AI Energy Demand to Drive Global Power Revolution

On June 25, 2025, Mike Schroepfer, former Chief Technology Officer at Meta and now founder of venture capital firm Gigascale Capital, made a bold prediction during a conversation on a podcast. Schroepfer asserted that the energy demands of Artificial Intelligence (AI) will be so significant that they will drive a major breakthrough in how power is generated and delivered globally. This prediction is not just about the need for more servers and data centers; it is about a substantial paradigm shift in global infrastructure.

Schroepfer's remarks highlight that as AI models become more complex and capable, the energy they require becomes impossible to ignore. This energy demand is expected to motivate the world to rethink its energy systems. The growth of AI is tied to energy innovation because the computational load required for AI applications, such as self-driving cars and predictive medicine, is growing exponentially. This means more electricity, more servers, and more power-intensive infrastructure.

Schroepfer believes that the race to scale AI will soon hit a wall due to energy bottlenecks, not compute limits. The faster AI innovates, the more pressure is placed on outdated grids and traditional sources of electricity. This is where AI energy innovation becomes a central requirement for the industry to thrive. The demand generated by AI will drive serious investment into energy technology, from advanced battery systems to new forms of low-cost renewable energy.

Schroepfer compares the current moment to the dot-com boom, when a technology surge triggered secondary revolutions in hardware, internet speeds, and digital services. Today, AI could force a similar secondary surge in how energy is produced and distributed, especially as regulatory pressure around carbon emissions intensifies. Schroepfer’s venture firm, Gigascale Capital, is already backing startups that are building for this future. The next great innovation might not be the next AI model, but the breakthrough that allows data centers to run on sustainable, efficient power without collapse.

This shift in funding attention from pure software to energy technology is already visible. Startups focused on heat reuse, hydrogen storage, and modular nuclear reactors are gaining traction, not just for their eco-friendliness, but because they solve a critical bottleneck in AI infrastructure. On a global level, energy-rich countries are likely to become the new global tech centers. Areas with abundant sun, wind, or geothermal energy will be significantly more appealing for the establishment of large AI training facilities. For countries that are clean energy leaders, they could see new demand for power exports or infrastructure investments specific to AI.

For Schroepfer, this intersection of AI and energy is more than a business opportunity; it’s a necessity. He stated, “We are going to run out of power before we run out of ideas,” hinting that solving the energy piece is the only way to unlock AI’s full potential. If Schroepfer is right, AI energy innovation could define the next two decades of progress, combining clean power solutions with the brainpower of next-gen AI systems. It’s a moment that calls for new thinking, bold investment, and tech that respects the planet while accelerating intelligence.