The Cotton Mill and the Smartphone: Two Machines, One Business Model

Read enough industrial revolution history and you start to notice that the machines change but the arrangement around them rarely does. A new device concentrates something scattered across millions of ordinary lives, pours it into a single mechanism, and hands the output to whoever owns that mechanism. Two hundred and fifty years ago the machine was the water-powered cotton mill on a river in the north of England. Today it is the glass slab in your pocket. They look nothing alike, and they do, at their core, exactly the same job. Both are engines for turning a human resource into concentrated profit. Understanding the first one is the fastest way to understand the second — and to see what is coming.

What the cotton mill actually changed

Before the mill, cloth was made in homes. Spinning and weaving were domestic crafts, done on a spinning wheel or a hand loom, often by a whole family working at its own pace between other tasks — tending a garden, minding animals, sleeping when tired. It was often poorly paid. But the worker set the rhythm, owned the tools, and understood the whole trade from raw fibre to finished thread.

The powered mill dismantled that world piece by piece. Richard Arkwright’s water frame in the 1770s, and the mills built around it, moved production out of the cottage and into a single large building where the machines lived. The machines needed a power source — first water, later steam — that no household could supply. So the work went to the power, and the workers went with it. This is the first and least reversible change: production moved into the factory, and with it went any pretence that the worker controlled the terms.

Inside the mill, three things happened to the nature of work itself. First, clock discipline. A waterwheel and a steam engine run continuously and expensively, so idle machinery is waste. The mill therefore ran on the clock, not the sun or the season. Bells and gates enforced fixed hours; latecomers were fined or shut out; the factory day was measured, timed, and supervised in a way domestic work never had been. Historians like E. P. Thompson argued that this was a genuine rewiring of how people experienced time — labour became something you sold in measured units, and the clock on the mill wall belonged to the owner.

Second, the division of labour. Adam Smith had already described, in 1776, how splitting pin-making into distinct tiny operations multiplied output enormously. The mill industrialised the idea. A trade that one skilled artisan had understood end to end was broken into fragments, each fragment a repetitive motion a machine could pace. Output soared. But the worker who once knew the whole craft now knew only a sliver of it.

Which leads to the third change: deskilling. When the machine holds the skill, the person tending it does not need much. That was often the point. A task simple enough for an unskilled hand — or a child — is a task the owner can fill cheaply and replace easily. The bargaining power that a skilled spinner once had over the terms of the work drained into the machine. This is not a side effect of the factory system; it is close to its purpose.

The mill took a resource spread thin across countless homes — human labour — and funnelled it into a mechanism owned by someone else. The genius was never the cloth. It was the funnel.

The misery decades

We should be honest about what this cost, because the honest version is grimmer than the tidy one. The output gains were staggering and real: cotton cloth that had been a luxury became cheap enough for ordinary people to wear, and Britain’s cotton output multiplied many times over across a few decades. But for the first two or three generations, the people producing that abundance did not share in it. The evidence on real wages is contested among economists, yet even the optimistic readings concede long stretches in the late eighteenth and early nineteenth centuries when working-class conditions were flat or falling while national wealth climbed.

And the conditions themselves were brutal. Mills ran long days — commonly twelve hours or more — in hot, deafening, dust-choked rooms. Child labour was not an aberration but a structural feature; children were small, cheap, and could crawl under running machinery to clear jams. Parish orphans were shipped to mills as “pauper apprentices.” Accidents maimed and killed. The industrial towns of that era had life expectancies that would horrify us now. The first British factory acts of the early nineteenth century, limiting the hours of children, are usually dated to around 1802 and 1819 — and their weakness and near-total lack of enforcement tells you everything about who held power. It took decades of further agitation before the reforms had teeth.

Hold that timeline in mind, because it is the single most important thing the period teaches. The productivity arrived first. The sharing arrived much, much later, and only because it was forced.

The smartphone runs the same play

Now look at the machine in your hand. It does not spin cotton. It spins something less tangible and, it turns out, just as monetisable: your attention and your behaviour.

The parallel is precise if you state it carefully. The mill took two inputs — human labour and raw cotton — and turned them into cloth it could sell. The smartphone, and the platforms running on it, take two inputs — your attention and your behaviour — and turn them into data and, ultimately, advertising revenue. Different inputs, different outputs, identical business model: gather a resource that is spread thinly across millions of ordinary lives, concentrate it inside a mechanism you own, and sell what comes out the other end.

The resource this time is unusual because we do not experience ourselves as producing it. Nobody clocks in to scroll. But that is what makes it efficient. The mill had to march workers through a gate at six in the morning; the platform needs only to keep you engaged, and it has engineered its machine — the notifications, the infinite feed, the variable rewards — to do precisely that. I have written before about how technology gets captured: a tool that could serve you is quietly re-pointed to serve its owner’s balance sheet, and the interface is designed so you barely notice the switch. The feed that feels like it is showing you the world is a mill wheel, and you are both the labour and, increasingly, the raw material.

