The taboo of our time
There are questions no society likes to ask in public. "What is work for?" is one of them. For 200 years, the subject has been carefully avoided. Work is taken as a given: we work to live. It is the foundation of the entire industrial society, the cement of the social contract, the thing that structures our days, our weeks, our lives.
And then suddenly, within a few years, that certainty starts to crumble. Not through anyone's intention. But because machines are beginning to do in our place what we did to earn a living. Not just repetitive manual tasks, as in the industrial revolution. This time, it is intellectual professions that are tipping over: journalists, lawyers, graphic designers, translators, as we documented in our article on graphic designers and our analysis of journalism.
The question rises, muffled. You can feel it in professional conversations, in the anxieties of thirty-somethings, in the eyes of students wondering what the point is of five years of study. If tomorrow AI does our work better and more cheaply, what do we do?
This article tries to address that question honestly. Not by speculating about distant futures. By asking precise questions about what work really means.
Work has not always existed as we know it
To think about work, we must remember that it has not always taken its current form. The French word "travail" comes from the Latin tripalium, an instrument of torture with three stakes. For centuries, working meant suffering, toiling, carrying out a painful and necessary activity. Work was not glorified, it was endured.
In ancient Greece, manual labour was reserved for slaves. Free citizens devoted themselves to politics, philosophy, art, sport. Aristotle considered that only a man free from the necessity of work could truly think. It is a perspective that seems shocking today, but it structured Greek civilisation for centuries.
In the Middle Ages, work was both material necessity and spiritual fulfilment. Monasteries practised the Benedictine rule "Ora et labora" (pray and work). But the idea that everyone had to work 40 hours a week did not exist. Most people worked when they had to (sowing, harvesting), rested when they could, and alternated with religious festivals that sometimes occupied up to 100 days a year.
It was industrialisation in the 19th century that invented work as we know it: factories, fixed hours, clocking in, monthly wages, measured productivity. This model is less than 200 years old. Before it, humanity lived differently for tens of thousands of years.
This historical perspective changes everything. Work as we organise it is not an eternal inevitability. It is a historical construction, suited to a particular stage of economic development. If that stage changes, the construction can change too.
The term "métro-boulot-dodo" entered the French vocabulary in 1968. Before the Trente Glorieuses (the period of economic prosperity in France from 1945 to 1973), few people talked about their lives in those terms. For less than a century, French humanity organised its life around a rhythm imposed by factories, offices and transport. If you can do it for 80 years, you can also organise it differently. That does not mean it will be easy. But it is not impossible either: the proof is that it has not always existed.
Why work really matters to us
To understand what is happening with AI, we need to distinguish the different functions that work fulfils in a human life. Work is not just a way of earning money. It is also:
A structure for time. Without work, a person's day is a hole. Many people who have experienced long-term unemployment, an abrupt retirement or a forced convalescence describe it this way: having no reason to get up is psychologically very hard. Work provides a framework, hours, a before and an after.
A social status. "What do you do for a living?" is one of the first questions asked when getting to know someone. Your job places you in the social hierarchy, gives you an identity, indicates your cultural references. Without that answer, many people feel lost. A few societies (notably certain religious communities) have tried to think about identity differently, but they remain very much in the minority.
Social relationships. The majority of adult friendships are formed at work. That is where you meet colleagues, share projects, go through difficulties together. Without work, many adults end up isolated. The phenomenon is widespread and well documented: the happiest retirees are often those who maintain a collective commitment (associations, volunteering, sport).
A sense of usefulness. Feeling useful to something or someone is a fundamental psychological need. Work, even unrewarding work, often provides that feeling: you have produced something, you have helped someone, you have moved a project forward. Many people experiencing burnout say that they first suffered when that feeling disappeared.
A cognitive challenge. Our brain needs challenges. Without intellectual effort, it atrophies. Work often offers (not always, admittedly) that challenge: solving problems, learning skills, progressing in a trade. Studies on brain ageing show that people who remain professionally active for longer preserve their cognitive abilities better.
Here is the crucial point: AI could free us from work as a "means of earning a living", without necessarily replacing all those other functions. And that is precisely where it gets complicated.
The experiments under way around the world
Several societies are already trying to answer the question. Not out of ideology, but out of pragmatism: what happens if you give people a guaranteed income, with no work requirement?
Finland, 2017-2018. The Finnish government paid 560 euros a month to 2,000 randomly selected unemployed people, for 2 years, with no conditions. Published results: recipients did not work significantly less. But they reported markedly higher psychological well-being (less stress, better mental health, more trust in institutions). The experiment was not continued, but it left its mark on the European debate.
Kenya, since 2017. The American NGO GiveDirectly pays around 22 dollars a month to several thousand Kenyans across entire villages, for 12 years. It is probably the longest basic income experiment in the world. Published results show a significant improvement in nutrition, children's education, and micro-businesses created. With no negative effect on motivation to work.
The United Arab Emirates, since 2010. Emirati citizens receive massive state support (free housing, free education, free healthcare, various allowances), funded by oil revenues. Many do not work at all, or only part-time. The model has raised questions about motivation and identity, but has also enabled a high quality of life and remarkable social stability in the region.
Alaska, since 1982. All Alaska residents receive their share of the state's oil revenues annually, around 1,600 dollars a year. Not enough to live on, but enough to make a difference. It is the oldest partial universal income in the world, and it is completely accepted politically.
