‘Misanthropic and evil’: Elon Musk’s SpaceX may get up to $120.5 billion from Anthropic, the AI lab he once slammed
Key Points:
- Anthropic, the AI company behind the chatbot Claude, began trading on Nasdaq with a valuation around $1.5 trillion after raising approximately $81.3 billion, surpassing SpaceX's $75 billion IPO earlier this year.
- Elon Musk, who has criticized Claude for alleged biases, ironically has a significant financial stake in Anthropic through a massive compute capacity deal worth up to $120.5 billion with SpaceX's AI unit, SpaceXAI, extending through 2029.
- Anthropic has committed to using all compute resources at SpaceX's Colossus 1 data center, involving over 220,000 Nvidia GPUs, with monthly costs potentially reaching $3.1 billion; this arrangement already accounts for a substantial portion of SpaceX's revenue.
- The compute agreement includes non-cancelable periods until early 2027, after which either party can terminate with 90 days' notice, and Anthropic is also developing Anthropic Direct to lease data center capacity from partners, reflecting total infrastructure commitments nearing $900.9 billion.
- SpaceX reports strong demand for compute resources, with new deployments financially self-sustaining within a year, indicating robust growth in AI infrastructure needs amid rising market interest.