The Switch

Your hand finds the switch in the dark because it has done this ten thousand times. You are not thinking about it. You are thinking about whether you left the milk out.

The room fills with light before your hand has finished moving. Nothing about this feels like an achievement. It feels like the floor being where you left it.

For almost all of human history this was the hardest problem in the house. Every evening ended in fire. Candles, oil, gas: something had to burn in the room with you, and it could burn the room down. Then it stopped being a problem, and within one lifetime it stopped being noticeable.

The interesting question is not how the bulb works. It is what had to be true, at that exact second, for the room to light.

Behind the plate: two wires and a screw. Behind those, a loop of wire running back to a panel, and a breaker built to fail before the wire does.

Here is the part that should stop you. There is no tank. Your building holds essentially no electricity. Nothing was set aside earlier in the day in case you wanted light at nine at night.

Compare it to the faucet. When you turn on water, you are drawing from water that already exists, sitting in a pipe or a tower, waiting for you. Electricity does not work that way. The electricity you just used did not exist a second earlier. It was made at the moment you asked for it, by a machine somewhere else that was already spinning.

So the switch is not a valve on a tank. It is closer to a question, asked instantly, of a system that has to answer instantly, and that has never once been allowed to say wait.

Out at the pole, a metal can about the size of a kitchen trash bin. That is a transformer, and its job is to take the voltage on the street and lower it to something your house can survive.

Voltage is the pressure behind electricity. To push power a long way you want that pressure very high, because high pressure means less is lost along the road. But high pressure is lethal in a kitchen. So the system raises it to travel and lowers it to be used, over and over, all the way down to your wall.

None of this equipment knows about you. It is not reacting to your switch. It was placed, sized, energized, and maintained years before tonight, by people who will never learn your name and were not thinking about you when they did it.

They were thinking about load. Somewhere in a forecast, you are a fraction of a number nobody reads out loud.

Your switch is now a signal inside a machine covering six states. One organization runs the grid across Connecticut, Rhode Island, Massachusetts, New Hampshire, Vermont, and most of Maine, and its entire job is keeping supply and demand equal, all day, every day, without a pause. ¹

Every generator on that grid spins in step with every other one, all turning at the same rate: sixty cycles per second. That shared rhythm is what operators actually watch, because it works as a scoreboard for whether supply is matching demand.

The mechanism is simpler than it sounds. Picture one enormous flywheel with engines pushing it and brakes dragging on it at the same time. The engines are power plants. The brakes are everything that is switched on. If the brakes win, the wheel slows and the frequency drops below sixty. If the engines win, it speeds up. Nobody is measuring your kettle. They watch the wheel.

The wheel never stops turning. What almost never changes is how fast. Under normal conditions the whole eastern half of a continent holds that speed within about two hundredths of a cycle per second, with formal correction required if it drifts much past three hundredths. ^{2 3}

Nobody conducts this. There is no one deciding your room in particular should light. Power plants offer electricity, prices move, machines respond, and the balance holds because thousands of separate decisions keep landing close enough to right. When you flipped the switch, the wheel dipped by an amount too small to see, and something, somewhere, pushed very slightly harder.

Follow the power back past the plant and it stops being a story about wires. It becomes a question about storage: how do you keep sunlight until nine at night?

Sunlight has one great weakness. It is everywhere and it stays nowhere. Enormous amounts arrive, spread impossibly thin, and then it is gone. To light a room after dark, something has to catch it and hold on.

Plants were the first thing that could hold it, and here the numbers need care. A leaf in a field converts roughly one to two percent of the light landing on it into stored plant matter, against a theoretical ceiling near five percent. ^{4 5}

That sounds like a failing grade. It is not. It is a narrow measurement of one specific thing: light in, chemical energy stored. Only about half the incoming spectrum is the right color to drive photosynthesis at all. Of what is usable, a leaf in full sun is already saturated, and it deliberately dumps the excess as heat, because light beyond what the machinery can process does damage. That dumping is not a defect. It is a safety system, and plants that lacked it died.

And look at the result. One to two percent of an enormous number is still an enormous number. That sliver is what covers continents in forest, fills oceans with plankton, and feeds every animal that has ever lived, including the one reading this. Plants are not bad at catching sunlight. They are astonishingly good at surviving on a small, deliberate cut of an overwhelming supply.

So the first step captures a sliver. The second step was geology, and it is slower than anything a person can really picture. Between roughly 359 and 299 million years ago, in a stretch called the Carboniferous, swamp forests grew and fell into water and did not rot. Mud buried them. Pressure and heat squeezed them for a span of time next to which all of human history is a rounding error. More coal formed in that window than in any comparable stretch since. ^{6 7}

That is what a lump of coal really is. A thin slice of ancient sunlight, captured by plants that were not trying to help anyone, buried by accident, pressed for hundreds of millions of years, and then removed in a couple of centuries.

