Generative AI and public reason – democratic Implications of LLM

The project investigates the potentials and challenges for democratic systems that come with the proliferation of generative AI applications (gAI) within the realm of public spheres and political deliberation. It focuses on three central aspects: First, the epistemic dimension of gAI both as a source of knowledge in political debate as well as its increased potential to further disinformation. Second, the ethical dimension of the quality of training data scraped from public sources and its implications for the representation of social groups and political opinions within these data and thereby also within answers provided by gAI models. And third, the democracy theoretical dimension of gAI applications as “participants” of political discourse and its repercussions for public deliberation and public reason.

With a genuinely interdisciplinary work plan, the project is able to analyze these three dimensions from a philosophical as well as a machine learning perspective and to discover crucial limitations as well as conditions for institutional and technology design to reap possible benefits of the technology. The overarching goal of the project is to develop and submit an interdisciplinary grant proposal (VW, ERC) for a larger project on the topic of “Generative AI and Public Reason”.