Prime Highlights :
- Musk reversed his earlier criticism of Anthropic, saying the company’s commitment to humanity-focused AI impressed him after a meeting with its leadership.
- Cherny said AI’s capability is already here, the only gap remaining is how fast the industry puts it to work.
Key Facts :
- Anthropic is an AI safety company and maker of the Claude AI, known for products including the fast-growing coding tool Claude Code.
- SpaceX’s Colossus 1 facility in Memphis houses over 220,000 Nvidia processors and will supply Anthropic with 300 megawatts of computing capacity.
Background :
Anthropic announced a major computing agreement with Elon Musk’s SpaceX in the first week of May, granting the AI company access to the full processing power of SpaceX’s Colossus 1 facility in Memphis, Tennessee. The site runs more than 220,000 Nvidia processors and will deliver 300 megawatts of additional capacity to Anthropic within a month.
The deal gives SpaceX, which is preparing for a public offering, a high-profile customer as it pitches investors on its AI ambitions. For Anthropic, it eases growing capacity pressure driven by rising demand for products like its AI coding tool, Claude Code.
Anthropic announced a developer event in San Francisco, where it also unveiled a new AI feature called “dreaming.” The tool enables artificial intelligence systems to assess their performance during intermission periods by detecting recurring patterns that lead to modifications of their maintained settings to achieve better results.
With the added capacity, Anthropic doubled Claude Code’s rate limits for paid plans, removed peak-hour caps for Pro and Max accounts, and raised request volumes for its Claude Opus models. The company also expressed interest in working with SpaceX to develop space-based orbital data centres, a capital-intensive ambition central to SpaceX’s IPO strategy.
Musk said he made the leasing decision after meeting Anthropic’s leadership, adding that their focus on building AI that benefits humanity left a strong impression. The remarks marked a notable reversal from February, when Musk publicly criticised Anthropic’s AI for alleged bias.
Anthropic’s head of Claude Code, Boris Cherny, closed the event by noting that AI’s productive capability already exists; the remaining challenge is the speed of adoption.
