The Builders Are Dying: How the Populations That Made the Modern World Are Disappearing — And What Happens Next

In 2018, the Swiss alpine village of Albinen made international headlines by offering families 25,000 Swiss francs per child to move there. The commune was desperate — its school was dying, its population ageing, its future evaporating. Across Germany, over 200 primary schools closed between 2020 and 2024. In South Korea, 3,800 schools now operate at less than half capacity. In rural Japan, entire school districts have been consolidated into single buildings, with children bussed in from depopulated villages where the median resident is over seventy.

Meanwhile, six thousand miles to the east, a Foxconn factory in Zhengzhou — the one that assembles most of the world’s iPhones — is replacing production lines with robotic arms. Not because robots are fashionable, but because there are no longer enough young Chinese workers willing to do the job. China’s working-age population has been shrinking since 2012. The factory that once employed 300,000 people at peak production is quietly automating its way through a labour force that no longer exists.

Between them, Europeans and East Asians invented practically everything you are using to read this sentence — the semiconductor, the internet protocol, the LCD screen, the alphabet, the printing press, and quite possibly the chair you are sitting in. They are also, collectively, going extinct.

This is not a lament. It is an inventory, a set of projections, and an uncomfortable question. But first, the numbers — because the numbers are where the argument begins and where polite evasion ends.

The Arithmetic of Extinction

The word “extinction” is not hyperbole. It is mathematics.

For a population to sustain itself, women must bear an average of 2.1 children — one to replace the mother, one the father, and a fraction to account for mortality before reproductive age. Every European nation and every East Asian nation is below this line. Most are far below it.

In Europe, the figures are uniformly bleak. Italy: 1.24. Spain: 1.19. Greece: 1.30. Germany: 1.35. Poland: 1.29. The United Kingdom: 1.49. France, long the continent’s fertility champion thanks to generous family policies, has dropped to 1.68 and is falling. Not a single European country is at replacement level. Eastern Europe is emptying through a double blow: low births compounded by mass emigration of the young to the West.

In East Asia, the collapse is even more extreme. South Korea’s total fertility rate of 0.72 is the lowest ever recorded by any nation in human history. Taiwan: 0.87. China: 1.02, with several provinces reporting rates below 0.8. Japan: 1.20. Hong Kong: 0.77. Singapore: 0.97.

The share of global births makes the trajectory visceral — and the true figures are worse than the geographic data suggests. The UN Population Division tracks births by country, not ethnicity. But births in Europe are not the same as births to European-heritage peoples. In the United Kingdom, 34% of births in 2024 were to foreign-born mothers (ONS, 2024). In Germany, 29% (Destatis, 2024). In Sweden, 27.5% of under-21s have a foreign background (SCB, 2024). Across the EU, roughly one in five births is now to a mother born outside Europe. Conversely, European-heritage births in the United States (49.6% of 3.6 million — now below half for the first time), Canada, Australia, Russia, and Latin America are not captured in the European geographic total at all.

When you adjust for ethnic composition — counting European-heritage births wherever they occur and excluding non-heritage births within Europe — the builder populations accounted for roughly 37% of all births on Earth in 1960: more than one in three (UN WPP 2024; CDC NCHS; Eurostat; ONS; Statistics Canada; ABS; Russia 2021 Census). By 2025, that share has fallen to approximately 14%. By 2100, on current trajectories, it will be around 7% — roughly one in fifteen. The European-heritage share alone will be just 2.2%. Meanwhile, sub-Saharan Africa’s share of global births has risen from 8% in 1960 to over 31% today and is projected to reach 43% by mid-century. The civilisations that invented the scientific method, industrial manufacturing, and the internet are on course to produce a smaller share of the world’s children than a single sub-continent.

The population projections that follow from these numbers are not speculative. They are arithmetic. A fertility rate of 1.0 means each generation is half the size of the one before. South Korea’s current rate of 0.72 means each generation is roughly a third of the previous one. The UN Population Division’s median variant projects that China’s population will fall from 1.4 billion today to approximately 800 million by 2100 — a loss of 600 million people, equivalent to the entire population of Europe. Japan will fall from 125 million to roughly 75 million. South Korea, from 52 million to perhaps 24 million.

