Meta is Telling Us What We Already Knew

Meta recently announced that they will offer cloud space with “excess AI compute capacity.” I, and many others, have suspected the AI spending would come to a head sooner or later with the expectation that the market will experience a shock wave since it’s been carried by the optimism of AI spending for a few years now. What’s happening now is perhaps the catalyst we’ve been looking for.

When Meta made the announcement, the shares jumped 9%. I expect Wall Street to react to this type of news from a Mag 7 player but I rarely believe the movement signifies well thought out investment decisions rather than people taking advantage of the short-term momentum.

The story is, we’ve watched big tech commit billions at unprecedented levels to fund the AI infrastructure buildout without a clear roadmap to what the return on investment will be and when it will materialize. This “pivot” from Meta is a sign that spending has been out of hand.

One thing you have to remember, in the tech world there are only a few ideas circulating with many copy cats. Pay attention. Every move you see a big tech firm make is one that others have made or will begin to mimic. They are smart people leading these firms, but they are figuring it out in real time. Many times, they are investing in an area they know they should, but without a definite target on how much spending is enough. When your competitors are spending billions, you’d better be as well.

There has always been questions about the return on investment with this AI buildout. We either received an ambiguous response when the executives of these firms were questioned or they painted a picture so rosy, common sense told us to be cautious.

Now we have Meta joining Uber, Microsoft, and Amazon, admitting to overspending, only Meta has a slick way of controlling the narrative. Let’s should analyze what’s going on but remain neutral.

Until now, Apple and Meta were the only hyperscalers who didn’t provide cloud servers to external customers. Meta is now joining, IBM, Oracle, Microsoft, Amazon and Google as cloud server providers. This is all find and dandy but I want you to focus on the key word - excess. What we don’t know is how much excess Meta is admitting to. Meta has spent approximately $111 billion since 2025 with at least $125 billion committed for 2026 alone.

The other hyperscalers should be in good shape even though they are all spending significantly more than Meta. For one, they are spending with the goal of offering AI cloud services along utilizing AI to enhance they’re existing and future in-house product development. Also, Amazon is building a robotics empire within their warehouse and logistics operations. The spend makes sense for Amazon Web Services and their robotics ambitions.

I believe most of Meta’s committed future spend will be considered in the pile of excess spending. My assumption is based on the perceived timing of the benefit of the pivot. When I read the numerous articles discussing this issue, it’s implied that selling compute capacity will result in a ROI that will hit sooner than Meta using all of the cloud space for themselves. Implying a quicker return makes me think Meta had to justify the rest of the spending.

One of the main jobs of a tech CEO is to predict a positive future. You have to remember that every time they speak. Communicating a positive narrative is imperative to the success of the firm and their tenure as the CEO. Remembering this helps you decipher their messages to find the truth.

Is Meta pivoting to keep control of a positive narrative or is this truly a positive change in strategy? I think it’s a little of both. Zuckerberg said this pivot is on the table if they “overbuilt”. With $125B+ committed for this year alone, will Meta scale back any future commitments? Will they be able to utilize all the AI infrastructure through this internal and external use split?

Meta’s announcement is a master class on producing more questions wile keeping the narrative on the positive side.