Oil and gas are the same trick with different organisms and different rock. Burn any of them and you are releasing light that arrived on Earth before there were flowers, before there were dinosaurs, before the Atlantic Ocean existed.

It is worth knowing how much of it there is, because the time step alone can mislead in either direction. Coal we can currently reach and afford to mine comes to somewhere around a trillion tonnes. ⁸ That is an absurd quantity, more than enough to feel infinite from where anyone is standing.

Then put the other number beside it. We take out roughly nine billion tonnes a year, and rising. ⁹ At that pace, what we can currently reach runs somewhere in the range of a century or so. ^{8 9}

So both things are true at once, which is the honest version of abundance. There is a staggering amount of it. Tens of millions of years went into laying it down, and we are drawing it out on the step of a human lifetime or two. The ground filled that account very slowly and we are spending it very fast, and no part of that process runs in reverse on any timescale that matters to us.

And the demand side is not holding still. There are more people every year, and each of them reasonably wants what you have: light after dark, a cold fridge, a warm room, a device that answers. Global electricity demand rose about 4.3 percent in 2024 and is forecast to keep climbing at a similar pace for the rest of the decade. ¹⁰ Every year, the world asks the wheel to spin against more brakes.

That is the real problem, and it is worth stating plainly because it is where the optimism actually comes from. It is not that we are running out. It is that a finite, slowly made, one-way supply is being asked to serve a demand that grows. Which means the interesting question stops being how much is left, and becomes: how do we do more with each unit, and where do we get supply that does not run down?

There is a second route, and it does not involve our Sun at all.

The problem with sunlight was never supply. It was delivery. Almost all the energy in this solar system is in the Sun, which holds about 99.86 percent of everything here by mass, and it sits ninety-three million miles away. ¹¹ What reaches us is only the portion of its light that happens to strike a small planet at that distance, which works out to a few hundred-millionths of one percent of the total.

That sliver is still roughly 173,000 terawatts landing on Earth continuously. ¹² Put that next to ourselves. Everything humanity does, every car and furnace and factory and phone on the planet, runs on something like 19 terawatts on average. ¹³ All of human civilization is operating on about one ten-thousandth of the sunlight that falls here, and that sunlight is itself a rounding error on what the Sun puts out. We are not straining the supply. We are barely touching it.

Fission collects differently. Instead of gathering light a star is radiating now, it releases energy that other stars already packed into matter, which has been sitting quietly in rock ever since.

The density is the part that sounds invented. Burn a kilogram of coal and you get about 30 megajoules. A kilogram of reactor fuel holds closer to 3,500 gigajoules, more than a hundred thousand times as much. ^{14 15} Split uranium-235 completely and one kilogram is worth two to three million kilograms of coal. Real reactors never come close, because they use only the small fraction of uranium that splits easily. Even so, one ceramic pellet stands in for a wheelbarrow of coal.

And you could hold it. A fresh fuel pellet is a small grey ceramic cylinder, cool to the touch, about the size of a pencil eraser. Uranium is a rock. It sits in the ground all over the world doing nothing, because splitting atoms is not something matter does on its own at any useful rate. It takes a reactor: the right isotope, concentrated, arranged in a precise geometry, surrounded by a material that slows neutrons enough to keep the reaction going.

That is worth sitting with, because it inverts the usual intuition. The danger is not that the rock is angry. The rock is inert. Everything alarming about nuclear power is a property of the machine we build around it and the waste that machine produces, not of the substance itself. We went looking for the densest energy in the ground and found something that will not release it unless you ask correctly.

So there are two ways to get a star into a room. One runs sunlight through leaves and then through geology, and it needs hundreds of millions of years and a planet willing to bury the results. The other skips both steps, because the concentrating already happened somewhere else, under conditions no planet could ever produce.

Which answers half the question from the last step. Where do you get supply that does not run down? From matter that was already concentrated, in quantities that make the annual draw look trivial.

The other half of the question was how to do more with each unit, and that is what electrification is actually for. It is not simply swapping one fuel for another. Burning things to do work is wasteful in a way that is easy to miss: a gasoline engine turns something like 12 to 30 percent of the fuel into motion and sheds the rest as heat, while an electric drivetrain puts well over 70 percent of the electricity into moving the car. ¹⁶ A furnace is capped at burning its fuel once, near 100 percent at absolute best. A heat pump does not make heat at all, it moves heat that already exists, which is why it commonly delivers two to four units of warmth per unit of electricity. ¹⁷

So electrifying a task usually means the same comfort for less primary energy. That is the lever: not asking people to want less, but making each unit go further. And there is enormous room left, because electricity is still only around a fifth to a quarter of final energy use in most large economies. ¹⁸

But that lever has a consequence, and it is the reason the grid matters more every year. Move heating, driving, cooking and industry onto electricity, and you concentrate a growing share of all human demand onto one machine. The wheel from a few steps back stops being one system among many and becomes the system.

