A.I. explained to humans

The Maison Suger is pleased to host a presentation of the book A.I. explained to humans by Jean-Gabriel Ganascia, published in September 2024 by the Éditions de Seuil.

This talk will first echo AI explained to humans, which, through a fictional dialogue with four imaginary schoolchildren, presents AI, its epistemological status, the various techniques to which it has given rise, their applications and the societal and ethical issues raised by their massive deployment. This will be an opportunity to review the evolution of AI from its origins to the present day, with a particular focus on the very recent emergence of large-scale language models and generative AI. We then turn to the meaning and originality of generative AI productions. First, we'll look at the semiotic status of the images, sounds and videos generated. We'll see that while some of them depict in an analogical mode and are therefore symbolic, others seem more akin to photos or videos, suggesting that they are what C.S. Peirce calls indisignia. However, the realities they designate are totally fabricated, with potentially disastrous political consequences. We will then turn to the aesthetic status of the works generated, showing that we are witnessing a shift similar to that which W. Benjamin foresaw in the 1930s, except that the status of the work of art must now be considered not in the era of its simple technical reproducibility, but in that of its technical producibility, which, as we will show, has major consequences.

• Jean-Gabriel Ganascia is a professor at Sorbonne University, where he conducts research on artificial intelligence at LIP6. He has chaired the CNRS Ethics Committee and has already published various works with Le Seuil, including Le Mythe de la Singularité, which won the Roberval prize in 2017.