Three Landmarks
There is a stretch of motorway south of Budapest where, if you know to look, you can see the lock-in for yourself.
The road is the M5. It was built in the 1990s on top of the older Route 5, which was paved in the 1960s on top of the post-war reconstruction, which was laid on top of the imperial highway that ran south from Budapest to Belgrade for two centuries before that. Underneath all of it, in places, you can still find the basalt blocks of the Roman road that connected Aquincum to Sirmium under the Antonines. About a metre and a half below the truck lane, in a stratum no one will ever uncover voluntarily, there is concrete laid in the year a particular Hungarian prime minister wanted to be photographed cutting a ribbon. Above it is concrete from a different prime minister. Above that, asphalt. Above that, more asphalt.
The road is not, in any meaningful sense, going away. If Hungary collapsed tomorrow and the forests came back, the M5 would still be there in the year 3000 as a long, low ridge through the soil, with trees growing in its cracks. The asphalt would oxidise. The reinforcing bars would rust through. But the aggregate — the crushed rock that makes up two-thirds of every road by mass — would still be a crushed-rock layer, indistinguishable from a stratum laid down by a glacier, and as durable. Roads decay slowly. Crushed rock essentially does not decay at all. It is one of the most patient materials on Earth.
I want you to hold that picture, because it is the thing the discourse keeps missing about the moment we are living through.
In late 2020, a paper appeared in Nature with a chart on it that briefly went viral. The chart had two lines. One was the total mass of every living thing on the planet — bacteria, fungi, plants, animals, all of it summed — and the line was roughly flat at around 1.1 trillion tonnes. The other was the total mass of human-made objects on the planet — buildings, roads, machines, infrastructure — and the line was a near-vertical climb across the twentieth century. The two lines crossed somewhere around the year 2020.
This is the moment that has, ever since, been treated as the headline. By weight, the planet is no longer carbon-based. The green world has been overtaken by the grey one. The throne is empty, or full of a stranger.
The headline is correct. It is also, I am going to argue, the least interesting thing on the chart.
The crossover is what statisticians call a milestone. A milestone is a moment when a smooth process happens to cross a round number — when a country’s population reaches a billion, when a fund crosses a trillion under management, when a single tree’s mass equals the weight of a blue whale. The milestone is striking because of what it lets you say in a sentence. It is not striking because something physically changed at that point. The two curves did not bend at 2020. They had been on this trajectory since at least 1900, and the year of intersection is just where the lower one happened to pass under the upper one. Move the goalposts by ten percent in either direction and you get 2017 or 2024. The crossover is a number we picked because it is easy to hold.
What is actually happening on the chart is something else. Let me describe it.
The grey line — anthropogenic mass, the human-built world — has been doubling roughly every twenty years for over a century. In 1900 it was about three percent of the green line. In 1940, six percent. In 1980, a third. In 2020 it caught up. Compound growth has the property that a process which looks like it is barely moving for a hundred years suddenly looks vertical from the other side. We are on the vertical side of it now. Even on a conservative estimate — and the production rate has, encouragingly, been slowing since around 2014 — the doubling time across the next century looks more like twenty-five or thirty years than the historical twenty. That is still doubling. That is still the kind of curve that, in any other context, would frighten anyone who had ever taken a high-school biology class about runaway populations.
The green line, by contrast, is roughly flat. It has been roughly flat for centuries. It will be roughly flat for centuries more. Living biomass is constrained by sunlight and water and chemistry; it cannot grow at three percent a year because there is nowhere to put the extra. The throne, in other words, was not stormed. The throne was approached, at speed, by something whose growth was never bounded the way biology’s was. We did not lose the crossover. We were always going to cross.
That is the first thing the headline gets wrong. It treats a doubling time as a single moment.
The second thing it gets wrong is much worse.
Living things decay quickly.
This is so basic that we forget it is a property of biology, not of matter in general. A tree falls in a forest and is, within a few decades, soil. A whale beaches and is, within a few years, sand. The average atom of carbon now sitting in the trunk of an oak will, within twenty-three years, be in the atmosphere, or in a fungus, or in a bird, or in another oak — turned over by the metabolism of the biosphere as the biosphere does what biospheres do. The whole green machine is a flow. Its mass at any moment is a balance between fast production and fast destruction. Stop the production and the standing stock collapses in a human lifetime. Resume the production — give the forests their land back — and the standing stock returns, slowly, over centuries.
Living things grow back. That is the consolation we usually grant ourselves about ecological damage. Forests can be replanted. Wetlands can be restored. Even after the worst of the bombing of European cities, the trees came back. Carbon is patient with us.
Concrete is patient with no one.
