What decisions do we want to delegate to machines, at the risk of transforming our societies and weakening our capacities for judgment and expression?
Visible debates, silent transformations
Each era is absorbed by its own controversies. The debates that dominate public space always give the feeling of being the most urgent and the most decisive. However, history teaches us that the most profound transformations are not always the ones we talk about the most. They often progress on the periphery of political debate, until their effects become irreversible.
We may be in such a period today.
For several years, public debate has been largely structured around societal battles: gender equality, inclusion, representativeness, diversity. These questions have their importance and their legitimacy. They reflect an aspiration for greater justice and have enabled real progress.
But while these debates occupy most of the collective attention, a much more profound transformation is unfolding: the gradual irruption of artificial intelligence into almost all dimensions of human life.
An anthropological transformation
Artificial intelligence is not just another technical innovation in the history of tools. It marks a much deeper shift. For the first time, artificial systems are beginning to reproduce certain faculties that were thought to constitute the human mind: analyzing complex situations, establishing correlations, producing texts and images, formulating reasoning, anticipating situations.
This shift is considerable.
For centuries, human uniqueness was largely based on our cognitive abilities: understanding the world, interpreting situations, expressing ideas, deciding and judging. However, these faculties gradually become shareable with technical systems.
Therefore, a fundamental question arises: how much of our decisions are we ready to delegate to machines?
A delegation already at work
In many areas, this process is already underway. Algorithms participate in the allocation of credits, academic guidance, sorting of applications, the prioritization of information or the optimization of public policies. They assist with medical diagnoses, guide financial investments and begin to intervene in certain administrative or legal processes.
Faced with this transformation, the political debate remains fragmentary.
Artificial intelligence does appear in programs and speeches, but almost always in partial form. For some political leaders, it constitutes above all a question of technological sovereignty in competition between great powers. For others, it mainly concerns military power or national security. Some see it above all as a lever of economic productivity capable of improving the competitiveness of organizations. Others are primarily concerned about its environmental footprint or the bias and discrimination that algorithms can produce.
All of these concerns are legitimate. But they remain peripheral.
Because they bypass the essential question, the one that should structure any democratic debate on artificial intelligence: which decisions do we want to continue to make ourselves, and which are we ready to entrust to machines?
Economic decisions, administrative decisions, medical decisions, judicial decisions, military decisions. At what point does an algorithmic system cease to be a tool and become an actor? When does technical assistance become delegation of authority?
The risk of an atrophy of human capacities
But the delegation poses another question, less visible and yet essential: what happens to our own capacity to think, to express and to decide when we gradually cease to exercise these faculties?
All technology changes the capabilities of those who use it. When writing became widespread, some ancient philosophers already feared a weakening of human memory. More recently, digital technologies have profoundly transformed the way we read, write and concentrate.
Artificial intelligence could produce a transformation of another magnitude.
If we gradually delegate to machines the writing of texts, the synthesis of information, the analysis of complex situations or the formulation of arguments, it becomes legitimate to ask whether these faculties will continue to develop in individuals, or whether they risk gradually atrophying.
A society in which machines write, analyze and decide for us could be a faster and more efficient society. But it could also become a society where individuals exercise less and less of their own judgment.
The place of emotions in the decision
Another question then arises, perhaps even more subtle: what place do we want to continue to give to human emotions in our collective decisions?
Artificial intelligence systems often promise more rational, more efficient, less biased decisions. They calculate, compare and optimize at a speed and scale inaccessible to the human mind. However, political and social life has never been based on rationality alone.
Compassion can guide a legal decision. Empathy can inspire social policy. Outrage can spark reform. Doubt, caution or intuition can temper the cold logic of efficiency.
Emotions are not just imperfections that technique would correct. They also contribute to our capacity for judgment.
A society in which decisions are increasingly produced by algorithmic systems could be more predictable, faster, perhaps even more efficient. But it would also risk becoming a society in which that part of humanity which introduces compassion, responsibility and lived experience into political decisions gradually disappears.
A shift in power
The question of artificial intelligence cannot therefore be reduced to a technological problem. It is deeply philosophical and political.
Because the entire history of human societies can be read as a question about the location of power: in a monarch, in an elite, in a sovereign people. Artificial intelligence introduces a new possibility: that of power mediated by technical systems whose logic often remains invisible to those who use them.
Major technological transformations have always redesigned power structures. Printing revolutionized the diffusion of knowledge. Electricity has transformed economic structures. The Internet has profoundly changed the circulation of information.
Artificial intelligence could transform the way decisions are produced even more profoundly.
The real democratic debate
In this context, the way in which our public debates are structured sometimes appears singularly out of sync. While we divide around questions of identity or symbolism, a silent transformation redefines the very conditions of human action.
The essential question is perhaps no longer just how to distribute places more equitably in the world as it exists.
It becomes: what world do we want to build in the era of artificial intelligence?
Because AI can be a formidable instrument of emancipation: accelerating scientific research, improving medicine, optimizing the use of resources, broadening access to knowledge.
But it can also become an instrument of concentration of power, generalized surveillance or massive automation of collective decisions.
Everything will depend on the principles that we choose to place at the heart of its development.
The issue is therefore not only technological. He is deeply democratic.
If we want the future of our societies not to be determined solely by the strategies of a few technological players or by the sole logic of efficiency, then it becomes necessary to open a real public debate.
A debate which does not simply question the uses of artificial intelligence, but which asks the central question: the place we want to give it in the organization of our societies.
Because behind the question of artificial intelligence lies perhaps the most decisive question of our time: what part of our decisions, our judgment and our humanity are we ready to delegate to machines?
