Three people working at glowing computer terminals in a dark room
Image credit: @delta_sauce

We are entering one of the most transformative periods in human economic history. At the centre of this shift is a force quietly reshaping what we know about labour, value and productivity: zero marginal cost intelligence, or ZMC-I.

The term may sound technical, but its implications are deeply human. Imagine a world where many services people perform—teaching, designing, diagnosing, coding and writing—can be replicated by artificial intelligence at virtually no cost. Once trained, an AI does not charge by the hour. It does not tire, demand higher wages or need rest. It can design millions of buildings, write legal documents, offer therapy or teach quantum physics to every student on Earth simultaneously, all at a marginal cost approaching zero.

This is not science fiction. It is starting to happen.

We have seen similar transitions before. The Industrial Revolution replaced human muscle with machines. The digital revolution made the cost of copying and distributing information essentially zero. Now intelligence—the thing we thought made humans irreplaceable in the economy—is becoming abundant and cheap. History suggests that when the marginal cost of something approaches zero, the economy reorganises around it.

But this time we are not only talking about faster factories or cheaper books. We are talking about the foundation of how people earn a living. For centuries, we have relied on a basic assumption: if you want an income, you work. In a world where intelligence is increasingly automated, that link begins to weaken.

What happens to labour?

Human labour has long been a measure of economic value. Physical labour dominated before machines; cognitive labour—knowledge work—then became the new gold standard. ZMC-I is beginning to automate both.

For much of the twentieth and early twenty-first centuries, we believed cognitive jobs were safe. Lawyers, architects, doctors and engineers were treated as irreplaceable, with expertise playing a central role in the social contract. Now even those roles are being unbundled and absorbed by AI. And AI is not only replacing tasks; it is scaling them. A system that can perform tax planning or generate code can, in principle, do so for everyone, everywhere and almost instantly.

As this happens, the economic value of routine or replicable human labour declines. Why pay for what AI can do better and more cheaply?

That raises profound questions. If most cognitive and service work can be handled by AI, what will humans do? How will we earn money? And perhaps more importantly, will we need to?

What happens to the economy?

Different economic traditions offer different answers and warnings.

From a classical capitalist perspective, ZMC-I is a triumph of productivity. But there is a twist: when the cost of producing intelligence-based services approaches zero, prices collapse. That creates a difficult question for markets: where do profits come from when scarcity disappears?

The Marxist tradition might argue that capitalism reaches a breaking point. If human labour no longer creates surplus value and machines do the work, how can capital continue to extract profits? Unless ownership changes—unless AI and its outputs are shared—automation could produce more concentrated wealth and widespread economic exclusion.

Keynesian economists worry less about supply than demand. Without jobs, how do people buy things? Even if AI produces abundance, someone must still be able to consume it. Policies such as universal basic income or shorter working weeks might distribute the benefits of AI and maintain economic stability.

Finally, information economists may see ZMC-I as part of a broader trend. Intelligence becomes like digital content: cheap, reproducible and prone to monopoly. On this view, the long-term goal should be to treat intelligence as a public utility, like water or electricity, and ensure everyone can access it.

From scarcity to abundance—and new inequalities

As intelligence becomes cheap, the economic narrative flips. The challenge is no longer scarcity, but distribution: not “How do we produce enough?” but “Who receives the benefits?”

The most important inequalities of the future may come not from income, but from access to genuinely scarce assets: prime land, unique experiences or control of key infrastructure. If we are not careful, the owners of those assets—and of the AI systems themselves—could accumulate extraordinary power in a world where everyone else relies on the commons.

Even if material needs are met, there remains a question of purpose. Work gives many people structure, identity and meaning. A post-labour world could be liberating for some and disorienting for others. Will we drift into passive consumption of AI-generated entertainment, or fill our time with creativity, community and lifelong learning?

That depends on the institutions we build.

Rethinking economic assumptions

ZMC-I challenges the labour theory of value, the sanctity of wage labour and the necessity of economic scarcity. For centuries, we have built systems around the idea that work is essential not only to the economy, but to a person’s moral worth.

That assumption is starting to weaken.

If AI can do most work better than us, our economic logic must evolve. Income may need to be decoupled from employment. New forms of ownership may be required to prevent runaway inequality. Education systems may need to prioritise personal growth over vocational preparation.

A post-labour economy would probably still contain markets, money and innovation. But it could also adopt new metrics, such as wellbeing, creativity and time sovereignty; new institutions, such as citizens’ dividends and open infrastructure; and new narratives, in which life is treated less as employment and more as exploration.

From survival to sovereignty

We are at the beginning of a transformation as radical as any in human history. ZMC-I promises a world where survival is no longer tied to labour, abundance becomes possible and intelligence—once a finite and guarded human monopoly—is widely available.

This future is not guaranteed to be fair, inclusive or meaningful. That depends on choices we make today: how we distribute the gains, redefine purpose and imagine life after labour.

Will we use zero marginal cost intelligence to liberate humanity, or to deepen inequality?

We will need the courage to ask the right questions and rewrite the rules before they rewrite us.