Leonie Tucker earned £65,000 a year as a graphic designer. Last year she earned £26,000. She is now on Universal Credit. She has a kit in her garage that she cannot afford to keep but cannot sell. Twenty years of skill, accumulated carefully, cancelled by software.

The recent Telegraph article about Tucker and others like her is worth reading. It quotes serious economists. It uses real data. Daron Acemoglu’s warning that anyone below the 75th income percentile faces redundancy, devaluation, or elimination is not scaremongering: it is a structural observation from a Nobel Prize winner.

Dario Amodei’s prediction that AI will be better than humans at essentially everything within a few years deserves to be taken at face value, not dismissed as tech-CEO hubris. The danger the article identifies is real.

But the piece has a shape that deserves examination.

It presents the AI disruption as a natural event, something like a flood or a crop failure. Workers are threatened. Politicians are grappling. Experts are urging action. The companies deploying the technology appear briefly, usually as sources of alarming quotes about their own products, before disappearing again.

What goes unasked is who benefits from the displacement and by how much. What goes unmentioned is that every pound stripped from Leonie Tucker’s income becomes a pound added to someone else’s margin.

This is a structural feature of how the mainstream press covers technology. The disruption is real; the beneficiary is absent.

We have been here before, though not in the way the article suggests. Its historical frame reaches back to the Industrial Revolution, which is accurate but distant enough to feel abstract. The more useful comparison is closer.

What is happening to graphic designers and copywriters now is the same logic that was applied to miners, steelworkers, and textile workers from 1979 onwards. The mechanism is different. The direction of transfer is identical.

Destroying union power

The Long 1980s did not end in 1990, 1997, or at any of the points when political commentators declared it finished. It is a continuous project: the systematic dismantling of the institutional arrangements that gave labour a claim on productivity gains, and their replacement with arrangements that route those gains to capital.

The destruction of the unions was phase one. The privatisation of public assets was phase two. The financialisation of housing, which turned shelter into an investment vehicle and extracted wealth from renters toward owners, was phase three. What we are watching now is not a new disruption but the same project applied to a new category of worker.

The people now discovering they are expendable are not the workers who were told, in the 1980s, that they were the past. They are the workers who were told they were the future. The graphic designers, the copywriters, the junior lawyers, the marketing directors: these are the people who followed the advice. They trained.

They accumulated skills. They bought into the idea that education and professional development were a form of insurance. They were not told that the insurance policy included a clause.

The article briefly mentions Universal Basic Income. Lord Jason Stockwood, the investment minister, says the Government will “undoubtedly” have to consider it for impacted workers. Acemoglu, to his credit, warns against it: if 60 per cent of the population live off UBI without jobs, he says, they become an underclass regardless of the payment level, because class is about social status and not just purchasing power.

This is a correct and important observation. But the alternative he offers, steering companies toward AI that complements rather than replaces labour, is a policy lever so weak it barely constitutes a suggestion.

Two options get no mention at all.

Two possible options

The first is a high, universal, unconditional income: not the means-tested, targeted, politically hedged variant that governments discuss when they want to appear forward-thinking while doing nothing, but a payment sufficient to decouple survival from employment.

The political objections are real. The fiscal architecture would require a transformation of tax policy that no current government is willing to contemplate. But those objections are arguments about power and distribution, not about feasibility. The money exists. What does not exist is the political will to take it from capital and give it to labour. The absence of this option from the article is not an analytical failure. It is a political one.

The second is a reduction in working hours without a reduction in pay. The productivity gains from AI are going somewhere. If they go entirely to shareholders and executives, you get Leonie Tucker on Universal Credit and a rising return on capital. If they are distributed through shorter working weeks at existing wages, you get the same output, fewer workers displaced, and a different answer to the question of who the technology is for.

The four-day work week is not a utopian demand. Pilot programmes in Iceland, Britain, and elsewhere showed that productivity either remained the same or improved. The resistance to it is not technical. Companies do not want to give up the surplus AI generates. That is the argument, plainly stated.

What the Telegraph article actually offers, beneath its genuine alarm, is a menu of options all drawn from the same side of the ledger. Retrain. Save aggressively. Do AI training. Sell up and move outside London.

These are individual responses to a structural problem, and they load the cost of the transition onto the workers bearing the disruption while leaving intact the conditions that produced it. This is the ideological work the piece does without appearing to do so: it naturalises a political choice as a technical inevitability and then asks individuals to adapt to it.

This has been the dominant mode of British political economy since 1979. There is no alternative. Adjust or be left behind. The market has spoken. Each phase of the Long 1980s came with its own version of this argument, and each time the workers who lost were told that the disruption was regrettable but irreversible, that the correct response was individual adaptation rather than collective demand, and that the productivity gains would eventually filter down to everyone.

They did not filter down. They accumulated. That accumulation is now visible in asset prices, in the wealth gap, and in the gap between what Leonie Tucker earned three years ago and what she earns today.

A new underclass?

Acemoglu’s most useful observation is also his most politically uncomfortable one: the Industrial Revolution took eighty to ninety years before workers saw any benefit.

The standard reassurance, that technological disruption always creates new jobs in the long run, holds only if you are prepared to write off the people caught in the transition. Weavers who lost two-thirds of their real earnings did not benefit from the eventual abundance.

They died poor. Their grandchildren did better. This is cold comfort if you are Leonie Tucker, with a kit she cannot sell and a mortgage she cannot defer. It is cold comfort, too, for the millennial graphic designer in the article who was raised to believe that hard work led somewhere, and is now concluding that it does not.

That conclusion is correct. Under the current distribution of power, it does not. What is being built, one displaced professional at a time, is not a temporary disruption on the way to a better equilibrium. It is a new settlement, with a smaller and better-rewarded tier at the top, a vast stratum of precarious or unemployed workers below it, and no institutional mechanism to shift the balance, because those mechanisms were dismantled forty-five years ago and have not been rebuilt.

The underclass Amodei and Acemoglu warn about is not a future possibility. It is a class relation being created now, in real time. It is being created not by AI as such but by the decision to deploy AI in the service of capital accumulation rather than human flourishing. That decision is political. It is also consistent with decisions made in 1979, 1984, 1986, 1997, and 2010.

The Long 1980s did not produce this moment. But it cleared the ground for it, stripped out the defences, and ensured that, when the technology arrived, there was nothing in place to prevent the gains from going anywhere other than upward.

Tucker is angry, she says, not at anyone in particular. That generalised anger is understandable. It is also exactly what the beneficiaries of this transition are counting on.

Art Book Review Books Capitalism China Climate Emergency Conservative Government Conservative Party COVID-19 EcoSocialism Elections Europe Fascism Film Film Review France Gaza Imperialism Israel Italy Keir Starmer Labour Party Long Read Marxism Marxist Theory Palestine pandemic Protest Russia Solidarity Statement Trade Unionism Ukraine United States of America War