Europe’s trajectory is slower but equally irreversible. The continent’s population is projected to fall from approximately 450 million to 350 million by 2100. But the working-age population will fall faster, as the proportion of elderly rises. The median age in Europe is already 44.4, compared to 19.7 in sub-Saharan Africa.

The ethnic European share of global population tells the story most starkly. In 1900, people of European heritage — counting not just those living in Europe but the diaspora in the Americas, Australasia, Russia, and southern Africa — constituted roughly 25% of the world’s population. By 1950, it was approximately 22%. Today it is around 12%. By 2100, on current trajectories, it will be approximately 5%. East Asian populations are on a similar trajectory, delayed by a generation but falling faster.

No country in history has reversed a sustained fertility rate below 1.5. Augustus tried tax incentives in 18 BC. They failed. Singapore has spent billions on baby bonuses since the 1980s. Its fertility rate is 0.97. Hungary, which has implemented the most aggressive pro-natalist policies in Europe — income tax exemptions for mothers of four, subsidised housing, free IVF — has nudged its rate from 1.23 to 1.52. Heroic effort. Still half a child short of survival.

The builders are dying. The question is why it matters — and the answer lies in what they built.

Why This Matters: The Builder Ledger

Civilisation is not a birthright. It is an inheritance — a body of knowledge, institutions, and practices that must be actively transmitted from one generation to the next. When the chain of transmission breaks, the inheritance is lost. Libraries burn. Skills die with their holders. Institutions hollow out. The Romans knew how to build aqueducts that supplied fresh water to a million people over distances of ninety kilometres using nothing but gravity and precision stonework. After Rome fell, nobody knew how to fix them. It took a thousand years before European cities again had reliable water supply.

The modern world’s inheritance is overwhelmingly the product of two civilisational traditions: European and East Asian. This is not a claim of inherent superiority — this site has argued at length that the rise of the West was based substantially on luck. It is a factual accounting of who designed and built the systems on which eight billion people now depend.

The European record. The scientific revolution — Copernicus, Galileo, Newton, Kepler. Germ theory: Pasteur and Koch. Antibiotics: Fleming. The periodic table: Mendeleev. Electromagnetism: Faraday, Maxwell, Tesla. Nuclear physics: Rutherford, Bohr, Fermi, Einstein. Evolution: Darwin. Genetics: Mendel. Computing: Babbage, Turing, von Neumann. Through 2024, roughly 87% of all Nobel Prizes in the sciences have gone to European-heritage laureates. The Fields Medal: 84%. The Turing Award: 88%. The Abel Prize: 100%. Charles Murray’s Human Accomplishment catalogued every significant scientific breakthrough between 800 BC and 1950 using multiple independent historiographies: European and European-diaspora populations produced approximately 97% of all significant accomplishments in the sciences during the period 1600–1950.

Beyond science, the institutional operating system of the modern world is European in origin: constitutional democracy, the independent judiciary, the university, patent law, double-entry bookkeeping, central banking, the scientific journal, the hospital, international law. The electrical grid, water treatment, the internal combustion engine, aviation, nuclear power, the internet — all European-heritage inventions. Every functional state on Earth, including China, Japan, and South Korea, runs on institutional architecture designed in Europe.

The East Asian record. For most of recorded history, China was the most technologically advanced civilisation on Earth — paper, printing, gunpowder, the compass, the civil service, canal engineering, blast furnace iron smelting, and thousands more innovations documented across the twenty-seven volumes of Joseph Needham’s Science and Civilisation in China. When East Asian societies decided to industrialise, they did so with a speed that stunned the West. Japan went from feudal agrarian society to industrial power in forty years. South Korea’s GDP per capita was lower than Ghana’s in 1953; seventy years later it is the world’s tenth-largest economy. China now produces approximately 30% of all global manufacturing output — more than the United States and the entire European Union combined. East Asia dominates solar panels (~80%), electric vehicles (~60%), batteries (~75%), and rare earth processing (~90%). China, South Korea, and Japan collectively build 93% of all new commercial ship tonnage. Taiwan’s TSMC manufactures 90% of the world’s most advanced semiconductors.

