Artificial intelligence (AI) is profoundly transforming the world of work. Long perceived as a simple support tool, it is now becoming a central actor in intellectual production. Capable of learning, analysing, and producing at unparalleled speed, AI is calling into question the traditional value of knowledge and many so-called “intellectual” professions.
According to Satya Nadella, CEO of Microsoft, we are entering an era in which knowledge has become widely accessible, instantaneous, and almost free. What now makes the difference is no longer what one knows, but the ability to ask the right questions, to steer tools effectively, and to exercise critical judgement.
Three major families of professions are particularly affected:
Diagnostic and certification professions (doctors, chartered accountants, auditors)
Synthesis and advisory professions (lawyers, consultants, analysts)
Intellectual production professions (journalists, artists, developers)
In medicine, AI excels at pattern recognition from images, symptoms, or biological data. It is already used for diagnostic support, patient triage, automated report generation, and medical imaging. The OECD estimates that around 40% of medical time is devoted to administrative tasks, making this an ideal area for automation.
Financial and legal professions are undergoing a similar evolution. Transaction analysis, fraud detection, financial modelling, and legal research are now largely automatable. Studies conducted by Goldman Sachs indicate that approximately 300 million full-time equivalent jobs worldwide could be exposed to automation by AI.
In strategic consulting, AI is capable of instantly synthesising decades of reports, generating complex economic scenarios, and producing structured recommendations in real time. This mechanically reduces the value of services based solely on the collection and synthesis of existing information.
Software development is not immune to this transformation. Writing code is gradually becoming a commodity, while real value is shifting towards architecture, tool orchestration, and the supervision of intelligent systems. The emergence of so-called “agent” AI marks a new stage: these systems can analyse a situation, make decisions, execute tasks, verify results, and self-correct, at an extremely low marginal cost.
However, despite these spectacular advances, AI still presents significant limitations. Several studies show that it can produce between 15% and 50% errors, fabrications, or hallucinations, particularly in legal and factual domains. Human judgement, ethics, empathy, and relational capacity remain irreplaceable.
Paradoxically, skilled manual professions are proving more resilient than many intellectual professions. Physical robotics remains costly, difficult to deploy, and poorly adaptable to non-standardised environments. By contrast, once developed, a software-based AI can be replicated at near-zero cost, leading to sustained pressure on the economic value of intellectual labour.
The world of work is not disappearing; it is transforming. As with previous major technological revolutions, the central challenge will be adaptation. Learning to work with intelligent assistants, developing critical thinking, and strengthening human skills will become essential conditions for remaining relevant by 2030.
References
MoneyRadar and FutureRadar – “L’IA vient de tuer tous les métiers !”
OECD – Artificial Intelligence in Society (2019)
Goldman Sachs – The Potentially Large Effects of Artificial Intelligence on Economic Growth (2023)
McKinsey Global Institute – Generative AI and the Future of Work (2023)
Microsoft – Work Trend Index Report (2023)
Harvard Business Review – How Generative AI Changes Productivity (2023)