Time, contribution, and the question AI is really asking
By Tony Callaghan, Founder, Better Thinking
Framing paper — part of the If not work series
Start with the hours
Every human being alive today has exactly the same daily allocation: twenty-four hours. Eight billion people, twenty-four hours each, no exceptions and no extensions. What varies, enormously and consequentially, is what those hours contain.
A significant portion is non-negotiable. Sleep takes roughly a third. Eating, physical maintenance, and the basic biological overhead of staying alive takes more. Call it ten hours of fixed cost, give or take, across an average human life. The remaining fourteen or so are available: available for work, for care, for play, for connection, for creation, for worship, for rest, for contribution of any kind.
That available portion is not fixed. It has been expanding for most of human history, slowly at first and then with increasing speed, as technology has reduced the time required to secure the basics. The shift from foraging to agriculture freed time previously spent in search of food. The industrial revolution, for all its costs, eventually produced labour-saving technologies that reduced the hours required to sustain a household. The twentieth century compressed domestic labour through machines and reduced working hours through collective bargaining. Each transition released more available time into the human day.
Artificial intelligence is likely to be the next and possibly the largest such release. Automation of cognitive and routine work, combined with advancing robotics, will reduce the hours required for a significant portion of what currently occupies human working life. How much, how fast, and how unevenly distributed: these are genuinely uncertain. But the direction is not. More available time is coming.
The question this paper is asking is not whether that will happen. It is what humans will do with the time when it arrives. And the answer to that question, it turns out, is not primarily economic. It is social.
What humans have always done with available time
Look across the full span of human history and a pattern emerges that is easy to miss if you start, as most AI and work discussions do, from the industrial present.
Hunter-gatherer societies, which represent the overwhelming majority of human existence, did not organise available time around individual productivity or income accumulation. They organised it around contribution to the group and the recognition that contribution generated. Hunting skill, ecological knowledge, the capacity to resolve conflict, the ability to maintain social cohesion: these were the currencies of status. The person who contributed most to collective survival was the person who was most valued. The recognition was immediate, visible, and social. Your contribution was seen by the people whose opinion of you mattered most.
Agrarian societies reorganised available time around land, collective stewardship, and seasonal labour. The contribution that generated recognition shifted: what you grew, what you built, what you maintained, the reliability with which you showed up when the community needed you. Recognition was still local, visible, and tied directly to observable contribution. Your neighbours could see what you had done.
Industrial societies made a significant change. They inserted a mediating layer between contribution and recognition. Rather than contribution generating recognition directly, contribution generated income, and income generated the consumption that signalled recognition. The factory worker’s contribution was not visible to anyone outside the factory floor. What was visible was the house, the suit, the holiday: the consumption that income made possible. Recognition became mediated through markets rather than direct through community.
This shift was not inevitable. It was constructed, over roughly a century and a half, through specific historical processes: the reorganisation of daily life around the wage, the displacement of older community-based contribution systems, the gradual equation of human worth with economic productivity. Historians and anthropologists have traced this construction in detail. The point here is simple: it was built. And because it was built, it can in principle be rebuilt differently.
The industrial recognition system is not the culmination of human social organisation. It is the latest in a series of arrangements, each calibrated to the economic conditions of its time, each appearing permanent to those living within it, and each eventually giving way to something different when those conditions changed.
The conditions are changing again.
The C/R framework
These observations point toward a framework that this paper uses throughout, and that the companion papers in this series apply to specific cases.
Contribution is the act of giving time, skill, or energy toward something that produces value for others, whether or not it also produces value for yourself. Contribution can be paid or unpaid, formal or informal, large or small. What matters is that it generates something others can recognise.
Recognition is the social signal that your contribution has been seen and valued. Recognition can be formal (salary, title, promotion) or informal (gratitude, visibility, community standing). What matters is that it is legible: others can see it, and it reinforces the contributor’s sense of being useful and belonging.
The two are coupled, and the coupling is the point. Recognition reinforces contribution: people sustain effort when they can see it is valued. Contribution enables recognition: you cannot be recognised for something you have not done. A system with high contribution but low recognition exhausts people. A system with high recognition but low contribution produces empty performance. Both are needed, and the balance between them determines whether a contribution system sustains or collapses.
Beneath the C/R system sits infrastructure: the processes, tools, technology, and administration that enable or constrain how contribution happens and how recognition flows. Infrastructure is not the point. It is what makes the point possible. This distinction matters because much of what passes for a solution to C/R problems is really an infrastructure proposal: better software, better platforms, better management systems. Infrastructure that does not serve the C/R system is, at best, neutral overhead. At worst, it is the thing that drives people away from contribution entirely.
