AI’s impending social crisis
LSE’s Professor Nick Couldry reflects on the contribution of artificial intelligence (AI) to society, how far we are prepared to consider intelligence itself as a utility, and the possibilities for resistance.
In January this year, Dario Amodei, CEO of Anthropic, the company behind the Claude AI chatbot, made a significant proposal. ‘We simply need’, he wrote in an essay titled ‘The Adolescence of Technology’, ‘to break the link between the generation of economic value and self-worth and meaning.’
In the essay, Amodei argued that in the future, there won’t nor should there be any connection between humans’ sense that they matter as individuals, and their ability to earn money in the economy. Although he calls the process ‘simple’, Amodei knows otherwise, for, as he admits in his next sentence, ‘that is a transition society has to make, and there is always a risk we don’t handle it well.’
Amodei is talking here, with rare honesty, about the implications for society for adopting artificial intelligence just to suit capitalist ends, rather than in the interests of the embodied humans that make up society.
The value of self-worth
There are, of course, plenty of reasons why individuals’ sense of self-worth is, and should be, about more than their paid employment; Amodei suggested as much in his long 2024 essay Machines of Loving Grace.That is lucky, when so many people’s work, especially in the neoliberal era, qualifies as what anthropologist David Graeber irreverently called ‘bullshit jobs’, and when so many other people lack any job.
Does that mean those two rather large groups of people wouldn’t want a job that both pays them and gives them a sense of individual value? Almost certainly not, although there are, no doubt, a few serene individuals for whom ‘just being’ is sufficient reward, and a slightly larger group of driven people whose sense of vocation is so intense that they would pursue it, even without being paid to do so.
In other words, the connection between participation in the economy and people’s sense of personal and moral worth is important, not just to the people themselves but to society more widely. Insofar as capitalism involves consent, it is hard to see it persisting if this connection collapsed.
Yet at the start of 2026, Amodei is proposing we sever that connection because of the unstoppable progress of AI, while admitting personally that he has much financially to gain from the very same transformation. Whether we see this as disarming honesty or arrogant insouciance matters little. The point is that Amodei has spotted the fissure in capitalist societies that the latest AI developments open up.
Amodei’s insight, such as it is, is based on his own deeply pessimistic view of Generative AI’s likely impact on white-collar jobs: the loss of 50 per cent of entry-level positions in the next one to five years. While Amodei’s case is carefully argued, much more so than similar Big Tech boss pronouncements, it is hotly disputed. Many, on left and right, dispute the idea that AI will produce high job losses, even as it transforms most white-collar employment.
But what if familiar disputes about how to predict the impact of a new technology will impact the economy is not the most crucial point? What if the key point goes deeper even than Amodei admits?
Intelligence as a utility
If we see AI and its contribution to society only in terms of economic benefits—how much money it can save, how much production time it can cut, how much it can reduce labour costs—then there is certainly no longer a reason to believe in a link between ‘the generation of economic value’ and the personal value that human workers get from performing tasks that AI can also perform. If, as Amodei wrote in January, ‘in the end AI will be able to do everything’, how much role does that leave for human labour in the economy, if intelligence is just an economic utility to be supplied at the lower possible cost?
As long as we proceed only according to economists’ logic, there is a stark choice here: live in societies where most humans do productive work, which is how the capitalist economy and most other societies have operated until now; orbuild a very different type of society whose rules we barely understand. Amodei is not the first tech boss to flirt with the logic of universal basic income—OpenAI CEO Sam Altman did too. But more important is that his own more elaborated thinking leaves him no choice but to do so: not because he has a radical vision of a participatory society, but because, like all tech bosses, he is wedded to what I call ‘the management concept of mind’.
This way of thinking about intelligence, which dominates the tech sector, evaluates intelligence like everything else in society—solely in terms of economic values, or its contribution to maximising economic value. It treats intelligence as utility. At an industry summit in March this year, Altman said this out loud: “We see a future where intelligence is a utility, like electricity or water, and people buy it from us on a meter”.
Microsoft too has talked, if more mutedly, of ‘intelligence on tap’. This is just the expanded corporate capture of mind that Marxist authors Nick Dyer-Witheford and colleagues nervously anticipated in their 2019 book Inhuman Power: for them, AI as intelligence repackaged for profit would mean the end of the resistant proletariat of which Marx once dreamed. Whether or not you are Marxist, the business dream of substituting corporate-supplied AI for human thought raises serious questions about where political resistance will come from in the future.
For now however, the idea that society’s cognitive resources will come increasingly coming under corporate control, rather than the control of humans and their embodied intelligence, is something that can be socially resisted, provided its deadly seriousness is understood.
Resistance and rebellion
That’s why we must not rush to dismiss as journalistic excess the Wall Street Journal’s unusual use of the term ‘rebellion’ for the resistance that is mounting against data centres across the US. This is not just a North American phenomenon: the same day as the WSJ article, a survey conducted by Kings College London reported that one in three UK university students think that AI-driven job losses will trigger civil unrest. Though there is widespread AI use by university students, signs of student resistance to AI are also beginning to emerge.
No one yet knows how such tensions will play out. As with every emerging social crisis, said the anthropologist Victor Turner, everything depends on whether a larger narrative emerges, encapsulating all aspects of the crisis into a single inspiring narrative of broader change and social resistance.
There are certainly many narrative strands involved in how contemporary societies currently are adopting AI:
the injustice in how potential contributions to society are distributed (social inclusion);
the injustice in how actual contributions to the economy are rewarded in economic terms (inequality);
the sustainability for our planet before AI’s vast and growing energy needs (environmental sustainability);
the sustainability, under conditions of indiscriminate AI adoption, of society itself; or at least the human-labour based institutions on which capitalism’s legitimacy as a social structure has depended for nearly two centuries (put bluntly, is a society where everyone works as a one-person company – popular in China this year in the wake of the latest GenAI model advances – really a society at all?).
It would be foolish to assume that a societal consensus against commercially-driven AI adoption will coalesce soon. Too much money, and too many global forces, are at stake on the other side: not just the self-interest of the investment-hungry tech sector, but perhaps also, as Payal Arora has argued, the ambitions of Global South countries, who may see AI as their best chance to rebalance the highly distorted global economy that is the legacy of colonialism. Perhaps new and radical possibilities of social transformation could arise via this alternative, more globally distributed route?
But it would be equally foolish to dismiss the growing resistance against AI data centres, and the more uneven resistance against AI in the workplace, as just noise. Because it speaks of such resistance, the publication in late May of Pope Leo XIV’s first Encyclical—on the topic of Artificial Intelligence—may mark the beginning of a genuinely new chapter in the global debate about AI.
How that new chapter will unfold is excitingly uncertain.
Nick Couldry’s next book, Predatory AI: How We can Resist Corporate Capture of the Human Mind, will be published by Polity in November 2026. This post gives the views of the author and not the position of the Media@LSE blog, nor of the London School of Economics and Political Science.
Featured image: Photo of Dr. Tanja Bosch (left) and Dr. Nick Couldry (right) at ICA, Cape Town, June 2026.