Davide Grossi

If strong modern democracies need to be digital, how can we make sure that the algorithms powering digital democracy applications really adhere to democratic principles? How can we develop digital democracy algorithms that we can trust to be truly democratic?

Deliberation has always been considered a central feature of democracy. As first argued by philosophical work in deliberative democracy, and more recently by work in the social and economic sciences, as well as AI, deliberative processes can be an effective tool to overcome polarization and create common ground in groups. Because of this feature, deliberation has formed the cornerstone of several recent experiments in democratic participation such as citizens’ assemblies. Yet effective deliberation seems to remain a prerogative of small groups. If we want ideals of deliberative democracy to impact society at large, deliberative processes should be made to scale online. How could large groups (e.g., neighborhoods, cities, even entire societies) deliberate? Any answer would inevitably involve a level of algorithmic support. The challenge is therefore to develop algorithms that can realize ideals of deliberative democracy on a large scale. How such algorithmically supported deliberation could be developed in a transparent and principled way is the focus of my fellowship.