![](\"https:\/\/www.newsbeep.com\/ca\/wp-content\/uploads\/2025\/09\/7a7414080e993fa22f82a9f9f0c56f82cb-altman-AIpays.rhorizontal.w1100.jpg\")
<\/p>\n <\/p>\n
\n Photo-Illustration: Intelligencer; Photo: Getty Images\n <\/p>\n
No two companies are as important to the recent AI boom as OpenAI and Nvidia. OpenAI\u2019s journey from research lab to chatbot juggernaut reoriented the entire tech industry around the thesis that large-language models will change everything. Nvidia\u2019s chips have been almost synonymous with scaling, on which OpenAI and others are collectively spending hundreds of billions of dollars. For the last few years, the companies\u2019 fortunes have been aligned, and their relationship has been, for the most part, an obvious one: OpenAI, which has become one of the biggest start-ups in the world, is a major customer for Nvidia, which has become the largest public company in the world.<\/p>\n
In 2025, their twin trajectories have left much of the rest of the economy behind, and their stated ambitions \u2014\u00a0necessary to keep the momentum going \u2014\u00a0have become extreme. In a recent blog post, Sam Altman said his company wanted to \u201ccreate a factory that can produce a gigawatt of new AI infrastructure every week\u201d to \u201cmaybe\u201d be able to \u201cfigure out how to cure cancer\u201d or \u201cprovide customized tutoring to every student on Earth.\u201d OpenAI hasn\u2019t had any trouble raising money so far, but when your investment projections start crossing into the trillions<\/a>, financing starts to get tricky, particularly if you aren\u2019t a company like Google or Meta, with an existing business throwing off tens of billions of dollars a year. Which might help explain<\/a> this:<\/p>\n OpenAI and NVIDIA today announced a letter of intent for a landmark strategic partnership to deploy at least 10 gigawatts of NVIDIA systems for OpenAI\u2019s next-generation AI infrastructure to train and run its next generation of models on the path to deploying superintelligence. To support this deployment including data center and power capacity, NVIDIA intends to invest up to $100 billion in OpenAI as the new NVIDIA systems are deployed.<\/p>\n Emphasis mine! The issue of circular financing in the AI world has been bubbling up for a while. As the Information reported<\/a> in May, Nvidia has been engineering disorienting deals with customers for a while, most obviously with a company called Coreweave, which rents compute \u2014\u00a0basically, access to Nvidia hardware \u2014\u00a0to AI firms:<\/p>\n The chip giant invested $100 million in [Coreweave] in early 2023. It funneled hundreds of thousands of high-end graphics processing units to CoreWeave. And it agreed to rent back its chips from CoreWeave through August 2027.<\/p>\n In related news, just this week<\/a>, Coreweave announced it had \u201cexpanded its partnership with OpenAI in a new deal worth up to $6.5 billion, bringing the total value of their agreements to $22.4 billion.\u201d These announcements aren\u2019t exactly attempts to obfuscate what\u2019s going on here \u2014\u00a0\u201cNVIDIA intends to invest \u2026 as the new NVIDIA systems are deployed\u201d \u2014\u00a0but it\u2019s worth stating even more plainly what\u2019s happening here: Nvidia is investing money in one of its largest customers, which will use some of that money to buy or lease capacity from Nvidia. On Tuesday, we got a little more insight<\/a> into how that would actually happen. OpenAI, Oracle (a cloud provider that depends on Nvidia), and Softbank (a major investor in Nvidia and OpenAI) fleshed out plans to build out more data centers, which would one day be full of Nvidia hardware paid for or leased by \u2026 OpenAI.<\/p>\n If you\u2019re a layperson, this probably sounds a bit weird. Maybe our friends in finance can explain a hidden logic here? Rich Privorotsky of Goldman Sachs attempts<\/a> to summarize the arrangement: \u201cNvidia invests up to $100bn for non-voting shares, and OpenAI uses the cash to buy Nvidia chips, with a plan to deploy what could be at least 10GW of Nvidia systems.\u201d It is what it looks like, in other words: a small group of large companies handing money back and forth.<\/p>\n With a number of caveats \u2014\u00a0every cycle is different, and the earnings multiples aren\u2019t as crazy yet \u2014 he also draws a parallel to the late 90s, which have been coming up a lot lately. \u201cVendor financing was a feature of that era, when the telecom equipment makers (Cisco, Lucent, Nortel, etc.) extended loans, equity investments, or credit guarantees to their customers who then used the cash\/credit to buy back the equipment,\u201d he writes. Indeed, vendor financing features prominently in tech-bubble postmortems<\/a>. In 2002, Newsweek explained<\/a> the strange role they played at the tail end of the \u201990s tech-investment boom:<\/p>\n Swept up in the free-money spirit of the time, they were financing not only the customers\u2019 purchase of their wares \u2014 the switches, routers and other nuts and bolts of the Internet\u00a0\u2014 but the growth of the customers\u2019 businesses, too. The standard practice of vendor financing thus became another wretched bubble excess: as if Ford were loaning customers money to buy cars \u2014\u00a0and boats, jewels and houses, too. \u201cVendor financing was the sickest sign of how things spun out of control,\u201d says Tero Kuittinen, technology adviser to Finnish investment bank Opstock in Helsinki.<\/p>\n The parallels are obvious \u2014\u00a0or, at least, would look extremely obvious if the AI industry\u2019s fortunes turned \u2014\u00a0but they\u2019re not determinative. During the telecom buildout, vendor financing surged at a particular time. \u201cVendor-financing deals account for a relatively small fraction of the total $1 trillion in debt held by telecom carriers worldwide,\u201d Newsweek noted at the time, but investment continued well after the early 2000 market collapse \u201cwiped out the silliest ideas for pet, toy and trinket Web sites,\u201d on the thesis that \u201cdot-coms might be dead, but the Internet itself would thrive,\u201d a perfectly reasonable argument if timing doesn\u2019t matter, which of course it does, quite a bit. Last month, Kuittinen argued that this time, things really are different, but not necessarily in a good way:<\/p>\n The fascinating thing about AI bubble popping is that it\u2019s not gonna play out like Lucent in 2001. Because this crew does not believe in free markets. They\u2019re gonna try to prop up this thing like Soviet television industry in 1980\u2019s. Tax payer money is gonna FLOW.<\/p>\n \u2014 Tero Kuittinen (@teroterotero) August 21, 2025<\/a><\/p>\n The constellation of companies raising, spending, and sharing money on the broad technological and ideological project<\/a> that is artificial intelligence in the 2020s is betting that, given enough resources, they can build the most profitable and powerful companies of all time, and that early concerns over profitability and circular financing will, in hindsight, seem blinkered and silly. (Or, according to Peter Thiel, like complicity<\/a> in the summoning of the Antichrist.) In the process of trying, the industry is willing to drift into a strange and potentially risky arrangement \u2014\u00a0although some of the implied risk might be mitigated, as Alphaville mockingly points out<\/a>, by vagueness and lack of binding obligations \u2014 which, in any case, given the potential upside, investors still seem comfortable with for now: a hypercapitalized financial polycule in which everyone is everyone\u2019s investor, partner, vendor, and customer.<\/p>\n Sign Up for John Herrman column alerts<\/p>\n Get an email alert as soon as a new article publishes.<\/p>\n Vox Media, LLC Terms and Privacy Notice<\/p>\n