Resumes filtered by AI, ghost jobs, poverty-level wages, and qualifications demanded for roles that don’t need them: the job market has become a shared fiction
There is a pantomime that repeats itself every day, millions of times. On one side, a candidate carefully refines their resume, choosing every word and fixing every comma. On the other hand, an algorithm that discards it in 0.3 seconds because they wrote “I dealt with daily challenges at work” instead of “problem-solving.”
Neither side truly knows what the other is looking for. And often, deep down, neither does the company itself.
The interview as theatre
Job interviews are losing any connection to reality. Recruiters ask questions to which they already know the expected answers; candidates provide those answers, knowing they are performing; and everyone pretends that the goal is to find the right person for the right role. But the real objective, almost always, is something else: profit—for the company, for the recruitment agency, and often for both.
If we add that, even before the interview, the resume is filtered by automated systems—the so-called Applicant Tracking Systems—then the disconnect from reality becomes enormous. A survey of one thousand American job seekers found that the vast majority of applications are rejected by software before any human being ever sees them. Candidates no longer optimize their professional profile: they learn to think like an algorithm. And so the resume stops being a portrait of a person and becomes a document engineered to pass a filter.
Ghost jobs and the illusion of abundance
But there is an even more fundamental problem: many job listings do not correspond to real positions. The so-called ghost jobs—postings published with no real intention of hiring—serve to collect data, monitor the market, and keep a pipeline of candidates open “just in case.” The result is a false perception: it looks like there is far more work available than actually exists.
And even when a job does exist, we must ask: Is it really a job? A position that does not allow for basic self-sufficiency—that does not cover rent, food, or essential expenses—is not a job: it is subsistence disguised as employment. If we removed all underpaid and precarious positions from the equation, the amount of real employment would shrink dramatically.
The poverty trap that blocks access to work
There is also a contradiction that is rarely named: to look for work, you need to already have resources. You need a car for the interview and the job, a computer to apply online, a smartphone to be reachable, and a stable address. Those in financial difficulty find themselves in a paradoxical trap: no job because they lack the means, and no means because they lack a job.
In recent decades, no structural measures have been taken to adapt work to real life: shorter working hours, higher minimum wages, and policies to combat burnout. The preference has been to let the market self-regulate, and the market has chosen to offload the cost onto workers.
The elephant in the room: artificial intelligence and robotics
Yet the biggest problem remains largely underestimated. Artificial intelligence and robotics are already reshaping the job market — not in some distant future, but right now. The Boston Consulting Group has estimated that in the next five years, between 10% and 15% of jobs in the United States could disappear, affecting between 17 and 25 million people.
Elon Musk has recently proposed a “Universal High Income“—a universal income aimed not at bare survival but at genuine well-being—as a response to the wave of technological unemployment. The idea is that AI and robotics will produce goods and services in such abundance that the cost of living will collapse, making even a modest monthly payment enough to live well. It is a compelling vision, but one that assumes a fair redistribution of gains—something that, as the economic history of the past fifty years reminds us, does not happen automatically.
More cautious researchers argue that the priority should be professional retraining, not subsidies: giving people the tools to adapt, not just a check to survive on. It is an understandable position, but one that conceals several serious flaws.
The first is that not everyone can retrain for any job. Skills are not infinitely transferable: a fifty-year-old factory worker, someone with cognitive difficulties, a person who never had access to a solid education—they cannot simply “upskill” and become data analysts or IT technicians. There is a real human diversity that the job market continues to ignore.
The second is time. Training takes months, often years. In the meantime, people need to eat, pay rent, and keep their families afloat. There is no official pause between “your job has been automated” and “you are ready for the next one.”
The third, perhaps the most overlooked, is that we cannot ask people to reinvent themselves continuously to chase a market that changes faster than any human being can adapt. Work is not just a source of income: it is identity, routine, and social connection. Asking someone to start over every five years is not flexibility — it is precarity dressed up as opportunity.
Qualifications as an ideological filter
Meanwhile, requirements for qualifications keep multiplying for positions that would not need them at all. Manual or routine jobs that require a university degree. Entry-level positions that demand years of experience. Simple tasks for which six different software tools must be mastered. This is not a genuine necessity: it is an additional filter, often unconscious, that excludes those who never had access to certain educational paths—and feeds a narrative that blames the individual: “If you don’t have a job, it’s because you don’t have the right skills.”
It is a convenient narrative because it shifts responsibility onto the individual and relieves institutions of the task of building a fairer system.
Stop pretending
The job market as we know it rests on a series of shared fictions: that the resume reflects the person, that the interview evaluates talent, that job listings correspond to real work, and that anyone who wants a job can get one. Dismantling these fictions is not nihilism: it is the first step toward building something more honest.
Because as long as we keep pretending the system works, we will have no incentive to change it. And in the meantime, algorithms keep rejecting perfectly good resumes, and millions of people keep believing the problem is them.
