The AI revolution and our aging power grid are on a historic collision course, threatening to stall innovation and raise energy costs for everyone. Physicist and AI grid futurist Varun Sivaram reveals how we might flip this looming crisis into a once-in-a-generation opportunity — turning AI into a flexible user of power that actually strengthens the grid. Learn how this innovation could unlock new capacity for AI data centers while lowering costs and accelerating the energy future we’ve been waiting for.

How AI can solve its own energy crisis

The explosive growth of AI data centers threatens to overwhelm an aging power grid and drive up energy prices, with U.S. data center demand set to triple by 2030. Yet, a demonstration in Phoenix showed that AI servers can cut their power use at peak times by 25 percent while still meeting key performance needs. By making AI data centers flexible about when and where they draw power, vast amounts of unused grid capacity can be unlocked. This shift could lower costs, speed clean energy adoption, and turn AI from a risk into a vital grid ally.

AI Can Strengthen The Grid If We Make It Flexible

The explosive growth of AI data centers threatens to overwhelm an aging power grid and drive up energy prices, but a new approach flips the script. By making AI data centers flexible in when and where they use electricity, they can relieve grid stress and even help integrate more clean energy. In a Phoenix test, AI servers cut their power use by 25 percent during peak demand without sacrificing performance. If widely adopted, this strategy could unlock trillions in new AI investment while keeping energy reliable and affordable.

AI Can Rescue The Power Grid While Powering Its Own Boom

*Taming the Sun: Innovations to Harness Solar Energy and Power the Planet*

"How AI Factories Can Help Relieve Grid Stress"

Emerald AI

energy

electricity

technology

science

innovation

Countdown

TED Stage Talk

Varun Sivaram: How AI can solve its own energy crisis