Si les algos de recommandation sont les biais cognitifs de l'humanité, en faut-il de nouveaux ?

For over a decade, our attention has been controlled by algorithms that are dangerous to humanity, and we are looking the other way. Every second, highly sophisticated artificial intelligences choose which messages, images, or videos we will see. But their sole objective is to capture as much of our available brain time as possible to extract the greatest profit, regardless of the now well-documented individual and societal consequences: isolation, compulsive use of social media, deterioration of mental health, mass cyberbullying, political polarization, information warfare, hate speech... Using concrete examples, we will see how these recommendation AIs, now controlled by multinational corporations or authoritarian states, have severely weakened our democracies and human rights. We will also discuss the limitations of an approach based solely on "individual critical thinking" and how to try to promote "critical thinking on the scale of our information systems." In our screen-saturated world, designing more democratic algorithms that incorporate greater critical thinking has become urgent. But there is a way to transform this digital dystopia into a resilient society. There's still time to take back control! #REC25 *Saturday-Lauragais1