The same mechanisms reappear. There is a clock discipline to the attention economy, only now the clock is inside your own nervous system — the pull to check, the dread of missing out, the sessions that stretch past any intention. There is a division of labour in how the work of holding you is split among specialised systems: one model to rank, another to predict, another to auto-play the next thing. And there is a kind of deskilling too, in the way the machine absorbs judgment that used to be ours — what to read next, whom to trust, what deserves an hour — and hands back a stream optimised for engagement rather than for us. If you want the full anatomy of that extraction, I have laid it out in the attention economy.

The shared logic: gains flow to whoever owns the machine

Strip both stories to the frame and you get the same sentence. Enormous productivity is unlocked, and the gains are captured by whoever owns the machine.

This cuts against the story we like to tell about technology — that a better machine lifts everyone, that the rising tide is automatic. It is not automatic. A machine multiplies output; it says nothing about who receives that output. That is a separate question, decided by ownership and power, not by the technology itself. The cotton mill made Britain richer and immiserated the people running it, at the same time, for decades, with no contradiction. Both were true because productivity and distribution are two different things, and the machine only settles the first.

The attention economy is running the identical split. The productivity is astonishing — never in history has so much information, coordination, and communication been available so cheaply. And the returns have concentrated to a degree the mill owners could only have dreamed of, into a handful of firms that own the funnels. The gains are real. The question of who gets them is, once again, wide open — and being answered, by default, in favour of whoever owns the machine.

A machine settles what can be produced. It never settles who gets to keep it. That second question has always been decided by a fight.

The lesson the reformers paid for

Here is the part of industrial revolution history that gets sanded down in the triumphant retellings, and it is the part that matters most for us.

The industrial gains were eventually shared. Wages rose, hours fell, children came out of the mills, the worst conditions were outlawed, and over the long run the descendants of those mill workers lived lives of a comfort their ancestors could not imagine. All of that is true. But none of it was delivered by the machine, and none of it was gifted by the owners. It was won — slowly, unevenly, and against fierce resistance — through organising, reform, and regulation. Trade unions that were illegal for much of the early period and organised anyway. Factory acts fought for over generations. Public-health laws, the vote extended, inspectors given real power. Distribution was not a natural consequence of progress. It was a political achievement, and it took the better part of a century.

It is also worth remembering the people who saw the capture early and were mocked for it. The Luddites are remembered as fools who hated machines, but they were skilled workers who understood precisely what the new arrangement did to them — and who were, on the narrow question of their own livelihoods and bargaining power, largely right. Their mistake was not their analysis; it was believing they could stop the machine by breaking it. What actually shifted the balance came later and looked different: collective organisation and law.

And this was not the first time ordinary people were separated from a resource they had relied on. The mills were fed, in part, by workers pushed off the land — a story I tell in the enclosure of the commons, where common fields that villages had shared for generations were fenced into private property, and the dispossessed had little choice but to sell their labour in the new factory towns. The pattern is old: enclose the common resource, own the machine that exploits it, and let the dispossessed line up to work it on your terms.

What that predicts for the attention age

So what does the mill tell us about the slab in our pocket? Three things, I think, and none of them is comfortable.

First, do not wait for the machine to share its gains. It won’t, any more than the water frame did. The productivity of the attention economy is not going to distribute itself; expecting the owners to voluntarily hand back the value they have concentrated is a misreading of the entire history. The gains from the Industrial Revolution were not a dividend that eventually matured. They were extracted, from the top, by pressure from below.

Second, the sharing, when it comes, will come from the same three sources: organising, reform, and regulation. That might look like data rights and genuine ownership of what we produce. It might look like limits on the manipulation of children — the modern echo of the child-labour acts. It might look like antitrust breaking up funnels that have grown too concentrated, or institutions we have not invented yet, the way the union and the factory inspector once had to be invented. The specific forms are unknowable. The category of solution is not: it is political, not technological, and it will be resisted.

Third, and most soberingly, the misery decades are a feature of the pattern, not a bug we might skip. There was a long, grim gap between the mill’s arrival and the reforms that tamed it, and a great deal of avoidable human damage happened inside it. We may be living in the equivalent gap right now — the stretch after the machine has arrived and before we have built the institutions to make it serve us rather than mine us. Whether it lasts decades, as it did last time, depends on how quickly we recognise the pattern for what it is.

The reassuring thing about history repeating is that we have seen how it ends: the gains did get shared, eventually, and life did get better. The warning is in that word eventually. Nothing in the record suggests it happens on its own, or gently, or soon. Same business model, new machine. The cotton and the cloth have become attention and data — but the funnel, and the fight over who keeps what comes out of it, is exactly the one our ancestors already fought. The only open question is how long we take to remember that we are in it.