Together, European-heritage and East Asian populations — currently about 21% of the world’s people by ethnic heritage — account for roughly 95% of all Nobel Prizes in the sciences, over 90% of Fields Medals and Turing Awards, approximately 71% of global manufacturing output, 93% of global shipbuilding, and the overwhelming majority of global patent filings. They designed, built, and currently operate the infrastructure of the modern world: the power stations, the semiconductor fabs, the pharmaceutical laboratories, the fibre-optic networks, the hospitals, the universities.

These populations constitute roughly 12% of the world’s population when measured by ethnic heritage rather than geography. They account for approximately 87% of the world’s most prestigious scientific and mathematical prizes. The disproportion is not a talking point. It is a structural feature of who builds the knowledge frontier — and it is the reason their disappearance matters to everyone, not just to them.

What Happens After the Builders Leave

Civilisation as Cultural Inheritance

The critical insight is that civilisation is not carried in blood. It is carried in culture — in the institutions, the training pipelines, the apprenticeships, the universities, the professional standards, and the tacit knowledge that allow complex systems to function. A child born anywhere on Earth, raised within a functioning institutional framework, educated in its methods, and integrated into its professional culture, can operate and extend that framework. The question is not whether any individual can learn to run a power grid or design a semiconductor. Of course they can. The question is whether the institutions that transmit that knowledge survive the demographic transition intact.

The evidence, from multiple contexts and centuries, suggests they often do not.

The Post-Colonial Record

When European colonial powers withdrew from Africa and Asia after the Second World War, they left behind physical and institutional infrastructure: railway networks, ports, power stations, hospitals, universities, legal codes, parliamentary systems, civil services. The question of whether these systems could be maintained by the populations that inherited them is not theoretical. We have seventy years of data.

The broad pattern across sub-Saharan Africa is one of institutional and infrastructural decay — documented extensively by the World Bank, the African Development Bank, Freedom House, and Transparency International. Africa’s total railway network has shrunk since independence, with many colonial-era lines abandoned or inoperable. Freedom House’s democracy index shows persistent democratic backsliding. Transparency International’s Corruption Perceptions Index consistently ranks the majority of sub-Saharan African nations in the bottom quartile globally.

A caveat that is analytically essential, not merely rhetorical. Colonial rule was extractive by design. Infrastructure was built to move resources from the interior to the coast for export, not to develop integrated national economies. Cold War interference destabilised governments. Structural adjustment programmes gutted public services. The resource curse distorted economies. Debt traps compounded the problem. To attribute post-colonial decay solely to the characteristics of the inheriting populations would be historically illiterate.

And yet, the caveat does not explain everything. The most instructive case study is not a poor country crippled by extraction. It is the richest and most industrialised nation on the continent: South Africa.

In 1994, when Nelson Mandela took office, South Africa possessed the most sophisticated economy in Africa. Eskom, the state electricity utility, provided reliable power with surplus capacity. The road and rail networks functioned. The universities were internationally ranked. Thirty years later, Eskom has subjected the country to rolling blackouts for over a decade, with 2023 seeing more than 300 days of power cuts. Transnet, the state freight rail operator, has seen volumes collapse by approximately 40% between 2015 and 2024. Multiple municipalities can no longer maintain water treatment systems. Cholera — controllable with basic sanitation since the nineteenth century — returned in 2023. Unemployment stands at approximately 33%. GDP per capita has declined since 2011.

The skills flight is quantifiable. Between 800,000 and one million skilled South Africans — disproportionately white — have emigrated since 1994. These were the engineers, doctors, managers, and technicians who operated the systems. They were not replaced at sufficient scale. The cultural inheritance — the professional knowledge, the institutional habits, the maintenance culture — left with them.

