This paper explores how artificial intelligence, particularly large language models (LLMs) and the "Habermas Machine," can address the inherent trade-offs in democratic deliberation between broad participation, meaningful discussion, and political equality. It examines AI's potential to enhance these aspects and discusses the challenges that remain in its application to democratic processes.

This paper investigates how fair voting methods can revitalize democracies by addressing long-standing challenges and new threats from AI. It presents a trilogy of findings showing these methods enhance legitimacy, accelerate novel impacts, and safeguard against AI biases, using real-world evidence from participatory budgeting.