← Explore Research Questions
Resource
Generative Social Choice: The Next Generation
Extends the Generative Social Choice framework with theoretical guarantees for approximately optimal queries and budget limits on overall slate length, addressing the challenge of producing representative slates of statements from open-ended user opinions. Uses GPT-4o to demonstrate effectiveness on city improvement measures and drug review datasets. Accepted to ICML 2025.
Related Capabilities
Aggregate perspectives
Representativeness
Ability to aggregate votes and distill more complex forms of open-ended input into outputs and decisions, in fair and understandable ways, such that participants feel their contributions are meaningfully taken into account (and can ideally see how).
Enumerate scenarios
Urgent Informedness
Ability to generate lists of likely scenarios, including edge cases, in which decisions will be applied, to help participants better understand the issue space.