Zimbabwe amplifies the pattern. Once the “breadbasket of Africa,” Zimbabwe became a net food importer within a decade of the land reform programme that seized white-owned farms and redistributed them to politically connected individuals with little farming expertise. The technical knowledge of commercial agriculture — soil management, irrigation, crop rotation, mechanisation — was a cultural inheritance that walked off the land with the people who held it.

The Western Evidence: Detroit and the Deindustrialised City

The pattern is not confined to the developing world, and it is not confined to racial categories. The most striking evidence comes from within the West itself.

Detroit peaked at 1.85 million residents in 1950, when the city was 82% white and the arsenal of American manufacturing. By 2013, the population had collapsed to 680,000 — a 63% decline — and the city filed the largest municipal bankruptcy in American history. The racial composition had inverted: 82% Black, 10% white. Over 140,000 homes had water services shut off. The poverty rate exceeded 40%, the highest of any major American city. Vast neighbourhoods stood abandoned.

The conventional narrative attributes Detroit’s collapse to deindustrialisation and the decline of the automobile industry. This is partly true. But deindustrialisation hit many cities — Pittsburgh, Cleveland, Milwaukee, Sheffield, Essen — and most recovered or stabilised. What made Detroit different was the simultaneous collapse of its tax base, its professional class, its institutional capacity, and its maintenance culture. White flight was not merely a demographic shift; it was a transfer of human capital — the engineers, the managers, the tax-paying professionals — out of the city. The cultural inheritance that knew how to run a modern city left with them, and the population that remained, impoverished by decades of segregation and discrimination, had been systematically excluded from that inheritance.

The same pattern repeats across American cities. Baltimore: population fell from 950,000 (1950) to 570,000 (2024), with a homicide rate of 35.6 per 100,000 — ten times the national average. St. Louis: population halved since 1970, homicide rate of 52.9 per 100,000 — the highest in the nation. Both cities rank in the bottom fifteen of America’s seventy-five largest cities for fiscal health (Truth in Accounting, 2024).

But here is what makes the argument civilisational rather than racial: the same pattern occurred in ancient Rome, involving Europeans inheriting from Europeans. When the Western Roman Empire collapsed in the fifth century, the Germanic peoples who inherited its territory — Franks, Visigoths, Ostrogoths, Vandals — were unable to maintain the infrastructure they acquired. The aqueducts fell into disrepair within a generation. Roads deteriorated. Literacy collapsed. Public sanitation disappeared. Rome’s population fell from over a million to fewer than 30,000. The Germanic tribes were European. The failure was not genetic. It was a failure of cultural transmission — the inheriting population had not been trained in the institutional knowledge required to operate the systems they had acquired.

And consider the counter-example. London is now a majority-minority city — 53.8% of residents identified as non-white British in the 2021 census. Yet London is the wealthiest city in Europe, its infrastructure functions, its institutions remain world-class. The difference is not racial composition. The difference is that London maintained its institutional continuity, its professional training pipelines, its tax base, and its investment in infrastructure throughout its demographic transition. The cultural inheritance was transmitted to the incoming population. The chain was not broken.

The lesson is civilisational, not racial: complex systems require the specific cultural capital that built them. When that capital disappears — whether through emigration, demographic collapse, institutional failure, or civilisational discontinuity — the systems decay. They do not maintain themselves. The question is whether the cultural inheritance of the builder populations can be transmitted fast enough and broadly enough to survive their demographic decline.

The AI Question: Can Machines Replace the Builders?

If the populations that built the modern world are disappearing, and the evidence suggests that complex systems do not survive the departure of the human capital that created them, then the obvious question is whether artificial intelligence and robotics can serve as a substitute.

