The AI power crisis has an obvious answer. The market just hasn’t priced it yet.

Every week brings another headline about data centers and the grid. A single large AI campus can now draw as much power as a mid-sized city, and the hyperscalers have made a promise they are quietly struggling to keep: that all of it will be carbon-free, around the clock.

https://www.cindrigo.com/geothermal-in-germany

That last part is where the story gets interesting, because most of what we call clean energy cannot actually deliver it.

Solar stops at night. Wind stops when the air goes still. Both are cheap and both are essential, but neither gives a data center operator what they actually need, which is firm power at three in the morning in February. The industry’s answer so far has been batteries, gas with offsets, and increasingly small modular nuclear reactors that remain years from commercial reality.

There is an older answer sitting underneath all of this. Literally underneath.

Geothermal produces constant baseload power from the heat of the earth. It runs day and night, it is unaffected by weather, and once a well is drilled the fuel cost is effectively zero for decades. Its output profile is almost perfectly matched to what a data center consumes: flat, continuous, predictable. For a buyer who has committed to 24/7 carbon-free energy, that match is not a nice-to-have. It is the entire game.

So why does geothermal still attract a fraction of the capital flowing to solar and wind, even as a few capable operators quietly build real positions?

One reason, and it is an honest one. The cost is front-loaded. You spend heavily to drill before you generate a single megawatt, and early wells carry real subsurface risk. Capital that is comfortable funding a solar farm, where the cost curve is gentle and the technology is boringly proven, gets nervous at the drilling phase. This is the genuine constraint, and pretending otherwise is how amateurs lose serious money in this sector.

But two things have changed, and they are why I am spending my time here.

First, the resource is no longer confined to volcanic geographies. A decade of horizontal drilling and reservoir engineering imported from the shale industry has begun to unlock geothermal in places that were previously uneconomic. Central Europe, with its deep sedimentary basins, is one of the most promising.

Second, and more important for anyone deploying capital, governments have started absorbing the part of the risk that scared private money away. Germany now offers state co-funding that can cover half of geothermal pre-development costs and a substantial share of construction. That single policy change rewrites the risk calculus. It moves the hardest, most uncertain dollars off the private balance sheet and onto the public one, precisely at the stage where private capital was unwilling to go.

When the state de-risks the drilling and a hyperscaler is contractually desperate for exactly the output profile you produce, you no longer have a science project. You have an asset with a public subsidy on one side and a price-insensitive buyer on the other.

That is a rare shape for an energy deal, and it is why I think geothermal deserves a serious second look from anyone thinking carefully about where European energy goes next.

The locations that combine a real resource, a supportive state, and proximity to demand are not infinite, and the most capable operators are securing them now. The opportunity is not that geothermal is new. It is that the conditions around it have quietly changed while most of the market was looking elsewhere.

If you are thinking about European energy and where the durable opportunities sit, this is a conversation I find worth having.

https://www.linkedin.com/in/mollen

R.J. Mollen is Founder and Managing Director of Gleam Capital Partners. The views above are his own perspective on the European energy market and are offered as general commentary, not as investment advice, a recommendation, or an offer of any security.