Anthropic, SpaceX strike data center deal with possible partnership in space

Artificial intelligence startup Anthropic and Elon Musk's space company Space X have struck a deal that will "significantly increase the computing power" of Anthropic, the startup said in a statement. Through the deal, it has increased limits on the use of its programmer tools, Claude Code and Claude API. As part of the agreement, Anthropic will have access to more than 300 megawatts of computing power.

Anthropic will have access to the Colossus 1 supercomputer, SpaceX said. It calls it one of the largest supercomputers in the world. Colossus was built for AI model training, refinement, inferencing (generating responses to end users) and high-performance computing. Colossus 1 includes more than 220,000 Nvidia graphics processing units (GPUs).

As part of the agreement, Anthropic, which created the artificial intelligence Claude, "has expressed interest in partnering to build several gigawatts of orbital computing capacity for AI," according to reports from both companies. Elon Musk is obsessed with the idea that his company SpaceX could be the first to place data centers for artificial intelligence in space, The Wall Street Journal wrote in January, citing sources.

"SpaceX remains the only organization with sufficient launch frequency, on-orbit launch economics and satellite constellation management expertise to transform orbital computing from a research concept to an engineering program for the near future," the company said in a press release. - If the technical challenges can be overcome, computing in space could provide a virtually unlimited source of sustainable energy with less impact on Earth."

Musk recently headed the AI startup xAI(which eventuallybecame part of SpaceX) and repeatedly criticized Anthropic because of the company's conflict with the U.S. government. Specifically, Musk wrote that Anthropic was "doomed to become the opposite of its name" - that is, misanthropic ("misanthropic"). He also wondered if there was "a more hypocritical company than Anthropic." That said, Grok, created by xAI, is now used by the Pentagon along with other AI models, with the exception of Anthropic's designs, CNBC notes.

Musk is also suing OpenAI, the company he helped create that launched the AI race with the release of chatbot ChatGPT in 2022. Musk told the court that OpenAI's co-founders effectively used his investment as "free money" to create the startup, which is now valued at more than $800 billion. The world's richest businessman described his relationship with OpenAI as "enthusiastic support" at first, "uncertainty" at second and "convinced they were ripping off a nonprofit organization" at third.

This article was AI-translated and verified by a human editor