xAI & Anthropic: Why We’re Not Buying This AI Deal

KEY TAKEAWAYS

• xAI’s Compute Pivot: Anthropic has acquired all the compute capacity at xAI’s Colossus 1 data center in Tennessee, signaling a shift for xAI from frontier AI development to becoming a “neocloud” provider.
• IPO Heat Check: This move is widely seen as a strategic “heat check” for SpaceX ahead of its public offering, aiming to present a more immediately believable, revenue-generating business model to investors, even if less exciting than pure AI innovation.
• Innovation vs. Monetization: The deal highlights xAI’s struggles to launch a competitive AI product like Grok and raises questions about its long-term vision and capacity for genuine AI innovation, especially as it prepares to dissolve as a separate entity within SpaceX.

xAI’s Great Pivot: From Frontier AI to Compute Rental Ahead of SpaceX IPO

In a move that caught many by surprise, Anthropic, a leading AI safety and research company, this week announced a significant partnership with xAI, Elon Musk’s artificial intelligence venture. The deal sees Anthropic taking over the entirety of the compute capacity at xAI’s Colossus 1 data center in Tennessee. This development immediately sparked intense discussion about its implications, particularly for xAI’s future and its parent company, SpaceX, as it navigates the path toward a much-anticipated initial public offering (IPO).

On the latest episode of TechCrunch’s Equity podcast, Kirsten Korosec, Sean O’Kane, and I delved into the intricacies of this partnership. We explored what it truly signifies for xAI, especially as reports circulate that the organization is slated to dissolve as a separate entity, integrating more closely with SpaceX itself. The consensus formed around a fascinating tension: is this a shrewd business pivot for xAI, or a desperate measure to shore up its financials and public perception before a major market debut?

The Compute Conundrum: Anthropic’s Need, xAI’s Solution

“I always love a surprise, especially when everybody’s eyes [are] on another ball, a major trial that’s happening,” Sean O’Kane quipped, setting the stage for the unexpected announcement. The core of the deal is straightforward: Anthropic, which has been actively seeking additional compute resources to fuel its enterprise-focused AI products, gains critical access to the substantial infrastructure at Colossus 1. For Anthropic, it’s an “escape valve,” providing the necessary horsepower for its ambitious AI training.

But the flip side of this coin reveals xAI’s predicament. The enormous compute capacity they had been building seemed destined for their own AI models, primarily Grok. However, as Sean observed, Grok “outside of X, is not burning up the world as far as becoming the new hot consumer chat bot.” This left xAI with a significant, underutilized asset. Enter the “neocloud” model.

Kirsten Korosec clarified the term for listeners: “And we should say that in terms of what a neocloud is, for those who don’t know, this is the idea of buying GPUs from Nvidia and the like, and renting those out as opposed to using those for their own AI, training their own AI models.” This distinction is crucial. While many companies are indeed building out massive data centers, the priority for most remains internal AI model training. xAI’s decision to rent out its capacity therefore suggests a departure from this norm, hinting that its own AI model training might not be as robust or advanced as initially anticipated.

Dual Perspectives: Optimism vs. Cynicism on xAI’s Strategy

The podcast discussion quickly veered into the two primary interpretations of xAI’s new direction.

The Positive Spin: A New Revenue Stream. Kirsten offered “a positive view” on the partnership. From this perspective, xAI has simply “figured out a way to make money.” In a climate where AI companies are constantly seeking viable business models, transforming an underutilized asset into a revenue generator is undeniably a smart move in the short term. It provides a more tangible and reliable income stream, which could be appealing to investors looking for stability.

The Cynical Take: An IPO “Heat Check.” Sean, however, immediately countered with a more skeptical outlook: “Why be positive when you can be cynical?” He viewed the deal as “a major heat check before the IPO that we’re about to see get rammed into the markets with SpaceX.” The implication is that this isn’t primarily about genuine innovation, but about presenting a more “believable business in the near term” to potential public investors. A “neocloud” service offers a clearer, more immediate path to profitability than the speculative, long-term returns of a frontier AI lab, especially one facing an environmental lawsuit over its Colossus 1 facility.

The Innovation Dilemma: Grok’s Limitations and xAI’s Identity Crisis

A recurring theme in our discussion was the performance and perception of Grok, xAI’s flagship chatbot. “Most people would not necessarily think of Grok as — not only that it’s known for some pretty unpleasant, if not downright illegal, content, but also it’s not necessarily super cutting edge,” I noted. The challenge for xAI is particularly acute in the enterprise AI space, where reliability, accuracy, and ethical content are paramount. “You don’t hear a lot about people using Grok for work-critical tasks,” I added.

This leads directly to xAI’s identity crisis. As Kirsten highlighted, “when you are positioning your company — in this case, SpaceX-slash-xAI — as a forward-looking, innovative company, that’s tougher to sell if you are simply just renting out your GPUs and not using them for that innovation.” If xAI isn’t actively pushing the boundaries of AI model training with its own infrastructure, its claim to being a “frontier lab developer” becomes tenuous.

Sean further elaborated on the internal turmoil at xAI, citing reports that employees themselves “were using other models, they weren’t even using [Grok] internally.” This internal skepticism reportedly led to a major shakeup, with all co-founders except Elon Musk departing. Musk’s declaration that he’s “starting from scratch on xAI,” despite SpaceX’s reported $250 billion valuation for it, paints a picture of a company struggling to find its footing.

The ultimate sign of this internal reevaluation is the impending dissolution of xAI as a separate entity, with Musk reportedly beginning to refer to the integrated operation as “SpaceXAI.” This move, while perhaps simplifying corporate structure, also dilutes the brand of a dedicated AI research company. As Sean put it, “this man loves nothing but to ruin a brand that has some value to it — see Twitter.”

While a neocloud business might indeed offer “a bit more reliable and certainly more real” financial footing for an IPO, it fundamentally changes the narrative. “It’s also not the kind of business that’s going to draw the same — at least, in a normal environment — outside investment that we’re seeing go into all the frontier labs,” Sean concluded. This tension between short-term reliability and long-term, high-growth innovation is perhaps one of the biggest challenges facing SpaceX’s IPO process.

BOTTOM LINE

The Anthropic-xAI deal marks a pivotal moment, exposing the raw tensions between the pursuit of cutting-edge AI and the hard realities of monetization and investor appeal. While it provides xAI a much-needed revenue stream and bolsters SpaceX’s IPO prospectus with a more tangible business model, it simultaneously casts a long shadow over xAI’s aspirations as a leading AI innovator. The pivot to a “neocloud” operator, coupled with internal struggles and Grok’s lukewarm reception, suggests a strategic retreat from the AI frontier, prioritizing financial stability over groundbreaking research in the immediate term. For investors, the question remains whether this pragmatic shift signals shrewd management or a loss of pioneering vision.