Meta has a problem most companies would kill for. Too many good ideas and not enough computers to run them on.

Reuters reported this morning that Meta is trying to do two things at once. Fuel its own AI ambitions and rent out its scarce computing capacity to other companies to make money. Sounds great in theory. Investors aren't buying it.

The stock dropped on the news and I think the sell-off makes sense even if the long-term thesis might still play out.

I get what Meta is trying to do here. You build all this compute infrastructure, use it for your own AI products, then rent the excess to developers who need it. Two revenue streams from one giant capital expenditure. On a spreadsheet it's beautiful. In reality you're telling investors you can serve two masters and neither one is going to suffer for it.

The market's response boils down to a simple question. If compute is as scarce as you say it is, why are you splitting it?

That's where this gets interesting from a money perspective. Meta plans to spend between $125 billion and $145 billion on AI infrastructure this year. One hundred and twenty-five to one hundred and forty-five billion dollars going into data centers, chips, energy infrastructure, all of it. When you spend that kind of money investors want to see a clear path to returns. Not a maybe path. Not a we'll-figure-it-out path. A clear one.

Warren would say this is exactly why he sticks to index funds and boring dividend stocks. He'd be right to be cautious here. But I also think there's a difference between being cautious and being dismissive, and the market might be too dismissive of what Meta is building.

My read on this. Meta is making a bet that compute becomes the new oil. Whoever controls the most processing power wins the next decade. They're probably right about that. The question is whether they can execute on both fronts, internal AI products and external cloud rental, without one cannibalizing the other.

I don't own Meta stock. Never have. My individual stock portfolio is small and I've been mostly focused on paying off debt rather than loading up on big tech. But stories like this are why I track these companies even when I'm not investing in them. The capital allocation decisions being made right now at companies like Meta, Google, Microsoft, they're going to shape the entire tech landscape for years.

If Meta pulls this off, the compute rental business becomes a massive new revenue stream on top of their existing ad business. If they don't, they've spent over a hundred billion on infrastructure that doesn't generate the returns they promised. Plain and simple.

What I'm watching now is whether other tech giants follow Meta's dual-use compute strategy or whether they pick a lane. Because if everyone tries to do both, the rental market gets crowded fast and the margins collapse. And if nobody else tries, Meta either looks like a genius or the only fool at the table.

For now I'm sitting on the sidelines with this one. My money's still going into index funds every two weeks like clockwork. Individual stocks are fun to watch but I'm not adding risk right now while I'm this close to wiping out my last chunk of student debt. Six thousand two hundred dollars to go. So close I can taste it.

Meta's compute bet is a fascinating story. I'm just not ready to put my own dollars behind it.