Artificial Intelligence: 'A New Form of Creativity Emerges, Accessible to All'
By Guillaume Chevillon (Economist) and Julien Malaurent (Professor of Information Systems)
Published in Le Monde on January 4, 2025
AI can be used to create texts, images, and videos that we cannot imagine, argue Guillaume Chevillon and Julien Malaurent, professors at ESSEC, in an opinion piece for "Le Monde". This is possible provided we understand the machine, its capabilities, and its limitations.
The surge of hopes, fears, fantasies, and technological allegories linked to the emergence of generative artificial intelligence (AI) would be fascinating to study for researchers like us, if these concerns didn't also affect us somewhat.
Because AI often frightens people. It is indeed normal to worry about machines surpassing our cognitive abilities, after having long ago exceeded our physical capabilities. The fear arising today is that of losing control, of humans over machines, which affects our cognition, our personality, our identity.
It is not unrelated to the vertigo we feel when faced with neurodegenerative diseases, as the super-powerful machine also threatens our ability to think, invent, and act. Does this signal the end of novelists, journalism, advertising, and design? The end of filmmakers, photographers, researchers, teachers? What will be left for us?
Deviating from Patterns
Some will say this was predictable, arguing that certain studies have shown creativity can be understood statistically, and thus be modeled and reproduced. Nevertheless, the ease of access and use of tools for generating text, images, music, or video strongly materializes these issues. And this will continue because generative AI benefits from massive investments that continuously multiply its processing capabilities. The net is tightening. But are we talking about the same thing? Isn't there a fundamental difference between human creativity and algorithmic creativity?
Just a few months ago, an OECD study indicated that the use of AI-based tools would allow humans to highlight some of their unique qualities, namely empathy and... creativity. Creativity is our ability to be imperfect, unpredictable, and sometimes to deviate from predefined patterns.
Creativity is indeed this space where subjectivity, perceptions, influences, and experimentations intertwine to give rise to productions of all kinds. It is also the fruit of chance encounters between humans, but also between disciplines, where contradictions, ambiguities, uncertainties, and mixtures transform into resources or results, sometimes unexpected, even unhoped for.
But today with AI, a new form of creativity emerges, accessible to all. Anyone can create an image, a play, a piece of music, with just a few instructions (or prompts). Human artistic creation, which until now required technical mastery of an art acquired through experimentation and encounters, is now confronted with a new form of creation, accessible through a few keywords.
New artists are emerging thanks to their mastery of prompts. Galleries represented in recent months around Art Basel Paris and Paris Photo have indeed presented works by many artists working with and about AI. This will also be the theme of an exhibition at the Jeu de Paume in Paris starting in April.
Tool of Democratization
Isn't generative AI simply a new tool allowing us to experiment more? The appearance of photography in the 19th century and its democratization revolutionized the arts. It allowed both to change the perspective of all artists, while previous advances – camera lucida or obscura, for example – were reserved for privileged individuals like the painter Vermeer, and perhaps even constituted one of the great "secret knowledge" of the Renaissance, if we believe the essay published by David Hockney on this subject. Photography made available to the masses not only helped the development of contemporary art, but it allowed, through cinema among others, the invention of new forms of narrative.
Today, generative AI constitutes a democratization tool that allows everyone to explore and invent. We are at the dawn of a proliferation that will bring forth new forms and objects that we do not yet suspect. The technological diffusion, experiments, and exchanges it will bring about rely on the strength of human societies, that of collective intelligence which we must encourage. For open societies and democracies, it's about capitalizing on their assets to strengthen and progress.
To achieve this, education is paramount. Creating original works that reflect the artist's vision with generative AI tools requires many unsuccessful attempts, as one must understand the machine, its capabilities, and its limitations – and this without a user manual. It's not about improving productivity here, but about using AI to extrapolate, to go beyond what exists in order to reach the inaccessible or unimaginable.
Specialists, whether they are artists or in any other occupation, whose "creativity" appears threatened today, will need as always to be at the forefront of experimentation and innovation, of art and tekhnê. Internet and social networks have, for example, modified the role of journalists and strengthened their role in curation, hierarchization, and putting information into perspective.
Photography and cinema didn't make painters and actors obsolete but opened new ways for them to flourish. AI thus constitutes an apparatus reinforcing the need for perspective, for dialogue between the collective and expertise, because the mass of data and images on which it relies makes it essentially a tool of social collaboration.
In line with the invention of printing, which accelerated access to writing and knowledge, AI carries hope if we know how to accompany it collectively. We must massify education, as recommended by the AI Commission report. Our democracies have the opportunity to prove that our educational system constitutes the best means of liberating human intelligence, sensitivity, and creativity.
Guillaume Chevillon is an economist, professor at ESSEC Business School, Media and Digital Chair, and academic director of the Metalab for Data, Technology and Society. Julien Malaurent is professor of information systems, deputy director general of ESSEC Business School.
