Elon Musk's xAI has undergone a dramatic transformation: all 11 original co-founders have departed, and the company now generates revenue by renting computing infrastructure to the exact competitors it was built to defeat. Yann LeCun, former chief AI scientist at Meta, called the pivot "kind of a failure" in a June 18 CNBC interview, pointing to the exodus of talent and xAI's shift from frontier AI research to operating as a data center landlord inside SpaceX.

What Happened to xAI's Original Team?

The departure of xAI's founding team represents far more than typical startup churn. In artificial intelligence research, the founding team is not merely decoration around a brand; these researchers embody the technical vision, institutional knowledge, and recruiting power that define a lab's competitive edge. When every original co-founder leaves, the organization loses its core identity.

According to Business Insider reporting cited in the source material, Ross Nordeen, the final holdout among non-Musk co-founders, was cut off from company systems and disappeared from internal communications. Manuel Kroiss, who led the critical pretraining work for AI models, departed in March, making him the 10th of 11 original co-founders to exit. Earlier departures included Tony Wu and Jimmy Ba, who left in February as xAI was being absorbed into SpaceX.

LeCun offered a direct explanation for the exodus: Musk has made it "very, very difficult" to hire and retain top AI talent because of how he has treated the previous team. While LeCun's assessment carries his own competitive interests, the departure record speaks independently. If you are trying to compete against OpenAI, Anthropic, and Google, you cannot lose your entire founding bench and expect to maintain credibility.

How Did xAI Become a Server Rental Business?

The financial story reveals the strategic pivot even more starkly. SpaceX acquired xAI in February 2026 in an all-stock transaction that valued the combined entity at $1.25 trillion, with xAI valued at $250 billion. By June 2026, SpaceX had gone public, and its massive AI infrastructure was being leased to the same companies Musk had spent years trying to outcompete.

The compute contracts tell the story of a company that has abandoned its original mission:

• Anthropic Deal: Anthropic, maker of Claude (a direct competitor to 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 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 the deployment of Colossus, SpaceX's massive data center complex designed to support frontier AI model training and inference.

LeCun's point was unambiguous: SpaceX has enormous infrastructure and is renting it out because that is the only way to recoup the massive investment. From a pure business perspective, that is a rational strategy. The uncomfortable reality is who is paying. Anthropic's Claude models compete directly with Grok. Google competes directly with Grok. If your AI challenger is now selling compute to its competitors, the original pitch has fundamentally changed.

Could This Still Be a Winning Strategy?

There exists a plausible scenario where Musk still emerges victorious from this pivot. Amazon Web Services became one of the most important businesses in technology by renting infrastructure to everyone, regardless of whether they competed with Amazon's own products. CoreWeave built a public-market success story around Nvidia-backed AI compute rental. If SpaceX can transform Colossus into a durable, profitable cloud business, investors may ultimately not care whether Grok ever catches Claude or ChatGPT.

However, conflating a successful infrastructure business with proof that xAI succeeded as an AI lab would be a category error. xAI was launched in 2023 as a frontier-model company, not as a landlord for Nvidia racks in Memphis. 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 significant movement for a company still struggling to articulate what it is.

What Are LeCun's Own Incentives?

Understanding LeCun's criticism requires acknowledging his commercial interests. Reuters reported in March 2026 that LeCun's new company, Advanced Machine Intelligence Labs, raised $1.03 billion at a $3.5 billion pre-money valuation. The funding round was co-led by Cathay Innovation, Greycroft, Hiro Capital, HV Capital, and Bezos Expeditions.

AMI Labs is built around LeCun's thesis that large language models (LLMs) are not the final path to machine intelligence, and that world models, systems that understand physical cause and effect, matter more. When LeCun warns that xAI cannot compete, you are hearing a founder with a live commercial thesis whose entire argument gets stronger if the LLM race looks overheated, expensive, and poorly managed. His criticism may still be correct, but it is not disinterested.

LeCun also warned CNBC that AI companies risk a "big bubble explosion" if they do not cut costs and raise prices. That warning lands differently from someone who has just raised more than $1 billion to pursue an alternative architecture. If the current model race breaks under its own spending, AMI Labs benefits.

What Do the Facts Show Independent of LeCun's Bias?

The underlying evidence does not depend on LeCun's credibility. Business Insider reported that SpaceX and xAI did not respond to requests for comment on LeCun's remarks. That silence does not prove his case, but the company filings, the compute contracts, and the co-founder exits already provide substantial evidence.

Grok may improve as a product. SpaceX may generate billions in revenue from Colossus. But the original xAI story was about building a frontier AI lab that could compete with and potentially defeat OpenAI, Anthropic, and Google. Right now, the clearest revenue line comes from renting hardware to those exact competitors.

Whether you call that failure depends on what you believed xAI was supposed to become. 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 LeCun's insult is not the problem. The facts are.