We don't have enough power.
AI chip-driven electricity demand is projected to surge to a record ~315 gigawatts globally by 2033.
That would represent an increase of more than +1,100% since 2025.
The US is expected to account for ~64% of this new AI electricity demand, or ~200 GW.
Meanwhile, AI data centers are also putting huge strain on power infrastructure.
During model training, hundreds of thousands of GPUs can power up and down simultaneously, causing electricity usage to spike as much as 50% above design capacity.
These sharp power swings can accelerate wear on batteries, generators, cooling systems, and other critical infrastructure, increasing maintenance and replacement costs.
We need new energy infrastructure.
Texas, Pennsylvania and now Ohio.
This is a powder keg and has the potential to unwind 200-300 basis points of annual GDP if it metastasizes.
How did it get this bad?
The collective leadership of frontier AI has failed miserably in doing the basics:
1. Paint a positive picture of what is possible with AI. Instead of spilling the ink on the glass half full, stories abound of pearl clutching, catastrophizing or arcane technical jargon that, together, breed mistrust, fear and skepticism.
2. Visibly and disproportionately enrich the local communities AI data centers operate in. Tell these stories with a megaphone and have local leaders lead with the financial proof so it’s irrefutable.
3. Stop infighting and stop trying to influence the government or trick them because of the asymmetric knowledge the industry has relative to outsiders and observers. This further makes it seem like a shady money grab to many bystanders.
Data centers have unfortunately become THE symbolic representation of the asymmetric upside for a very narrow tech elite and a class that are untrustworthy.
Inasmuch, the safer bet may be that the pushback and resistance grows…
People still don’t get that $PLTR is basically an operating system for physical AI and robotics. I’ve seen zero mentions of this on 𝕏, so let this be number one.
The robot itself is only one layer of physical AI. It can see, move, and follow instructions. But it still needs to understand the world it operates in.
That’s where Palantir’s Ontology comes in.
The Ontology is basically a live digital map of a business. It tells the AI what everything is, how things are connected, what state they’re in, and what actions are allowed.
Take a factory robot.
The robot may be able to identify a broken machine. But the Ontology can tell it:
This is Machine 27. It makes this component. These orders depend on it. This replacement part is in inventory. This technician is available. Production can be moved to this other line. And these are the actions the AI is allowed to take.
Now the robot isn’t just an intelligent machine. It’s a machine that understands its role in the business, like a well-trained human worker.
As physical AI spreads through factories, warehouses, hospitals, defense, and infrastructure, intelligent machines will increasingly need this kind of connection to the real world.
That’s why I think Palantir will become the OS for physical AI, with an Ontology that gets richer, more precise and more useful as more of the physical world gets connected to it.
(Not investment advice.)
$IREN
$40M + per MW on the newer short-term deals is a very good signal. Nvidia’s 15%+ server price hike is simply getting passed through and customers are still lining up with 45% prepayments. When demand for energized compute is this strong, higher input costs don’t compress margins… they just reprice everything altogether at the end of the day Power is a real bottleneck and customers will pay premium prices for it.