Anthropic teams up with SpaceX as the AI ​​race turns weird

Anthropy and Elon Musk’s company SpaceX said on Wednesday that the two entities had signed an agreement for Anthropic to use computing resources from an xAI data center in Memphis, Tennessee. It’s the latest connection in an industry scrambling to find enough computers to run complex AI programs. SpaceX and xAI were previously separate companies, but the two companies merged earlier this year. The combined entity, also owned by Musk, is called SpaceXAI.

Anthropi executives made the announcement on stage at the company’s annual developers conference in San Francisco. SpaceXAI too Post a blog post Sharing more details about the deal, which will see Anthropic draw power from xAI’s Colossus 1 supercomputer.

This partnership comes at a pivotal time for SpaceXAI, which seeks to go public as soon as next month. A relationship with a leading AI lab could boost SpaceX’s credibility, as it encourages investors to the potential gold mine of more data centres, including in space.

Notably, Space XAI He said In its blog post, Anthropic expressed “interest” in partnering to develop “orbital AI computing capability” — essentially, data centers in space. Having Anthropic as a potential customer could help SpaceXAI boost investor confidence that there will be buyers for the expensive supercomputing project.

Earlier this year, Musk criticized Anthropic’s AI models In a post on XHe called the company’s policies “misanthropic” and “evil” and claimed without evidence that AI models showed racial and sexist biases.

Now, he has noticeably changed his tone, Write on X He “spent a lot of time last week with senior members of the Anthropic team to understand what they are doing to make sure Claude is useful to humanity and was impressed by him.”

Musk’s SpaceXAI first began work on Colossus 1 at a former Electrolux site in 2024. The company claims it is one of the largest and fastest AI supercomputers in the world. It includes approximately 220,000 Nvidia GPUs, ranging from Nvidia’s H100 and H200 to the newer GB200 chips. The company loves To brag She built the supercomputer in 122 days.

Emissions from gas turbines The data center raised many complaints from local residents. Last month, environmental protesters took part Proven At a nearby gathering of SpaceX investors ahead of the IPO.

Anthropic plans to use its newfound computing power to “directly improve capacity for Claude Pro and Claude Max subscribers,” the companies said. The deal will give Anthropic access to more than 300 megawatts of new capacity, or approximately 220,000 Nvidia GPUs. In recent months, software developers have increasingly complained about price caps and service outages for Cloud Code, due to the physical limits of available computing resources.

As more and more people use services like OpenAI’s Codex and Anthropic’s Claude Code — sometimes running programming programs and proxy tasks for hours on end — the services are faltering. According to Anthropic, the average developer now spends at least 20 hours a week running Claude Code.

Yesterday information I mentioned Anthropic has committed to spending $200 billion on Google’s AI cloud services and TPU chips. Anthropic has also used cloud computing services and AI chips from Amazon since 2023. Last month, Anthropic committed “More than $100 billion over the next 10 years for Amazon technologies.” According to the Information Report, contracts with Anthropic and OpenAI now account for “more than half of the $2 trillion backlog at major cloud providers” like Amazon, Microsoft, and Google.

Paresh Dave and Zoe Schiffer contributed to this report.