AI is growing fast. That growth needs a huge amount of electricity. Felix Prehn says this is becoming one of the biggest market shifts in the United States. While many people focus only on AI chips and software, a different story is building in the background. America’s power grid is under pressure, and that pressure is creating new interest in energy and infrastructure.
Felix Prehn, founder of Felix Prehn Goat Academy, points to a simple idea: follow where the money is already going. In this case, money is moving into power, grid repair, and energy supply.
The power grid is the network of wires, substations, transformers, and power plants that delivers electricity to homes, offices, and factories. Much of this system in the US is old. Many parts were built decades ago. That was fine when demand was lower. It is not fine now that AI data centers need round-the-clock power.
A data center is a building full of computers that store and process large amounts of digital information. AI systems use many of these computers at once, which means they use a great deal of electricity. Some estimates suggest that power use from AI data centers could more than double in the coming years.
That creates a major problem. The grid was not designed for this kind of nonstop demand. In simple terms, the system is being asked to do far more than it was built to handle.
This is why grid upgrades matter. When the government supports grid reliability, it can help keep older power plants running longer and speed up new energy deals. One example is a power purchase agreement. This is a long-term contract that says a buyer will purchase electricity from a power producer for many years. These agreements matter because they can create steady income for energy companies.
Another key term is backlog. A backlog is work a company has already won but has not finished yet. It is not a guess. It is signed business. When a company has a large backlog, it can show that real demand already exists.
Felix Prehn highlights several areas that may benefit from this shift.
First, there is off-grid power. This means power made directly at or near a data center instead of coming through the main grid. Large tech companies are looking at on-site energy because waiting years for a normal grid connection may slow down expansion.
Second, there are the firms that build the physical grid. These companies work on transmission lines, substations, and plant upgrades. A transformer is a piece of equipment that helps move electricity at the right voltage. It is one of the many basic parts needed to keep power flowing safely.
Third, there are component makers. These companies produce cooling systems, circuit boards, switchgear, and other key tools. Switchgear is equipment that controls, protects, and isolates electrical systems. It helps prevent overloads and failures.
Finally, raw materials also matter. Copper, aluminum, and uranium can all play a role in power expansion. Copper and aluminum are used in wires and transmission systems. Uranium is used as fuel for nuclear energy.
The main point in Felix Prehn’s view is clear. AI is not only a software story. It is also a power story. As electricity demand rises, the market may continue to reward companies tied to energy supply, grid repair, and industrial support.
For investors and readers trying to understand what comes next, this is the bigger lesson: when a major trend is real, the strongest clues often appear in contracts, backlogs, and spending before they appear in headlines.