In this session Mohammad Mahdi and Bernhard Rieder will discuss the transformative potential of so-called “foundation models” by focusing on a series of perspectives including their technical mechanisms, promised applications, ethical considerations, and economic implications. Rather than taking the transformative potential of these technology for granted, this session will examine how this transformative potentials are discursively, economically, and technically produced. The session will also examine how AI technologies are becoming “platforms” and why this matters for critical analysis.

About the Speakers

Mohammad Mahdi is a PhD student at the University of Amsterdam's VIS Lab, focusing on multimodal foundation models. His research involves visual-language models capable of processing images to generate textual descriptions, as well as hybrid multimodal architectures combining autoregressive models with diffusion (flow matching) objectives. His work encompasses data preparation, model pretraining, instruction tuning, and comprehensive evaluation.
Bernhard Rieder is Associate Professor of New Media and Digital Culture at the University of Amsterdam and a collaborator with the Digital Methods Initiative. His research focuses on the history, theory, and politics of software and, in particular, on the role algorithms play in the production of knowledge and culture. This work includes the development, application, and analysis of computational research methods and the investigation of political and economic challenges posed by large online platforms.

Programme