So go back to it. It has engines pushing and brakes dragging, and the speed has to hold every second of every day, now against more brakes each year. Some engines only push when conditions allow: the wind has to blow, the sun has to be up, the water has to be behind the dam. Those are real and useful and they are not steady. A wheel carrying everything also needs engines that simply do not stop.

That is the job fission is already doing here, and it is not a hypothetical. On the same six-state grid from a few steps back, nuclear supplied about 23 percent of the electricity in 2024, second only to natural gas at 51 percent. Two plants, Millstone in Connecticut and Seabrook in New Hampshire, produce roughly a quarter of the electricity New England uses in a year. ¹⁹

Which means the answer to what lit your room is probably mixed. Some of the push came from gas burning tonight. Some came from water falling. And with reasonable odds, close to a quarter of it came from atoms splitting in a building a couple of hours down the road, in a machine that has been running steadily since long before you reached for the switch, and that will still be running steadily long after you have gone to bed and forgotten you turned anything on.

Put the whole shape together and it is not a story about shortage. It is a story about engineering: a finite store being spent, a demand that keeps rising, and a species answering by making each unit do more work and by learning to tap supplies that do not run down on any timescale we will ever care about.

Hold the clocks next to each other for a second. A human life is roughly eighty years, and we are getting better at making it longer. The uranium in that pellet has been in the ground for four and a half billion years. The Sun has about five billion years of steady burning left in it. ²⁰ Against those numbers, everything our species has ever done is a flicker, and the fuel is not the constraint. It never was. Our own cleverness is.

That matters because the alternative to getting cleverer is ugly. More people living longer and wanting more has usually meant taking more: clearing more ground, burning more, crowding out the other living things we share the place with. People are right to be upset about that. It is a real cost and it has been paid by creatures who never had a say.

But that trade is not a law of nature. It is what happens when demand grows faster than efficiency does. Get more work out of each unit of energy, and get that energy from something dense enough that you barely have to disturb anything to reach it, and growth stops requiring the same bill. That is the whole case: not fewer people wanting less, but more people wanting more while taking less to get it. Fix the place first, then go further out. There is a great deal of galaxy, and nobody is using it.

We followed two roads here, and they arrive in the same place. So does every other road, which is the part worth noticing. There is really only one source, wearing different clothes.

Coal, oil, and gas are sunlight that living things caught and the ground kept. Hydro is the Sun lifting water into the sky so it can fall and be caught coming down. Wind is the Sun heating one patch of air more than the patch beside it.

A solar panel is the same star collected directly, with none of the waiting: no plants, no burial, no three hundred million years. What it moves instead is the work, forward onto us, into mining and refining and factories and the people who build and install them. The energy was always free. Every method is only a different arrangement of who does the labour, and when.

It does that at a rate worth sitting with. Every second, the Sun fuses about 600 million tonnes of hydrogen, and roughly 4 million tonnes of that stops being matter at all and becomes energy. ²¹ Every second. It has done this for four and a half billion years and is nowhere near finished.

There is a stranger detail. Energy released in the core does not fly straight out. It is absorbed and re-emitted over and over by the crushing material above it, wandering a random path, and estimates of that journey run from ten thousand to well over a hundred thousand years. ²¹ Only the last eight minutes are a straight line. The warmth on your face left the center of the Sun before anybody had planted a crop.

The uranium is older still. Elements that heavy are not made in ordinary stars. They need atoms to be hit with neutrons faster than they can fall apart, and that requires somewhere extraordinarily violent. The best-supported site is two neutron stars colliding, confirmed in 2017 when astronomers caught one both as a ripple in spacetime and as a flash of light. Whether certain rare supernovae also contribute is still being worked out. ^{22 23}

So a fuel pellet is debris from a catastrophe that happened before the Sun existed, scattered across the galaxy, mixed into the cloud that became this solar system, folded into the rock of a forming planet, and left alone for four and a half billion years until somebody dug it up to boil water.

The light in your room is starlight either way. Some of it left a star eight minutes ago. Some of it was made in a collision so violent it briefly outshone whole galaxies, and has been waiting in the ground since before Earth had a surface to stand on.

Sit with the sizes for a moment. One star, holding very nearly everything. A planet catching a sliver of its light. A species using a ten-thousandth of that sliver. And inside that species, one person, in one room, flipping one switch for a few seconds without looking up.

That is not a small thought because it makes us insignificant. It is a large one, because look what the insignificant thing built. Nobody knows how to make the light in your room. No person alive holds the whole procedure. It happens anyway, every night, for almost everyone, and it is so reliable that you built a life on top of assuming it.

That is the thing I find hard to be pessimistic about.