14 September 2026
Heard In AI

If robots do the work, who has the money to buy what they make?

On The Diary of a CEO, economist Steve Keen and commentator Konstantin Kisin agreed that machine production would not by itself give ordinary people an income — and then disagreed about where that income would come from. Keen argued only government money creation could supply it; Kisin, who says his anti-communist credentials are well established, said redistribution becomes unavoidable once AI does the work.

A briefing reports one development when it happens. We correct or clarify it later; a new development gets a new briefing. How our formats work

Steven Bartlett left AI to the end of his conversation with economist Steve Keen and commentator Konstantin Kisin on The Diary of a CEO, and asked simply what impact they thought it would have.

Keen began with the pleasant version of the future. AI is "a potentially transformative technology, without a doubt," he said, and the dream outcome is the one he attributed to Elon Musk: universal high income, the machines doing the work, "Star Trek style."

Then he put the question that occupied the rest of the exchange. "To get there, how are you going to give people income if they're not working anymore?"

The purchasing-power problem

The abundance story usually stops at production. Suppose factories and offices are run by machines owned by a relatively small number of companies and shareholders. The goods exist. But most people do not own capital — shares, land, machines, the assets that pay an income without work — and in Keen's description, if they are no longer paid wages either, they have no income at all. Without income, he said, "they can't buy the goods being produced by the robots in the first place."

So something has to put money in people's hands before the shelves can be cleared. "You need a mechanism that enables people to pay for the basic goods and services to stay alive," Keen said. Above that floor, in his sketch, you would have to become an entrepreneur to become wealthy.

His own answer was blunt: "The only way to do that is the government has to create the money." That, he added, is something he does not think Musk has thought through.

Why Keen says you can't just tax the AI companies

Bartlett offered the obvious alternative. Could you tax the AI companies instead of printing money?

"No," Keen said. "Taxation takes money out of the economy. It doesn't pay for services. It takes money out."

In his account, the government does not collect money first and then spend it; its spending is what creates the money, and taxation is the instrument that removes some of it again. The purpose of the tax is to stop too much money accumulating in the system. If the government spent and did not tax, Keen said, the money supply would grow at 20% or 30% a year, and "you would have hyperinflation without a doubt."

Bartlett pressed with a concrete proposal anyway: take one of the big AI companies — call it company A — tax 50% of its profits and "give it out as a UBI or something." A universal basic income is a payment made to everyone regardless of whether they work.

Keen agreed a system of that kind would be needed, on his own reasoning about where the jobs go. If AI works, he said, manual labour and "all the clerical, boring clerical jobs will disappear." And if they disappear and people have no incomes, "then we end up in the Hunger Games. And that's the fear that I have." A purely private-sector approach to AI, in his view, leaves two destinations: the Hunger Games, or a government-provided income that gives people a basic standard of living, with entrepreneurship the route above it. His jobs claim is conditional on AI working as advertised, and his remedy is a proposal about how a state would have to act.

An anti-communist reaches a communist conclusion

Kisin's response began with a disclaimer. "I think my anti-communist credentials are pretty well established," he said — Bartlett agreed — "but I did say that I thought that if we get to the position where AI is doing all the work, communism is an inevitability."

He glossed the slogan as he went: "from each according to his ability, which will be zero, to each according to his needs, which will be whatever our needs are." On the mechanism — taxation or something else — he was open, and he said plainly that he struggled to understand the economic concept Keen had been presenting, and that there was no time to get into it.

What he did insist on was that some form of redistribution becomes unavoidable, for reasons of self-interest rather than justice. If six people in the world end up owning all the revenue-generating assets, he said, "they're either going to have to share it or they're going to find their heads on spikes." Powerful robots guarding big walls would not save them, in his telling. "So I would not recommend that approach to the big tech AI oligarchs."

What Marx made it conditional on

The phrase Kisin adapted comes from Marx's Critique of the Gotha Programme, written in 1875. Marx separated two stages. In the first, what an individual gets to consume still depends on the labour they supply. Only in a later phase does distribution follow need — and he made that later phase conditional on greatly expanded productive capacity, on work itself being transformed, and on the end of the narrow division of labour that ties people to one occupation. He also assumed that before anything is shared out to individuals, the social product has to cover replacement equipment, new investment, reserves, schools, healthcare and people unable to work.

That is a different condition from Kisin's. Marx's higher phase arrives through changed human work; Kisin's arrives when human contribution to production falls to zero because machines do it. The slogan is the same; the reason for reaching it is not.

The interim, which may last a while

Kisin was careful about timing. Nobody knows what short, medium and long term mean in AI, he said, because nobody knows how quickly we reach artificial general intelligence — in his description, a machine that can think like humans, probably faster and better — and, as Bartlett added, the robotics to go with it, machines that do physical things faster and better than people. "But we are not there at the moment," Kisin said.

Asked whether this was the most consequential issue of our time, he said it was, because it runs through everything else the three of them had discussed: immigration, socialism and capitalism, economic stagnation, private debt. All of those change in such a scenario.

His advice for the meantime was narrower than his forecast. What he tells the people who work for him, he said, is: "you will not be replaced by AI... you will be replaced by someone who uses AI better than you do." For a young person making their way in the world, these are tools, and the point is to use them as efficiently as possible to succeed in the current environment. "And I think that situation will carry on for some time."

Share this article

Go to the original

Sources & further reading

  1. 01

Connected ideas and articles

From the conversation

Podcast episodes

Article history

Updates to this article

Tags

Could AI be the next Newton? Brian Greene considers what physicists would lose

A colleague told Brian Greene that physicists should choose their final research problems before AI takes over. Greene finds the prospect both exciting and terrifying. His account of three kinds of scientific creativity explains why he thinks machines could make foundational discoveries—and why collaboration might change, rather than simply erase, the human scientist’s role.

6 min read

What a town can ask for before the data center arrives

Meta says its infrastructure investments generated tax revenue that paid $50,000 teacher bonuses in Richland Parish. Its community commitments prompted a bargaining argument on Moonshots: towns hosting AI data centers should negotiate benefits, with Alex Wissner-Gross proposing “universal basic electricity.” The unresolved question is how to turn an attractive offer into lasting local gains.

5 min read

If nobody can stop AI, what is left worth doing?

On The Diary of a CEO, Konstantin Kisin tells Steven Bartlett that nobody — not him, not the host, not Trump — can stop AI, and that what remains is living well with family and community. Bartlett refuses to give up on public pressure and expects AI on the ballot in 2028; Kisin expects voters to blame whoever is in office instead.

6 min read

Hinton's maternal AI meets an objection: protection is still control

On The Diary of a CEO, Steven Bartlett offered Geoffrey Hinton's proposal for a protective, mother-like AI as the most hopeful answer to Konstantin Kisin's forecast that humans become pets or cattle. Kisin argued that maternal care runs on a genetic incentive a machine would not share; Steve Keen answered that it runs on empathy — and then pointed out that an AI determined to keep us safe might forbid war or cut energy use, which is protection by way of control.

6 min read