AI Mining Software: The Picks and Shovels of Discovery
AI mining software is the quiet protagonist of the resource boom story, and it is the technology behind the most striking mining narrative of the past two years. In Whitney Tilson’s Mount Rushmore presentation for Stansberry Research, the mechanism is described precisely: “Geologists at this firm got access to 145 years of historic records for the area, old maps, drill results, surveys, that kind of thing. It had just been collecting dust in archives and storage lockers for decades. It was way too much information for humans to sift through… but not AI. The firm put all this historic data into AI-powered mining software… And it uncovered 114 tons of gold.” For the full backstory, see our Pentagon MinMod: The AI Mapping America’s Minerals article.
That single paragraph is the whole thesis in miniature, and it explains why “AI in mineral exploration” has become a keyword investors are searching. Here is how the technology works and where it is showing up in returns.
How the software finds what humans missed
Mineral exploration has always run on data that already existed somewhere: state geological surveys, century-old drilling logs, claim filings, university theses, satellite imagery. The problem was never the absence of data; it was that the data was unindexed, unstructured, and physically scattered. Machine reading changes that. An AI system can ingest the full archive of a mining district, correlate old drill results with modern geophysical surveys, and flag anomalies that indicate where ore bodies sit. The Pentagon’s MinMod program did it nationally, “mapped out 670,000 mineral sites across America… Including 184,000 gold sites.” Dakota Gold did it locally, and found “114 tons of the highest-grade gold,” “roughly a $22 billion payload,” in the Black Hills.
The economics explain the stock moves. Exploration is the mining industry’s biggest money incinerator, billions spent drilling holes, most finding nothing. Software that improves targeting from the industry’s historical single-digit hit rate toward Earth AI’s reported “150 times the industry average” does not just find more mines; it collapses the cost per discovery, which reprices every acre of explored ground.
Where the returns are showing up
The pattern is already visible across the sector, and Tilson’s presentation assembles the evidence wall. The Randall Atkins story is the flagship: “A man named Randall Atkins bought a 100-year-old coal mine” in Wyoming, DOE scientists arrived with “a new kind of AI that predicts where ‘unconventional’ deposits of natural resources are,” and “he found 1.7 million tonnes of rare earths, worth about $37 billion… Enough of these essential resources to supply the entire U.S. for more than 100 years.” Ramaco Resources made “773% in a little more than six months.”
The private-sector layer: KoBold Metals, the AI exploration company backed by “Jeff Bezos, Bill Gates, Mark Zuckerberg, Sam Altman, Marc Andreessen, Michael Bloomberg, and Jack Ma,” is “now valued at nearly $3 billion,” which tells you what the best-informed technology capital thinks the discovery edge is worth. The listed-sector results: Ero Copper, a Terra AI partner, “ran up 318% in less than a year”; Intrepid Metals, using TerraEye on Arizona copper, “went up 277% in six months”; K2 Gold “soared 711% in a little more than a year” on Death Valley exploration; Equinor “nearly doubled in a little over three months” on Arkansas lithium.
The demand pull
The software story connects to two demand curves. The first is national security: “We are 100% reliant on imports for 16 of our most important commodities,” a gap the White House is closing with “12 executive orders” and “Project Vault,” “a $12 billion mineral stockpile.” The second is AI itself: data centers and electrification need copper, rare earths, lithium, and the energy to process them, which is why one of Tilson’s four AI Miners picks is “a West Texas AI driller that just penned a deal with a massive tech firm to supply power to one of its data centers for the next two decades.” The market for the technology itself is growing fast: Tilson cites the AI mining market as “set to soar 23 TIMES bigger in the next seven years.”
Considerations
The honest caveats: “150 times the industry average” is a private company’s reported figure, not an audited statistic; the stock runs quoted are maximum historical gains, with the presentation’s own hedge, “Past performance is no guarantee of future results”; and the Dakoka Gold case study is pre-revenue, with production not expected until 2029. The durable part of the thesis is the mechanism itself, archival data plus machine reading, which is real, deploying, and already repricing the sector.
Where to learn more
For surrounding coverage, see critical minerals stocks, rare earth mining stocks, and gold mining companies. Tilson’s full presentation, including the AI Miners report, is at the Stansberry order page with the 30-day refund guarantee.
NewsletterVetter is an independent publication. We receive compensation from some of the services we review through affiliate links. Nothing on this site is investment advice. Always do your own research.