Humanism and artificial intelligence

This is the simplified and abridged text of the speech I gave on 28 October 2025 at PUC-RS in Porto Alegre to master’s and doctoral students in the humanities, accompanied by their professors.

Let us first define humanism as a reflection on the essence of humanity, characterised by its abstraction and situated within a universal horizon. Secondly, based on this reflection, humanism is concerned with the good of humanity, meaning that it has a normative, ethical purpose.

Karl Jaspers called the middle of the first millennium BCE the « Axial Age », that moment in history when Confucius in China, Buddha in India, Zarathustra in Persia, the Hebrew prophets in Israel and Socrates in Greece each founded, in their own way, great humanist traditions. It should be noted that this was always a matter for scholars, based on the use of the alphabet or a system of standardised characters, as in China. At that time, oral traditions were beginning to be written down, and manuscripts, rewritten with each copy, were fluid, fragmented into multiple versions. As for the actual authors, anonymous and plural, they often hid behind the authority of great mythical ancestors.

The Bible and Greco-Roman literature are the two great roots of Western humanism. I will leave aside the Bible, which I dare not discuss in front of Marist brothers who know more about this subject than I do, and will content myself with discussing Greco-Roman humanism. Greek paideia and Roman humanitas (which is its translation) are based on three main pillars: literature, open-mindedness and a sense of human dignity.

Literature here includes mastery of language and writing (grammar), the science of reasoning and contradictory dialogue (dialectic), and finally, the art of persuasion, which was essential in this culture of political orators and lawyers (rhetoric). A well-rounded education required knowledge of the sciences of the time and, above all, immersion in the works of the classical authors: poets, playwrights and philosophers.

Open-mindedness is expressed in this famous maxim from a play by Terence (2nd century BCE): ‘Nothing human is alien to me.’ The phrase itself was inspired by Menander, a playwright of the Hellenistic period.

The third point, which still defines the basis of humanist moral attitudes today, is the primacy of human dignity. It could be argued that the Romans and Greeks, who practised slavery, did not live up to their own principles. This is undoubtedly true. But it should be remembered that almost all societies practised slavery or serfdom, which were only abolished in the 19th century. However, despite their inferior legal status, slaves could be treated “humanely” or not. The playwright Terence, whom I mentioned earlier, and the Stoic philosopher Epictetus were born slaves and were freed by masters who admired their talents.

The history of symbolic technologies mirrors that of humanism. During the Renaissance, printing, by mechanising the reproduction of texts, made copies and translations available. Publishing became an industry and modern literature developed. This resulted in the birth of the modern original author, which materialised at the end of the 18th century and especially in the 19th century with the emergence of copyright.

The Renaissance ‘humanists’ edited, fixed, translated and printed ancient texts belonging to the biblical and Greco-Latin traditions. This led to the emergence of textual criticism, i.e. the establishment of texts based on divergent copies. The studia humanitatis then brought together knowledge of Hebrew, Greek and Latin. Beyond linguistic competence, the profession of humanist required familiarity with the great texts of literature and philosophy, a new sensitivity to philology, history and the contexts in which texts were written, which would lead to the birth of modern hermeneutics in the 19th century.

Textual criticism gradually led to critical thinking. Luther initiated the schism in Latin Christianity by challenging the authority of the Church, which he shifted to the Holy Scriptures, now available in vernacular languages: this was the famous slogan ‘Sola scriptura’. The leading figure among European intellectuals, Erasmus of Rotterdam made a living from his writing thanks to the printing press, navigated a transnational intellectual network, did not hesitate to criticise the society and elites of his time (as in his famous ‘Praise of Folly’), and established himself through his monumental work as one of Europe’s leading publishers, philologists, translators, theologians and educators. Faced with rising religious hatred (and unlike the firebrand Luther), Erasmus defended a peaceful Christian humanism.

At the beginning of the 19th century, a debate, particularly illustrated by the educator Friedrich Niethammer, divided opinion in Germany. Should education – which was increasingly aimed at the entire population – focus on “useful” subjects such as science and technology, or rather on developing the mind, taste, independent moral judgement and the ability to participate in a shared culture through the study of ancient texts? The first option, which was more immediately practical, was known as philanthropy. The second option, which emphasises personal development or “bildung”, is called humanism. In the Western world, this debate continued into the 20th century, until humanistic education was reserved for a small minority of professional specialists and no longer formed the backbone of education for the majority, or even for the elites.

In the second half of the 19th century, historian Jacob Burckhardt redefined humanism (which he saw as a product of the European Renaissance) as a philosophical and practical orientation towards the autonomy of the human spirit, emancipated from the family clan, social class and ecclesiastical authority that stifled individual freedom. Burckhardt’s ideas would have a major influence on Nietzsche, himself a philologist by profession and highly sensitive to the historical nature of ways of living and thinking.

The result of a development that began during the Renaissance, between the 19th and 20th centuries, humanism focuses on the value and dignity of human beings, adopts a universalist ethic, and takes a general perspective of emancipation or the gain of autonomy. Finally, it places particular importance on literary and artistic studies for personal development. This approach has been the subject of much criticism from Christian theologians, socialist thinkers, and detractors of conventional morality. However, I will not dwell here on these numerous challenges, which became particularly heated after the end of the First World War, perceived as a collapse of European humanism.

