The anti-AI backlash is an anti-elite backlash

The problem with treating the plebs poorly is that you may need them one day

I define AI populism as a worldview in which AI is viewed not only as a normal technology but as an elite political project to be resisted. It regards AI as a thing manufactured by out-of-touch billionaires and pushed onto an unwilling public to achieve sinister aims like “capitalist efficiency” (layoffs) and “population management” (surveillance). AI populists don’t really care whether ChatGPT is personally useful, or if Waymos eke out some safety gains: AI’s utility as a tool is immaterial relative to the unwelcome societal change it represents.
Jasmine Sun, 14/4/26

I don’t think most people are formulating a new worldview in which AI is a boogeyman political project hatched by billionaires. I think they’re more likely to understand AI as an extension of an already inequitable system, and as an accelerant of that inequality… AI has become the avatar of the ills of unrestrained capitalism. “AI populism” is really just “21st century populism” or just, “populism”…

Fear and anger at AI are often best understood as fear and anger at how AI will function within capitalism… That’s why the students are booing [AI], I think. They’re experiencing AI in realtime as a forecloser of futures; as the cruel new face of hyper-scaling capitalism, as the prime agent moving a world that’s become a deck stacked against them.
Brian Merchant, Blood in the Machine, 16/5/26

One reason Americans seem to despise A.I. more than people in other countries is that they know our government is too sclerotic to handle it…

We simply lack the political infrastructure in America to distribute A.I.’s benefits to the public. With the systematic evisceration of the labor movement that started during Ronald Reagan’s presidency, said Ramamurti, “the institutions that many other countries have for mediating these kinds of technological advances don’t exist in the United States.”

Tech billionaires might be less likely to announce that their inventions will cause mass unemployment if they felt constrained by public sentiment. The fact that they don’t shows how broken America’s democratic feedback loop has become.
Michelle Goldberg, NYT, 18/5/26

It’s old news that a fierce backlash to AI is developing.

As Goldberg notes above, this is hardly surprising on one level, given the alpha tech bros’ AI sales pitch has been, “Embrace this technology that will likely stop you earning an income and possibly destroy human civilisation”.

As Goldberg also points out, the fact alpha tech bros demonstrated so little concern for the fate or likely reaction of the little people shows just “how broken America’s democratic feedback loop has become”. (I’d add the democratic feedback loop hasn’t been functioning well anywhere in the Anglosphere for many years, but let’s not get ahead of ourselves.)

But that raises the question of why AI isn’t on the nose across all classes and locations. It’s not like China doesn’t have a youth unemployment problem, but the Chinese are considerably more upbeat about AI than Americans. And it’s not like working-class and middle-class Anglosphereans wouldn’t like the much-promised – if still yet to manifest – upsides of AI, like a cure for cancer.

So why are those in the Anglosphere, especially those lacking the insulation of wealth, so opposed to AI?

I’d argue it’s down to elite failure. Or perhaps more accurately, half a century of short-sighted elite greed and hubris.

Noblesse oblige vs meritocracy
Elites are inevitable. Fortunate societies have elites that put the general welfare of their society above their own class interests. Unfortunate societies have elites that put their class interests above the general welfare of their society.

In the three decades following WW II, Anglosphere societies were fortunate societies where wealth was shared around relatively equitably. Famously, even blue-collar workers could enjoy the trappings of a middle-class life, such as marriage, home ownership, children and stable work offering a living wage.

Around 1980, the Anglosphere decided to embrace what might be labelled meritocratic neoliberalism. One of the many unfortunate consequences of this was the evaporation of a sense of noblesse oblige among Anglosphere elites.

Old-school elites at least had a vocabulary of obligation, however hypocritically observed. The new meritocratic elite has a vocabulary of just deserts. In a meritocracy, everyone gets what they deserve. If that means wealth creators prosper while those in the bottom four quintiles of the income distribution struggle, well, what are you going to do? The market has pronounced its verdict, and no correspondence will be entered into.

The problem with elite self-dealing – at least in the modern democratic age – is that there are a lot more non-elite citizens than there are elite ones.

To some extent, elites can override the wishes of the mass of voters – there’s a famous 2014 study that shows economic elites had substantial independent influence on US policy outcomes, while average citizens and mass-based interest groups had little or no independent influence.

But if you end up in a society where there’s a small (and shrinking) group of people living an elite lifestyle and a much larger (and growing) group of people who are despairing of their futures, eventually those in the latter group will take umbrage.

