Descriptions:
Peter Diamandis and his Moonshots co-hosts cover an unusually dense week of AI industry developments centered on the open-source versus closed-source debate. The episode leads with Jensen Huang’s (Nvidia CEO) first-ever tweet — a public letter arguing the world needs both frontier closed and open models, signed by 77 companies and accompanying the launch of the Open Secure AI Alliance. The panel unpacks Anthropic’s three-day silence before Dario Amodei released a nuanced new position on open weights, and the notable news that OpenAI and Anthropic — longtime rivals — are reportedly coordinating Washington lobbying around a federal review process for the most powerful AI models.
On the geopolitical front, the hosts analyze China’s emerging AI diplomacy strategy, with Xi Jinping deploying AI as statecraft leverage across the Global South. They also dig into the Claude Opus 5 release, debating benchmark performance against Fable 5: HLE scores moved narrowly from 63.9% to 64.7% with tools, while some domains showed regression — raising suspicion of benchmark optimization skewed toward code generation and vision tasks. The panel positions Opus 5 as Anthropic’s direct answer to Grok on cost-performance curves.
Additional segments cover Anthropic’s chief security officer projecting AGI in two to three years, Kimi K3 from Moonshot AI, and Starship 13’s successful launch. A useful weekly digest for anyone tracking frontier model competition, AI policy, and geopolitical dimensions of the AI race.
📺 Source: Peter H. Diamandis · Published July 29, 2026
🏷️ Format: Podcast
