At the Saudi-US Investment Forum in November, Elon Musk sketched a future in which AI and humanoid robots will do almost all the work.
Money, he suggested, will become almost irrelevant. Jobs will be “optional,” more like hobbies such as gardening.
Machines will have ended poverty, because everyone will receive a “universal high income” from the state.
Musk is hardly the only tech titan with this vision of the future. Demis Hassabis of Google DeepMind looks forward to an era of “radical abundance” in which AI will deliver extraordinary productivity and prosperity, with the gains all “fairly” distributed.
Mustafa Suleyman of Microsoft AI argues for a “universal basic provision” that would treat access to powerful AI systems and digital services almost as an entitlement.
And Sam Altman of OpenAI has proposed an “American Equity Fund” that would tax large companies and private land at 2.5% per year to pay every US adult an annual dividend.
Closet socialists
Put simply, the leading architects of AI are open about the fact that they are creating systems whose success in generating material abundance could also obliterate large swaths of the labor market.
In their envisioned future, the “springs of cooperative wealth” will flow so abundantly that people will receive “according to their needs,” not according to the hours they clock in a factory.
If that last sentence sounds familiar, that’s because it comes from Karl Marx.
Are capitalism’s most fêted poster boys actually closet socialists? In one sense, yes.
Look closely at tech leaders’ proposals and you will find that their apparent affinity for socialism fades fast
The people building advanced AI are unusually candid about the distribution of wealth. They accept that if machines perform tasks more cheaply than humans, labor’s share of national income will shrink.
If wages disappear, people will need another way to feed and house themselves, and the economy will need new mechanisms to sustain purchasing power.
But look closely at tech leaders’ proposals and you will find that their apparent affinity for socialism fades fast.
Altman does not argue for workers’ control of OpenAI, nor for public ownership of the infrastructure.
He wants governments to socialize only the returns. While a “universal high income” could help share the spoils, the chips, models, and platforms that generate those spoils would remain firmly in the hands of an extraordinarily wealthy few.
This would not be socialism as we know it. A tiny elite would own AI’s “commanding heights” and hand everyone else a check or some form of digital ration. The sum would be enough to live on, but not enough to challenge those in power.
But if the universal income being offered is high enough to deliver a comfortable life, some will argue, it doesn’t matter who owns the algorithms and data centers.
AI dividends
There are at least three reasons to be skeptical.
First, we are told that generous AI dividends will come whenever the productivity gains have been fully unleashed.
But history suggests that once wealth and ownership are locked in, the beneficiaries rarely volunteer to dilute themselves.
Already, a handful of AI and platform businesses account for an astonishing share of global corporate value.
By the time any serious AI-funded income scheme arrives, much of this value will have been converted into concentrated equity and dynastic wealth.
Asking today’s AI barons to retrofit egalitarianism onto this structure is like asking Victorian-era mill owners to invent the welfare state.
Even if some sort of distribution scheme did materialize, what about the vast majority of countries that host no frontier AI firms
Second, even if some sort of distribution scheme did materialize, what about the vast majority of countries that host no frontier AI firms?
If local jobs are automated away while the profits pile up in California, Seattle, or Shenzhen, who exactly is going to fund their citizens’ income? The AI founders have been strikingly quiet on this question.
Third, a monthly payment – however generous – is no proxy for a meaningful life.
Work has long been one of the main ways that we contribute to society. It is how we prove to ourselves and others that we matter. It gives our lives purpose, structure, and validation.
Without it, we risk becoming a society of passive spectators – well-fed, permanently entertained by AI-generated content, and tended to by humanoid robots, but stripped of the dignity that comes from caring for others and being needed.
A stipend can pacify; it can also be tinder for revolt. A population that is materially provided for but politically powerless is unlikely to remain forever docile.
Socialism from above
So, even if governments do figure out how to deliver a universal high income and secure a meaningful social stake in AI’s upside, the answer to large-scale automation cannot simply be to tax robots and buy everyone a new Tesla. Income matters, but so does agency.
This means ensuring that governments and civil society remain in control of the evolving AI terrain.
Rules, restrictions, and safeguards cannot be left to powerful private-sector architects.
The tech titans are offering us a future of socialism from above: they keep the means of production, we get allowances - Sam Altman
Moreover, a significant share of any future AI bounty would need to be spent on specific goods belonging to the “human economy”: care, education, the arts, local democracy.
The purpose would not be to manufacture pointless jobs, but to sustain the idea that citizenship is predicated on contribution.
And finally, we would need global mechanisms to spare countries without frontier AI champions from becoming collateral damage.
One option would be an International AI Dividend Fund, financed by a modest levy on the largest AI and cloud companies’ profits or compute usage, with payouts directed to the countries hit hardest by automation.
Such a scheme would be imperfect and politically difficult to pull off; but at least would provide an answer to the question that Musk and his peers have ignored: Who pays for everyone else?
The tech titans are offering us a future of socialism from above: they keep the means of production, we get allowances.
Our task is to promote democracy from below. That means demanding not only a share of the AI wealth, but also the power to shape and control the means of generating it.
Noreena Hertz is an honorary professor at the UCL Policy Lab, where she leads research on AI.
