The New Financial Union of 3 Tech Giants
Michael Burry of ‘Big Short’ fame has made a $845 million bet against the American AI industry. The question is, why? Well, for a number of years, analysts have alleged that the AI industry is vastly overvalued and exists in a financial bubble, and this month has seen an ongoing and deeper confirmation of this in a striking cycle. Three of the world’s most valuable companies: OpenAI, Nvidia, and Oracle, have devised for themselves a closed web of chummy deals amounting to literally trillions of dollars, leading to even further concerns of a bubble waiting to pop.
The $5 trillion Nvidia, the world’s most valuable company, has become an integral part of the AI ecosystem. They have a near-monopoly on the crucial tech used to build the data centres that power AI models, namely their GPUs. Oracle is the foremost company that builds these data centres, a line of work that has made them billions, and briefly made their co-founder Larry Elleson the richest person in the world. Nvidia is their main source for GPUs, deals worth $40 billion. OpenAI has utilised these Oracle datacentres time and time again, with the latest deal of an eyewatering $300 billion. The problem lies with the financial viability of AI companies, and how the ecosystem has sought to rectify it. Even previously, AI companies have been in a strange position. There is currently no financial model for profitability in AI, so companies often run at a loss, kept afloat only by lofty investments. If those investments dry up, no more AI. The sheer scale of cost is also near unprecedented for its non-existent profitability. Hundreds of billions are being sunk into solving the mystery of super-intelligence with no real returns in sight. Clearly this is a bet big tech believe will pay off in the end, but to maintain its position, they have had to master the art of perceived value and hype. Even if it is not financially viable, everyone is talking about AI, and the most recent deals have
invigorated investors, with stock prices of all three of these companies skyrocketing.
OpenAI has been searching for a way to fund their “trillion dollar plan” for a few years now. Last week, their CEO Sam Altman said at a press conference: “There’s always a lot of focus on technological innovation, but what really drives progress is when people also figure out how to innovate on the financial model, the financial instruments. I don’t think we’ve figured out yet the final form of what financing for compute looks like.” He has also said he is willing to run at a loss. Currently OpenAI is around $5 billion dollars in the red. Yet the money has to come from somewhere, which is where investors like Nvidia come in. Their latest investment into OpenAI, to the tune of $100 billion, is seen as deeply problematic and brings with it more questions than answers. This is a case of the supplier investing in the ecosystem. By funnelling huge investment into OpenAI, that cash will eventually make its way back to Nvidia through Oracle datacentres. Although, who else would invest such money into a business that has no current promise of returns? Nvidia probably views themselves as one of the only possible candidates to help OpenAI. Even if there are no real returns on investment, at least there will be more GPU sales later down the line.
Be this as it may, it does not excuse the reality of a bubble. If, for whatever reason, one company cannot pay their expenses, or their deals fall through, then the cycle breaks down and the bubble pops. Perhaps the most likely member for that is OpenAI, since they are fully dependent on outside investment. As it happens, this is not the first time Nvidia has been involved in a scheme like this. The company CoreWeave, which builds data centres, needed a lifeline investment. Nvidia gave them 6.3 billion. The results? CoreWeave now uses Nvidia GPUs. Then there is the London based Nscale, a unicorn startup that received a $1.1 billion Nvidia-backed funding round. Most recently announced, Nscale will be working as an AI infrastructure partner for Microsoft, OpenAI, and, you guessed it, Nvidia. This is a sort of fake growth, which serves only to inject cash into a small cycle, rather than fund a venture that will result in profits and returns. At an Nscale conference, Nvidia’s Jensen Huang reportedly told founders: “I look forward to making a fortune off you.”
There are huge similarities here with the 2000 dotcom bubble. This crash was caused by something called vendor financing, which is an arrangement where the seller provides a loan or credit to the buyer to complete the purchase. Imagine if Ford were loaning customers money to buy their cars. On the surface, Ford’s numbers would look great. They were selling cars after all. But the raw finances of it, quietly obviously, don’t work. Essentially you are paying your customers to buy your product, so there is no revenue, let alone profits, and your business will eventually crash and burn. In this dotcom bubble, the makers of internet hardware flooded the market with investment, which ballooned the stock prices far above their true valuation. When many of these companies eventually went bankrupt, stocks plummeted, and the bubble burst. It was an era of “get big fast” and “growth over profits”, remind you of anything?
In fact, the AI industry leaders have spoken on this, like OpenAI’s Sam Altman himself: “Someone is going to lose a phenomenal amount of money. When bubbles happen, smart people get overexcited about a kernel of truth.” In March, the chairman of Alibaba also warned of a swelling bubble in AI data centres: “People are talking, literally talking about $500 billion, several 100 billion dollars. I don’t think that’s entirely necessary. I think in a way, people are investing ahead of the demand that they’re seeing today, but they are projecting much bigger demand.”
One theory of why these three companies are running such a risky scheme is their faith that super intelligence (AGI) is possible and will be developed in the near-future. This is essentially artificial intelligence that is as smart or smarter than a human being. If this does happen, it will create untold amounts of profit, and the losses will all be worth it. But this is speculation, an uncertain bet. Amazon’s Jeff Bezos has said as much, suggesting that bubbles are a natural part of the industrial market: “Every experiment gets funded. Every company gets funded. The good ideas and the bad ideas. And investors have a hard time in the middle of this excitement… Industrial bubbles are not nearly as bad. It could even be good because when the dust settles and you see who are the winners, society benefits from those inventions… And that’s what’s going to happen here too. The benefits to society from AI are going to be gigantic.”
Although there are good reasons to be skeptical of an opinion from Jeff Bezos, considering his company Amazon’s $38 billion deal with OpenAI, and the fact Amazon survived the dotcom bubble as a victor, there is some merit to this. Companies like YouTube, Google, and of course Amazon, started out life running at massive losses. Yet as financial models were developed, they later became the most profitable businesses in the world.
Meanwhile, the AI stock prices just keep going up. Oracle for example has not announced higher profits, nor higher revenue. All they announced was their $300 billion dollar deal with OpenAI, tripling their Remaining Performance Obligations (RPO). This took their stock to the moon. It has left some analysts thinking, is this part of the real gameplan? To create hype for stock market investments and push their stock up? Those that get out early will make their millions, but this is how a bubble forms, and how you crash a market.
After all this hype, all these billions, OpenAI is ready to seize the moment and has prepared an IPO (Initial Public Offering) to become a publicly traded company on the stock market. I’m sure after watching their companions of Oracle, Nvidia, and Amazon surge with stock market investments, OpenAI cannot wait to join the bounty of public trading. Yet my concerns remain. Just because you make bank on the stock market does not make your company profitable or viable. It remains to be seen if they truly will change the world, or are just prolonging the inevitable.
