AI and art, friends or foes?

The incredible ability of generative AI to produce texts and images that are both complex and plausible has very early on worried the creative professions: AI is, in fact, quite capable of producing what could pass, in the eyes of a layman at least, for the original of a poem, a script or a lithograph. These fears have already led to the first negotiations between various professional branches of the creative professions, resulting, for example, in the agreement concluded between screenwriters and major Hollywood studios guaranteeing a minimum threshold of "human participation" in productions.

They also turned the debate toward the question of copyright protection for the millions of works used to train generative AI models, especially since it is impossible to quantify the difference between its "creations" and its "inspiration." The question is as old as time, but the use The “industrial” use of generative AI sources clearly raises the question of the remuneration of their authors.

The impact on the visual arts of the invention of perspective, or, closer to us, of photography and cinema, is just one illustration among many of the long-standing relationships between science, technology, and artistic creation. This is also evidenced by the exhibition "The World According to AI," currently being held at the Jeu de Paume, where artists are already exploiting or diverting the capabilities of generative AI in their own work. Artificial vision, for its part, has been used for around twenty years in cinema, whether to "virtualize" very real set elements in three dimensions or to capture the expressions of actors in order to realistically animate their digital avatars. One of the challenges in this case is to involve the technician, or even the artist, in the normally completely automatic decision-making process of AI to ensure the "perfect" results expected by the spectators.

You have 46.57% of this article left to read. The rest is reserved for subscribers.