Fueling the AI Evolution Across America

The United States is witnessing a significant shift in its energy landscape, with the construction of dozens, and soon potentially hundreds, of AI-centric data centers. Each of these data centers requires substantial amounts of electricity, with power needs ranging from 200 MW to 1000 MW. This surge in demand is driving growth in regional and national electricity demands at a rate not seen in a half-century.

According to the Federal Energy Regulatory Commission (FERC), the need for additional generating capacity to meet the demand from data centers is estimated to be between 50 GW and 130 GW. The market, however, seems to be overweighting the pursuit of natural gas to meet the energy needs of data centers.

The advance of knowledge, including the development of artificial intelligence (AI), has led to the creation of new products and services, which inevitably increase energy consumption. Companies such as Anthropic, OpenAI, Google, and other tech giants are investing in national energy plans to ensure power for AI and data center growth. A single data center hosts hundreds or thousands of refrigerator-sized AI server racks, each weighing as much as a car and using as much energy annually as 100 cars.

The issue of powering the AI and data-center revolution is a concern that has attended the emergence of all radically new technologies over history. It's no longer about the exhaustibility of resources but about the need to weigh choices and tradeoffs, while minimizing government friction. Some companies' plans include accelerating geothermal, natural gas, and nuclear permitting to support energy and data center buildouts.

The telecommunication network supporting these data centers is itself power hungry, made up of hundreds of thousands of miles of physical cables and wireless "roads". Caterpillar announced a partnership with a Utah developer to supply some 4 GW of diesel engine generation for data centers, while the three major vendors for utility-scale gas-fired turbines are sold out through 2030.

In contrast to the U.S., the main actors actively involved in planning and financing renewable energy facilities to supply AI and data centers in Germany include the federal government, which mandates data center operators to cover 100% of their electricity consumption with renewable energies from 2027, EU institutions supporting clean technology development and decarbonization programs, and energy infrastructure operators facilitating grid connection capacities and digitalization.

The global race to "dominate" AI is likened to a modern "Sputnik moment" in terms of its potential impact. However, the debate around Big Tech and Big Energy is becoming embroiled in naked politicking. The FERC has more than tripled its forecast for growth in U.S. power demand by 2030, suggesting a detailed national survey found a total of some 55 GW of new data-center projects already committed or under construction.

Energy-constrained "digital sobriety" would merely choke expansion, as these hundreds of mega-scale data centers under construction in the U.S. are joining thousands of existing smaller data centers. The race for AI dominance is not just a technological challenge but also an energy one, with the stakes high for those who can navigate this complex landscape.