Musk와 Anthropic, 우주로 전기를 찾으러 간다

Original | Odaily Planet Daily (@OdailyChina)

Author | Azuma (@azuma_eth)

Musk and Anthropic have finally joined forces!

On the early morning of May 7th Beijing time, a joint announcement from Anthropic and SpaceX instantly shook the entire AI industry. According to the announcement, the two parties have signed a cooperation agreement where Anthropic will utilize the full compute capacity of the SpaceX Colossus 1 data center. This will provide Anthropic with over 300 megawatts of new compute power (equivalent to over 220,000 NVIDIA GPUs) monthly, directly improving the user experience for Claude Pro and Claude Max subscribers.

This marks the first direct, formal, and large-scale cooperation between Musk's business empire and Anthropic. Over the past years, Anthropic's core partners have been Amazon and Google. For cloud infrastructure, chip supply, and model training, Anthropic has long been deeply tied to the AWS and Google TPU ecosystem. Musk has not only never publicly invested in Anthropic but has also repeatedly criticized its AI safety approach and political leanings. There were virtually no public records of infrastructure cooperation, model collaboration, or business alliances between the two parties before this.

Now, the most capable large model company and the world's most talked-about richest man have finally connected. Such a partnership is inherently newsworthy.

Musk's Hidden Agenda

For Musk, the timing of this cooperation is particularly intriguing.

Just ten days ago (April 27th), Musk's lawsuit against OpenAI officially entered the trial phase in the U.S. District Court for the Northern District of California. Throughout the recent court sessions, both sides have been trading barbs, creating a highly charged atmosphere. This lawsuit, viewed by outsiders as the "biggest case of the AI era," has essentially brought Musk and OpenAI's long-standing feud into the public eye.

Anthropic, coincidentally, is currently one of OpenAI's most direct and core competitors. This lends an additional layer of meaning to the partnership – the enemy of my enemy is my friend; if it makes OpenAI uncomfortable, Musk will do whatever it takes.

Looking deeper from the perspective of the AI competitive landscape, this collaboration also signals an evolution in Musk's strategic thinking in the AI era.

On the surface, this is a standard compute transaction – SpaceX provides GPU clusters, Anthropic gains more inference resources, and both sides get what they need. However, the situation is clearly more complex.

Because what Musk is doing around AI now goes far beyond just "building models." In the past two years, since Musk personally launched xAI (now renamed to SpaceXAI), outsiders often viewed large model companies like OpenAI and Anthropic as Musk's potential competitors. However, with the completion and operation of the supercomputing cluster Colossus, its data center capacity has begun to spill over, subtly changing Musk's role and positioning.

Today, Musk increasingly resembles an "arms dealer" in the AI world – anyone who needs compute power can come to him, even former potential competitors.

Anthropic, Plagued by "Dumbing Down" Controversy, Finally Finds Its Savior

For Anthropic, the importance of this cooperation might be even greater than outsiders realize.

Over the past few months, Claude's reputation has been undergoing subtle shifts. On one hand, Claude Opus 4.7 and the mysterious Mythos are still regarded as top-tier models; but on the other hand, voices about Claude "getting dumber" are increasingly frequent in the community.

This sentiment is particularly strong among heavy developer groups. Some have found that Claude's reasoning ability experiences a "cliff-like drop" when handling long code or complex engineering tasks. Research reports also indicate significant cuts to the "thinking budget" or response length of some Claude models. More casual users are reporting increasingly severe hallucinations, with the model more prone to sounding confident while being completely wrong when processing complex information.

Following the escalation of the "dumbing down" controversy, Anthropic's official team released a technical post-mortem report, acknowledging that due to product-level adjustments and bugs between March and April, Claude models experienced "performance degradation" on complex tasks.

However, this explanation failed to convince the market. Public opinion still generally believes that high inference costs and compute shortages are the main reasons for the performance fluctuations of models like Claude in real-world applications.

The stronger the model, the higher the inference cost; the larger the user base, the more terrifying the GPU consumption. Every AI company faces an unavoidable commercial contradiction – users want the model to run at "full power" forever, but the company must control costs. Consequently, mechanisms like dynamic throttling, inference budget adjustments, response length control, and priority scheduling are almost inevitable, and the user's final perception is that "the model has gotten dumber."

This is precisely why this compute partnership with SpaceX is so critical for Anthropic.

In its announcement, Anthropic stated that with the cooperation in place, the experience for core Claude users will be directly improved:

• First, Anthropic will double the five-hour usage limit for Claude Code for Pro, Max, Team, and per-seat Enterprise plans.
• Second, Anthropic will lift throttling restrictions on Claude Code for Pro and Max accounts during peak hours.
• Third, Anthropic will significantly increase the API rate limits for the Claude Opus model.

AI model evaluation expert Alex Finn commented on this partnership, saying, Anthropic has been somewhat subdued for the past few months – quotas dropping, models getting dumber... and now Musk comes to the rescue, giving Anthropic access to the world's largest supercomputing cluster. Anthropic's compute crisis has been the Achilles' heel of the entire company, causing a continuous decline in user reputation and market sentiment. Musk solved this problem with a single cooperation deal.

Alex Finn used a metaphor more relatable to sports fans: "Musk's help is like handing Anthropic Victor Wembanyama!"

The Final Frontier – Power from Space

Within the cooperation announcement, a small piece of text was missed by many: "The parties are also interested in collaborating to develop multi-gigawatt-scale orbital AI compute power." In simpler terms, Musk and Anthropic want to move AI data centers into space. This sounds like science fiction, but the underlying issue it reflects is very real.

Aakash Gupta, one of the most influential analysts in the AI space, explained: "Earth's power, land, and cooling capacity simply cannot meet the demand fast enough."

Anthropic has already secured around 15 gigawatts of compute capacity, equivalent to the electricity consumption of 11 million households, but even that isn't enough... Nvidia can produce chips, and Anthropic has enough money, but what truly can't be produced on time is the power, land, and cooling capacity – and the growth rate of model demand far outpaces the construction speed of this infrastructure.

The frontier of the compute race is now beginning to leave Earth. And only one company in the world truly has the capability to launch gigawatt-scale solar arrays into orbit at scale – SpaceX.

If this interstellar-travel-like story eventually becomes reality, the person capable of achieving it might only be Elon Musk.