Big tech’s newest appetite: electricity
At a G20 tech meeting, Mark Zuckerberg and Elon Musk basically delivered the same message: AI is hungry, and the buffet needs to get bigger. They argued for more data centers and the massive power supply needed to keep them humming.
That matters because the AI story has quietly evolved from "who has the best model" to "who can actually power the thing without tripping a breaker." If you’re an investor, that shift points to a very practical theme: the next phase of AI spending may be as much about grids, turbines, transformers, and data-center buildouts as it is about semiconductors.
The bottleneck nobody can ignore
The more AI workloads explode, the more the boring parts of the economy start looking exciting. Think:
- utilities that need to expand generation and transmission
- data-center landlords and developers racing to build capacity
- power-equipment companies supplying the gear that keeps everything online
That’s the weird magic trick of tech cycles: the flashiest trend often creates demand for the least flashy stuff.
Why investors should care
If the AI buildout keeps accelerating, the winners may not just be the companies making the models. The winners could also be the ones laying the pipes, stringing the wires, and keeping the lights on.
Big picture: AI may be “software eating the world,” but first it has to eat a lot of electricity.