Pat Gelsinger, the former Intel chief executive who led the chipmaker from 2021 to 2024 and now serves as a general partner at venture firm Playground Global, has warned that the United States lacks the electricity needed to power its ambitions in advanced computing. Appearing on a podcast from a16z, the venture firm Andreessen Horowitz, on October 9, 2026, Gelsinger argued that US energy capacity has been effectively flat for 15 years—a stagnation he views as a hard ceiling on economic growth in an AI-driven economy. His remarks arrive as companies across the technology sector race to build AI infrastructure.

America, in his framing, wants to build the future of computing but may not have the electricity to plug it in. His sharpest line cut straight to the dilemma facing every company in that race: "Why build a new data center and buy the million GPUs if I can't power them?"

Fifteen Years of Running in Place

Gelsinger's core point is about net growth rather than gross additions. The United States has been adding renewable energy capacity, but it has also been retiring coal plants. According to Gelsinger, those two trends have nearly cancelled each other out—meaning the country's total generating capacity has barely moved in a decade and a half.

There has been some recent movement. Gelsinger acknowledged that capacity growth has picked up to about 4% annually, but he said that pace falls well short of what AI and data center operators will need—at that rate, doubling the country's generating capacity would take close to two decades.

The timelines for adding new power make up the other half of the problem. Gelsinger said gas turbine deliveries now take approximately eight years—an interval that would push equipment ordered today into the mid-2030s. He also noted that the last new US nuclear reactor came online around two decades ago.

Defaults on the Horizon

Gelsinger did not stop at diagnosing the bottleneck. He said the industry should expect more defaults on data center projects as power needs go unmet on schedule.

He is not alone in seeing a gap between what AI requires and what exists. Executives at companies including Cerebras and Nebius have made similar observations, saying demand for AI compute is running far ahead of available supply. Remarks from Oracle have also been read as a signal of potential underwriting caution among operators, with the central concern being commitments to new chip purchases made before energy has been secured.

Gelsinger has been involved with energy-focused ventures including Alva Energy and PowerLattice, which work on efficiency and new capacity. That background gives his warning a practitioner's perspective, though it is worth keeping in mind that he has backed companies positioned to address the very gap he describes.

What This Means for the AI Buildout

The immediate pressure falls on hyperscalers and chipmakers that have made large capital expenditure commitments. Those plans assume the power will be there when the hardware arrives. Gelsinger's comments have raised concerns in the tech community about whether those commitments remain viable if generation and transmission cannot keep pace with project timelines.

Recent policy discussions around permitting reform and grid modernization are closely tied to these capacity concerns. Faster approvals for generation and transmission projects would directly affect how quickly new power can reach data center sites.

The indicators worth watching are concrete: whether permitting reform actually advances, whether turbine lead times shorten, and whether more operators follow Oracle in tying chip purchases to secured power. Any reported defaults on data center projects would be the clearest test of Gelsinger's prediction.