According to TheRegister.com, in late 2025 Nvidia agreed to invest up to $100 billion in OpenAI to finance a massive AI data center buildout requiring millions of Nvidia GPUs. Weeks later, OpenAI struck a $300 billion cloud deal with Oracle, who plans to spend $40 billion on roughly 400,000 Nvidia GB200 chips to power the project. AMD also joined the frenzy, granting OpenAI warrants for a 10% equity stake in exchange for deploying 6 gigawatts of AMD hardware. CoreWeave secured $22.4 billion in OpenAI contracts while Nvidia both invested in and pre-purchased $6.3 billion in services from the cloud provider. These circular deals created instant paper wealth, with Oracle stock spiking 36% in a day and Nvidia adding $170 billion to its market cap.
Sponsored content — provided for informational and promotional purposes.
The trillion-dollar feedback loop
Here’s the thing about this AI economy: it’s basically a giant closed system where everyone invests in everyone else. Nvidia puts money into OpenAI, who buys Nvidia chips. Oracle gets a massive contract from OpenAI, then turns around and buys Nvidia hardware to fulfill it. CoreWeave gets investment from Nvidia, then buys more Nvidia chips to serve customers like OpenAI. The money just keeps circulating among the same players.
And honestly? It’s working brilliantly for them on paper. When that Oracle deal hit, Larry Ellison’s net worth jumped $88 billion overnight. Nvidia added another $170 billion in market cap. These aren’t small numbers – we’re talking about wealth creation that would have seemed impossible just a few years ago.
Dot-com déjà vu?
So is this sustainable? Critics are definitely getting flashbacks to the late 1990s dot-com bubble where companies basically just bought each other’s ads. The difference here is that actual physical infrastructure is being built – data centers, chips, the whole nine yards. But the self-reinforcing nature of these deals does make you wonder.
Think about it: Nvidia investing $100 billion in a company that then has to buy Nvidia products. That’s essentially bankrolling your own future sales. It creates the appearance of explosive growth, but how much of it is organic demand versus just money moving in circles?
Where Washington fits in
Now here’s where it gets really interesting. The government has become part of this loop too. Through the CHIPS Act, the US took a 9.9% stake in Intel worth $8.9 billion. Traditionally, governments don’t play the stock market like this, but chips have become a national security priority.
Basically, public money is now feeding the same ecosystem. If Intel succeeds in making advanced AI chips domestically, they become another player in this circular economy. And Washington’s restrictions on China mean all this money has to flow inward, potentially intensifying the loop pattern.
Where does this end?
Jensen Huang isn’t shy about his thinking – he said on this podcast that he believes OpenAI will be the next multi-trillion-dollar hyperscale company. So from his perspective, investing $100 billion is a no-brainer. But what happens if one link in this chain breaks?
The sheer scale of these deals is unprecedented. OpenAI was valued around $30 billion in early 2023 – now we’re talking about single investments that dwarf that. The planning security is great for these companies, but the systemic risk if demand doesn’t materialize could be massive. We’re building an entire economic ecosystem on the assumption that AI demand will continue growing exponentially. Let’s hope that assumption holds.
