Scalable Deliberation Meetup

​How can collectives make decisions about what actions to take, when a collective's constituents might disagree?

​Effective self-governance relies on determining the best course of actions for a given group of people, in spite of potentially complex interactions and conflicts between their preferences. Voting systems can capture these preferences, but frequently fail to handle the implications of their interactions – such as cyclical preferences resulting from ranked-choice voting, where there is no clear option that a majority finds superior to all others.

​While not a complete solution, deliberative processes can surface some such conflicts of opinion, helping constituents share relevant information and find the cruxes of their disagreements. Deliberation best practices are clear for time-bounded, face-to-face interactions – but how could collectives that exist in mediated spaces adopt similar processes for themselves?

​AI Objectives Institute is taking the first steps in this direction with an LLM-based tool for discourse visualization and simulated deliberation. Colleen McKenzie will give a brief presentation with background, followed by an open discussion of tools for mediated collective governance.

​Colleen McKenzie is a researcher, designer, and product strategist, interested in exploring and improving humans' interaction with technology in a variety of contexts. She co-founded the Median Group, a scientific research nonprofit, where she explores technological trends and their societal impacts. Other past roles include Product Manager and Software Engineer at Google, Chief of Staff at the Center for Humane Technology, and Head of Product for distributed computing startup Kalix Systems. She holds a B.A. in Computer Science and Neuroscience from Columbia.