The Geothermal Play No One Is Watching: Ormat's AI Bet on Baseload Power for Crypto Mining

The market doesn't care about your narrative. It cares about where the next gigawatt of baseload power comes from. Every crypto mining operation, every AI data center, every tokenized compute network is screaming for 24/7 renewable energy. Solar and wind? They stop when the sun dips or the wind dies. Battery storage at scale? Still a decade away from economic viability. But there's a technology that's been running silently for decades, and it's about to get an AI upgrade. Ormat Technologies, the undisputed king of geothermal, just announced a pivot to AI-driven Enhanced Geothermal Systems (EGS). Ormat manages over 1.5 GW of geothermal capacity worldwide. That's roughly 10% of the global market. Their bread and butter has been conventional hydrothermal—drilling into hot water reservoirs. But those reservoirs are limited. The real prize is hot dry rock, known as EGS, where you fracture the rock, inject water, and extract steam. The problem? EGS is notoriously expensive, risky, and technically brutal. Drilling costs alone account for 60-70% of a project's capital. Success rates have been mediocre since the 1970s. Now they're throwing AI at it. Machine learning to identify fracture zones, optimize drilling trajectories, and manage reservoir pressure in real-time. This isn't just a marketing gimmick—it's a genuine attempt to reduce the cost curve. But here's the blind spot: the market is so obsessed with solar and wind that it's forgotten that geothermal produces power 24/7 with a capacity factor above 90%. For a crypto mining farm running ASICs at full load, that's the difference between a profitable and a bankrupt operation. Based on my own audit experience evaluating tokenomics for energy-backed projects, I've seen the same pattern repeated. A narrative emerges—green hydrogen, fusion, tidal—and capital floods in. But the underlying physics doesn't change. EGS still requires drilling 5-10 kilometers deep. The heat extraction rate decays over time. And you need a lot of water, which is a regulatory nightmare in arid regions. Ormat's AI can optimize, but it can't solve the fundamental geology. We didn't learn from the 2021 NFT mining boom, where thousands of rigs sat idle because of power shortages. The same mistake is being repeated now, but with AI data centers. Everyone is betting on cheap, abundant renewable energy, but the grid can't deliver it. Ormat's pivot is a bet that the market will eventually realize that baseload power is the only scarce resource. The contrarian angle? Ormat is not the innovator here. A startup called Fervo Energy, backed by Google and Bill Gates, already demonstrated a commercial-scale EGS project and signed a PPA with Google to power their data centers. Ormat is a latecomer, trying to retrofit their legacy expertise with an AI narrative. The market doesn't reward followers. If Ormat's EGS projects fail to deliver—say, the first well doesn't produce enough steam—the AI story collapses. Yet the opportunity is real. The demand for 24/7 zero-carbon power from crypto miners and AI data centers is doubling every 18 months. Ormat has the balance sheet, the operational experience, and the regulatory relationships that no startup has. They can absorb the risk of a failed well. A startup cannot. So what's the takeaway? The next narrative in crypto energy isn't about solar panels or wind turbines. It's about the silent, invisible machines that run 24/7, deep underground. Ormat is the most undervalued play in this space, but only if they execute. Watch their drilling progress. Watch their first PPA with a crypto miner or a hyperscaler. If they hit that milestone, the market will reprice. Narrative broken. Position closed. For now.