The OpenAI Bubble

Log In<br>Subscribe

Sign up

Sign in

Log in

Subscribe

The OpenAI Bubble

Ed Zitron<br>Jul 15, 2026<br>59 min read

Table of Contents

Thanks for reading this week’s free Where’s Your Ed At newsletter. As I said last week, I’m taking the rest of this week off, so there won’t be a premium on Friday. That said, if you aren’t already a member, now’s a great time to subscribe.<br>To celebrate the one year anniversary of the premium newsletter, I’m offering a sale on one-year subscriptions. Between now and midnight July 22, you can get a permanent annual rate of just $60— a $10 discount on the usual price of $70 - for life. Click here for the offer.<br>In addition to getting access to the entire back catalog of premium posts, you’ll also receive one additional post each week — usually anywhere between 10,000 and 20,000 words — covering the most pressing topics in the AI bubble - the best value in tech analysis. Highlights include last week's Hater's Guide To The Memory Crisis - a guide to how AI made everything more expensive - How OpenAI Kills Oracle (which pairs nicely with the Hater's Guide To Oracle), The Hater's Guide To NVIDIA, The Hater's Guides To Private Credit and Private Equity, and how the entire AI Compute Demand Story Is A Lie.<br>Today’s piece is one of the largest free newsletters I’ve ever written, and pulls together the last six months of my work.<br>And it all starts with a question: how much do you trust Sam Altman? The stock market and (to some extent) the global economy rests on your answer.<br>You see, OpenAI has become one of the largest liabilities in recent economic history. You can argue that OpenAI’s no longer the focal point of the AI bubble — you can talk all you want about open source models or Anthropic or any number of other elements — but without OpenAI, the AI industry doesn’t exist, and the justification for trillions of dollars of capex evaporates.<br>The AI bubble isn’t a result of any actual return on investment — whether that be in purely monetary terms, like revenue or profitability , productivity gains, or anything tangible or measurable. Rather, it’s an episode of cult-like psychosis that infected the brains of some of the most powerful and wealthy individuals and institutions, where the powerful mythology of a company inspired — and been used to inspire — the greatest capital misallocation in history.<br>As much as this’ll piss some people off, I fully believe that the only reason this has kept going so long is that OpenAI has yet to collapse. Its failure would be a watershed moment — the Lehman Brothers of the AI bubble, and an event that would define the end of one epoch, the start of another, and that would shake the afflicted out of that psychosis. Absent this wake-up call, NVIDIA has continued to sell GPUs, the coffers of the semiconductor industry have continued to swell, and more and more spending commitments have been made.<br>Look. OpenAI intends to burn over $852 billion by the end of 2030. It accounts for $748 billion of the remaining performance obligations of Microsoft, Amazon, and Oracle, on top of at least another $70 billion of RPOs across Cerebras, CoreWeave, Nebius, IREN, Lambda, and Nscale (per Kakashii), and plans to spend indeterminate billions’-worth of Broadcom “Jalapeno” chips. It intends to spend $50 billion or more on compute this year, which I estimate is more than 50% of all global AI compute spend (with OpenAI taking up 50%+ of all AI compute infrastructure).<br>OpenAI can only afford to pay that as a result of its latest (assuming it fully closes) $122 billion funding round, of which it has received at least $50 billion, with $20 billion from SoftBank (of $30 billion, with the third tranche due October 1, 2026). NVIDIA mentioned in its latest quarterly earnings report that it “estimate[d] that one AI research and deployment company contributed to a meaningful amount of [its] revenue by purchasing cloud services from [its] customers in the first quarter of fiscal year 2027,” referring, of course, to OpenAI.<br>The AI Bubble Is An OpenAI Bubble — To A Mortal End<br>OpenAI is the reason anyone cares about AI. In March 2019 (per JustDario), NVIDIA bought a company called Mellanox that made the high-speed networking tech necessary to create AI GPU clusters, and four months after that, Microsoft invested a billion dollars in OpenAI and started buying AI GPUs and building AI infrastructure for it. By March 2020, NVIDIA would ship its A100 GPU, and in May 2020, Microsoft would announce it had built a supercomputer just for OpenAI with “more than 285,000 CPU cores [and] 10,000 GPUs.”<br>The launch of ChatGPT in November 2022 came at the perfect time for a tech industry that had run out of ideas and was flirting with a prolonged depression. The IPO market had collapsed, interest hikes killed the Zero Interest Free era dead, pandemic era overhiring began to unwind with some of the worst layoffs in the history of the industry, global venture funding dwindled after...