How Geothermal Energy Could Power the AI Boom

I-Pulse is highlighting a $250 million U.S. investment aimed at accelerating geothermal drilling technology that could provide around-the-clock clean power for AI infrastructure. Robert Friedland argues geothermal, rather than wind or solar, should anchor America's future energy system and is calling for government-backed industrial policy. The piece is constructive for geothermal and AI power-supply themes, but it is primarily a strategic commentary rather than a near-term market catalyst.

Analysis

The important second-order effect is not “geothermal vs. wind/solar,” but whether firm power becomes the new bottleneck for AI capex. If hyperscalers keep signing load growth faster than grid interconnects and gas turbine supply can scale, investors should expect a premium to accrue to technologies that can monetize baseload, not just electrons. That creates a potential re-rating path for the entire subsurface stack: drilling automation, high-temp materials, power electronics, and grid infrastructure that can actually absorb 24/7 generation.

The near-term winner is likely the enabling supply chain rather than the pure-play developers, because commercialization risk sits in project execution, not headline technology. If the drilling breakthrough is real, it could compress the “cost of discovery” curve for deep geothermal in the same way horizontal drilling changed shale economics; that would be bearish for incumbent utility-scale generation monopolies that rely on long-duration permitting moats. The clearest competitive loser is not wind or solar outright, but anything whose economics depend on intermittency being papered over by storage subsidies or curtailment assumptions.

The key risk is time. Even with policy support, this is a years-not-quarters story, and the market may overprice the optionality before it is bankable at scale. A failure mode is that demonstration wells work technically but not economically, or that permitting, water management, and seismicity concerns slow deployment enough that AI demand is temporarily met by gas peakers and grid upgrades instead.

Contrarian angle: consensus is likely underestimating industrial policy willingness here because AI has become a national competitiveness issue, not just a climate one. That said, the trade is likely best expressed as a basket of “picks and shovels” with multiple monetization paths, rather than a concentrated bet on early-stage geothermal names.

Key Decisions for Investors

• Long XLI vs. short ICLN for a 6-12 month horizon: if AI-driven baseload demand stays tight, industrial/electrification beneficiaries should outperform intermittent renewables by 300-500 bps as capital flows toward firm-power infrastructure.
• Build a basket long on drilling enablers and high-temperature industrial suppliers (e.g., SLB, HAL, CAT) over the next 3-6 months; risk/reward favors the equipment layer because it gets paid before geothermal becomes a utility-scale winner.
• Buy medium-dated calls on ET or KMI, 6-9 months out: if geothermal remains slow to scale, gas infrastructure remains the bridge fuel and benefits from data-center load growth; upside is capped but the policy optionality is attractive.
• Avoid chasing speculative geothermal developers on headline risk; wait for a bankable project pipeline or commercial offtake contracts before taking directional exposure, since execution disappointment can re-rate these names down 30-50% quickly.
• For higher-risk positioning, consider a small starter long in GEVO-style deep-tech energy infrastructure proxies only after a financing milestone or DOE-backed award; otherwise the probability-weighted IRR is dominated by dilution.