Coreweave’s CEO says the AI cloud’s wild ride is just getting started

According to TechCrunch, Coreweave CEO Michael Intrator defended his company’s performance and strategy at Fortune’s AI Brainstorm summit in San Francisco. The AI cloud infrastructure provider had a major IPO in March that didn’t meet hype, saw a planned acquisition of Core Scientific falter in October, and has a stock price that debuted at $40, climbed over $150, and now sits around $90. Intrator blamed post-IPO economic “headwinds” and argued they are creating a “new business model,” even using their valuable Nvidia GPU inventory as collateral for loans. He also dismissed concerns about “circular” deals between major AI players like Nvidia, a key investor and supplier, saying companies are just “working together” to manage supply and demand. Since March, Coreweave has acquired companies like Weights and Balances, OpenPipe, Marimo, and Monolith, while expanding a cloud partnership with OpenAI by up to $6.5 billion and announcing plans to enter the U.S. federal market.

The meme stock of AI cloud?

Look, you can’t talk about Coreweave without talking about that stock chart. It’s been a rollercoaster. Debuting at $40, shooting past $150, now back down to around $90? That kind of volatility gets you compared to a meme stock, and for good reason. Intrator calls critics “myopic,” and points to the tough economic timing of their IPO. And sure, that’s a factor. But here’s the thing: the recent 8% drop came right after they announced more debt to finance data centers. The market is clearly jittery about their leverage. They’re borrowing against their GPUs to build more capacity to host… more GPUs. It’s a high-stakes, capital-intensive bet that the AI demand surge is permanent and not a bubble.

The circular economy defense

This is where it gets really interesting. Intrator swatted away concerns about the incestuous, “circular” nature of AI investing—where Nvidia invests in Coreweave, which buys billions in chips from Nvidia, and both service the same big clients like OpenAI. His argument? “Companies are trying to address a violent change in supply and demand. You do that by working together.” Basically, he’s saying the old rules don’t apply when you’re building infrastructure for a technological gold rush. But is that collaboration, or is it a dangerous feedback loop? Wall Street is definitely asking questions. If demand from a few mega-projects slows, does this whole house of cards shudder? It’s a valid concern he’s basically telling us to ignore.

Acquisition spree and federal dreams

While the stock bounces and critics chatter, Coreweave isn’t sitting still. Their acquisition spree is telling. They’re not just buying more servers; they’re snapping up AI developer tools (OpenPipe, Weights & Balances), open-source notebooks (Marimo), and even industrial AI firms (Monolith). This isn’t just about raw compute power anymore. It’s about building a full-stack platform to lock developers in. And the move into the federal market is a huge signal. Government contracts are long-term, stable, and massively lucrative. If they can become the go-to AI cloud for the U.S. government and defense base, it would be a massive validation and a hedge against private sector volatility. It’s a smart, albeit expensive, play.

Building on borrowed time and chips

So what’s the real takeaway? Coreweave is executing a blitzscale strategy in the most capital-intensive sector imaginable. They’re using debt, strategic investor money, and a volatile public stock to fuel a land grab. Intrator’s right that disrupting the static cloud oligopoly of AWS, Google, and Azure requires a new model. But their model is incredibly risky. It depends on unrelenting AI demand, continuous access to Nvidia’s scarce chips (and Nvidia’s continued goodwill as an investor), and the ability to out-build everyone else. When you’re in a race this fast, on a track this expensive, every bump feels like a canyon. They might be building the future of AI infrastructure. Or they might be the perfect case study of what happens when the hype cycle meets the debt cycle. Only time, and probably a few more earnings calls, will tell.