The optimistic case is powerful. Japan — the nation most advanced in its demographic decline — has the highest robot density in the world: 399 industrial robots per 10,000 manufacturing workers, compared to 285 in Germany and 255 in the United States. AI systems can already diagnose diseases, write legal briefs, generate code, manage supply chains, and operate power grids with increasing autonomy. General-purpose humanoid robots — Tesla’s Optimus, Figure.ai’s Figure 01, several Chinese competitors — are approaching the $20,000 price point at which they become cheaper than human labour in virtually every context. If AI can maintain power stations, operate water treatment plants, and run semiconductor fabrication plants, then the demographic decline of the builders becomes a solvable problem.

But the pessimistic case rests on a recursive dependency that is not easily dismissed.

Who builds and maintains the AI? The answer, today, is overwhelmingly the same two populations whose demographic decline we have been documenting. The major AI laboratories — OpenAI, Anthropic, Google DeepMind, Meta AI in the United States; Baidu, Tencent, Alibaba in China — are staffed almost entirely by engineers of European and East Asian heritage. The semiconductor fabrication plants that produce AI chips are located in Taiwan, South Korea, the United States, and the Netherlands — operated by workforces drawn from the same demographic pools.

The physical infrastructure that AI requires is vast and fragile. Data centres consume enormous quantities of electricity — which must be generated and distributed by power grids maintained by skilled engineers. Those data centres require cooling systems, fibre-optic connectivity, and a supply chain stretching from rare earth mines in Inner Mongolia to precision optics laboratories in Oberkochen, Germany.

Here is where the Eskom parallel becomes instructive. If a country with South Africa’s resources cannot maintain a fleet of 1970s-era coal power stations, the notion that it could build and maintain a hyperscale AI data centre at the frontier of semiconductor technology is not credible. The skills gap is not smaller with AI. It is larger.

The optimists respond that AI can eventually maintain itself — that once the system reaches sufficient capability, it can manage its own infrastructure. This may prove correct. But it depends on a critical assumption: that the build-out of self-sustaining AI infrastructure is completed before the populations capable of building it decline below the threshold of competence. If the builders disappear before the machines are ready, the window closes.

The Window

Here, then, is where we stand.

The two civilisational traditions that designed, built, and currently operate the infrastructure of the modern world are in demographic collapse. Europe will lose a quarter of its population by 2100. China will lose nearly half. South Korea and Japan will lose a third to two-fifths. These are the UN’s median projections, not worst-case scenarios.

No policy intervention in history has reversed a sustained fertility rate below 1.5. Augustus failed. Singapore failed. South Korea is spending $200 billion over the next five years on pro-natalist policies. If past evidence is any guide, it will fail too.

The evidence from contexts where populations have inherited complex systems they did not build — from post-Roman Europe to post-colonial Africa to post-apartheid South Africa to deindustrialised American cities — suggests that institutions and infrastructure do not maintain themselves. They require the specific cultural capital that created them. When that capital departs, the systems decay. Not because of who the inheritors are, but because the chain of cultural transmission was broken. London shows this can be avoided. Detroit shows what happens when it is not.

AI and robotics represent the most plausible technological solution — perhaps the only one. But AI itself is a product of the very populations now in decline, and its physical infrastructure is more complex and more demanding than anything those populations have previously built. The tool that might save the builders is itself being built by the builders, and it is not yet finished.

The window is perhaps twenty to forty years. That is roughly the time before European and East Asian working-age populations begin to decline catastrophically, and roughly the time that optimistic projections give for AI to achieve the kind of autonomous capability that could sustain complex infrastructure without large-scale human oversight. Whether these two timelines intersect is the most consequential question of the twenty-first century.

If they do, then the demographic collapse becomes a transition, not a catastrophe. The machines carry forward what the humans can no longer sustain.

If they do not, then the modern world — its electricity, its medicine, its food systems, its communications, its clean water — may follow the Roman aqueducts into ruin. Not because anyone chose to destroy it. But because the people who knew how to keep it running were no longer there, the cultural inheritance was not transmitted, and the machines were not yet ready.

The builders are dying. The inheritance is not yet secured.

The clock is still ticking.

---