Can Artificial Intelligence govern society better than politics? The hypothesis of an “objective” government, guided by data and free from human passions and errors, may appear seductive. But it is a dangerous seduction, because it confuses the efficiency of the means with the meaning of an end.

A set of values ​​to govern

The misunderstanding lies in thinking that politics is a technical problem to be solved, while in reality it is a set of values ​​and needs to be governed. An algorithm can optimize a process; cannot establish a destiny. He can suggest the most efficient solution to reach a goal, but he does not have the moral capacity to decide what that goal should be. And politics is precisely this: the collective determination of objectives, the profound sense that leads us to go in one direction rather than another. Every public decision implies a hierarchy of values. Establishing whether to allocate more resources to healthcare or defense, whether to favor economic growth or environmental protection, whether to protect a production sector or liberalize it is not a mathematical operation. It is a normative choice, which reflects a vision of justice, equity and the common good. No Artificial Intelligence system, however sophisticated, can independently generate a scale of values. It can only apply criteria that someone has defined before. And that “someone” can only be politics, that is, the community that decides which goals to pursue. It is objected that AI could identify the “best” solution for the community on the basis of advanced predictive analyses. But better according to what parameter? GDP maximization? Reduction of inequalities? Social stability? Individual freedom? Each parameter incorporates an ethical choice.

There is no true neutrality

And as many scholars of the phenomenon observe, the algorithm is not neutral: it reflects the priorities that are assigned to it. If he decides based on maximizing efficiency, it is because someone has established that efficiency must prevail over other values. But efficiency does not exhaust justice, and what is efficient is not necessarily what is right. Then there is a second limit that makes the replacement of politics with AI impossible: the historical and transformative nature of political action. Artificial Intelligence works by statistical continuity, that is, it learns from the past to predict the future. But politics, in its decisive turning points, has been discontinuous. The great democratic achievements – the extension of political rights, the recognition of new freedoms, the construction of welfare systems – were not the most probable outcome of the previous data. They were counterintuitive choices compared to the established order. Politics is the ability to say: “It has been like this until now, but it is not right that it continues.” A third element, equally decisive, concerns responsibility. Democracy is not just a decision-making mechanism; it is a system of attribution of decisions. Whoever governs must be able to be judged, confirmed or replaced. Political responsibility implies transparency, accountability, the possibility of electoral sanction. If a choice causes damage, someone must answer for it to the citizens. An algorithm cannot be called to the polls, it cannot be questioned in Parliament, it cannot assume moral responsibility for its decisions. Added to this is the dimension of empathy and understanding of the human context. Politics does not only administer resources; governs social relationships. Behind every number there are people.

How do you understand the humiliation of marginality?

Artificial Intelligence can analyze indicators of poverty, but it cannot understand the humiliation of marginality. Politics requires the ability to put oneself in the other’s shoes, to evaluate not only the efficiency of a measure, but its social and symbolic sustainability. Then there is an even deeper aspect: politics is public dialogue. It is a comparison between different positions, it is mediation between opposing interests, it is the construction of compromises. It is not a calculated output, but a deliberative process. Replacing the debate with the result of an algorithm would mean reducing citizenship to a mere user. The citizen would no longer be a co-author of collective decisions, but a passive recipient of technical solutions. This does not mean rejecting Artificial Intelligence but placing it in its correct role. It can give more information more quickly but cannot decide the goals of a community. Because the State is the expression of a collective will that is recognized in a set of rules, symbols, rights and duties. It is not an equation but a cultural and moral construction.

A silent pact

It is the silent pact between the grandparents who built and the grandchildren who will inherit; it is the embrace of a community that decides not to leave anyone behind, not because it is efficient, but because it is right to do so. The future cannot be written by a cold sequence of codes, but must continue to be traced by our hands, imperfect yet creative. We must have the courage to claim the primacy of the human over technology, of the compass of values ​​over the calculation of probabilities. Because politics, in its truest sense, is the art of remaining human in a fast-paced world. And the answer to the challenges of tomorrow does not live within a circuit, but in the beating of a heart capable of still dreaming of a world that does not exist, and then getting to work and building it together.