AI does not destroy jobs: it transforms work. Tasks become automated, skills shift, profiles gain value, and the tension now focuses on adaptation.
Announcements of job cuts in tech have been coming one after the other for two years. Each wave seems more spectacular than the previous one, and the same argument always comes up: artificial intelligence would make it possible to do better, faster, with less. This discourse, often presented as an inevitability, masks a more complex reality. Because AI does not destroy professions: it redefines them, moves them and, sometimes, revalues them.
A silent recomposition
Behind the figures, the job market is undergoing a silent restructuring. Automation does not eliminate work: it changes its nature. Artificial intelligence takes care of repetitive tasks, documentation, technical writing, support and even software integration. But at the same time, it gives rise to new needs: model supervision, data validation, intelligent interface design, ethical and security risk management, and more importantly, scaling.
In industry as in services, uses are increasing. According to the latest survey by the Society of Engineers and Scientists of France (IESF), 73% of engineers already use AI in their work. Among them, 37% use it for design, 34% for data analysis, and a large majority expect it especially for the automation of complex tasks. In other words, artificial intelligence is now a work companion, not a substitute.
Growing pressure on skills
In France, the market for engineers is at almost full employment. There are more than 1.28 million working professionals, with an unemployment rate of less than 3% between 30 and 54 years old. Industry remains the main employer, and digital sectors – electronics, telecommunications, cybersecurity, software engineering – account for almost a fifth of young graduates. Yet companies continue to talk about a “talent shortage”.
This paradox is explained by a structural tension. Jobs are evolving faster than training, and the needs of companies now go beyond traditional frameworks. 65% of recruiters say they encounter difficulties in hiring, mainly due to a lack of suitable profiles. Design engineers, project managers and technical experts are the most sought after – and the rarest. Result: salaries increase by an average of 4.7% in one year, reaching nearly €68,000 gross per year. But this dynamism hides a growing misalignment between market needs and the structure of professions.
A polarization of the job market
The job market is polarizing. On the one hand, very senior profiles, capable of understanding, guiding and piloting AI. On the other, more standardized, automated, outsourced or offshored positions. Between the two, the intermediate zone, that of qualified execution functions, is gradually reduced.
This polarization is not only technological: it is economic and organizational. Companies are looking to combine immediate productivity gains with an upgrade in supervisory functions. They increasingly value hybrid profiles, capable of navigating between technique and management, between engineering and strategy. It is a logical development, but it requires an in-depth transformation of professional career paths.
The turning point in public organizations
This phenomenon is not limited to the private sector. The State itself is beginning its transformation. The Interministerial Digital Directorate (Dinum) recently launched, with Mistral AI, an experiment with an artificial intelligence assistant intended for 10,000 public agents. The objective is not to replace agents, but to help them work better: assist in writing notes, simplify documentary research, streamline certain procedures.
This initiative illustrates a conviction shared by many players: AI should not be a replacement tool, but a lever for efficiency. It can improve the quality of public service, reduce the administrative burden and secure uses, provided it is supervised and understood. It is a pragmatic approach, which also deserves to inspire the private sector.
Towards a skills economy
The current transformation is no longer based on the logic of the position, but on that of competence. It is no longer the title that counts, but the ability to adapt, to understand augmented systems and to derive value from them. Companies that have understood this are now recruiting profiles capable of reasoning with the machine, not just operating it. They are looking for engineers who know how to interpret a model, identify a bias, and communicate with an algorithm.
This shift does not herald the end of human work. On the contrary, it opens a new balance between the power of technology and the singularity of human reasoning. Because the more AI advances, the more the value of judgment, creativity and discernment increases.
Think about work differently
The waves of layoffs observed in tech do not reflect a collapse of the market, but a poorly understood transition towards an economy of complementarity. In San Francisco, this ambiguity is visible: while Marc Benioff affirms that Salesforce will no longer hire software engineers, the company continues to advertise dozens of positions – proof that jobs do not disappear, they move.
This shift, accelerated by AI, requires an active skills development policy and a more detailed reading of what technology really transforms in work, especially since the real challenge of the coming years will be that of a shortage of assets linked to aging and non-demographic renewal. Artificial intelligence does not signal the end of professions: it requires us to reinvent the value of work, no longer as a series of tasks, but as a chain of skills to be developed. In this new world, the challenge is not to “adapt to the machine”, but to build with it an economy that is more fluid, more agile, and above all more human.
