Déjà Vu? AI Bubble Fears Intensify Amid Nvidia-OpenAI Collaboration
Three years after OpenAI and Nvidia helped ignite the global AI frenzy, the two companies are joining forces in a bold move aimed at paving the way for a more capital-intensive growth phase, but one that also fuels mounting concerns about an AI "bubble."
Nvidia recently announced a potential investment of up to $100 billion in OpenAI, coupled with plans to supply the ChatGPT creator with millions of market-leading AI chips. This move, virtually unprecedented in the tech sector, has prompted some analysts to question whether Nvidia is artificially propping up the market to sustain corporate spending on its products.
"This action will clearly exacerbate 'circular' concerns," wrote Stacy Rasgon, an analyst at Bernstein Research, in an investor note following the deal announcement. These concerns have shadowed Nvidia to varying degrees throughout the AI boom. According to PitchBook, the chipmaker has participated in over 50 venture capital deals with AI companies in 2024, and is on track to surpass that number this year. Some of these firms, including AI model makers and cloud service providers, then use the funds to purchase Nvidia’s pricey graphics processing units (GPUs).
But Rasgon argues that the OpenAI investment "dwarfs all other investments". He adds that this deal "may exacerbate these concerns more acutely than we’ve seen before and (perhaps rightfully) raise worries over the motivations behind the action." Nvidia has stated, however, that the OpenAI investment will not be used for any “direct purchasing” of its products.
Wall Street's 'Perpetual Motion Machine'
Earlier this month, Oracle announced it had signed billions of dollars in contracts to provide cloud computing services to OpenAI and a few other large clients, sending its shares soaring and making co-founder Larry Ellison one of the world’s richest people.
With Nvidia's investment in OpenAI, a complete “triangular relationship” is forming. Under the agreement, OpenAI will use Nvidia technologies to build and deploy AI data centers with a capacity of at least 10 GW, equivalent to 4 to 5 million Nvidia GPUs – nearly Nvidia's total shipments this year. This means:
* OpenAI purchases massive cloud services from Oracle.
* Oracle buys a significant number of GPU chips from Nvidia to meet OpenAI's needs.
* Nvidia returns a portion of its profits to OpenAI in the form of investment to support infrastructure construction.
This seemingly perfect circle allows all parties to get what they want: OpenAI solves its extreme thirst for computing power, Oracle gets huge infrastructure orders, and Nvidia improves its sales, forming a “perpetual motion machine.” A picture also circulated on social media showcasing this “triangle” built by the three giants, sparking heated debate.
Bubble Risks Loom
However, this is a no-fail scenario. Although this collaboration has driven up Nvidia and Oracle shares, the underlying risks are substantial. First, although OpenAI is valued at hundreds of billions of dollars, it has yet to become profitable, and continues to record losses. Estimates suggest its losses in 2025 could exceed $5 billion, and its annual spending on data centers and cloud services alone is $60 billion, far exceeding its current $10 billion in annual revenue.
Second, Oracle also faces challenges. In order to keep up with the AI craze, its high spending has led to prominent debt and negative cash flow problems. Compared to competitors such as Microsoft, Oracle's debt-to-equity ratio is 427%, indicating a heavy burden.
With competition from strong rivals such as Google and Anthropic, the “ONO” (OpenAI, Nvidia, Oracle) alliance looks more like a high-stakes game that allows only success. Any problem in any ring could lead to a domino effect, or even ignite the market.
The Nvidia-OpenAI deal also comes at an uncertain moment in the industry. More people inside and outside the industry are acknowledging the risk of an AI bubble in the US stock market, similar to the dot-com bubble burst 25 years ago. OpenAI CEO Sam Altman has stated that some valuations of AI startups may be irrational, though he claims he believes in the long-term potential of AI and needs to invest "trillions" of dollars in infrastructure to support it.
By building a closer relationship with this most valuable company in the world, OpenAI may be able to unlock funding and more compute power that it cannot currently obtain on its own as a loss-making business.
"It’s a bit like having your parents co-sign your first mortgage," says Jay Goldberg, an analyst at Seaport Global Securities, who holds a rare “sell” rating on Nvidia stock. Goldberg also believes the deal smacks of circular financing and could be a sign of “bubble-like behavior.”
"When things are good, it makes things better. We’re going to grow faster, the stocks are going to go up a lot faster," he says. "But when the cycle turns, and it always does, it’s going to make things worse."
Further Analysis
It is important to recognize that the AI market is still in its early stages, and many factors could affect its future course. These factors include technological advancements, regulatory changes, and global economic conditions. A critical assessment of the potential opportunities and risks associated with AI is essential, rather than relying solely on media hype and optimistic projections.
KO
