Greater productivity could give families more money and more time - and make help for those who still need it easier to afford.

The most appealing promise of universal basic income is economic breathing room.

The promise of a dependable monthly payment is that it could help people survive an emergency, leave a miserable job, care for a relative, or spend less of their lives worrying about bills. Those are understandable ambitions. The difficulty is financing them.

Artificial intelligence offers another possible route to that freedom: an economy productive enough that ordinary earnings buy more, families can live comfortably on fewer paid hours, and government can afford better assistance for people who remain vulnerable.

Jeff Bezos recently described that possibility, suggesting AI could eventually allow some households to rely on one income or choose a three-day workweek. His forecast deserves consideration, although it remains a forecast rather than a commitment employers have made.

The economic logic is straightforward. Productivity measures how much we produce for the resources we use. When a worker can accomplish more in an hour, the economy has the potential to generate more wealth without requiring an equivalent increase in labor.

AI could help businesses reduce administrative burdens, develop products faster, improve logistics, and put sophisticated tools within reach of small companies. A person who previously needed an entire support staff might eventually operate a viable business with much less overhead.

If those efficiencies spread, their benefits could reach households through higher wages and lower prices. Families would gain purchasing power. Some might save more. Others might decide that a parent can stay home, a second job is unnecessary, or fewer working hours are worth the income forgone.

This resembles the freedom UBI advocates envision, although the mechanism is different. A basic-income program distributes money. A more productive economy expands what society can afford.

The distinction is especially important when discussing national costs. Giving every adult a substantial monthly payment requires enormous continuing revenue. Productivity growth could improve living standards while also expanding the resources available to finance public services.

But there is no automatic conveyor belt carrying corporate efficiency into the family budget.

A business can use AI to expand, lower prices, increase compensation, or reduce staffing. It can also retain more profit. Which outcomes dominate will depend on competition, labor demand, ownership, and the opportunities available to workers.

Amazon illustrates the tension. The company announced approximately 30,000 job reductions across October 2025 and January 2026. Its chief executive, Andy Jassy, had also warned that AI efficiencies would reduce its corporate workforce over time. These developments do not prove permanent economy-wide unemployment, but they show why an optimistic destination can involve a difficult journey.

Someone who chooses a shorter workweek while maintaining a comfortable life enjoys greater freedom. Someone whose position disappears before suitable employment becomes available faces insecurity. Any serious account of AI prosperity must address both.

Nor will cheaper digital services necessarily solve the expenses that strain households most.

Housing requires land, materials, infrastructure, and permission to build. Medical care requires facilities and trained professionals. Electricity requires generating capacity and transmission. AI can help improve these systems, but its promise will be constrained if society refuses to expand physical supply.

An abundance agenda therefore requires construction and investment alongside innovation. Faster software has limited value to a family whose rent keeps climbing.

The guaranteed-income experiments we discussed offer another useful lesson: Additional money helps, but it does not reliably solve every underlying problem. The large American OpenResearch trial found modest reductions in work and initial improvements in stress and food security that faded. Germany’s trial found no detectable employment decline, while Kenya’s early results showed stronger business activity and income. Circumstances and program design mattered.

AI will likewise produce different outcomes across households. People who can use it to increase their value may prosper. Others may need help adapting. People unable to work will need support regardless of how efficient businesses become.

That is where a larger economy could strengthen targeted assistance.

Higher taxable incomes and profits can expand the revenue base, although the result depends on tax policy and where gains accrue. Policymakers could use some of that capacity to support displaced workers, people with disabilities, or families facing exceptional hardship.

Targeted assistance could concentrate resources where they accomplish the most, while avoiding the expense of sending an identical payment to every affluent household. It would still require competent administration and careful eligibility rules.

None of this guarantees a three-day workweek. Productivity gains must be substantial, sustained, and broadly accessible.

Still, AI could become the closest practical route to the economic freedom UBI promises. Its success should be measured in households that can pay their bills, withstand setbacks, and choose how to spend more of their time. That is an ambitious goal — and a worthwhile reason to build.

(Contributing writer, Brooke Bell)