

330 | Petter Törnberg on the Dynamics of (Mis)Information
127 snips Sep 29, 2025
Petter Törnberg is a computational social scientist and professor at the University of Amsterdam, specializing in complex systems. In this engaging conversation, he explores how individual interactions in online communities can lead to large-scale emergent behaviors like polarization. Törnberg discusses the dynamics of information flow and segregation, the potential of using LLM-driven agents for better understanding online dynamics, and the societal impacts of social media on journalism and political misinformation. He also reflects on ways to redesign platforms for improved community engagement.
AI Snips
Chapters
Books
Transcript
Episode notes
Platforms Create A New Epistemology
- Platforms embed a new epistemology: society seen as self-organizing interactions rather than machine-like design.
- That hides power exercised via rules of interaction and algorithms that shape large-scale outcomes.
Tiny Preferences Produce Big Segregation
- Small individual preferences can produce large-scale segregation through cascades.
- Integrated states are unstable and neighborhoods tend to tip toward homogeneity.
Online Groups Are Prone To Segregate
- Schelling-style segregation is even stronger in group-based online communities than on lattices.
- Echo chambers can emerge from structure alone, not only from algorithms or preferences.