Lecture on January 20, 2025
The Future of Prediction. Algorithmic Forecast in Science and Society
Prof. Dr. Elena Esposito (Universität Bielefeld, DE & University of Bologna, IT)
About the lecture [poster, pdf] [recording]
For the past few decades, with the advent of big data and the rapid proliferation of advanced machine learning algorithms, the status of prediction in research and social practices seems to be changing. Is a new form of prediction being developed? And how is it used? The fundamental distinction between algorithmic predictions and traditional probabilistic predictions lies in two key features of recent digital forecasts: their opacity and their individual reference. The talk argues that these two characteristics should be seen not merely as challenges but as the foundation of the unprecedented performance of algorithmic predictions that combine them. Algorithms excel in generating highly specific predictions with speed and precision precisely because they do not need to understand the content they process. However, this very feature often renders them incapable of explaining their results or generalizing them to cases beyond the immediate context. As a result, we require evaluation criteria distinct from those applied to probabilistic techniques, which are still rooted in human forms of information processing and rely on explicability and generalizability.
About the speaker
Elena Esposito is Professor of Sociology at Bielefeld University and the University of Bologna. She has published extensively on the theory of society, media theory, memory theory and the sociology of financial markets. Her current research on algorithmic prediction is supported by a five-year Advanced Grant from the European Research Council (PREDICT). Her latest books are Artificial Communication. How Algorithms Produce Social Intelligence (2022); Kommunikation mit unverständlichen Maschinen (2024).
photo: Elena Esposito
Monday, 20. January 2025, 18:15-19:45 (CET)
Address for joining us on-site:
Flügelbau Ost, 2. OG, Raum O221
Edmund-Siemers-Allee 1
20146 Hamburg