Letter 166 asked which islands are ready today, and found the honest answer was almost none. This letter asks a different, longer-horizon question: given real, sustained political commitment and enough time — the actual decades-long window a serious infrastructure investor should be thinking in — which islands have the underlying geology to eventually earn their way in. One small Caribbean nation has been pursuing exactly this since 1969, has finally started building, and sits on an estimated resource fifty times larger than anything it could use itself.
Letter 166 concluded that almost no small island is ready today to host serious data-center infrastructure — grids too small, power too expensive, or resource potential still years from being tapped. That conclusion was about the present. A genuinely long-horizon investor — the kind actually willing to underwrite a 20-to-30-year infrastructure bet — is asking a different question: given real political commitment sustained over decades, and given the underlying geology, which of these places could still become the greatest opportunity, once enough time and enough will are actually applied.
Dominica is a genuinely remarkable case, and the remarkable part isn't the geology — it's the political timeline. The Commonwealth of Dominica has been formally pursuing geothermal energy since 1969, with United Nations-assisted studies from that era determining the island could plausibly power itself entirely from its own volcanic heat. The government created a dedicated Geothermal Development Corporation in 1974 — fifty-two years ago — specifically to pursue that ambition. It then took nearly four more decades to secure the funding for serious exploratory drilling, which finally confirmed commercial viability. That is not a company chasing a trend. That is a government that decided on a resource strategy before most of today's hyperscalers existed, and stayed with it across multiple changes of government, for over half a century.
The payoff is now visibly under construction. A 10MW geothermal plant near the village of Laudat, in the Roseau Valley, is being built by Ormat — an established, publicly-traded geothermal developer — under a signed 25-year power purchase agreement with Dominica's state electricity provider. Financing combines a $34.8 million Caribbean Development Bank loan, a $25.4 million Green Climate Fund commitment, and direct support from the governments of the US, UK, Japan, and New Zealand. Dominica's energy minister has stated the explicit goal of eliminating diesel generation entirely by 2030.
The scale of the underlying resource is the real story for anyone thinking in decades rather than quarters. Dominica's current peak electricity demand is only about 18 megawatts. Government and independent estimates place the island's total geothermal potential as high as 500 to 1,000 megawatts — fifty times current peak demand, on an island of just 75,000 people. That is not a resource sized for domestic consumption. It is a resource sized for export, and the government has said so explicitly: long-term plans include selling surplus power via undersea cable to neighboring Martinique and Guadeloupe, positioning Dominica as a genuine regional energy hub rather than merely a self-sufficient island.
Dominica decided what it wanted to become in 1974. It has taken fifty-two years, four decades of that spent simply finding the funding, to build the first ten megawatts of a resource estimated at up to a thousand. That is what "long-term" actually means when a government means it.
The honest complication is the same one that constrains the wider region: geothermal development requires scale to be economical, and scale requires demand large enough to justify it. Dominica's own population cannot use anywhere near the full resource, which is precisely why the undersea export plan matters — but that plan depends on cooperation with neighboring French territories, financing for cross-border cable infrastructure, and a regional market that doesn't fully exist yet. The Organisation of Eastern Caribbean States is pursuing similar geothermal exploration in Grenada, St Lucia, St Kitts and Nevis, and St Vincent — meaning Dominica's real long-term opportunity may be less "the one island that wins" and more "the anchor of a regional Eastern Caribbean geothermal grid," which is a genuinely different, larger, and slower thesis than a single data-center campus.
A comparable but more structurally limited case exists in Portugal's Azores. São Miguel Island already generates roughly 50% of its own electricity from geothermal — a real, proven, decades-old success, not a projection. But a direct academic assessment of the archipelago's geology found that submarine cable connections between the Azores' own islands are not feasible due to seafloor topography, meaning each island operates as a fully isolated grid. That caps any single Azorean island's project scale at whatever its own local demand can absorb, regardless of how much geothermal potential sits underneath it — a genuine structural ceiling Dominica's flatter regional-export ambition doesn't share in the same way.
Setting Paraguay's near-term, already-financed opportunity from Letter 166 alongside this longer bet produces a useful two-tier way to think about this category. Paraguay is investable now, on a hydro surplus created by a fifty-year-old treaty nobody originally built for this purpose. Dominica is a genuine, credible decade-plus bet on a resource whose scale would matter enormously if the regional export infrastructure gets built — a bet with fifty-seven years of demonstrated political continuity behind it, which is a real, unusual, and honestly rare form of evidence for this kind of long-horizon claim. Most islands in this research had either the resource or the political will, rarely both, sustained over decades. Dominica is the one case where both are genuinely present, verified against the actual historical record rather than a five-year strategic plan.
On a long enough time horizon, and specifically with the kind of sustained, multi-decade political commitment most jurisdictions never actually demonstrate, Dominica is a genuinely credible answer to this letter's reframed question — not because it's ready now, but because its fifty-seven-year track record of pursuing this exact resource is the closest thing to real evidence that a small island government's "long-term ambition" claim can actually be trusted. The dependency is real and worth stating plainly: the payoff scales with regional undersea export infrastructure that doesn't fully exist yet, not with Dominica's own tiny domestic demand. The Azores prove the geothermal technology works at real scale but are structurally capped by island-to-island grid isolation — a genuinely different, more limited case. The risk to this thesis: a fifty-year political commitment is unusual specifically because it is hard to sustain, and Dominica's export ambitions depend on cooperation with French territories and cross-border financing that remain unbuilt; this is a bet on continuity holding for another decade, not a bet already realized the way Paraguay's is.