An AI investment review usually counts the hours a company expects to save. The more interesting question arrives afterwards: who gets them, and what becomes possible that was not possible before?
At company scale, this is a decision about capacity, income and authority. Taken further, it becomes the old question of what people would do if earning a living no longer organised most of their days.
In Economic Possibilities for our Grandchildren, published in 1930, Keynes imagined technology and accumulation eventually relieving the struggle for subsistence. He also wondered how people would use that freedom. The second question is worth keeping even if we reject his timetable or doubt that the first problem can ever be fully solved.
This is a philosophical exploration with a few bounded empirical anchors. Post-scarcity is its thought experiment, not its forecast.
Abundance does not decide who receives it
Making something cheaper is not the same as making it available to everybody. Production, ownership and access are different arrangements. A more productive technology can coexist with insecurity for people whose bargaining position weakens.
That distinction is already enough to change an AI discussion. A company can lower the effort required for a service and use the gain to improve access, increase margin, reduce working hours, absorb more demand or eliminate roles. Technology permits several responses; it does not choose among them.
At wider scale, the corresponding choices involve competition, taxation, public services, bargaining power and the distribution of ownership. There is no single technical milestone after which those institutions stop mattering. Even an extraordinary increase in productive capability would leave the question of who can make a claim on it.
I would therefore be cautious about a story that moves directly from abundant intelligence to universal security. The missing link is a political and institutional choice. Equally, it would be a mistake to treat concentration as an inevitable result. The point is that neither outcome follows automatically from the capability curve.
Some scarcities change shape
Fred Hirsch’s Social Limits to Growth provides a useful distinction between material abundance and goods whose value depends on relative position. We can make more useful products; we cannot give everyone the same exclusive rank.
Consider the difference between a comfortable house and the most sought-after address. Construction can expand the first. The second retains a positional element. Access to education can improve widely while the status attached to a particular selective institution remains competitive.
Attention and time introduce another limit. An abundance of generated books, music or advice does not expand the hours available to read, listen and act. A person still has to choose. More output can make trusted selection more valuable without making the selection easy.
These are conceptual distinctions, not an argument against abundance. Less material hardship would be an enormous achievement. They explain why a society could become much better at production and still find itself arguing about status, place, trust and access.
The corporate version is familiar. Faster analysis does not settle who decides. More recommendations do not make the accountable manager’s time unlimited. Removing one bottleneck exposes another. A serious abundance argument should say which constraint is relieved and which remains.
What cash experiments can tell us
The OpenResearch unconditional-cash study followed 3,000 participants in Illinois and Texas over three years: 1,000 received $1,000 a month and 2,000 received $50. It studied a particular population and period, not a society in which work had disappeared. Recipients worked an average of 1.3 fewer hours per week than the control group.
The team’s health findings report early improvements in stress, mental distress and food security that faded, and no significant average improvement in several physical-health measures. Cash also changed some patterns of medical-care use.
This evidence matters because it lets us examine particular outcomes instead of treating income as an all-purpose explanation. It does not show that recipients lacked meaning, or that a meaning deficit caused the fading effects. Continuing housing, care, health and social constraints offer possible explanations that need their own tests.
Nor does a three-year individual-transfer study settle what a permanent, universal programme would do to wages, prices, public services or community life. Other cash-transfer designs can inform different questions, but their outcomes should not be collapsed into one verdict on post-scarcity.
My conclusion is modest: material security and a satisfying life deserve separate attention. That is not a reason to withhold material support. It is a reason to ask what people can do with the security they gain, without diagnosing their inner lives from an aggregate chart.
Freedom needs something worth choosing
Work can provide income, structure, relationships, competence and a place in which effort matters. It can also be exhausting, coercive or humiliating. Defending the first set does not require preserving the second.
A common worry about a world with less necessary labour is that people would lose purpose. I find that too quick. It assumes paid work is the only credible source of purpose and that people freed from it would have nothing worthwhile to attempt. Neither is a premise I would accept without examining the opportunities and institutions around them.
Ryan and Deci’s account of Self-Determination Theory examines autonomy, competence and relatedness as needs relevant to motivation and wellbeing. It is a useful lens here, not a validated blueprint for a future civilisation.
My own question is what an automated institution leaves people able to choose, learn and contribute. An employee might gain time while losing discretion over the work. A junior might receive polished answers while losing the difficult practice through which judgment develops. A service might become efficient while offering people fewer meaningful ways to influence it.
Those are design possibilities, not measured consequences of every deployment. They suggest what to examine alongside output per hour. Ask people what changed in their work, which skills they are developing and whether they have more or less room to act. Participation is not something a leader can infer solely from productivity telemetry.
The deeper thought experiment
Nick Bostrom’s Deep Utopia asks us to imagine a successful future in which human effort is no longer needed for practical purposes. This is a philosophical premise, not a capability forecast. It goes beyond freedom from paid work to the unsettling possibility that a machine could do nearly anything we try to do better.
The distinction forces a question about the value of participation. Is playing music worthwhile only if nobody else can perform it better? Does raising a child matter only because care is a production bottleneck? Would curiosity lose its appeal if an answer were already available?
My answer is that relationships, craft and learning contain forms of value that are not exhausted by outperforming an alternative. The effort may matter to the person, the relationship or the practice even when it is not economically necessary. That is a philosophical position I find persuasive, not a conclusion demonstrated by the cash-transfer evidence.
There is a real objection. A person might experience a designed challenge as hollow if they believe their participation changes nothing. Freedom can also become lonely if it arrives without access to people, shared projects or the means to pursue an interest. A comfortable environment does not guarantee a worthwhile life.
That is why I would resist an institutional answer that simply assigns everyone a synthetic purpose. People need room to disagree about what is worth doing. The freedom to refuse an offered path is part of the freedom being promised.
Three possibilities to examine, without odds
An abundant future could distribute gains widely and strengthen people’s practical agency: time for care, education and chosen work, with institutions that let people participate. It could instead distribute material support while concentrating decisions elsewhere, leaving people secure but with little influence. It could also produce uneven gains, with some enjoying both time and authority while others absorb displacement and uncertainty.
These are illustrative possibilities, not exhaustive scenarios or a probability distribution. They are useful only if they identify choices and observations. Who owns the productive systems? Do prices and access improve? Can people influence the institutions on which they depend? Are opportunities to learn and contribute expanding?
Fertility, distress or declining trust cannot by themselves identify which path a society is on, let alone diagnose a general loss of meaning. Those outcomes have multiple possible explanations. A responsible account would investigate mechanisms rather than use them as a mood board for a future already chosen.
The more immediate evidence is available inside organisations: whose work changed, where the gain went, and which decision rights moved with it.
The choice inside the next business case
Consider a hypothetical service team whose AI-assisted preparation releases capacity. Leadership could absorb a backlog, give customers more time, protect learning for junior staff or reduce an avoidable expense. It could also leave workload targets unchanged while quietly removing discretion from the team.
The alternatives will have different economics. They should also have different accounts of who benefits and who bears the transition. A finance case can make the cost visible while a workforce discussion establishes whether the proposed development opportunities are real. Neither conversation replaces the other.
Ask the sponsor to put the distribution next to the productivity estimate: who gains time, income, authority or optionality, and who loses it? Name the decision being made about the released capacity and the people affected. Then check what actually happens after implementation.
We do not need a forecast of utopia to ask that question. It applies when a few hours are freed, long before the economic problem is solved. The future may be much less abundant than its advocates hope, but the choice about what we do with each gain is already ours.