The historical pattern described above is, in C/R terms, a story of how different societies have organised the coupling between contribution and recognition, using different infrastructure, in response to different economic conditions. Hunter-gatherer societies had direct C/R: contribution was immediately visible, recognition followed naturally, infrastructure was minimal. Industrial societies have mediated C/R: contribution flows through employment, recognition flows through income and consumption, infrastructure is vast and complex.
The question AI raises is what happens when the industrial C/R mediation weakens. Not whether people will stop wanting to contribute and be recognised: that need is as old as the species. But whether the specific system through which it has been met for the past century and a half will remain adequate, and if not, what could replace it.
How humans spend available time: the full picture
To answer that question properly, it helps to map all the ways humans currently spend available time, not just the ones that look like work or formal volunteering. The table below is an attempt at that mapping. It is diagnostic rather than exhaustive: the point is to see the C/R profile of each category clearly, not to produce a definitive taxonomy.
Each row represents a meaningful category of time use. The columns assess: contribution level, internal recognition (within the activity itself), external recognition (visible to wider society), how well internal R translates to external R, and the infrastructure currently supporting the activity.
| Time use | Contribution (C) | Internal R | External R | R translation | Infrastructure |
|---|---|---|---|---|---|
| Paid employment | High: structured, skilled, purposeful | Moderate: peer respect, development, professional identity | High: salary and consumption signals universally legible | Complete: salary directly converts contribution to visible status signal | Strong: organisations carry the overhead |
| Care (children, elderly, family) | Very high: sustained, skilled, socially essential | High: direct relational reward, deep personal meaning | Very low: almost entirely invisible to wider society | Broken: contribution is real and high; external legibility near zero | Weak: falls almost entirely on the individual carer |
| Youth volunteering (Scouts, Guides) | High: skilled leadership, sustained, real community impact | High: internal award systems, leadership development, personal reward of watching young people develop | Very low: invisible beyond the immediate group | Broken: sophisticated internal R does not translate outward | Very weak: compliance burden falls on volunteers; systems built for professional organisations |
| Civic contribution (environment, community) | Low to moderate: practical, local, visible output | Low: informal peer acknowledgment only | Low to moderate: physical output visible locally but not amplified | Partial: natural visibility exists but no amplification mechanism | Weak: dependent on individual convenors; collapses when they leave |
| Leisure and sport (active) | Low to moderate: physical skill, team contribution, personal development | Moderate: team belonging, competition, personal achievement | Low: visible within sporting community; invisible beyond it | Weak: translates to external R only at elite levels | Moderate: clubs, leagues, facilities provide reasonable infrastructure |
| Arts and creative practice | Low to moderate: skill, expression, cultural contribution | Moderate to high: creative satisfaction, peer recognition within creative communities | Low to moderate: visible only where distribution infrastructure exists | Weak to moderate: professional routes translate; amateur practice largely doesn’t | Weak for amateur practice; moderate for professional routes |
| Social connection (church, WI, clubs, informal networks) | Low to moderate: relational, supportive, socially cohesive | High: belonging, mutual support, shared identity | Low: valued within the community; almost invisible outside it | Very weak: social contribution has no external legibility mechanism | Variable: established institutions have reasonable infrastructure; informal networks have almost none |
| Digital gaming | Low to moderate: skill development, team coordination, strategic thinking | High: genuine progression systems, skill recognition, team belonging, community identity | Low for most players; meaningful within gaming communities | Weak: gaming achievement translates to external R only at elite levels | Strong: engineered and well-resourced; low barrier to participation |
| Social media (commercial platforms) | Low: consumption, reaction, and broadcasting; rarely generative | High (artificial): likes, metrics, reach, algorithmically curated belonging | High (artificial): immediately and continuously visible to wide audience | Engineered: continuous instant translation of activity into visible signal | Very strong: billion-dollar investment; designed to maximise time-on-platform |
| Intellectual and educational pursuit | Low to moderate: knowledge generation, problem solving | Moderate: personal satisfaction, peer recognition in learning communities | Low to moderate: credentials translate; independent learning largely doesn’t | Weak to moderate: formal qualifications translate; informal pursuit doesn’t | Moderate for formal education; weak for independent pursuit |
| Disengagement (isolation, dependency, withdrawal) | None | None, or negative: stigma within peer group | None, or negative: societal stigma | None | None, or negative: systems designed to process rather than support |
Table note 1: C/R profiles reflect broad patterns in contemporary Western industrial societies. The authors recognise that in many cultures and communities, institutions such as the church, extended family care networks, and community organisations remain strong C/R systems with well-developed infrastructure and high external legibility. Where strong C/R systems exist outside employment, they are the model, not the exception.
