AI Stocks: A Dubious Doubling?

So, everyone’s talking about these AI stocks, CoreWeave and Nebius. Apparently, they’re up… something. Twenty-five percent for CoreWeave, seventeen for Nebius. Fine. Numbers. What gets me is the expectation. The breathless pronouncements that they could “double” by the end of the year. Double! Like it’s some kind of right. It’s a stock, not a guaranteed outcome. And honestly, the whole premise is… irritating. It’s like they’re trying to create hype. Like subtlety is dead.

Look, they’re building data centers. Good for them. Everyone and their mother is building data centers. It’s the new thing. But the idea that this justifies a doubling of the stock price… it just feels… aggressive. And don’t even get me started on the phrasing. “Could double.” It’s so wishy-washy. Either it will, or it won’t. Why can’t anyone just be direct? It’s the ambiguity that gets me.

The Infrastructure, It’s Just… a Lot

These companies, CoreWeave and Nebius, they’re renting out computing power. It’s a simple concept, really. Buy a bunch of fancy chips, let people use them, charge them money. It’s not rocket science. But the whole thing is predicated on the assumption that there’s an infinite demand for this computing power. Which, let’s be honest, is probably not true. There’s always a limit. And when that limit is reached… well, then what? Then you’re left with a lot of expensive chips and a lot of disappointed investors.

Nebius, they’re claiming they’ll go from $1.25 billion in annual revenue to seven to nine billion. Seven to nine! That’s a pretty wide range, isn’t it? Are they just throwing numbers at the wall to see what sticks? And they’re doing this by adding… sites. More data centers. It’s just… relentless. And the fact that they had only two sites running at the end of 2024? Two! That’s practically a hobby.

CoreWeave is no better. A 134% revenue increase? Fine. But they have a $55.6 billion backlog. A backlog. It’s like they’re deliberately creating a problem for themselves. And 40% of that backlog is supposed to materialize in the next 24 months? That’s… ambitious. It’s like they’re setting themselves up for failure.

So, yes, they could double. Theoretically. But here’s the thing: they’re losing money. Burning cash. It’s like they’re deliberately trying to annoy me. I mean, you’re building a business, shouldn’t you, you know, make money? It’s a novel concept, I realize. But seriously, the lack of profitability is… unsettling. It’s like they’re operating on hope and hype.

CoreWeave is “not far from breaking even.” That’s what they say. It’s always “not far.” It’s like they’re perpetually on the verge of something. And Nebius? They’re expanding so rapidly, they don’t even bother with profit margins. They just… grow. It’s like a weed. A very expensive weed. And the market is supposed to be okay with this? Really?

Look, I’m not saying these companies are bad investments. I’m just saying… the whole thing feels… precarious. It’s a lot of expectation built on a foundation of… well, not much. And honestly, I’m starting to suspect it’s all a giant scheme to make someone rich while everyone else loses their shirts. It wouldn’t be the first time. It’s just… irritating. They could both see strong demand, sure. But double? I’ll believe it when I see it. And even if they don’t, they’ll probably still outperform the market. Which, frankly, is a pretty low bar these days.

2026-02-23 09:32

The Infrastructure, It’s Just… a Lot

Read More