Industry

I. Where we start

Today the world is still living with the consequences of the pandemic, the war in Europe, the US–China tariffs, and the discovery that global "just in time" was fragile. Semiconductors, pharmaceuticals, fertiliser, rare earths, and batteries have left the spreadsheet and moved into the cabinet room. China remains the workshop of the world, with competitors attempting "de-risking" without knowing how to pay the price. Automation is already high in specific sectors; most of the world's manufacturing is still people plus classic machines.

II. The decade in between

Over the decade, industry does not return to the naive hyper-globalism of 2005, nor does it become autarky. It becomes fortified interdependence: states fence chips, weapons, energy equipment, and critical minerals, while ordinary trade keeps crossing supposedly hostile lines. Each bloc tries to keep a costly vital minimum at home or with an ally and discovers how much of the rest still comes through China.

Materials and energy are the ceiling. Whoever has copper, lithium, nickel, rare earths, uranium, gas, and cheap sun has an industrial policy; whoever does not has a speech. Recycling of batteries and electronics grows and does not close the loop. Mining becomes visible geopolitics again, including on the contested seabed.

III. Ten years from now

The productive map has four facts:

1. China remains the planet's largest complex manufacturer, even while decelerating and even while losing some low-value lines to Vietnam, India, Mexico, Indonesia, and now Brazil. In electric vehicles, batteries, solar panels, drones, and much of consumer electronics, it sets the world price.
2. The United States reindustrialises selectively and expensively: chips, defence, energy, some pharmaceuticals, some cars. It does not recover the industrial employment of 1970. It recovers strategic capacity and profit at points of the stack, and gains a southern flank of cheap minerals it did not have a decade ago.
3. Brazil arrives as the third pole at the manufacturing table. Cheap midday electrons, lithium, copper, rare earths, niobium, ore, an Atlantic coast, and a deliberate industrial policy pulled data centres, green hydrogen, battery assembly, refining, bioplastics, agroindustry automation, and a serious share of the EV supply chain south. Embraer moves from regional jets into autonomous cargoliners; Brazilian lithium producers supply regional refineries, while Vale anchors iron ore and base-metal chains. The third pole does not catch China on volume; it owns the southern hemisphere's stack and rents its surplus to the blocs that want it.
4. The rest fights over slices. Mexico gains from American nearshoring. India gains a little less than the marketing promises, but gains. Europe defends capital niches (machine tools, fine chemicals, luxury, premium autos) and loses volume; the EU–Mercosur deal finished, then widened, became a third leg of the tripolar system, and let European firms re-platform into Brazilian territory without giving up the European headquarters. Argentina and Chile reindustrialise inside the bloc, not for export alone.

The factory ten years from now is a software object that happens to melt metal. Digital twins run the shift; humans walk the exceptions. Smaller batches, customisation, overnight SKU changeover. This does not eliminate scale: bread, cement, steel, plastic, and basic chemicals remain volume businesses, now with fewer people per tonne. The process engineer and the high-voltage electrician are fought over like surgeons. A surprising number of new plants are designed from day one for a lights-out night and a thin day crew.

Agroindustry computerises: seed, drone, credit, insurance, traceability. Productivity rises. Climate pulls the ground from under some regions and hands it to others. Food-exporting countries (Brazil, the US, parts of Africa in potential) become more strategic and more volatile. The rural workforce shrinks in number and rises in capital per head wherever the machine gets in.

Construction becomes a repeat-order industry. Regional factories deliver panels, bathrooms, and service cores to sites already cleared for housing. Standard interfaces let several suppliers compete for the same block; design variety sits on a small number of certified parts. Long initial orders finance equipment, then private demand fills the lines. In adopting cities, the combination of cheaper assembly, more building rights, and funded utilities halves comparable real rents (see cities and housing). Factories alone never deliver that result. The most valuable factory is the one with a steady order book, not the most photogenic printer.

Industrial work. Housing throughput grows faster than labour per home falls, so installation, commissioning, and maintenance expand even as repetitive site crews shrink. Inspectors and electricians become the critical path. Factories publish component histories and fund defect reserves; a bad joint repeated ten thousand times is a balance-sheet event. Cement, copper, and skilled connections briefly bottleneck the boom. Reusing buildings and standardising repair matter as much as opening another line.

IV. Uncertainties

What does not happen in this scenario: the end of China as the factory; the return of 1970s industrial employment to the United States; Industry 4.0 as an inclusive miracle; Brazil catching China on volume. What does happen: producing physical things is a matter of state again across three blocs, a surprising number of night shifts have no one in them, the supply-chain spreadsheet has a general staff in every capital, and Brazil's rise is the difference between a global industry organised around two suns and one organised around three.