Artificial intelligence: only humans can save humanity

Within the space of a few days, two stories have made international headlines regarding artificial intelligence, and together they paint a picture that deserves more attention than the sum of the individual stories. Dario Amodei, CEO of Anthropic, has published an appeal in which he does not call for a halt to the development of the most advanced models, but rather for the pace of development to be slowed down, because the systems’ ability to improve themselves risks outpacing our ability to understand and control them.

Almost at the same time, a young researcher at the same company resigned, stating that those working on this technology sincerely believe it could put the whole of humanity at risk by the end of the decade, and a colleague confirmed that he believed the probability of a catastrophic outcome within a decade to be greater than ten per cent.

These are voices from within the very sector that develops and sells this technology, and for that very reason they carry more weight than warnings from outside. What is striking in this debate, teetering on the brink of apocalyptic scenarios, is how the responsibility for saving humanity is systematically shifted elsewhere – either to technology, which is expected to stop itself, or to governments, which are expected to regulate it from the outside. It is a subtle form of nihilism, not because it denies values, but because it treats the individual as just another variable, an externality to be balanced in the equation of risk and opportunity, whilst extinction, the loss of control, war, become debatable scenarios discussed in the same tone as one might discuss an investment.

When the irreducibility of the person ceases to be the criterion guiding choices and becomes merely one of the possible outcomes to be calculated, something has already been turned on its head. There is a starting point: the place of the person in the reasoning process, not as an outcome to be protected following a decision taken elsewhere, but as a guiding principle that must shape the decision from the very beginning. Those who call for rules and independent checks are right, but regulation alone remains blind unless it is equipped with a tool to give it direction, and it is here that Pope Leo XIV’s encyclical, ‘Magnifica Humanitas’, must be clarified and valued for what it offers: a definitive anthropological criterion for assessing every technological choice – not merely one option amongst others, but a fixed point of reference. The individual must not be sacrificed on the altar of technical power, nor entrusted entirely to the prudence of those who govern.

Léon Harmel, a textile entrepreneur rather than a theorist, approached the Second Industrial Revolution by taking a gamble that innovation would only generate value if it involved the empowerment of labour and the individual carrying it out; and Adriano Olivetti, in the post-war period, took the same gamble on a larger scale, the factory as a place where human fulfilment and innovation coincided, fed off one another, and did not contradict one another.