Ed Zitron: Big Tech Is 'Handing Money To Itself' To Inflate The AI Bubble

The AI trade is getting another vote of confidence from Wall Street.

JP Morgan says big tech's massive AI investments are starting to pay off, while NVIDIA is turning to private equity and credit giants like Apollo and Blackstone to help fund the next phase of the build-out.

But as big tech takes on more debt to keep up with the AI spending race, questions are growing over whether the returns will justify the cost.

Now, to help break it all down, I'm joined by tech industry analyst Ed Zitron. Ed, thanks so much for coming back on. Thanks for having me.

So I want to start on this headline coming out of JP Morgan.

Wall Street for years has been asking, when will we finally see AI CapEx spend pay off?

And now we're seeing JP Morgan is raising its S&P 500 target, saying that it's seeing evidence of those returns.

Do you think that they're declaring victory too early? I think they're declaring victory when they should be declaring failure.

Right now, analysts and actually Bloomberg reported as well that 70% of AI revenues are from OpenAI and Anthropic, two companies that can't afford to actually pay for the compute.

And that's across Microsoft, Google and Amazon.

Their future revenues are based on their ability to harvest money from two companies that can quite literally not afford to pay them.

On top of that, their actual compute spend is going to be 100 billion or something this year.

That may sound like a lot, but already big tech has spent over a trillion dollars in CapEx. They're going to spend another trillion dollars next year.

Trillions more if they're allowed to have their way. I'm not sure how the mathematics works here.

Microsoft's AI revenue for fiscal year 2026 was $34.33 billion.

Of that, OpenAI represented $24.1 billion. Even with OpenAI's money, that's pretty pathetic considering they've spent $261 billion in CapEx since 2022.

I don't know what JP Morgan is talking about, but I will say stock values are not a representation of reality.

And as we've seen with the AI bubble, because the media and the markets have been giving credit to big tech for AI paying off,

despite the fact that AI really hasn't, and it definitely hasn't even put a dent in the amount of capex they're spending.

And so with the stock returns, obviously situational awareness, that hedge fund imploded. They had a huge bet on AI.

Do you think that that is a sign of the AI bubble or was it just a hedge fund taking on too big of a risk? I think we're going to have a few more of those.

I think we've got Bougalbara as well. We've got more to come, I think, with those.

Because what those are is, it's funny, on top of being very risky, it's attempted to apply a logical-ish thesis to an illogical market and an irrational market.

I mean, Leopold Aschenbrenner has, I think he's married to someone who works at Anthropoc. He clearly has an MPI.

also he has two and a half billion maybe five billion dollars worth of anthropic stock in situational awareness but his bet was that semiconductor stocks would just run infinitely

of course that wasn't going to happen nothing runs infinitely even things

that go up seemingly forever have down periods and

he was clearly over leveraged the problem is is that increasingly retail investors are over leveraged too so at some point these violent shakes that are going to start happening that are going to start washing people out.

I mean, they're not going to be hedge fund sized people, but retail investors are going to face the same kind of things because right now, the stock market, the US stock market, actually a lot of stock markets are effectively casinos.

They're not based on whether the companies are good. They're based on some

wobbly version of news combined with vibes, combined with reading the tea leaves of companies, especially in the case of the AR companies that really only ever lost money.

And going back to that hedge fund last year, it had about 200 percent in returns.

Do you think that we could continue to see the AI trade work, even if you think these businesses won't?

I think for the AI trade to grow much further, the businesses have to show stuff that they're not right now.

I think that these tech stocks connected to AI can keep running because the market is completely insane.

CoreWeave, their losses are increasing with their revenues. Their debt is increasing with their revenues. No one seems to care about that.

They don't care that Corweave has $40 billion worth of debt. What they care about is number go up.

The thing about that is that can run very far, right up until the vibe cracks, right up until the market stops having confidence in this irrational thesis.

Who knows what that could be? Maybe it could be OpenAI, not IPO-ing.

Maybe it could be Anthropics S1 being, I don't know, looking like the inside of a funeral home. Maybe it could be something else.

Maybe it could be a surprising event from one of the AI companies. Maybe Corweave not being able to raise debt.

Either way, we are in a vibes based casino market. And so anything could happen.

Things can keep going up no matter how bad things get. But at some point, reality will dawn and at some point the money will run out.

And you think that could be when Anthropic or OpenAI IPOs?

I think that Anthropics S1, I think they're going to try and do some dodgy stuff with capital expenditures.

i think they're going to try and put their training costs as capex

if they do that welcome back to 2001 welcome

back to the dot-com bubble i also think that if their