Three Suns

On fusion, paint, and the field already in bloom — what is truly beckoning on the energy horizon, and what is already on the shelf.

Every age watches its own horizon. Ours holds two bright objects: a sun sealed in a magnetic bottle, and a sun you could roll onto a wall with a brush. Both are worth watching. But watching well is a discipline — the discipline of asking not only what is promised but who is doing the promising, in what venue, and under whose scrutiny. That question, it turns out, sorts the energy horizon more cleanly than any efficiency chart. And it leads somewhere unexpected: to a third sun, the ordinary one, already harvested at a scale our institutions have not caught up with.

I.The Sun in a Bottle

In June 2026, the Journal of Plasma Physics published something genuinely unusual: five peer-reviewed papers, plus an editorial, laying out the physics basis for ARC — the fusion power plant that Commonwealth Fusion Systems intends to build at a real address, in Chesterfield County, Virginia, in the early 2030s. The machine on paper is a high-field tokamak designed to produce roughly 1.1 gigawatts of fusion power and deliver at least 400 megawatts of net electricity to the grid. A sun in a bottle, with a county zip code.

What deserves our attention is not the rendering of the plant. It is the venue. A private company took the intellectual case for its flagship commercial product and submitted it to open peer review — to a literature where anyone, including its competitors and its critics, can check the math. The journal's editor responded in kind, publicly inviting debate, competing models, and challenge. This is the sun in a bottle placed on the open ledger.

And the open ledger does what it always does: it tells the truth, including the inconvenient parts. Read the papers and you find the uncertainties itemized rather than airbrushed. The highest-fidelity turbulence modeling predicts substantially less fusion power than the simpler models — roughly 677 megawatts against the nominal 1,130. Reasonable changes in assumptions about the plasma edge can swing the prediction by a factor of two. And nearly every section ends with the same honest refrain: SPARC will tell us. SPARC is the smaller experimental tokamak, still under construction, whose job is to test these models against actual burning plasma before ARC's design is final. The case for the power plant rests, candidly, on a machine that has not yet switched on.

So here is the calibrated claim, with no overreach in it: the most credible private fusion effort has published, in the open scientific literature, the case for a real power plant at a real site — and the same papers honestly display how much still rides on an experiment that hasn't run. That is not a guarantee of fusion by 2035. It is something rarer: a promise made in a venue where breaking it will be visible.

II.The Sun in a Can

Solar paint is the more seductive image — every wall, every roof, every car body quietly becoming a generator, deployed with a roller instead of a crane. The chemistry behind the image is real. But notice immediately that "solar paint" is not one technology. It is a marketing umbrella over at least three: a hydrogen-producing coating that uses sunlight and air moisture to split water, demonstrated in a lab in 2017 and not visibly advanced since; quantum-dot photovoltaic paints, still confined to laboratories; and perovskite-based coatings, the most promising family and the one with the most stubborn durability problems. Across the field, reported efficiencies run roughly three to eight percent — against around twenty percent for the silicon panels you can buy today, and below the ten percent threshold most analysts treat as the floor of viability. There is no can of solar paint for sale anywhere on Earth.

The strongest claim in the field belongs to Mercedes-Benz, which says it has developed a photovoltaic coating five micrometers thick — thinner than a human hair — weighing fifty grams per square meter, free of silicon and rare earths, achieving twenty percent efficiency, and capable of powering thousands of kilometers of driving per year. In late 2025 the company put it on an actual prototype vehicle. That is genuinely notable. It is also, strictly speaking, not paint: it is an ultra-thin applied module under a transparent protective coat. The word "paint" is doing marketing work.

And here is where the two horizons part ways — not on physics, but on epistemology. The fusion company said: here is our math, in a journal; please attack it. The solar paint field says: trust us, it's coming — and has been saying so for nearly a decade. Mercedes has disclosed almost nothing about the underlying cell chemistry. Its range figures carry the quiet asterisk "under ideal conditions." There is no published degradation curve, no independent verification, no peer review. The claims may all be true. But they live in press releases, and press releases are a ledger only the author can audit.

One horizon is written in the open literature with its uncertainties itemized. The other is written in marketing copy with its uncertainties omitted. Hold the dream of the painted wall — but hold it the way a careful person holds an unaudited promise.

III.The Sun on the Shelf

While we watch those two horizons, a third sun is already at work, at a scale that has quietly become the largest energy story in human history. In 2025 the world installed a record 814 gigawatts of new wind and solar capacity — 647 of solar, 167 of wind — bringing the global total past four terawatts. It was the twenty-third consecutive year that renewable deployment set a new record. Thirty different countries each installed more than a gigawatt of solar in a single year. None of this required a breakthrough. It required factories, financing, land, wire, and will.

That last word is the hinge of this essay. Because the question we set out to ask was not only what is beckoning but what could be accomplished right now, with no further technological development, if only there were sufficient cooperation. The data answer that question with uncomfortable precision.

Consider the United States. At last full count, roughly 2,600 gigawatts of proposed generation and storage projects were waiting in line for permission to connect to the grid — a line more than double the capacity of the entire existing American power fleet. Over ninety-five percent of that waiting capacity is zero-carbon: solar, wind, batteries. The typical project now waits around five years for its studies and approvals, and historically fewer than one in five projects that get in line ever reach operation. Let that arrangement sit in the mind for a moment: the machines exist, the developers exist, the capital exists, the demand exists — and twice a nation's worth of clean power stands outside a procedural door, knocking.

This is not a physics problem. It is a cooperation problem — a problem of grid governance, transmission planning, permitting, and the allocation of who pays for shared infrastructure. The proof that the pace can be different is not hypothetical. China alone commissioned nearly 500 gigawatts of renewable capacity in 2025 — roughly 370 of solar and 117 of wind — accounting for more than sixty percent of all global growth. Whatever one thinks of how that state organizes itself, the demonstration stands: when a society decides to build, the technology on the shelf builds at civilizational speed. The global pipeline of planned wind and utility-scale solar now stands near five terawatts; analysts note that even this falls somewhat short of the world's pledge to triple renewable capacity by 2030 — which is to say, the gap between pledge and pipeline is itself measured not in inventions needed but in projects approved.

So the honest hierarchy of the energy horizon runs opposite to the hierarchy of headlines. The least certain thing — paint — gets the dreamiest coverage. The middle thing — fusion — has earned real credibility precisely by submitting itself to scrutiny. And the most certain thing of all, the technology already deployed at terawatt scale and still the cheapest electricity humanity has ever produced, waits in line. Not for science. For us.

·Who Owns the Horizon

A reader of this publication will hear the deeper question underneath all three suns, because it is the question we always ask: not can it be built but who will own it when it is. The ARC papers carry a trademark symbol on the reactor's very name, even as they sit in the open literature — a perfect emblem of the tension between proprietary energy and common knowledge. Solar paint, if it ever ripens, will arrive first on the body of a luxury automobile. And the line for grid connection is, in its way, a story of enclosure by procedure: abundant energy held back not by scarcity but by the design of the institutions that govern access.

The sun delivers more energy to the Earth in an hour than civilization uses in a year. Sufficiency was never the question. The question is the deed — who holds it, who is waiting in line, and whether we will organize ourselves to let the harvest in. The horizon is real, and worth watching. But the field is already in bloom. The unfinished work is not technological. It is social, and it is ours.