LUM: UM's science and society magazine

Generative AI, symbolic AI, deep learning, neural networks... To better understand the contours of artificial intelligence, Lum invites you to delve into its methods, to explore its scientific and social uses, and to explore its limits.

Artificial intelligence is now an essential part of research. Not only because it enhances research capabilities by speeding up certain tasks, but also because it is itself an object of research that the University of Montpellier has fully embraced.

In this 22nd issue of Lum magazine, we'd like to give you a better understanding of artificial intelligence. First of all, in the first part devoted to methods, we explain what AI is, a question to which there are as many answers as there are people. What is a neural network, machine learning, federated learning, symbolic AI? In this section, Lum deciphers the ABCs of AI with the computer scientists and mathematicians at the heart of the reactor.

The second section looks at the uses of AI, which can prove an unexpected ally in artistic creation, or an innovative tool for better understanding communication in mandrills. At the UM, AI is also used to predict concrete cracking, detect fake news and study glaciers.

Finally, the third part looks at the social uses of AI, and its contribution to debate, for example. Its limits, too, as a technology that consumes a lot of energy and data. This part also raises the pressing question of AI's place in higher education and its proper use by students.

Contemplate these AI portraits as you read this 22nd issue of Lum magazine, in its expanded version, with access to the UM podcasts: A l'UM la science and Lumlu.

All Lum numbers

Download back issues
Download the cover wallpaper