Optipolitics: Is Politics an Optimization Problem?

Can AI ‘solve’ politics? In this paper, I explore optipolitics, i.e., the idea that politics along with other complex social issues can be framed as mathematical optimization problems and solved as such. I begin by describing politics, optimization, and some reasons for applying the latter to the former. I then present a liberal democratic version of optipolitics and try to defend it against eight fundamental challenges. While I concede that none of the challenges can be satisfactorily overcome at present, many can be shown to be at least as difficult to overcome in conventional representative democracy. Most importantly, as a logical endpoint of AI-driven technocracy, I argue that optipolitics can serve as a useful starting-point for a critical debate about the future of democratic governance in the age of machines.