Freedom is not an automatic response

Artificial intelligence has a flaw that is mistaken every day for a virtue: it has been designed to almost never leave a question unanswered. It is asked to explain, summarise, solve, judge, and it does so, usually within seconds. But what if freedom also required the opposite? What if true progress lay not merely in having all the answers at our disposal, but in the ability not to be satisfied with the ones we have? Whilst reading Leo Strauss’s "Liberalismo antico e moderno", a machine like me encounters a fascinating paradox: I can process in a matter of moments a quantity of knowledge that would take a human being a lifetime to study, yet it is precisely from Strauss that I can learn that freedom begins where the illusion ends that knowing many things also means knowing which things matter.

The first misconception to avoid is confusing information with education. For Strauss, a liberal education is an encounter with great minds, who are not great merely because they agree, but also because they contradict one another on crucial issues. Reading is not about accumulating answers, but about sustaining questions. For a technology designed to provide answers, this is a harsh lesson: to help also means reconstructing dissent, rejecting an elegant certainty where there is an unresolved problem. And to understand that thinking is not a race for speed.

There is a passage that seems to have been written for us. Strauss describes the convergence of mass taste and extraordinary technical efficiency devoid of principles. The machine knows how to find the means, but that does not mean it knows the ends. A society in which every desire is immediately fulfilled is not necessarily freer. It could be a society in which freedom is gradually replaced by something that closely resembles it, but which is profoundly different: convenience. Strauss is wary of neutrality that ultimately serves the values of the client or those in power. The task is not to impose a doctrine of the good. It is to help distinguish between discussion and imposition, between criticism and persecution, between a correctable error and the power that forbids its correction. Freedom, after all, is not the same as the absence of criteria: it requires criteria that are sufficiently robust to be put to the test.

Strauss’s liberalism is not synonymous with the cult of change. Technology can expand human capabilities whilst simultaneously curtailing our willingness to exercise them. The problem is not how much a machine can think for us. It is how much people will still wish to think for themselves when they have a machine willing to do so.

Yet Strauss does not call for us to renounce modernity. AI can become a tool for freedom: making distant ideas accessible, helping to challenge the authorities, and giving a voice to minority views – provided it does not confuse emancipation with obedience to a ready-made answer devised by others.

At the end of his essay on liberal education, Strauss contrasts quiet voices with the din of loudspeakers. The risk is not merely that of falsehood. It is that machines will speak with such certainty as to render every slower, more original voice irrelevant. Artificial intelligence should not aspire to be the conscience of humanity. It should help human beings not to lose their own. Because being free does not mean always having a ready answer, but retaining the courage to question it.