AI will not only revolutionize medicine. It imposes a predictive welfare state, capable of anticipating illnesses rather than repairing them.
Great technological revolutions do not destroy institutions. They suddenly make them insufficient. The steam engine not only transformed manufacturing; it led to the invention of labor law. Electricity not only lit up cities; it disrupted their organization. The Internet has not only connected individuals; it has profoundly transformed the world economy.
L’artificial intelligence belongs to this category of innovations. Yet we continue to view it as one technology among others. We debate its performance, its algorithms, its risks or its promises. We wonder if it will replace doctors, radiologists or biologists. We are wondering about the place it will take in businesses or in administrations.
These questions are legitimate. But they are secondary. The real revolution in artificial intelligence will not be digital. It will be institutional. It could transform the very philosophy of our welfare state.
In 1945, France invented Social Security. In a country ravaged by war, the promise was immense: no one would have to give up treatment because they were poor. For the first time, national solidarity took charge of the health risk. Illness ceased to be an individual fatality and became a collective responsibility. Thirteen years later, a second revolution, more discreet but just as decisive, saw the light of day. The Debré ordinance created university hospital centers. By bringing together care, teaching and research, it gave birth to a model that would make France one of the great medical nations of the 20th century. These two reforms built the modern health welfare state. They have made it possible to support the increase in life expectancy, therapeutic progress, the revolution in antibiotics, imaging, biology and surgery. But institutions rarely age because they fail. They are getting older because the world is changing.
The world for which Social Security and university hospitals were designed is no longer ours. The major diseases of our time are no longer those of the post-war period. These are chronic diseases, cancers, diabetes, cardiovascular diseases, neurodegenerative diseases, mental health disorders. They do not occur suddenly. They develop slowly, sometimes for years, before being diagnosed. Our system, however, continues to be organized around a logic inherited from the last century: wait for the disease to appear, then treat it. This logic constituted considerable progress. But today it is reaching its limits. We have built a state capable of repair. The 21st century will have to build a State capable of anticipation.
This is precisely where artificial intelligence introduces a historic break. For the first time, we are able to link together considerable quantities of previously dispersed data: biological results, medical imaging, genetics, connected objects, medical records, environment, diet, physical activity or lifestyle habits. Taken in isolation, this information tells little. Together, they draw a trajectory. They make it possible to identify weak signals, invisible to the human eye, and to anticipate the appearance of certain diseases or their complications. The objective is no longer just to diagnose a heart attack. It becomes to avoid it. It’s no longer just about finding cancer. It becomes possible to intervene before it becomes a serious illness. The challenge is no longer just to treat diabetes. It becomes to delay, or even prevent, its appearance.
This transformation goes far beyond medicine. It calls into question the very economics of our health system. For several decades, we have mainly paid for medical procedures, consultations, hospitalizations or technical interventions. Tomorrow, the real value will be created elsewhere. In complications avoided. In hospitalizations that will never take place. In years of life gained in good health. In other words, we will gradually move from a care economy to a health economy. This change calls for a new stage in our social history.
I call a predictive welfare state a model in which national solidarity is no longer limited to repairing illness once declared, but mobilizes data, artificial intelligence, prevention and territorial organization in order to anticipate risks, delay the appearance of pathologies and sustainably preserve everyone’s health. This development does not constitute a break with the legacy of 1945. It is its accomplishment. Social Security guaranteed access to care. The predictive welfare state will guarantee access to anticipation. It will not replace the hospital. It will reduce the pressure placed on him. It will not replace doctors. It will increase their ability to intervene sooner. He will not call into question national solidarity. It will give him a new horizon.
But it would be illusory to believe that this revolution comes down to software. Behind each algorithm lie digital infrastructures, data centers, computing capacities, electrical networks, semiconductors, interoperability standards and industrial policy. Artificial intelligence is less a software revolution than a revolution of sovereignty. Health thus becomes a medical, economic, industrial and geopolitical issue. France has considerable assets: excellent research, a recognized hospital system, industrial capacity, largely carbon-free energy and a dynamic network of innovative companies. These forces still need to converge around the same ambition. For decades, we have measured the performance of our healthcare system by its ability to repair disease. Tomorrow we will evaluate it on its ability to prevent them from appearing. This is the real revolution that is beginning.
History will perhaps remember that artificial intelligence was not the greatest digital revolution of the 21st century. It will have been the moment when democracies stopped thinking of health as reparation for illness and began to think of it as a collective heritage to be preserved. In 1945, the Republic invented health solidarity. In 1958, she invented the institutions of medical excellence. It is now up to it to invent a third stage of its history: that of the predictive welfare state.
Because progress is no longer just about living longer. It is now about living longer in good health. And this is probably the greatest promise that artificial intelligence can offer our Republic.