Table note 2: Gaming and social media are separated because they serve different human needs and have different C/R profiles. Gaming can involve genuine skill development, team coordination, and community belonging that maps reasonably to sports and leisure. Social media originally served genuine social connection needs (the same needs served by the WI, local clubs, and WhatsApp). Commercial social media has been progressively redesigned away from genuine connection toward engagement maximisation for advertising revenue: the underlying human need is the same; the infrastructure has been perverted away from serving it.
Reading the table as a whole, three things become visible.
First, the care row is the most striking omission from most discussions of work and contribution. Care for children, elderly parents, and dependent family members represents some of the highest contribution available to any human being: sustained, skilled, socially essential, and largely irreplaceable. Its internal recognition is real and meaningful. Its external recognition is almost zero. The C/R translation is entirely broken. And the infrastructure is the weakest of any high-contribution row: it falls almost entirely on the individual, usually a woman, with minimal organisational or financial support. The invisibility of care in economic and social accounting is not incidental. It is a structural feature of the industrial C/R system, which only values what passes through the market.
Second, the social media row is the most analytically important for understanding where available time will go by default. Commercial social media platforms do not win because their contribution component is high. It is, in most cases, very low. They win because their recognition architecture is the most engineered of any row in the table. The feedback is immediate, visible, designed to feel significant, and continuous. No other row comes close to this level of recognition infrastructure investment. The competition for available time is not between work and leisure or between employment and volunteering. It is between engineered recognition systems and un-engineered ones. That asymmetry explains most of what happens when people have unstructured time and no deliberate C/R system to receive it.
Third, almost every non-employment, non-digital row shares the same two structural weaknesses: broken R translation and weak infrastructure. The contribution is real. The internal recognition is often genuine and meaningful. But the contribution does not translate into external social legibility, and the infrastructure does not exist to sustain it at scale. These are not separate problems requiring separate solutions. They are the same problem in different contexts: the industrial C/R system captured the infrastructure investment and the external legibility, leaving everything else to manage with minimal support and minimal social visibility.
The baseline problem
Before available time can be directed toward any of the rows above, a prior question needs answering. Someone has to ensure that eight billion people have food, shelter, physical security, and the basic conditions for a functioning life. The industrial C/R system solved this, imperfectly and unequally, through the wage: employment generated income, income secured the basics. If employment declines significantly, the mechanism that currently secures the basics for most people weakens simultaneously.
This paper draws a deliberate boundary here. The baseline question, how societies ensure physical security for all if wage employment can no longer be assumed as the primary distribution mechanism, is a prior and larger question than the C/R design question this paper addresses. Answering it fully is beyond this paper’s scope. But two observations are worth making before moving on.
The first is that the baseline question is not unanswerable. Several credible mechanisms exist and are being seriously discussed. Universal basic income, a regular unconditional payment to all citizens regardless of employment status, has been piloted in Finland, Kenya, Stockton California, and Wales, with broadly positive results on wellbeing and social participation at pilot scale. The unresolved questions are fiscal: at living-wage level, UBI costs roughly 20 to 25 percent of GDP for most developed economies, which requires either significant tax restructuring or trade-offs with other public spending. A robot or automation levy, taxing the productivity gains generated by AI displacement and redistributing them through income support or social investment, is a more targeted instrument that several economists and some policy bodies now take seriously. Norway’s Government Pension Fund, which converted finite petroleum revenues into permanent intergenerational social capital, is the closest existing proof of concept for this kind of deliberate national wealth conversion. The capitalisation question, where does the money come from in a country without oil, remains the central practical difficulty.
The second observation is that none of the C/R systems described in this paper work without the baseline being secured. A person who cannot meet their basic needs cannot direct available time toward care, volunteering, arts, or civic contribution. They direct it toward survival. The C/R framework only becomes relevant once the baseline is in place. This is not an argument for waiting until the baseline is solved before thinking about C/R design. The two problems need to be worked on simultaneously by different actors. But it is an argument for intellectual honesty: the recognition architecture this paper describes is not a substitute for income security. It is what becomes possible once income security exists.
The further reading section at the end of this paper notes the most useful entry points into the baseline debate, including Bregman on UBI, Raworth on economic boundaries, and the Norway sovereign wealth model.
What the table reveals about the AI transition
If AI releases significant available time, and if the baseline problem is at least partially addressed, the question becomes which rows in the table are strong enough to receive that time productively.
The honest answer, reading the table, is that most rows are not currently strong enough. The infrastructure is too weak, the R translation is too broken, and the competition from engineered recognition systems is too well-resourced for most non-employment rows to absorb significant available time without deliberate investment.
