examines why the story spread and who might benefit. The hosts question percentage-based extinction estimates, explore the incentives of researchers, AI companies, politicians, and investors, and debate whether safety messaging can function as marketing. They ultimately shift attention toward concrete risks such as cyberattacks, sensitive data use, infrastructure failures, and data-center regulation, while acknowledging that improving capabilities make AI risks harder to dismiss. The conversation ends with uncertainty about long-term dangers, doubts about AI labs’ stewardship, and concern that public backlash could damage the industry.