在内华达的新发现代表数十年来首个由业界主导找到的可商业化「隐伏型」地热系统。Zanskar 透过大型地质资料的 AI 解析，提高从极低命中率的「针堆寻针」环境中定位深部高温储层的成功率。美国目前地热仅占总供应的不到 1%，而 1970 年代之后因联邦转向其他能源技术投资，隐伏系统的发现几乎仅源于偶然。Zanskar 的技术形成对比：其模型反复在传统产业未曾钻探的区域识别热点，并在今年透过深钻获得足以发电的水温，构成方法有效性的首个实证。

早期研究者（如 Faulds）在 2000 年代至 2010 年代建立隐伏系统特征资料库，并于 2018 年以低于 20 世纪方法的成本定位可发电系统，但因地理保护区无法商转。Zanskar 延续并扩展此类资料导向策略，其成果对比当前产业焦点——需额外注水并可能产生低阶地震活动的 EGS——显示传统地热若能有效寻获资源，其工程复杂度与成本皆更低。

2008 年美国政府推估尚未发现的地热系统平均可提供 30 GW，但 Faulds 指出此值可能低估一个数量级以上，认为美国隐伏系统潜力可能达数十至数百 GW。若更深、更高温的钻井技术持续进展，可利用的储层规模将同步扩张。

A new Nevada discovery represents the first industry-driven identification of a commercially viable “blind” geothermal system in decades. Using AI to process large geological datasets, Zanskar increases the hit rate in a historically low-yield “needle-in-a-haystack” search space. Geothermal provides under 1% of US energy, and post-1970s federal shifts left most blind systems found only by accident. Zanskar’s model repeatedly flagged hotspots in previously unused areas, and deep-drilling this year confirmed temperatures sufficient for power production, marking the first empirical validation of its method.

Earlier researchers such as Faulds built system attribute databases in the 2000s–2010s and located a power-capable system in 2018 at lower cost than 20th-century methods, though commercial use was impossible due to land protections. Zanskar extends these data-driven approaches. Compared with current enthusiasm for EGS—which requires external water injection and can induce low-level seismicity—successfully locating natural systems yields lower engineering complexity and cost.

A 2008 US estimate placed undiscovered geothermal potential at a mean 30 GW, but Faulds argues this may be underestimated by more than an order of magnitude, suggesting tens to hundreds of gigawatts in blind systems. As deeper, higher-temperature drilling technologies advance, accessible reservoir capacity will expand correspondingly.