Recent episodes of the All-In Podcast have centered on AI infrastructure economics, with guests like Brad Gerstner addressing whether revenue from large language models can cover massive data center capex, gigawatt-scale power constraints, and semiconductor demand. Discussions also covered data center backlash, local economic benefits such as job training and teacher funding from projects like Meta's facilities, and AI doomerism versus practical deployment. The September 18 episode features investigative journalist Nick Shirley examining California infrastructure fraud, high-speed rail inefficiencies, and government waste, aligning with recurring themes of regulatory capture, entitlement fraud, and tech-policy intersections. These verified guest appearances and topic patterns shape trader views on likely discussion points around AI buildout realities and public-sector accountability.