Abstract
Generative AIs today offer textual, musical or pictorial productions of "artistic" content, which are used by Internet users and professional artists alike. But beyond the new tool on offer here, with its advantages, biases and risks, is there any artistic addition on the part of the AI itself? Skeptics generally resist the idea of AI as an artist, citing the lack of creative intentionality of these disembodied algorithms with no referential link to the world. But the need for intentionality to qualify an object as artistic is questionable, as is the lack of intentionality in generative AIs. We propose to interpret the models we teach, the prompting process and the cultural mesh in which they are embedded as "mediated intentionality" or "dialogical intentionality". What emerges is not only a referential meaning, but a genuine artistic value, with new aesthetic effects.
