The article
Knowledge that serves the world. What role for universities in the age of AI?
Providing more training is not enough. We need to ask ourselves how to train students, now that a machine can produce an answer in a matter of seconds to the very problem we ask students to spend hours working on. Lorenzo Casini, Rector of the IMT School for Advanced Studies Lucca, writes to us

Photo by Nathan Dumlao on Unsplash
To the Editor – Although Francis of Assisi could read and write, he seemed to be wary of books. The Rule permitted them only to the extent strictly necessary. Eight hundred years after his death, that wariness has something to say to the university. Jacques Le Goff and Carlo Ginzburg saw Francis as an ‘interdisciplinary’ innovator, capable of speaking to a world of cities, merchants, money and new forms of poverty. Yet he wrote to Anthony of Padua: “I am pleased that you teach sacred theology to the friars, provided that […] the spirit of prayer and devotion is not extinguished”. The book served a purpose: that of preaching, of speaking to others with full knowledge of the subject. With Bonaventure, a few decades later, it would become as essential to the Franciscans as preaching itself. Not knowledge closed in on itself, but knowledge that serves the world. This remains, even today, a sound definition of higher education.
However, knowledge that is useful to the world must reach people. And this is where Italy has a problem. In 2025, according to Istat, just 31.1 per cent of 25–34-year-olds held a tertiary qualification, compared with the European average of 44.8 per cent. The country is second from bottom in the EU and falling further behind. Behind this average lies an injustice: 66.8 per cent of those with at least one parent who holds a degree have a degree themselves, compared with 12.9 per cent of those whose parents have at most a lower secondary school certificate. Talent can emerge anywhere. The opportunity to nurture it, even today, cannot. Universities cannot wait for students: they must go out and seek out those who cannot even imagine being able to enrol.
Providing more training, however, is not enough. We must ask ourselves how to teach, now that a machine can produce an answer in a matter of seconds to a problem on which we ask students to work for hours. Obtaining an answer and actually understanding are two different things. In the US, an MIT report published in August calls for a rethink of assessment and for strengthening learning through direct experience. The illusion of learning is the most insidious risk: at the first sign of difficulty, we delegate the task, no longer knowing how to judge the result. The most advanced technology thus brings us back to the simplest thing: one person explaining to another how they reasoned, where they went wrong, and why they changed course. This is why campuses, where people study and work together, are not a relic of the past: they are the most modern solution.
There is also a political dimension at stake. The American historian Jill Lepore, winner of the 2026 Pulitzer Prize, refers to the digital infrastructure through which governments and private companies automate public discourse as an ‘artificial state’. The risk is that decisions affecting the wider community will depend on powers that are not accountable to everyone. Nobel laureate Daron Acemoglu, in his latest book on liberal democracy, points out that the direction of technology is a choice, not a matter of nature, and proposes a ‘pro-worker’ AI that enhances the value of human labour. To make such choices, we need an intelligence capable of judgement, which must be cultivated. And this is cultivated through trial and error, experimentation, and, above all, by challenging one’s own certainties.
This is why investing in higher education pays off, now more than ever. A PhD teaches you to ask new questions and to account for your findings: invaluable skills in research, business and public institutions. Credit must be given to Minister Anna Maria Bernini for having defended the Ordinary Financing Fund in recent years – which will exceed 9.4 billion in 2026 – and for having done so whilst serving in a government from a political camp that has often viewed universities as merely an item of expenditure. However, these resources must enable universities to plan ahead, to make commitments to those who will come to study and work there, and to be able to honour those commitments.
Hannah Arendt wrote that a teacher is authoritative because they ‘take responsibility’ for the world and point it out to young people by saying: ‘This is our world’. Few have done so as Carlo Ginzburg, who passed away in June: from the miller Menocchio to the benandanti, he taught us to read the world through its details. When a colleague asked him when one should stop in a line of research, he replied before she had even finished her question: “Never”. The more machines are able to provide answers, the more we need people capable of asking questions. And of taking responsibility for the consequences. And this is what the university can and must be for.
Lorenzo Casini
Rector of the IMT School for Advanced Studies Lucca