Data centers are the physical infrastructure powering the digital economy. Every AI query, cloud application, streaming service, online payment, medical record, and software platform depends on servers housed in secure facilities that operate around the clock.

The United States is now experiencing a major expansion in data center capacity. Existing facilities continue to grow while thousands of megawatts of additional capacity are under construction or in development. Recent research suggests total planned capacity could rise dramatically from today’s operating level.

The physical infrastructure itself, however, is only part of the story. The true economic value lies inside the walls: semiconductors, networking equipment, storage systems, software, artificial intelligence models, and research and development. In other words, the data center is the container. The technology inside is the engine.

Data centers create ripple effects that go well beyond the technology sector. They increase demand for electric utilities, power generation, transmission lines, transformers, cooling systems, engineering firms, construction labor, and industrial equipment. In several states, data center investment is already becoming a meaningful source of local economic activity and tax revenue.

Of course, growth at this scale inevitably raises environmental questions. The rapid expansion of data centers is creating legitimate concerns about electricity demand, water usage, and carbon emissions. In the near term, AI’s growing appetite for computing power will place additional strain on power grids and require significant investments in new generation and transmission infrastructure.

History suggests, however, that technology rarely stands still. As processors become more efficient, cooling systems improve, and AI models require fewer computational resources, the amount of computing delivered per unit of energy should continue to increase. While total electricity demand may continue to rise as AI adoption expands, the environmental footprint per computation is likely to decline over time.

For investors, this transition creates opportunities not only in technology, but also in utilities, electrical equipment, grid modernization, power generation, cooling technology, and energy efficiency.

At the same time, every major technology boom produces winners and losers. During the automobile revolution, hundreds of manufacturers competed before only a few survived. The same was true during the internet boom. The internet changed the world, but many early companies disappeared.

Artificial intelligence is likely to follow a similar pattern. As more companies build AI models and data centers, competition should increase, prices may decline, and weaker competitors may fall behind. Recent signs of slowing pricing power suggest that this process may already be beginning.

Our view is that the data center boom is real, but it should still be approached with discipline. The opportunity is broader than a handful of popular technology stocks. It includes semiconductors, cloud computing, power infrastructure, industrial equipment, and utilities.

Technological revolutions create extraordinary opportunities, but they also create excesses. We believe the right approach is to focus on businesses with durable competitive advantages, strong balance sheets, sustainable earnings growth, and reasonable valuations.

While the data center boom may look like a construction cycle, it is in fact something much larger. It is the physical foundation of the next generation of computing, one that is likely to influence economic growth, capital spending, energy demand, and investment opportunities for years to come.