Three common root causes appear across almost every weak row.
The first is leadership and convening capacity. Every sustainable contribution system requires someone who identifies the need, builds the group, holds it together through difficulty, and maintains the social fabric of contribution. This is true of a Scout troop, a community garden, a church congregation, a sports club, and a creative collective. Leadership is not a nice-to-have. It is the load-bearing element. When leadership is absent or burned out, the contribution system collapses regardless of how willing the participants are.
The second is supporting infrastructure. Every contribution system requires the organisational overhead of scheduling, communication, record-keeping, compliance, and coordination. In employment, this overhead is carried by the organisation. In almost every other row, it falls on the leader or convenor personally. The cumulative weight of this overhead is the primary reason that good leaders leave contribution systems: not loss of commitment, but loss of capacity to carry everything alone.
The third is the visibility of contribution. For a C/R system to sustain, the contribution must be legible: visible to the contributor, to the group, and ideally to the wider community. Visibility is what makes recognition possible. Without it, the C/R coupling breaks. The contribution happens but the recognition does not follow, and without recognition, the contribution eventually stops.
These three root causes, leadership, infrastructure, and visibility, apply to every weak row in the table. They are not separate problems. They are the same structural deficit showing up in different contexts.
The engineered competitor
Any honest account of what humans will do with available time has to reckon with the fact that one row in the table is not weak at all. Commercial social media platforms have solved the leadership problem (algorithmic curation substitutes for human convening), the infrastructure problem (billion-dollar engineering teams build and maintain it), and the visibility problem (recognition is instant, continuous, and designed to feel significant).
The result is a system which wins by design superiority rather than by satisfying deeper human needs. What people do on commercial social media is rarely generative in the way that care, civic contribution, or creative practice is generative. But the recognition architecture is so well engineered that low-contribution activity feels rewarding in the moment, even when it is not meaningful over time.
Gaming sits differently. The gaming row in the table reflects genuine skill development, team coordination, and community belonging. These are real human needs and gaming addresses them, in some respects, as effectively as sport or creative practice. The concern about gaming is not that it satisfies the same needs as civic contribution: in some respects it does, and more effectively. The concern is whether the needs it satisfies are sufficient for human flourishing across a life, or whether they are a subset that leaves other needs, particularly the need to contribute to something beyond the screen, unmet.
This asymmetry is the central challenge for any alternative C/R system. It is not enough to say that care, volunteering, arts, or civic contribution are more meaningful than scrolling. They may well be. But meaning is not sufficient to direct available time if the recognition feedback from meaningful activity is slow, invisible, and unsupported, while the recognition feedback from commercial digital activity is instant, visible, and relentlessly engineered.
The implication is that strengthening non-digital C/R systems requires deliberate investment in the three root causes: leadership support, infrastructure reduction, and recognition visibility. Not because these activities need to compete with digital on digital’s terms, but because they need to be strong enough to receive available time when people are looking for something more than what screens provide.
The open question
This paper has argued that available time is the fundamental resource, that technology has always expanded it, that AI will expand it further, and that the question of what humans do with available time is primarily a question about C/R systems rather than about income or employment.
It has not argued that the answer is obvious or that the transition will be smooth. The baseline problem is unresolved. The infrastructure of most non-employment C/R systems is genuinely weak. The engineered competitor is genuinely formidable. And the industrial C/R system, though historically recent, has become deeply embedded in how most people understand their own worth and place in the world.
The question this paper leaves open, deliberately, is which C/R systems are strong enough to receive available time if deliberately supported, and what investment in leadership, infrastructure, and visibility would be required to make them competitive.
Two companion papers in this series examine specific cases where that question becomes practical and testable. The first examines the youth volunteering leadership crisis and asks whether AI can specifically address the compliance infrastructure burden that is driving experienced leaders out of the sector. The second examines civic contribution at street level and asks whether the collapse of community action groups can be understood as a C/R failure, and whether structured reasoning before action produces more durable change than the default intervention of addressing the most visible symptom.
Neither paper depends on this one to make its argument. Both are more useful having read it.
The question artificial intelligence is really asking is not: what will people do for work? It is: when the system that has organised human contribution and recognition for a century and a half weakens, what will we build to replace it?
That question deserves a better answer than the one we currently have.
This paper, and the two companion papers that apply its framework to specific cases, are offered as a complete analytical package for anyone with the standing and motivation to take the argument into policy, civic innovation, or voluntary sector reform. The author’s role is as research analyst and background support. The goal is to find one or two individuals or organisations willing to own the public argument and drive it forward. If that description fits you or someone you know, the conversation starts at betterthinking.co.uk.
