Like many others, I continue to be baffled and unnerved at how transactions that seem like definitional signs of a bubble (or possibly even fraud) are now so thoroughly baked into the very structure of the AI industry. This morning, Semafor’s Liz Hoffman starts a piece with this overview: “OpenAI and Anthropic need to borrow a lot of money but they don’t have profits or a track record of creditworthiness. Enter Jensen Huang.”
It goes on from there …
Nvidia’s $250 billion backstop to OpenAI will let the money-burning AI lab lease space at the largest data center ever built. OpenAI doesn’t have an investment-grade rating, so Nvidia is essentially lending its own. Broadcom did the same thing for Anthropic a few weeks ago; I wrote at the time that it was “like getting your parents to cosign the lease on your first apartment.” Nvidia’s backstop for OpenAI is literally that — OpenAI is trying to sign a lease and its landlord, SoftBank, doesn’t like the tenant risk. So Huang is cosigning.
This is not like getting your parents to cosign the lease on your first apartment. It’s more like getting your top employee to cosign your first lease because you pay that employee huge sums of money despite the fact that your company, which pays his salary, actually makes no money. A bank would likely see the problem with the top employee co-signing the mortgage on the boss’s fancy home. But it doesn’t seem clear to people in this case. Or rather, it seems completely clear. But we seem to have decided this is just how AI works: the technology is so amazing that it requires this kind of mutual leverage with no floor beneath it.
I’m reminded of some old Bugs Bunny or maybe Roadrunner cartoon where Bugs gets to the last rung at the top of his ladder. But he needs to go higher. So he grabs a rung from the bottom of the ladder and puts it on the top and keeps doing that and now he can just climb up without limit. As we know this only works with cartoon physics.
For what it’s worth, my own unlettered assumption is that AI is something like the internet circa 2000: both a massive bubble but also a transformative technology that would and did transform the future. The difference that worries me is that the dot com bubble was genuinely broadly based. The opportunity was obvious. Lots of companies were starting up to get a piece of that opportunity and lots of investors wanted to fund those gambles. It was a bubble. It did burst. But we now see the bursting of the dot com bubble as essentially a speed bump on the road to the tech boom that has defined in good ways and bad (increasingly bad) the economy and society of the early 21st century. The AI boom is not broad-based in the same way at all. It is driven overwhelmingly by the existing tech monopoly platforms. Even the new entrants — Anthropic, OpenAI, et al. — are heavily tied to the same sources of capital and often the same people. When you add in the fact that the tech monopoly platforms are mostly structured to keep control in the hands of CEO-founders, you have the fact that the decisions are being made by a very, very small number of people.
Of course, the stock market gains require widespread buy-in, and perhaps even more since they are heavily dependent, more than in the past, on retail investors. Still the core decision-making is very, very tightly held. And while it’s possible to interpret that fact in other ways, it certainly seems to me like it spells trouble.
