Barcelona Metropolis

The growing impact of artificial intelligence is coupled with a growing concern over the ethical, social and democratic risks of its uncontrolled deployment. One such example is the “Open Letter”, in which a thousand scientists requested a six-month break in AI development, as a result of the massive spread of ChatGPT. Nevertheless, there are many voices, including those of AI “manufacturers” themselves, calling for a hiatus, and for reflection and the regularisation of invasive procedures in all realms of life. Under these circumstances, however, the failure to tap into the cultural wealth of humanism, or the humanities in their educational form, would be both serious and irresponsible, given that they are to this day the repository of the best ethical, social and democratic knowledge – understood in a plural and evolved sense up to the present day and not limited to the foundations of Greco-Latin classicism – of an ethical, social and democratic nature. Disregarding it without further consideration would be a radical (anti)ethical, (anti)social and (anti)democratic choice. A commitment must be made to harnessing humanistic knowledge, to nurturing it – including with the help of AI – and to relating it to the deployment, adaptation and application of this technology.

Humanism, complex knowledge

This reflection is meant to facilitate the endeavour. However, to avoid confusion, clarification of the concept of humanism may be needed. Humanism, the cultural foundation of the humanities, is neither an ideology nor indeed a religion: it is a current of wisdom or full culture, initiated in classical Greece and Rome, revived in the Renaissance, rekindled in the Enlightenment and in permanent renewal: Edgar Morin, Martha Nussbaum, Nuccio Ordine, Joan Fuster and George Steiner, to name but a few, are very different contemporary humanists. It is a cultural nerve of our societies, which mainly consists of “faith in freedom, in intelligence, in the power of speech, in morality and in civic and social commitment”: a synthesis of an ancient philologist, ideal for understanding the value and significance of the profile of our intellectual “formatting”, and a catalyst for a free and committed dialogue with a technology that is considered to be intelligent.[1]

Humanistic knowledge resides more in the territory of symbols and values than in that of facts and applied functionalities. The Brazilian writer Mario Quintana dared to say that “facts are a secondary aspect of reality”, not because he primarily defended the unreliable “alternative facts” of these post-truth times, but because he considered understanding, interpreting, contextualising, criticising and evaluating any fact that, whether we like it or not, necessarily seeks a framework or format in which it can satisfy our need for meaning, to be a primary or fundamental aspect of human reality.

The Kantian idea that the autonomous rational being – which we claim to be – resides “in the realm of purposes” resonates in the background; and AI, as a tool, however powerful, fast and efficient it may be, resides in the realm of tools, incapable – except for falsifications – of achieving a purpose worthy of this name on its own. And it is only in relation to one freely chosen purpose or another that our everyday events can be interpreted and assessed. AI algorithms⁠⁠ do not think, they are thought. For, as Elvira Lindo said[2], “the algorithm is the parents”: it does nothing by itself and, above all, it is not magic, there are always “fathers and mothers” behind it who prepare the “gifts” that we give it.

Natural language under threat

Two thinkers as far apart as Yuval Noah Harari^[3] and Slavoj Žižek concur in their assertions – in different terms, but fundamentally they agree: the risk of the new wave of AI – chatbots like GPT – is that it will end up “hacking” humanity’s operating system, which is none other than the enigma of meaning that each natural language provides us with. An enigma composed of keys to interpretation, mental frameworks, purposes and values, in the form of epic tales, religions, myths, poetry, novels, theatricalisations, ideologies and similar concepts that provide us with various maps for life’s journey.

These chatbots, as generators of language with no basis other than a powerful but cold combination of an infinite amount of unverified data, cannot guarantee the truth of their assertions and consider that their mission is not to produce true or reliable texts, but rather plausible ones: in other words, satisfying simulations for each individual correspondent who addresses them. The line between truth and falsehood is blurred, and language becomes increasingly plain, undermining and potentially forfeiting the great variety of resources of natural language. Two prime examples: the capacity for metaphor and irony. The nurturing and care of natural language has always been at the heart of the humanities, which are being challenged by the threat of deterioration stemming from AI.

A call for ethics and democracy

A positive connection between AI and the humanities must be advocated for a number of other reasons, which are outlined below. The advent of a new technology ushers in new power, which is often in the hands of a few, even more so in the case of AI and the few big tech companies. AI manufacturers should learn the risk of hubris from the humanities, which can lead the most ironclad powers to self-destruction, like in Greek tragedies. Without so much drama, AI needs the humanities as a point of reference and a great repository of figures, myths and stories of the utmost human interest. On the other hand, AI users need the humanities for a reason as essential as the aforementioned natural language: the defence of freedom, which is inherent to the entire humanistic tradition, given the risk of AI’s attack via the invasion of privacy, manipulation based on the accumulation of data, and democratic destabilisation.

The humanities, with their promotion of critical thinking and the spirit of freedom, are the foundation required to meet the widespread and growing demand for an AI that respects the most fundamental ethical principles; ethics, as a practical philosophy or “knowledge of life” since Greek classicism, is neither a fashion nor an imposition of any authority, it is the attempt to meet the human need to learn to live a life in freedom, a good and happy life.