xAI has undergone a dramatic transformation from a frontier artificial intelligence lab to a computing infrastructure rental business, with all 11 of its original co-founders now gone and SpaceX renting its massive Colossus data centers to the very companies Grok was designed to compete against. The shift raises fundamental questions about whether Elon Musk's AI ambitions have fundamentally changed or simply failed to materialize as originally envisioned.

What Happened to xAI's Original Leadership Team?

The departure of xAI's founding team represents an unusual level of attrition for a company that launched in 2023 with significant resources and Musk's backing. Ross Nordeen, the final non-Musk co-founder to remain, was cut off from company systems and disappeared from internal communications, according to reporting from Business Insider. Manuel Kroiss, who led the company's pretraining work, left in March 2026, making him the tenth of eleven original co-founders to depart. Earlier departures included Tony Wu and Jimmy Ba, who left in February as xAI was being absorbed into SpaceX.

In a June 18 CNBC interview, Yann LeCun, the former Meta chief artificial intelligence scientist, called xAI "kind of a failure" and attributed the departures to workplace challenges. LeCun stated that Musk has made it "very, very difficult" to hire top AI talent because of how he has treated the previous team. While LeCun has his own commercial interests, having raised $1.03 billion for his competing startup Advanced Machine Intelligence Labs in March 2026, the factual record of departures speaks independently to the company's instability.

How Did xAI Become a Data Center Rental Business?

The transformation began when SpaceX acquired xAI in February 2026 in an all-stock transaction that valued the combined entity at $1.25 trillion. By June, SpaceX had gone public, and its AI infrastructure was being monetized through contracts with the exact competitors xAI was meant to challenge.

The financial arrangements reveal the scale of this pivot:

• Anthropic's Deal: Anthropic, whose Claude model competes directly with Grok, agreed to use SpaceX's Colossus One data center near Memphis for more than 300 megawatts of computing capacity across more than 220,000 Nvidia graphics processing units (GPUs). Anthropic is paying $1.25 billion per month through May 2029 for this compute capacity.
• Google's Deal: Google agreed to pay SpaceX $920 million per month from October 2026 through June 2029 for approximately 110,000 Nvidia GPUs, central processing units (CPUs), memory, and related infrastructure.
• Infrastructure Scale: These contracts represent massive computational resources, with combined monthly payments exceeding $2 billion, demonstrating that SpaceX's infrastructure business is now generating substantial revenue.

LeCun explained the business logic plainly: SpaceX has enormous infrastructure and is renting it out because that is how the company can recoup its costs. From a pure business perspective, monetizing idle computing capacity is more rational than maintaining it as an internal resource for a struggling AI lab.

Why Does This Matter for the AI Industry?

The uncomfortable reality is that xAI's infrastructure is now serving its direct competitors. If your AI challenger is selling computing resources to the companies it was meant to outcompete, the original strategic pitch has fundamentally changed. This creates a perverse incentive structure where xAI's parent company profits from the success of Claude and Google's AI models rather than from Grok's competitive advantage.

There is a version of this story where Musk still wins financially. Amazon Web Services became one of the most important businesses in technology by renting infrastructure to everyone else, and CoreWeave built a public-market story around Nvidia-backed AI compute. If SpaceX can turn Colossus into a durable cloud business, investors may not care whether Grok ever catches Claude or ChatGPT.

However, this outcome represents a fundamental departure from xAI's original mission. xAI was launched in 2023 as a frontier-model company focused on building the world's most capable AI systems. It absorbed X in 2025, folded into SpaceX in 2026, lost its original co-founders, and is now using Colossus to serve outside customers. That represents substantial movement for a company still trying to explain what it fundamentally is.

"Musk has made it very, very difficult to hire top AI people because he has not behaved well toward the previous team," said Yann LeCun.

Yann LeCun, Former Chief AI Scientist at Meta

What Are the Broader Implications for AI Competition?

LeCun also warned that AI companies risk a "big bubble explosion" if they do not cut costs and raise prices. This warning carries particular weight coming from someone who has just raised over $1 billion to pursue an alternative architecture based on world models rather than large language models (LLMs). If the current model race breaks under its own spending, his company benefits from that collapse.

The xAI situation illustrates a broader tension in AI development: the enormous capital requirements for frontier research may be unsustainable for companies that cannot monetize their infrastructure independently. Musk's decision to pivot toward infrastructure rental suggests that even with SpaceX's resources, maintaining a competitive AI lab while competing against OpenAI, Anthropic, and Google proved untenable.

SpaceX and xAI did not respond to requests for comment on LeCun's remarks, according to Business Insider. That silence does not prove his case, but the company filings, the compute contracts, and the co-founder exits already tell a substantial story. Grok may improve over time, and SpaceX may generate billions from Colossus. But the original xAI story was about building a frontier AI lab capable of beating the world's leading AI companies, and right now the clearest revenue line comes from renting hardware to them.

Whether you call this outcome failure depends on what you thought xAI was supposed to be. If it was meant to become a cloud infrastructure arm inside SpaceX, the strategy is working. If it was meant to build the world's strongest AI lab around Musk, then the facts, rather than LeCun's insult, are the real problem.