Who Decided the Job
The articles appear with reliable regularity. AI will take our jobs. These are the jobs AI cannot take. Here is how to make yourself AI-proof.
The genre is consistent across publications, platforms, and years: the threat is named, the vulnerable occupations are listed, the surviving occupations are identified, and the reader is directed toward whatever educational or professional reorientation will preserve their position in the labour market.
The articles assume, without examination, that the labour market is the appropriate frame through which to understand what is happening, and that the preservation of employment within it is the appropriate goal.
The assumption is so thoroughly embedded in the genre that questioning it feels like changing the subject. It is not changing the subject. It is identifying the subject that the genre has been avoiding.
Who decided that having a job was the primary function of being human?
The question is not rhetorical, and its answer is historically locatable. The identification of human worth with employment is not an ancient feature of the human condition. It is a product of a specific set of institutional changes that occurred between the sixteenth and nineteenth centuries in Western Europe and were subsequently exported, through colonisation and commercial expansion, across much of the world.
Before the enclosure of the commons, before the industrial organisation of labour, before the wage dependency that industrial employment created, most people performed tasks rather than holding jobs. The distinction is not merely semantic. A task is a specific piece of work with a beginning and an end. A job is an institutionalised role with a persistent identity, a regularised schedule, a wage that arrives regardless of whether any specific task has been completed, and—crucially—a social position that derives from the role rather than from the specific work performed.
The peasant who ploughed, harvested, gathered fuel from the common, tended animals, and repaired buildings did not have a job. They performed the seasonal tasks that their situation required. Their social identity derived from their community membership, their kinship relations, and their standing within the rituals and obligations of the local social order. It did not derive from their occupation in the modern sense.
The industrial transformation that produced the job produced it through a set of mechanisms that the earlier essays in this collection have traced in other contexts: the enclosure of the commons removed the distributed resource base that had allowed task-based survival; the factory system imposed the regularised schedule and the wage dependency that created the modern employment relationship; and the administrative systems of the emerging state required the population to be legible in ways that the task-based existence was not. The job was the institutional form that made the person legible to the emerging administrative and commercial systems. It was not the natural expression of what human beings are for.
The fusion that capitalism performed—the coupling of labour, moral worth, citizenship, and consumption into a single structure—turned the absence of employment from a practical inconvenience into an existential condition. The unemployed person is not, within this settlement, merely a person without current work. They are a person without recognised social legitimacy—without the structured participation in collective life that employment provides, without the moral validation that productivity supplies, without the consumer identity that wages enable. The welfare essays in this collection traced the compliance apparatus that the state built around the unemployed person, requiring them to demonstrate ongoing effort toward re-employment as a condition of support. The compliance requirement is not only fiscal management. It is the institutional expression of the moral weight that the job-as-identity fusion placed on employment.
This is why the AI job displacement discourse so frequently sounds existential rather than practical. People are not primarily afraid of losing income, though income loss is serious and concrete. They are afraid of losing the social identity, the moral validation, the sense of structured usefulness, and the recognised participation in collective life that employment provides within the current settlement. The job is not only how they earn. It is how they exist within the social order. The threat that AI will take the job is heard, within this settlement, as the threat that AI will remove the grounds on which their social existence is recognised.
The AI-proof jobs genre reveals its assumptions most clearly in what it selects as surviving occupations. The lists tend to cluster around activities that require physical presence—the surgeon, the plumber, the childcare worker—or around activities that require emotional attunement—the therapist, the nurse, the social worker. The selection is interesting not for what it includes but for the framework it uses to select. It asks: which tasks cannot be automated? It does not ask: which human activities are worth doing regardless of whether they can be automated? These are different questions, and the second question does not get asked because it would require stepping outside the labour market frame that the genre presupposes.
The labour market frame cannot ask whether an activity is worth doing unless it is also asking whether someone would pay for it. Care for elderly parents is among the most socially significant activities a person can perform. It does not reliably generate income. It therefore does not appear on most AI-proof jobs lists, because the lists are organised around market employability rather than social value. The lists are maps of the labour market, not maps of human flourishing, and the two maps do not have the same shape.
The question of who benefits from the preservation of compulsory employment is one that the AI discourse almost never asks, and its absence from the genre is itself informative. The compulsory character of employment—the fact that most people must work to survive, because the alternative mechanisms for accessing the resources required for survival are either inadequate or stigmatised—is the condition that makes the labour supply available to employers at the price employers are willing to pay. The universality of the compulsion maintains the labour market’s structure. If a significant portion of the population had access to adequate support without employment—through a universal basic income, through genuinely sufficient welfare provision, through the distributive mechanisms that several of the preceding essays examined—the labour supply would change in ways that would alter the employment relationship’s terms.
The employer who relies on the compulsion of necessity to maintain a workforce willing to accept available wages and conditions has an interest in the preservation of compulsory employment that is distinct from any concern about the employees’ wellbeing. The political systems that maintain inadequate welfare provision, that stigmatise non-employment, that treat labour market participation as the primary criterion for citizenship—these arrangements serve the employer’s interest as much as any other interest, and they do so through the moral and institutional settlement that identifies human worth with employment.
The AI that automates a task within this settlement is experienced as a threat because it removes a position in the labour market without altering the settlement that makes labour market participation necessary for dignified existence. The threat is real within the settlement’s terms. The settlement’s terms are not the only possible terms.
For most of human evolutionary history—a period measured in hundreds of thousands of years before the agricultural revolution and millions before the industrial one—human beings performed tasks within kinship and band structures that organised survival, care, ritual, and the transmission of knowledge across generations. The tasks were varied, seasonal, and not organised around persistent occupational identities. The hunter-gatherer did not have a job. They had skills, obligations, relationships, and the seasonal round of activities that their environment and their community required.
This is not a nostalgic claim. The conditions of hunter-gatherer existence were not uniformly good, and the transition to agriculture and eventually to industrial organisation produced genuine improvements in many dimensions of human life alongside genuine losses. The observation is more limited: that the identification of human worth with regularised employment in a wage-dependent labour market is a historically recent and geographically specific settlement, not a permanent feature of what human beings are or what they require to live well. The species managed for a very long time without jobs. The question of what it might look like to manage without them again—with the productive capacity that automation is genuinely developing—is not obviously less sensible than the question of how to preserve jobs within a settlement that automation is already disrupting.
The AI that takes the job is not taking the person.
The person existed before the job.
The job was an institutional form that organised the person’s time, identity, and survival within a specific settlement.
The settlement is disrupting itself.
The disruption is being discussed primarily as a question about jobs.
The question about jobs is a smaller question than the one the disruption has opened.
The larger question is what the settlement was for, and whether the disruption offers any possibility of organising things differently.
The possibility is not guaranteed.
Neither is it foreclosed.
The genre has not noticed it yet.
You are here: Department of Ontologenodynamology Research > Ingress