Key Takeaways

  • Meta Platforms is reportedly developing a cloud infrastructure business to sell access to its AI computing power and models, directly competing with industry titans like AWS and GCP.
  • This potential move could monetize Meta's “hundreds of billions of dollars” in AI infrastructure CapEx, leveraging idle compute capacity for a clear ROI stream.
  • Jordi Hays questions the strategic consistency of this pivot, arguing it clashes with Meta's stated goal of "personal super intelligence" and its lack of consumer AI products.
  • John Coogan highlights the market's mixed reaction: positive for Meta's revenue potential but negative for "neo clouds" (like CoreWeaver Iron) now facing a new, massive competitor.
  • The bearish take suggests Meta selling compute indicates they aren't constrained, potentially undermining the value proposition of companies built on selling scarce AI infrastructure.

The Disagreement

Jordi Hays, a self-described "cheerleader for Meta," voiced deep skepticism about the company’s rumored plan to launch a cloud business selling AI compute. While acknowledging the practicality of the move, Hays found the overall strategy confusing. “I think it is practical what they're doing,” Hays conceded, “But as somebody that that, you know, I would say overall has been a big, you know, cheerleader for Meta...it doesn't give you a lot of confidence in like the strategy overall.” He pointed directly to the elephant in the room: Meta's enormous investment in AI. “Well, I think that's what people have been wondering where's the ROI going to come from for this hundreds of billions of dollars to spend? And up until now there's been no obvious place that it's going to come from, right?” Hays implied this cloud business felt like a pivot born of necessity rather than a cohesive grand vision for consumer AI.

John Coogan offered a more pragmatic lens, seeing the cloud business as a way for Meta to extract tangible returns from its colossal AI CapEx. He described the "bullish take" as Meta building a serious cloud business, even if it means "spending more on CapEx to compete with GCP, AWS, and Azure." Coogan also laid out the "bearish take," observing that if Meta has "access compute that they are willing to sell," it suggests the market isn't as compute-constrained as some believe. This could be "really bad for neo clouds" like CoreWeaver Iron, as Meta might bypass deals with them and sell compute directly, cutting into their market. The market's reaction, Coogan noted, was telling: a positive bump for Meta, but a negative one for "neo cloud" providers anticipating a new, powerful competitor.

Who's Right (and When They're Wrong)

Both Coogan and Hays land on valid points, reflecting the tension every founder faces between audacious vision and practical monetization. Hays rightly flags the strategic whiplash. For a company like Meta, known for its long-term bets on the metaverse and "personal super intelligence," a sudden shift to selling raw compute power feels incongruous. It dilutes the narrative, potentially confusing employees, partners, and the market about Meta's true North Star. Founders, especially those in deep tech, understand the need for unwavering focus on a difficult, distant goal. A public company like Meta, however, also operates under relentless pressure for quarterly ROI on "hundreds of billions of dollars" in CapEx.

Coogan's perspective acknowledges this reality. Selling compute could be a necessary evil, a way to fund the longer-term vision by monetizing massive, existing infrastructure. It’s a cynical but pragmatic move. If Meta has excess capacity, why let it sit idle when demand for AI compute is sky-high? For a scrappy founder, this move is a playbook for surviving the capital demands of innovation. But for others, particularly those building "neo cloud" businesses like CoreWeaver or Iron, Meta's entry as a direct competitor signals a brutal new reality: a giant just entered their sandbox, potentially squeezing margins and access to critical resources. The "right" answer depends on your vantage point: strategic purity versus financial pragmatism.

What to Do With This

If you're building an AI-native startup, closely examine your current and projected compute spend. Ask: Are you over-provisioning, and could you temporarily monetize excess capacity to fund your core roadmap, even if it feels off-brand? If you're a "neo cloud" provider, immediately model the impact of a hyperscaler like Meta entering your direct market—what's your differentiation, and how will your pricing respond to this new, massive competitor?