Anthropic recently announced its decision to rent artificial intelligence (AI) computing power from a substantial data centre operated by SpaceX. This partnership is significant because it impacts the performance of Claude, Anthropic’s AI chatbot, which users appreciate for its ethical design and efficiency.
During the developer conference in San Francisco, Anthropic CEO Dario Amodei addressed previous criticism aimed at Sam Altman regarding numerous AI compute arrangements. He noted that the company is now adopting a similar strategy to that of its competitor, OpenAI. The shift comes as the demand for AI capabilities significantly outpaces the available computing resources. In recent months, Anthropic has faced operational challenges, including outages and usage limits, causing frustration among developers.
Amodei acknowledged these difficulties, stating, “We’ve had difficulties with compute.” He reassured attendees that the company is actively seeking to expand its computing capacities. Following the announcement of the SpaceX collaboration, Anthropic lifted various rate limits, particularly concerning Claude Code, its rapidly growing AI coding service. This adjustment is crucial as some developers previously migrated to OpenAI’s Codex due to fewer restrictions there.
The conference underscored the urgent demand for increased AI capacity among startups and tech firms. For instance, a startup CEO shared their attempt to secure more computing resources from Google, indicating a broader industry trend. Anthropic has access to SpaceX’s computing resources expected within the month, which may alleviate current pressures.
Looking ahead, if Anthropic maintains its impressive revenue growth—reportedly increasing at an annual rate of 80 times— it could establish itself as one of the highest-earning companies globally.
