What Ireland’s Basic Artist Income Experiment tells us about a new Arts Economy

In music, a ground bass is a repeating line in the lowest register — stable, unhurried, underneath everything — that gives performers freedom to improvise above it. It doesn’t dictate what you play, but it anchors it, giving shape to the music and making what’s above it possible.

Ireland just built one for artists.

After a three-year pilot that put €325 a week with no strings attached into the hands of 2,000 randomly selected artists, the Irish government made the program permanent. No applications for each project, no justifying your work to a committee, just here’s some money, go make things.

The results were startling. For every euro invested in the program, the Irish economy got €1.39 in return, with nearly €36 million flowing back through taxes. Recipients spent 11 more hours a week on their creative work than a control group, and the percentage of the control group who had stopped working in the arts altogether over the period rose from 6 to 13.5 percent over two years while for Basic Income recipients, the dropouts held steady between 4 and 5.5 percent.

Ireland demonstrated something more about a ground bass: economic insecurity doesn’t just force workers out, it diminishes the overall creative economy. That matters enormously right now, because we are entering a period when a lot of people across a lot of industries are about to lose their job security.

The fear is everywhere right now. Will AI take my job? You hear it across creative industries as the seeming inevitability of AI descends on us. The evidence hasn’t yet caught up to the fear, though. Economists report no measurable sign that AI is putting Americans as a whole out of work. Even Anthropic’s own researchers, who spent months mapping AI’s labor market impact, concluded with a call for prognostication humility, noting that a prominent earlier attempt to identify vulnerable jobs found that most of the categories had maintained healthy employment growth a decade later.

And yet companies are restructuring, using AI as the reason. Amazon eliminated 14,000 corporate roles. Salesforce cut 4,000 customer support jobs, with its CEO saying AI now handles half the company’s work. Block laid off 40% of its workforce. But AI-linked cuts accounted for only about 4.5% of total layoffs in 2025. That hasn’t stopped companies from increasingly using AI as the public explanation for restructuring that might have happened anyway, because it sounds like innovation rather than failure.

History suggests this is familiar territory. Every major technology advance has triggered fear of job loss, then embrace when new unforeseen jobs, even whole new industries emerge because of it. This historical pattern is reassuring, but it may not be the right frame this time.

Previous technology revolutions distributed their gains unevenly but eventually broadly. The factory owner got richer, but the factory needed workers. But AI productivity gains are different. The fear is that the benefits will wildly direct wealth to a handful of companies that own the models, the compute, and the data, impoverishing millions of jobless in the process. A company can replace ten workers with a system costing a fraction of their salaries, and the savings don’t have to get reinvested in wages. The firms that own the intelligence get richer at the rate that everyone else has less work, and that’s not an accident of the technology, it’s the actual business model.

Which brings me back to Ireland’ and to a reframe I think we need’s experiment’s experiment.

The case for some sort of Universal Basic Income has been gaining traction in recent years as technology advances and disparities in wealth grow. How do you maintain some sort of working social contract when the very fundamental pillar of a jobs-based economy shifts and millions of jobs are swept away? But the standard case for UBI frames it as charity: the state compensating people for being economically surplus.

That framing hasn’t won much political traction, and I think misdiagnoses what’s happening. The fundamental currency of value is shifting, from jobs to human activity. Every human who posts a review, searches for something, streams a film, or has a conversation online is producing data. That data trains the AI systems now capturing the productivity dividend — it’s extremely valuable at scale. So human activity has become the raw material of the new economy, but the humans generating it are currently compensated nothing for it.

A data dividend, a partial recapture of the value that human existence and behavior generates for AI systems, is a more honest description of what UBI could be. It changes the political and moral architecture of the proposition. You don’t need to prove your work is culturally valuable or economically marginal. You need only to exist and act in the world. (And before your privacy alarms go off: this doesn’t require tracking individuals. It requires taxing the companies that profit from what we collectively generate, more like a carbon levy than a surveillance program. You tax the fuel, not the driver.

And perhaps there’s a way to realign incentives for participation in civic society. You could reward civic participation at a higher rate — volunteering, caregiving, creating, teaching — recognizing that some human activity generates social value the market ignores but a healthier society would want to sustain. Contrast that with the current attention model algorithms that boost outrage, oddity and snark. This would be a ground bass built under contribution rather than employment.

Extend that logic to institutions, and it might be even more interesting.

A symphony orchestra’s audience data, a museum’s visitor engagement patterns, a theater’s ticketing and community relationships — all of this is currently harvested by the platforms those institutions use to reach their audiences. Ticketmaster, Meta, Google sit between cultural institutions and their communities, extract the data from these relationships and monetize it. The institution incurred the expense of creating the event and building the audience, but the platforms captured the value of the connections and its data.

This is what the destruction of cultural middleware looks like at the economic level. The connective tissue between institutions and communities was extracted and redirected, not just eroded. A data dividend framework creates the basis for rebuilding that tissue, not as charity, not as grant-funded infrastructure perpetually vulnerable to budget cycles, but as civic infrastructure with a legitimate revenue claim. This more properly relocates and encourages value where it’s generated.

We don’t fund bridges through philanthropy because we understand they’re important infrastructure for the economy, and because the people who benefit from crossing them ought to contribute to their maintenance. Cultural middleware is similarly load-bearing, but the platforms sitting on top of it have been collecting the toll without maintaining the bridge. A data dividend framework shifts the value proposition to value already extracted, not a request for more grant funding.

For individual artists, a ground bass would reorganize the whole ecosystem. The current arts funding model is structured around scarcity, a pyramid of competitive grants and fellowships where 80 percent or more of applicants fail, and the application process itself is a tax on the time that should go to making work. A ground bass changes the math. You don’t need to win the grant lottery to survive, you need it to thrive. That’s a different relationship to risk, and it shifts who holds power. If artists aren’t economically desperate, they have leverage. They can turn down the bad commission, the exploitive residency, the gig that requires sanding off what’s interesting about their work. Give artists a ground bass and you change who sets the terms.

Ireland started with artists because they were an easy case: chronically underpaid, culturally necessary, structurally abandoned by the market, and the political argument was winnable. But the experiment proves something more general: unconditional income produces more output, not less, generates return and not just expenditure, and retains people in socially valuable work who would otherwise give up.

In AI we are building a productivity machine that will concentrate gains in very few hands. The transition will fall hardest on people with the least flexibility to absorb it, and the current social infrastructure, designed for an employment-based economy, really doesn’t have an answer for what comes next.

The question isn’t whether we can afford a ground bass. Ireland showed it more than pays for itself. The more important question is whether we’re willing to think boldly to reframe the argument: from charity to infrastructure, from grant-seeking to revenue claim, from the arts-as-a-cost to the arts-as-a-civic-asset that has been generating value for platforms and algorithms and shareholders without adequate return to us all.

Ireland started with artists. But the architecture Ireland’s experiment suggests is considerably larger than a pilot program for 2,000 painters and poets.

---

To see all the stories on which these essays are drawn from, subscribe to ArtsJournal’s free daily and weekly newsletters.