Lecture: 'The Problem of Anthropomorphizing Artificial Intelligence'

In this lecture, we will analyze a deeply rooted trend: that of anthropomorphizing artificial intelligence, meaning attributing human qualities such as intention or understanding where they do not actually exist. Through examples like the Mechanical Turk or the ELIZA experiment, we will see that this is not a new phenomenon but a deep psychological mechanism that has accompanied us for centuries. Today, this trend is amplified by language models, capable of generating highly convincing responses that simulate understanding. But it is important to emphasize: this appearance does not imply that there is real thought or understanding. We are facing a phenomenon comparable to a perceptual illusion — like the Müller-Lyer — that persists even when we know it is false. In other words, it does not disappear with knowledge. Finally, we will see that this confusion is not innocent. It can lead to judgment errors, to delegating decisions where we shouldn't, and, above all, to a progressive deterioration of social trust. The proliferation of credible artificial content and identities threatens to erode the distinction between true and false, weakening fundamental institutions such as journalism, justice, and democracy.

Moderator: Senén Barro

About the speaker

Ramon López de Mántaras Badia is a Research Professor (emeritus) at the CSIC in the Institute of Research in Artificial Intelligence (IIIA), of which he was founder and director. He is also a Distinguished Visiting Professor at Western Sydney University. He has been researching Artificial Intelligence (AI) since 1975 and is a pioneer of AI in Europe. He holds a PhD in Physics from Paul Sabatier University in Toulouse (France), a Master of Science in Computer Science from the University of California-Berkeley (USA), and a PhD in Computer Science from the Polytechnic University of Catalonia. He was a Tenured Professor at the Faculty of Computer Science in Barcelona and a Full Professor at Pierre and Marie Curie University in Paris. He was the chief editor of the European AI journal and an Associate Editor of the Artificial Intelligence journal. He has authored over 300 scientific papers and the popular books "Artificial Intelligence" published by "Los Libros de la Catarata" and "100 Things You Should Know About Artificial Intelligence" published by Lectio. He has received, among others, the following accolades: Earle C. Anthony Award from the University of California-Berkeley in 1978, "Ciudad de Barcelona de Investigación" Award in 1982, "Robert S. Engelmore Memorial" Award from the Association for the Advancement of AI (AAAI) in 2011, "Distinguished Award" from the European Association for Artificial Intelligence (EurAI) in 2016, "Donald E. Walker Award" from IJCAI in 2017, and the National “Julio Rey Pastor” Award for Information and Communication Technologies in 2018. He is a Fellow of EurAI and a member of the Institut d'Estudis Catalans (Academy of Letters and Sciences of Catalonia). He chairs the External Scientific Committee of CiTIUS.