Descriptions:
The Moonshots podcast with Peter Diamandis convenes a panel—Dave Blondin (AI investing), Alex, and Immad Mustach—to analyze a dense week of AI news centered on the US government’s unprecedented intervention in commercial AI releases. The primary story is the executive branch’s national security hold on GPT-5.6 (OpenAI’s new three-model family: Soul, Tera, and Luna) and the partial reinstatement of Claude Mythos 5 access for roughly 100 approved organizations. Panelists frame this as the first time the executive branch has formally inserted itself into the AI release loop without clear legislative authority, debating whether it looks more like protectionism, legitimate security review, or a preview of a permanent licensing regime.
A second major story involves Anthropic publicly accusing Alibaba of conducting a large-scale model distillation campaign against Claude—systematically querying Claude outputs at scale to extract capability into competing models. The panel treats this as both a technical threat and a geopolitical signal, predicting it will be cited as justification for tighter US AI export controls targeting Chinese AI development.
The episode also addresses OpenAI’s apparent decision to pull back from near-term IPO plans, with speculation that the new regulatory environment has complicated the financial story for a public offering. Throughout, the panel applies an exponential-technology investment lens, treating all three stories as interconnected data points in an accelerating geopolitical realignment around who controls access to frontier AI.
📺 Source: Peter H. Diamandis · Published June 29, 2026
🏷️ Format: Podcast
