Two Dreams, One Machine

I. Two guiding images

The United States and China are usually described as opposites: here the open society, there the party state; here the freedom of the individual, there the order of the whole. The comparison mostly rests on institutions — elections, courts, the press. One level is almost always missing: the guiding image a system moves towards and uses to justify itself.

The American guiding image is old and remarkably stable. It is the “city upon a hill” that serves the world as an example, and the freedom of every individual to pursue happiness. The Chinese guiding image is harder to pin down. Officially the party holds on to socialism and has set the goal of becoming a modern socialist country by 2049. In the formula of the “Chinese Dream”, however, the rejuvenation of the nation stands in the foreground: strength and dignity after a century of humiliation. Perhaps it is less the dream of Marx than an older, Confucian one: a just order under wise leadership.

Both guiding images promise something good. The question of this essay is what becomes of them when a development arrives for which neither is prepared: that people are no longer needed as workers.

II. The well-provided have-nothing

In the second volume of The Obsolescence of Man, Günther Anders described a figure that captures the core of the modern economy: the bemittelter Habenichts, the well-provided have-nothing. He means the person who has enough means to keep the cycle of production and consumption going, but who never owns anything for good. He must be provided with means, otherwise he cannot buy. And he must remain a have-nothing, otherwise he stops buying.

Much that looks like waste turns out, on this view, to be a necessity. The phone that stops receiving updates after three years. The clothes that go out of fashion after one season. The device that cannot be opened because its parts are glued. Leasing and subscriptions complete the figure: there one no longer owns things even temporarily, one only pays for their use. The counter-image is the tool inherited from a grandfather and passed on to the grandchildren. For production it is a misfortune, because it is sold only once.

III. When work goes

So far, the well-provided have-nothing has drawn his means from wages. Henry Ford is said to have paid his workers double the usual daily wage in 1914 so they could afford the cars they built. The cycle was closed: whoever produced earned, and whoever earned consumed.

When machines take over the work, the cycle has a hole in it. In 1930 John Maynard Keynes predicted that his grandchildren would need to work only fifteen hours a week, and saw a liberation in it. He overlooked that in an order in which income depends on work, less work does not mean more freedom, but fewer means. The worker nobody needs any more is not liberated. He is excluded — as producer and as consumer alike.

IV. Three paths

For the role of the consumer without work, three paths are emerging, and they do not exclude one another.

The first: the means come from the state. A basic income or something similar keeps purchasing power alive, and consuming becomes the actual social function. That would be the well-provided have-nothing perfected — provided not by his own work but by allocation. Whoever allocates holds power, and whoever receives allocations is dependent.

The second: the masses become superfluous as consumers too. The economy orients itself towards the few who own the machines, data centres and models. And a growing share of demand no longer comes from people at all: data centres buy electricity and chips, programs buy services from other programs. The devices supply one another.

The third: the human being turns from buyer into raw material. Already today, many services are paid for not with money but with attention and data. If work and purchasing power fall away, what remains is what people supply in behaviour and traces of their lives.

V. The curse of not being needed

The gravest consequence is political. The power of the broad masses historically rested on their being needed: as soldiers, as workers, as taxpayers. It is no accident that in many countries universal suffrage followed universal conscription. Those expected to die for the country wanted a say in how it was run.

Economics knows the reverse case as the resource curse. States that live on oil or gas do not need their citizens to finance themselves, and they are strikingly rarely democratic. Some authors now speak of a looming intelligence curse: when machines produce everything a state needs, citizens lose their strongest lever. Whoever is no longer needed is no longer heard. This curse strikes both systems, only by different routes.

VI. China: the state above capital

For years China has installed more industrial robots than the rest of the world combined, and it has overtaken Germany and Japan in robot density. The state can plan and implement over the long term, as high-speed rail, power grids and the solar industry show. And the party has proven that it can discipline capital: in 2020 it halted the stock-market listing of Ant Group, and the large technology companies were cut down to size within months. If the gains of automation are to be distributed, a state that stands above capital has the tools to do it.

What China does with these tools so far tells a different story. Private consumption is well below forty per cent of economic output; the model channels returns into investment and exports, not into households. Rural pensions are minimal, and the household registration system excludes hundreds of millions of migrant workers from the services of the cities in which they work. The capacity to provide exists; the will to let everyone benefit from it, so far, does not.

And this is where the intelligence curse strikes hardest. A state that needs its citizens neither as workers nor as voters, and owes no account to an independent judiciary, provides for them only as long as it serves the state. In China political power stands above capital — but nothing stands above political power.

VII. The USA: capital above the state

In the United States it is the other way round. In 2014 the political scientists Martin Gilens and Benjamin Page analysed around 1,800 policy decisions: the preferences of economic elites show up clearly in policy, those of average citizens hardly at all once they diverge. Since the Citizens United ruling of 2010, corporations may channel practically unlimited money into election campaigns. In 2025 the richest man in the world sat for a time inside the government with a staff of his own. Power and capital are growing together, and one can watch it happen.

Yet the difference from China has not vanished entirely. Governments are voted out and lose in court. The powerful quarrel in public, and their fractures become visible. And a citizen can publish a text like this one under his own name, risking no more than that nobody reads it. That is not rule by the people. But it is more than nothing: in the West the fusion of power and capital is contested and, in principle, reversible; in China it is the foundation of the system.

VIII. The indifferent machine

Lewis Mumford found the megamachine among the pharaohs, in the Soviet Union and in the American military-industrial complex, each time under a different flag. It is indifferent to guiding images. The guiding image determines how it justifies itself, not how it works.

Anders described how freedom is abolished: not openly, but in its name. The same holds for justice. The West justifies surveillance by the protection of freedom; China justifies it by the protection of an order meant to serve everyone. Both guiding images provide enough to justify any intervention. And for the first time in history the tools of complete control are actually available: the recording of everything, analysis by machines that never tire, and provision that can be tied to good behaviour.

If both systems end in totalitarianism, it will not be because of their guiding images, but in spite of them. The guiding image would then be no more than the flag on a machine that does the same thing under every flag.

IX. How this essay could be refuted

The thesis would be refuted if a large country showed that it distributes the gains of automation broadly without making the recipients dependent — for instance through ownership of the machines rather than through allocation. It would be weakened if automation, like earlier waves, produced enough new work to preserve people’s role as producers. And it would be overstated if the tools of control proved less effective than feared, because people always find ways to evade them. There are signs of all three possibilities. None has been decided yet.

X. What remains

The decisive question is not which guiding image is the better one. It is this: who owns the machines, and what do their owners owe the people they no longer need? As long as the answer is “nothing, except what they give voluntarily”, it makes no difference whether the owner is a corporation or a party.

Large systems rarely change from within. They have changed through crises, and in a crisis people reach for whatever is ready to hand. That is a reason to ask the questions and work out the alternatives while it is still possible to publish them. It is more than nothing.