
Over the last three years the AI boom has become a giant circle jerk, with a private company at its center. Since OpenAI is private, we don’t know how much money they’re making, or whether they’re making any.
We assume, because we’re told, that OpenAI is currently losing money, but that the losses will narrow as revenue grows, ending in profit. We were told the same thing by Internet companies 30 years ago and know how that turned out.
TapTwice Capital has some graphs showing all this. It claims that losses will start to decline after 2026, after total losses of $44 billion. OpenAI “hopes” to be cash flow positive in 2029.
But that won’t happen because Large Language Models (LLMs) don’t get better as they get bigger. They get worse. Everyone in the field knows this, many of them are saying this. What’s called “vibe coding,” the use of LLMs to create new software, is already declining. One big investor calls it “a joke.”
OpenAI executives continue to insist that “Artificial General Intelligence” has arrived, is arriving, or is just around the corner. They’re lying.
The collapse of the sector happens when the numbers stop working. According to TapTwice, that will happen in early 2027. I believe it will happen earlier.
The House Collapses

What happens when we find it’s all a house of cards? Take out OpenAI and you’re pulling out a foundation piece of the industry. The structure collapses.
If OpenAI disappears, what happens to all those data centers the Cloud Czars and the Wannabes are building? They’re worth much less, if they’re not worthless. Write down those values and what happens to the value of public company stocks tied to that boom?
This is the problem with market cap wealth. It’s a company’s share price multiplied by the number of shares. If the last trade is priced at 1% less than the previous trade, that’s the new value of the company. If the last trade is 10% less, the same. What if it’s 20% less? What if no one wants to trade at all?
The only thing keeping the boom alive is belief, which includes the skepticism voiced in columns like this. Booms bust when everyone is all-in. We’re not all-in on this one. Stocks need a “wall of worry” to rise. It’s when that wall falls, when everyone is all in, that the collapse occurs.
This requires a precipitating event.
The Precipitating Event

There are many possible precipitating events regarding OpenAI. The most obvious would be a “down round,” where investors pay for new stock based on a lower valuation than a previous raise. To avoid a down-round, someone like Microsoft, Google, Amazon or even Nvidia might buy OpenAI.
Whatever their price, that will be seen by the market as the maximum value of AI assets, and the collapse will follow.
Right now, OpenAI is holding the entire economy to ransom. Keep supporting it, on the quiet, through funding rounds, or it all goes down. But that can’t last much longer. A lot of the “investment” being thrown around the AI space is, like Microsoft’s $13 billion deal in early 2023, in-kind. Most of the “money” wasn’t cash, but Azure compute. There have been a lot of deals like this, some involving Nvidia chips, or the promise of chips, or the promise of services.
What’s missing is cash.
Cash comes from investors, and when that spigot gets turned off bad things will happen. These will be very bad things, given the size of the bubble, that will have awful knock-on effects throughout the global economy.
That is why no one is talking about it. You’re lucky no one listens to me, or the Naked Emperor’s condition would already be revealed.