For almost half a century, Anglosphere elites have made the world more comfortable for themselves and more precarious for everyone else.

Wealth inequality has risen sharply, most spectacularly in the United States, where wealth concentration is again comparable to the Gilded Age. Labour’s share of national income has fallen while Capital’s has risen across the Anglosphere. Corporations have become more concentrated and powerful; unions have been weakened. Housing has been converted from shelter into a tax-advantaged asset class. Education has become a debt-financed credential race. Risk has been pushed down onto workers through casualisation, contracting, outsourcing and the gig economy, while asset owners have repeatedly been protected by monetary policy and bailouts. Labour income is taxed more reliably than capital gains, inheritances and asset inflation. The benefits of globalisation, automation and mass migration have flowed disproportionately to employers, property owners, professionals and investors, while the costs have landed on those with less bargaining power. And as I’ve pointed out once or twice, the language of morality – diversity, openness, flexibility, innovation, sustainability, global citizenship – was shamelessly deployed to launder material interest.

Depending on your politics, dear reader, you may think the balance of power progressively shifting from non-elites to elites from circa 1980 is a good or a bad thing. But it’s definitely a thing and, as always, there are no free lunches in life.

After decades of single-mindedly pursuing their self-interest, Anglosphere elites now expect the commoners to trust them when they say they’ll share the benefits of AI around equitably and definitely not throw any workers who’ve outlived their usefulness on the scrapheap.

Given their ‘lived experience’ of the last four decades, the hoi polloi simply aren’t buying it.

The AI backlash is proving more politically explosive than the tech industry expected. People are not assessing AI in a vacuum. They are assessing it through memories of deindustrialisation, housing inflation, wage stagnation, corporate consolidation and airy lectures about disruption from people who somehow never seem to get disrupted out of their own roles.

So, what happens now?
Even given the events that have unfolded since 2016, Anglosphere elites appear remarkably complacent. Having got their way so often for so long, perhaps that’s unsurprising. But the elite project of fast-paced AI transformation is encountering mounting popular resistance.

Let me count the ways.

• Local opposition to data centres over power use, water use, noise, land, grid strain and higher bills.

• Voters electing AI-sceptical politicians promising to protect jobs, schools, privacy and public services from automation.

• Union action and professional pushback, demanding limits on AI replacement, algorithmic management and synthetic use of workers’ voices, faces or output.

• Lawsuits and regulation over copyright, privacy, surveillance, discrimination and automated decision-making.

• Consumer resistance, favouring “human-made” journalism, art, teaching, customer service and advice.

• Cultural ridicule, turning AI “efficiency” into a synonym for cheapness, fakery and elite extraction.

Is a 21st-century New Deal feasible?
I’d argue that the Chinese people are sanguine about AI because they trust their elites to deliver on their side of the bargain. Say what you will about the CCP, but it has put a rocket under the living standards of Chinese citizens since 1979. In contrast, many workers throughout the Anglosphere believe their quality of life has been steadily decreasing since around the same time.

Fortunately, there’s a straightforward way out of the AI impasse – elites agreeing to share the cost and benefits of the AI transformation equitably. As opposed to the standard operating procedure of recent times, which has seen elites capture almost all the benefits of economic, political, technological and demographic changes while shifting the costs onto the lower orders.

Is such a settlement possible? It existed in the Anglosphere from 1945-1980 and has continued to exist in Scandinavian societies.

If elites want the AI transformation to proceed without triggering a full-scale popular backlash, they will have to start sharing the gains. Not vaguely promising to share the gains at some unspecified future date, but actually delivering the goods.

That translates to shorter working weeks, higher wages, generous redundancy payouts for displaced workers, taxation arrangements that substantially redistribute wealth (including windfall taxes on extraordinary AI profits), data dividends, some form of income floor and yes, even part-public ownership of AI companies and/or employee profit-sharing schemes, along with public equity stakes in foundational AI infrastructure.

The AI transformation will happen. The only question now is whether it’s a negotiated, somewhat orderly, win-win transformation or a bitter, bloody and chaotic one.

Are elites going to sign up to such a settlement? One where they get to keep many of their advantages but agree to share their wealth around a little more equitably?

A new settlement is not inevitable. The original New Deal required mass unemployment, elite fear of expropriation and a credible revolutionary alternative.

Those pre-conditions are not yet in place.

But if AI delivers mass disruption without mass compensation, elites may soon discover how much more like the 1930s the 2020s can become.