If humanism was born with the alphabet in a literate environment and was reborn with the printing press, what becomes of it when digital technology becomes the dominant symbolic technology? Let us determine the main characteristics of the metamorphosis of text in the 21st century. All symbolic expressions are gathered and interconnected in a ubiquitous universal digital memory. The manipulation of symbols (and not just their reproduction and transmission) is automated. Texts can be generated, translated and summarised automatically. Masses of digital data drive generative artificial intelligence (AI), which becomes the probabilistic voice of collective memory. Paradoxically, AI represents tradition all the better when questioned about texts from the humanist canon that have often been edited, translated, and commented on, such as the Bible, the Church Fathers, Homer, Plato, Aristotle, the great Western literary and philosophical works, not to mention the major works and sacred texts of other traditions. On the other hand, the closer we get to contemporary works and themes, the more AI expresses opinion: the rumours and echoes of Plato’s cave, now digital.

Humanism has never been as criticized as it is in the 21st century. Posthumanism denounces our illusions about the permanence of a humanity that is now obsolete, hybridised or surpassed by machines and biotechnologies. Environmentalism and anti-speciesism criticize our anthropocentrism: having become aware of the ravages of the Anthropocene, climate change and the collapse of biological diversity, we must renounce humanism, which sees Man as the ‘master and possessor of nature’. Finally, for proponents of a certain critical sociology (Marxism, anti-imperialism, intersectional feminism), universalist humanism masks the domination of one part of humanity over another.

But humanism should not be confused with its hypocritical invocation or caricature. Humanity is not obsolete. The latest technological developments confirm, if confirmation were needed, the terrible and wonderful uniqueness of our species. It is precisely because we – as human beings – have a symbolic capacity that opens us up to moral consciousness that we must take responsibility for the biosphere and defend the intrinsic dignity of all human beings.

In line with its historical evolution and the counter-currents that have opposed it while enriching it, I would now like to articulate my own version of humanism in the 21st century. I will set out a few very simple principles which, in my view, should guide the (now digital) community of the humanities.

At the root of this lies a certain relationship to speech and tradition. A humanist recognises the existential weight of speech and considers language to be the pre-eminent medium of meaning. In an age of demystification and widespread criticism, we must relearn how to cultivate a reverence for texts and symbols. Rather than blindly rejecting traditions in a spirit of ‘tabula rasa’, we should work to preserve them, not to reify them or keep them unchanged, but to bring them to life in the present, reinterpret them and pass them on.

The three quintessential humanist practices – reading, writing and thinking – are mutually dependent.

Reading is essentially a relationship with the library, whether its medium is ink and paper or screen and electronics. As a humanist, my vocation is to embrace, as much as possible, the virtually infinite source of meaning that is the library. When I read, I discover beneath the text a living word that speaks to me. In order to grasp the meaning of the text, I do not limit myself to a single methodology, but draw on philology, formal analysis, history and influences. Each text can be interpreted against the backdrop of a multiplicity of corpora (that of the author, the era, the genre, the subject, etc.), so that the unique figure of the text gives rise to several forms depending on the perspective. AI should never replace reading. Nothing can replace a direct relationship with a text. On the other hand, AI can enhance reading through explanations, comments, references, and even the evocation of secondary literature. To stop reading in the first person is to stop learning and give up on understanding.

Let us now turn to writing. Writing is inscribing oneself in time, maintaining a relationship with the past, the present and the future. In relation to the past, writing confronts canons and corpora. The solo author never sings alone, but is accompanied by the ghostly chorus of vanished generations. In the living present, I participate in a dialogue of scholars where collective memory (perhaps carried by AI) and personal memory intersect. I articulate a living word that addresses the other to bring forth contemporary meaning. In my relationship to the future, I add to a collective memory that contributes to training AI and that may touch the minds of future generations. What a responsibility! Except for administrative tasks, AI should never replace writing. But it can prepare for it by drafting files or organising notes, as an assistant would. It can also perfect a text by working on its editing or bibliography. To stop writing in the first person is to stop thinking.

And what exactly does it mean to think like a humanist? First and foremost, it means enriching our personal memory, which is the foundation of living thought. Just because ‘everything’ can be found on the internet does not mean we should stop cultivating our individual memory. This is precisely because thinking is a dialogue between memories. It is woven into a dialectic between the collective memory represented today by AI, the personal memory of each of us, and the open dialogue – contradictory and complicit – with our peers and contemporaries. The richer our personal memory, the better we can exploit the resources of AI, ask the right questions, spot hallucinations and shed light on blind spots. Under no circumstances can AI replace ignorance. But it can serve as an advisor and coach for our learning. If we are ignorant, we will be manipulated and misled by language models. In contrast, the more knowledgeable we are, the better we can master AI, which, although it is now the environment of thought or the new sensorium, is still only a tool.

Publié par Pierre Lévy

Assiociate Researcher at the University of Montreal (Canada), Fellow of the Royal Society of Canada. Author of: Collective Intelligence (1994), Becoming Virtual (1995), Cyberculture (1997), The Semantic Sphere (2011) and several other books translated in numerous languages. CEO of INTLEKT Metadata Inc.

Laisser un commentaire