The dominant material of the built world, by mass, is concrete — about forty-seven percent of the trillion-tonne total. The second largest component, at a third, is aggregate: gravel, sand, crushed rock dug from quarries and packed into roads and foundations and fill. Together, these two materials are eighty percent of the grey world. And both of them, on the timescales that matter for ecological recovery, are effectively permanent.
Reinforced concrete, abandoned and unmaintained, loses its structural integrity in fifty to a hundred years. That is what you can see in pictures of Pripyat or Detroit. But losing structural integrity is not the same as losing mass. The rebar rusts and the slab cracks and the building falls in on itself, and the result is a heap of broken concrete that takes centuries more to break down to a powder of calcium silicates, and even then the powder is rock. Roman opus caementicium, poured two thousand years ago in harbour walls at Portus and Caesarea, is still solid. The MIT team that worked out why, three years ago, identified the chemistry — the lime clasts in Roman concrete are self-healing crystals — but the takeaway for our purposes is simpler. Concrete that old is concrete that still weighs the same.
Aggregate is more patient still. Crushed basalt does not corrode, oxidise, or biodegrade. The rock used in roadbeds is, in a real sense, geology. Burying it does not destroy it. Time does not destroy it. There is no microbe that eats it. The blocks of the Roman road south of Budapest are still there because there was no mechanism by which they could go anywhere.
So the grey world, taken as a whole, behaves like this. Build it, and it stays. Stop building it, and most of it stays anyway. The model I have been running, with composition-weighted decay rates for the actual materials in the stock, suggests the bulk half-mass time for the built world — the time it takes for half of what we have built to physically disappear — is somewhere between three hundred and six hundred years, depending on which assumptions you make about concrete in particular. Some of the components decay faster. The asphalt rots in fifty years. The plastics are a wide range. But the great mass of it — the concrete and the rock — is on a half-mass timescale longer than the period from the fall of Rome to now.
This is the asymmetry. The green world decays fast and grows back. The grey world decays slow and accumulates. They are not playing on the same timescales. They never have been. The crossover was the moment a fast-growing slow-decaying biota caught up with a steady-state fast-decaying one — and on that chart, the only one of those two properties that has ever mattered is the second.
Once you see this, the chart reshapes itself.
There are three landmarks on the chart worth naming.
The first you have already met. The Crossover. Around 2020. The first moment the grey world outweighed the green one. A line on a chart, important mostly because it lets you say a sentence at a dinner party.
The second is harder to see and matters more. The Five-to-One. Somewhere around 2050, give or take a decade. The moment the built world weighs five times the biosphere. At five-to-one, biomass is no longer in the same order of magnitude as anthropogenic mass — it has gone from “comparable population” to “rounding error”. The model gives a window of 2047 to 2062 across every plausible production scenario, including ones where the growth rate continues to slow. By the time today’s twenty-year-olds are entering middle age, by anyone’s estimate, the chart will not be a comparison of two things. It will be a description of one thing with a green smear at the bottom.
The third is the one that costs to write. The Lock-In. Somewhere between roughly now and the end of this century — central estimate around 2066, but the range spans 2050 to 2100, and the location of that range depends mostly on how patient we believe concrete really is. The Lock-In is the first year past which, if all human production stopped at the stroke of midnight, the biosphere would not return to majority on the chart for a thousand years.
It is worth saying that sentence twice.
The first year past which, if all human production stopped at the stroke of midnight, the biosphere would not return to majority on the chart for a thousand years.
We are not at the Lock-In yet. We may already be at the Lock-In. The window is open and we are inside it, and we cannot tell, from inside it, exactly where the line falls. What we can tell is that the line falls somewhere roughly within a single human lifetime of now — and that it is being placed not by the crossover, not by the doubling time, but by the slowness of concrete.
This is the asymmetry I want you to keep.
The thing displacing the older biota is not the construction. It is the persistence. The construction is the symptom; the persistence is the disease. Every year we build, we add to a stock that decays on the timescale of empires. Every year we do not unbuild, we add to it again by default. Even if we changed our minds tomorrow, even if we stopped — and we will not stop — the road south of Budapest would still be there in the year 3000. The forests would have to grow around it.
The crossover was the moment we noticed. The five-to-one is the moment the comparison stops being close. The lock-in is the moment we lose the option to undo.
We are not heading toward a future event. We are inside a present one. The choice that will be remembered as the defining choice of this century is not being made at some distant point of crisis. It is being made in every road poured and every foundation set, day by day, while the discourse argues about a milestone that was already in the rear-view mirror by the time anyone learned its name.
The thing about concrete is that it does not have to win. It only has to stay.