Hélène Landemore: Can AI Bring Deliberation to the Masses?

One unsolved problem in democratic theory is how we can reconcile the twin goals of quality deliberation and mass participation. Both are arguably conditions for the full legitimacy of a democratic system. Quality deliberation, as a process through which laws and policies are generated, in theory promises good governance (output-legitimacy) as well as, at the very least, good reasons for the laws and policies put forward. Mass participation, by contrast, is a condition for the democratic input-legitimacy of the system, namely its capacity to take into account people’s needs and preferences. Unfortunately, thus far, it has proven impossible to reconcile those two goals as the quality of deliberation diminishes past a relatively low threshold of participants (a few hundreds, perhaps a few thousand people) and mass participation, on the other hand, is not conducive to the thoughtful, informed exchanges smaller numbers afford. In this presentation, Hélène Landemore explores the ways in which Artificial Intelligence may help bridge that gap, at least up to a point.
This HAI Weekly Seminar took place on October 19, 2022. Upcoming events can be found here: hai.stanford.edu/events