The race to deploy artificial intelligence infrastructure is clashing head-on with the limitations of America’s aging power grid, raising urgent questions about the future of technological growth and energy policy. In a recent episode of the “Curious by Nature” podcast, Professor Robert Cox—an expert in electrical and computer engineering at the University of North Carolina at Charlotte—spotlighted an AI data center built in Memphis in just 120 days. Despite this rapid construction, Cox revealed that connecting the new facility to the regional power grid could take the local utility several years, an alarming lag that forced the center to generate its own electricity on site.
The November 20, 2026 podcast episode illuminated what Cox described as a growing national challenge: While demand for AI-driven data processing accelerates, the energy infrastructure needed to support that demand remains slow-moving. “The power grid,” he emphasized, “can’t move at the speed of AI.” The grid itself was designed for steady, predictable usage patterns. In contrast, new data centers impose vast and rapidly fluctuating electricity loads, straining both local and national systems. (Newswise)
Why the Grid Lags Behind
According to Cox’s discussion and supporting materials from the U.S. Department of Energy, the utility sector’s limited risk tolerance and highly regulated processes hamper its agility. Utilities typically need years to upgrade transmission infrastructure, obtain regulatory approvals, and coordinate regional grid changes. By contrast, technology companies with ample capital can construct entire data centers in mere months, hoping to support the explosive growth of AI tools and cloud infrastructure. (DOE Challenge)
Modernization Is Underway, but Slow
Efforts are underway at federal and research levels to address the bottleneck. The Department of Energy is collaborating with grid operators and testing centers to develop AI solutions for the grid itself, aiming for faster decision-making and a 10% or greater improvement in reliability and cost. The hope is to apply AI not just as a demand source, but as a management tool for the grid’s complexity. Still, Cox and other experts caution that without sweeping modernization, these rapid technological advances risk stalling under their own power needs.
The situation in Memphis is emblematic of a nationwide issue—and a preview of the sort of stopgap solutions that may become common unless grid upgrades can keep pace with technology’s new demands.