If machines generate more future income, asking only how much to pay people each month after job loss may not be enough.
Another question deserves advance planning: who can own the assets that produce income?
A guaranteed payment helps with today's bills. An asset may provide a share in future growth. Both matter, but they solve different problems, carry different risks, and require different institutions.
I want to consider how someone can weather a transition and also share in its eventual success.
Cash first creates room to choose
Someone who loses an income still needs rent, food, and transport. They may know they need new skills without having the time or money to get through training.
Cash has an immediate value here: more time before accepting an unsuitable job, the ability to care for family, or money for training. These choices may not raise wages immediately, but can affect well-being and future opportunities.
OpenResearch's three-year randomized study in Illinois and Texas included 3,000 adults aged 21–40 whose household incomes were at or below 300% of the federal poverty level: 1,000 received $1,000 a month and 2,000 received $50. Its reports describe some gains in planning and choice, alongside lower average employment and working hours relative to controls. Responses varied among participants.OpenResearch study design and findings
The study helps explain how a specific group responds to extra cash. It does not settle every question about permanent nationwide universal basic income. Funding, changes to other benefits, taxation, and prices need separate analysis.
Fewer working hours should not automatically count as failure; freedom of choice should not erase fiscal costs either. Evaluation must consider life outcomes, sustainable funding, and the system's net effects.
Having money and being able to buy must go together
Where care workers, housing, or transport are scarce, cash payments cannot necessarily expand supply immediately.
That is why public services deserve separate attention. Reliable education, health care, basic transport, and care support reduce household expenses and make work and learning more accessible.
Physical AI could help here: public agencies or service providers could procure suitable robot services for cheaper handling, cleaning, and some support tasks, freeing people for judgment and relationships. This is a deployment direction to test, not an established fiscal saving.
Lower procurement prices may be offset by supervision, maintenance, or safety costs. Sustained, equitable service matters more than machine counts.
Cash and public services can complement each other. Cash respects individual choices; public services address supply and basic access. Both need budgets and delivery evaluation.
Shared AI ownership starts by defining the asset
Making everyone a robot shareholder sounds attractive, but a machine, software model, operating company, and service platform have different assets and revenue structures.
Equipment ownership brings depreciation, repairs, idle capacity, and obsolescence. Returns on operating-company equity depend on winning customers and delivering. A token's rights require specific examination.
Proken, proposed in the first two essays, is conceptual. Work records, future service credits, network tokens, and asset income rights need separate designs. Distributing digital certificates does not establish distributable profits behind them.
Possible routes include employee ownership, cooperatives holding service assets, and public funds holding diversified portfolios with rules for using returns. These are institutional options; collective language does not establish effectiveness.
| Arrangement | Primary purpose | Questions to answer |
|---|---|---|
| Income security | Current spending and freedom to choose | Who funds it, and who benefits net of costs? |
| Basic public services | Access to essential services | Can they be delivered, and how is quality monitored? |
| Broad capital ownership | Sharing long-term asset returns | What is owned, and who bears losses? |
Every arrangement should make rights and responsibilities understandable, beyond showing dividends in optimistic scenarios.
The strongest objection: why should the public bear risky investments?
Shared-ownership proposals must answer this directly.
If someone's wages, pension, and savings all depend heavily on one AI company, technological failure could destroy both employment and assets. Combining labor and investment risk does not necessarily make people safer.
Public funds can also invest badly, be captured by interest groups, or sacrifice the future to short-term political pressure. Claims of inevitable long-term gains cannot replace governance.
My preferred principles: basic security should not depend on one technology asset; participation should diversify risk; duties, fees, conflicts, and exit rules should be transparent. Public funds need independent oversight, and employees should understand concentrated-equity risk.
Shared ownership should widen opportunities to benefit, not justify asking people with little financial buffer to bear more risk.
Funding determines how far a proposal can go
Cash programs, public services, and public assets all have upfront costs.
Possible funding includes taxes, existing public-asset income, and returns negotiated through public investment, procurement, or resource licenses. Each has an opportunity cost and distributional effects across groups and generations.
A debt-financed asset fund must compare borrowing costs, investment risks, and payment commitments. Buying AI assets does not mean future returns have already covered the debt.
The IMF's fiscal analysis considers social protection and revenue mobilization together. That connection matters: promises to share AI gains must also explain how arrangements survive years of disappointing growth.IMF fiscal policy analysis
I favor small, verifiable arrangements first: specify eligibility, funding, rights, and risk limits, then assess outcomes net of related taxes and benefit changes. Nationwide funding, supply, and price effects still need separate evaluation. Ambitions can be broad; payment commitments must be precise.
Physical AI may move more work into capital-intensive systems. We can already ask how ordinary people can reliably participate and benefit, alongside being customers and workers.
This does not require everyone to own robots directly or one arrangement to fit every country. It requires designing security, services, and ownership together, with transparent evaluation of costs and outcomes.
A worthwhile future should help people survive the transition and share in its gains.
The next part follows the funding and employment questions: if robots earn no wages, what happens to the tax base, and where do new work opportunities begin?
Which gap should come first when sharing AI gains: income security, public services, or access to asset ownership?