
AI isn’t just a software story anymore
President Donald Trump took a victory lap for the AI trade on Friday, saying the technology could end up “bigger than oil” and warning that the U.S. can’t afford to let China win the race. Translation: the AI boom isn’t just about chatbots and semiconductors — it’s about who gets to build the massive physical machine underneath it all.
Follow the money, then follow the electricity
The article’s real thesis is simple: hyperscalers are still pouring hundreds of billions into AI infrastructure, and that spending doesn’t stop at chips. You need data centers to house the models, power lines to feed them, construction crews to build them, and enough electricity to keep the whole circus from short-circuiting.
That’s why ETFs like SMH, DTCR, PAVE, and URA are suddenly part of the same conversation as Nvidia:
- SMH is the cleanest silicon bet, with names like Nvidia, TSMC, Broadcom, and AMD.
- DTCR captures the landlords of the AI era — Equinix, Digital Realty, American Tower, and Crown Castle.
- PAVE leans into the grid-and-construction spillover with companies like Quanta Services and Eaton.
- URA is the “wait, AI needs nuclear power too?” trade, with Cameco and Oklo in the mix.
The market already gets the joke
These funds aren’t exactly hiding the punchline. SMH has ripped, DTCR has surged, and even the more “boring” infrastructure names are getting dragged into the AI vortex. Investors have gone from asking, “Who makes the chips?” to “Who keeps the data center from blowing a fuse?”
Big picture
Trump’s comments don’t change the economics of AI by themselves, but they do reinforce a market theme that’s already in motion: the winners of the AI race may not be just the model builders. They may also be the companies supplying the metal, power, real estate, and uranium that make the whole thing possible.
