TWON - Simulating Democratic Debates Online
TWON analyzes different platform designs by building digital twins and using agent-based simulations to research how dynamics of disinformation, radicalization, and polarization form on social networks.
Political debates online are often characterized by filter bubbles, echo chambers, and polarization. The TWON project researches these dynamics by building a digital twin of social networks and populating it with generative artificial intelligence. This can give insight into how people form radical political opinions online, how disinformation spreads, and how social networks strengthen tendencies toward a more polarised society.
A digital twin makes it possible to test how feed algorithms, recommendation systems and moderation rules shape collective discourse. The goal of the project is to produce evidence-based recommendations for regulatory innovations.