Discursive appendix: further reading
The argument in this paper draws on a body of work across sociology, anthropology, economic history, and civic research. The following are entry points for readers who want to go deeper. They are offered as reference points, not as precise citations: the argument in this paper is the author’s own synthesis, and these sources explore adjacent territory rather than proving specific claims.
On the baseline problem and what might replace the wage
Kate Raworth, Doughnut Economics (2017). Raworth argues that economic thinking needs to operate within a social foundation (ensuring basic needs are met for all) and an ecological ceiling (not exceeding planetary boundaries). She explicitly argues that GDP and employment are the wrong metrics for human wellbeing, and that care, community, and social contribution are systematically undervalued. A serious and accessible challenge to the assumption that employment-led growth is the only mechanism for securing the basics.
Rutger Bregman, Utopia for Realists (2017). The most accessible case for universal basic income, drawing on historical pilots and economic evidence. Bregman also argues for a fifteen-hour working week, connecting reduced working time directly to human flourishing and community life. Directly relevant to the time-first argument in this paper. The UBI chapters are the most evidenced and most relevant sections.
On AI, technology, and labour market disruption
Brynjolfsson and McAfee, The Second Machine Age (2014). MIT economists making the serious academic case for why digital technology, and by extension AI, produces genuine and sustained labour market disruption rather than the temporary displacement that previous technological transitions produced. Cautiously optimistic about long-run outcomes but honest about short-run pain and distributional consequences. The intellectual foundation for taking AI displacement seriously rather than dismissing it as hype.
On the invisibility of care and community contribution
The Care Collective, The Care Manifesto (2020). A short, accessible book arguing that care, broadly defined (for people, communities, and the planet), should be the organising principle of social and economic life. Directly addresses the invisibility of care in economic and social accounting. The diagnosis that care’s devaluation is a structural feature of market economies rather than an accident supports the care row analysis in this paper’s table.
On social connection, loneliness, and the failure of community
Noreena Hertz, The Lonely Century (2020). Examines the epidemic of loneliness in modern societies and its economic, political, and social consequences. Argues that withdrawal from community life is a structural outcome of how modern economies are designed, not a personal failing. Directly relevant to the social connection row and the argument about what humans actually need from their time. Includes analysis of digital platforms providing synthetic connection while deepening actual isolation.
Robert Putnam, Bowling Alone (2000). The foundational empirical work on the decline of social capital in the United States. Tracks the collapse of civic participation, community organisation, and informal social connection across the second half of the twentieth century. The evidence base for the claim that community contribution systems have been weakening for decades, independently of AI disruption.
On the history of work and identity
E.P. Thompson, The Making of the English Working Class (1963). Thompson’s landmark history documents how the identity, culture, and self-understanding of the English working class was actively constructed during industrialisation rather than naturally produced by it. The foundational text for the argument that the connection between work and identity is historically made rather than inevitable.
David Graeber, Bullshit Jobs (2018). Argues that a significant proportion of modern jobs are experienced by the people doing them as pointless, and that societies have organised around the moral imperative of employment rather than around the actual creation of value. Methodologically contested but intellectually stimulating. Read with appropriate scepticism.
NCVO, Time Well Spent (2023). The National Council for Voluntary Organisations’ major survey of volunteering in the UK. The primary quantitative source for evidence on the decline of the organiser layer and the rise of stress as a reason for leaving voluntary roles. Relevant to the companion papers in this series.
On sovereign wealth and social purpose
Norges Bank Investment Management, The Government Pension Fund Global (established 1990, ongoing reporting at nbim.no). Norway’s sovereign wealth fund is the world’s largest, holding over $2 trillion in assets accumulated from petroleum revenues. It is the most cited example of a society making a deliberate democratic choice to convert finite natural resource wealth into permanent intergenerational social capital, governed by strict ethical investment criteria including exclusion of tobacco, weapons, and companies with serious human rights or environmental records. Relevant to the baseline problem discussion in this paper as a proof of concept that national wealth can be managed for long-term social purpose rather than short-term consumption. The honest caveat: Norway’s model works because it had oil revenues to capitalise it. The question of how a social wealth fund might be capitalised in a post-AI-displacement economy (through a technology sector levy, robot tax, or transfer of public assets) remains genuinely open. The related concept of a social wealth fund, which directs returns specifically toward social welfare purposes rather than the state alone, is explored further in the Democracy Policy Network’s Social Wealth Funds framework at democracypolicy.network.
Two companion papers apply this framework to specific practical cases. They are written for practitioners in those spaces and are available at betterthinking.co.uk.