Nevada’s Hidden Heat Found by AI

AI uncovers a blind geothermal system in Nevada and hints at wider potential for U.S. geothermal growth

10 Dec 2025

A new discovery in Nevada has given the geothermal industry something to warm to. Zanskar, a clean energy start-up, says its artificial intelligence has identified a “blind” geothermal system, one with no visible surface clues. Further drilling and reservoir tests will be needed to confirm its size, but the find is the first of its kind in more than 30 years.

For decades developers have relied on surface hints such as hot springs or steaming ground to find geothermal resources. That approach missed many hidden systems. Zanskar’s algorithm, trained on subtle geological and geophysical patterns, pinpointed the Big Blind site, showing how data may uncover what human eyes could not.

The implications go beyond one discovery. Geothermal power offers round-the-clock electricity with no emissions, but exploration has been risky and costly. Many test wells come up dry. AI could lower that risk by improving predictions of where productive heat lies, cutting project delays and costs. This would make geothermal a more appealing option as the country seeks steady clean power for energy-hungry data centres and car factories.

Caution remains warranted. The Big Blind find, while impressive, is only a single case. Success will depend on the quality of geological data available and on regulatory approvals that often lag behind innovation. AI cannot conjure heat where there is none.

Yet the discovery points to a promising shift. As developers adopt digital tools, geothermal exploration may grow less speculative and more systematic. Zanskar’s experiment suggests that America’s hidden heat could be mapped not by intuition, but by machine.