Elon Musk's SpaceX has signed a $5 billion deal to provide Anthropic with compute capacity at its Colossus One data center. This is the same Musk who has been suing OpenAI, and the same Anthropic founded by ex-OpenAI researchers. The alliance is not ideological. It is infrastructural. Anthropic needs compute at a scale it cannot self-provision, and SpaceX has the facility to sell.
The episode also covers Anthropic's $200 billion partnership with Google, new AI templates targeting financial services workflows, and the ongoing Musk versus OpenAI trial. These are not separate stories. They are a single map of who controls the compute layer, the capital layer, and the application layer of AI in 2025. The Google number alone warrants attention.
What makes this worth reading in full is the framing: Musk's pivot from critic to infrastructure supplier is not a contradiction. It is a play. The hosts argue the Colossus deal positions SpaceX as a neutral compute provider to multiple frontier labs, including potential future rivals to xAI. If that thesis holds, the competitive implications extend well beyond Anthropic.
[WATCH ON YOUTUBE →]