Several other experiments are under way: in the United States (Stockton, Hudson, Compton), in Canada (a pilot experiment in Ontario), in India, in Brazil. None is conclusive on a large scale, but all document something important: receiving an income with no work requirement does not push the majority of people to stop all activity. Most continue to work, but often differently: more small businesses, more community involvement, more studying, more care for relatives.
The three great visions of a future without work
If AI really forces us to rethink our relationship with work, three great visions are competing in the current philosophical and economic debate.
Vision 1: mass dystopia. A few individuals own the AIs and capture all the value created. The vast mass of the population becomes economically superfluous. To prevent revolts, a barely liveable minimum income is introduced, screens to distract, AI companions to fill the loneliness. A society straight out of Wall-E or Brave New World (Aldous Huxley's novel, in which humanity is soothed by distractions and drugs). This vision is probably the one that worries serious philosophers most. Yuval Noah Harari (Israeli historian, author of Sapiens) regularly cites it as a real risk.
Vision 2: the creative renaissance. Freed from survival work, humans finally devote themselves to what they are passionate about. Art, philosophy, fundamental science, sport, meditation, parenting, social ties, care for others and for nature. This is the vision Aristotle would probably have recognised: humans free from the necessity of work can finally develop what is properly human. As we explored in our article on AI discoveries, science itself could be democratised on a large scale.
Vision 3: gradual transformation. Work does not disappear, it changes. Some jobs disappear, others appear. Humans work less (moving to 30, then 25, then 20 hours a week), better (on more creative, more relational, more human tasks), and with more meaning. This is the middle vision, probably the most credible for the next 20 years. Several economists (Daniel Cohen, Thomas Piketty) argue along these lines.
The three are not mutually exclusive: we could see a combination of all three, with different dynamics depending on countries, social classes and generations. Vision 1 is likely to dominate in highly unequal societies (the United States?), vision 2 in highly educated and mutually supportive societies (Scandinavia?), vision 3 everywhere on average.
The choice between these three visions is not technical. It is profoundly political. Technology leads us nowhere on its own: it opens up possibilities, and it is up to societies to choose which ones to activate. Universal income, taxation of AI companies, lifelong training, sharing out working time: these are active political choices. As we explored in our article on the limits of AI regulation, these decisions must be taken within the next 5 to 10 years, or economic actors will impose their own vision by default.
What can be done concretely, individually
Beyond the collective debate, several things are already actionable individually, and deserve to be named.
Diversify your sources of meaning. If your entire identity rests on your job, you are vulnerable. If you are also a parent, a friend, a neighbour, an amateur athlete, a music lover, a volunteer, you come through transformations far better. This is probably the most valuable advice for anyone who wants to get through the next 20 years without a major identity crisis.
Cultivate relationships independent of work. If all your friends are colleagues, your social circle collapses along with your job. Investing in relationships that go beyond work (associations, sport, neighbourhood, extended family) builds social resilience.
Learning to enjoy yourself without buying anything. Many adults have unlearned how to entertain themselves without spending. Walking, reading, cooking, gardening, playing cards, watching a sunset: all of it is free, and all of it does you good. If we imagine a future where money is less central, it is better to know how to live with less, by choice rather than by necessity.
Investing in your body and your health. When you work a lot, you often neglect your body. The more time you free up, the more you can look after yourself: exercise, diet, sleep, medical check-ups. The studies are clear: physical health is one of the biggest determinants of subjective happiness. And it is one of the areas where AI and medicine are advancing fastest, as we explored in our article on the medicine of the future.
Learning properly human skills. The more AI progresses, the more valuable the skills that machines do not do well become: deep listening, original creativity, inspiring leadership, passing on knowledge, care. If you invest in these skills now, you will be valuable in 20 years, whatever happens to the job market.
The underlying philosophical question
At bottom, the question AI forces us to ask is the oldest in philosophy: what is a fulfilled human life?
For millennia, several traditions have offered answers. Ataraxia for the Stoics (a state of tranquillity of the soul, free from the passions). Eudaimonia for Aristotle (happiness as the fulfilment of human potential). Salvation for the monotheistic religions. Emancipation for Marx. Consciousness for the Eastern traditions. Authenticity for the existentialists.
None of these answers is centred on work in the modern sense of the term. Work, in all these traditions, is a means, not an end. It is instrumentalised for something more important: thought, love, commitment, transcendence.
If AI genuinely frees humans from part of that instrumental necessity, then it could, paradoxically, make us freer to aim at the true ends. Provided we still know how to name them.
This is perhaps the greatest challenge of our era: learning again what we would want to do if we could choose. For those who were raised to aim for a career, to earn a living, to "succeed", this question is uncomfortable. It is even more so for those who have never had the luxury of asking it.
But it is coming. Slowly, unevenly, individually. It is coming for that 45-year-old graphic designer whose job is disappearing. For that 30-year-old journalist watching colleagues being made redundant. For that 18-year-old student wondering what the point of their studies is. For that 70-year-old retiree who has finally found time again. The question is rising from everywhere.
Answering it individually is probably the most important work anyone can do in 2026. Answering it collectively, through our societies, our policies, our institutions, is probably the great challenge of the next 20 years.
To go further, you can read our analysis of the 1, 5 and 30-year scenarios, which places these issues in a longer timeframe, and our article on what we pass on to our children, which asks a similar question for the next generation. The era now opening is probably the hardest to live through since the industrial revolution. It is also the richest in possibilities. It is up to each of us to make of it what we can.