AI-generated image. Un archivo inexistente, Felipe Rivas San Martín. Courtesy of the artist.
Seminar: “Decolonial Approaches to AI-Generated Imagery”, curated by Pietro Conte, Sofia Pirandello and Maria Serafini.
Speakers: Cristina Voto and Felipe Rivas San Martín.
The seminar will be held in English and in a hybrid format at the Aula Polivalente, Via S. Sofia 9, from 10:30 a.m. to 12:30 p.m.
The seminar brings together Cristina Voto, researcher at the University of Turin, and Felipe Rivas San Martín, visual artist, to reflect on the intersections between generative AI, visual culture, and the historical structures that shape representation. From their respective perspectives, they will examine datasets as archives marked by biases, exclusions, and asymmetries, and explore how artistic practice can engage AI as a critical tool to question dominant visual histories and bring marginalized perspectives into view.
Cristina Voto is an assistant professor at the University of Turin, where she teaches Digital Languages and Semiotics of Cultural Heritage: Perspectives on Intersectionality, Ethics and AI. She serves on the doctoral board of the Design and Creation program (Universidad de Caldas, Colombia) and the Art and Techno-Aesthetics program (Universidad Nacional de Tres de Febrero, Argentina), and is Vice President of the Latin American Federation of Semiotics. Her research explores semiotics at the intersection of philosophy of technology, feminist and queer theory, and digital/AI-based art.
ABSTRACT
The aim of my talk is to rethink AI-generated images as contingent events with a specific mode of existence: they emerge from a chain of operations—data, compression, prompting, and interfaces—so each image is a provisional settlement between archives, algorithms, and users. I treat compression as the key cultural technique of generative AI, reorganizing visual meaning through proximity and probability, while prompts act as triggers that steer an aleatory process—opening some pathways and rendering others structurally unlikely. This helps explain why bias persists: social hierarchies return not as mere mistakes, but as probabilistic boundaries of what the system’s world can plausibly produce. I close by proposing a semiotic method to trace these modes of existence and locate points of intervention, so AI imagery becomes more accountable in what it makes visible.
Felipe Rivas San Martín is a Chilean transdisciplinary artist, essayist, and sexual dissident activist. He is currently a postdoctoral researcher at the Universitat Autònoma de Barcelona. His works are part of public and private collections such as the Reina Sofía Museum, Spain; 21C Museum, USA; MUNTREF, Argentina; the Chilean Ministry of Culture Collection; the MAC Museum of Contemporary Art, Chile; and the Museo de la Solidaridad Salvador Allende, Chile; among others. He is a co-founder of the University Collective of Sexual Dissidence, CUDS (2002-2019), a Latin American group dedicated to activism, artistic experimentation, and critical reflection. He is the author of the book Internet, mon amour: infecciones queer/cuir entre digital y material (Écfrasis editiones, 2019) and co-author (with Jaime San Martín) of the experimental science fiction book Sagrada Biblia Artificial (ESM, 2023). His work has been featured in the monographs Estatutos de la disidencia (2020) and Un Archivo Inexistente (2024).
ABSTRACT
This presentation addresses three projects that experiment with computational algorithms from the perspective of political memory and Latin American techno-queer speculation: The Neoliberal Dream (2015), in which the iconic image of the 1973 Chilean coup is repeatedly interpreted by Google’s Deep Dream until it becomes unrecognizable; A Nonexistent Archive (2023), which constructs a denied Latin American queer photographic archive using generative AI; and Non-linear (2025), a research process on representations that challenge the notion of linear time using hybrid methodologies between manual drawing and generative text and image models.
For further information, please contact sofia.pirandello@unimi.it and roberto.malaspina@unimi.it.
Teams connection link –> https://teams.microsoft.com/meet/3625383713183?p=alVoalVH7Tlls697cD