Jump to Content

Social Simulacra: Creating Populated Prototypes for Social Computing Systems

Joon Sung Park
Lindsay Popowski
Meredith Ringel Morris
Percy Liang
Michael S. Bernstein
Proceedings of UIST 2022, ACM (2022) (to appear)

Abstract

Prototyping techniques for social computing systems often recruit small groups to test a design, but many challenges that threaten the norms and moderation standards do not arise until a design achieves a larger scale. Can a designer understand how a social system might behave when later populated, and make adjustments before the system falls prey to such challenges? We introduce social simulacra, a technique enabling early prototyping of social computing systems by generating a breadth of possible social interactions that may emerge when the system is populated. Our implementation of social simulacra translates the designer’s description of a community’s goal, rules, and member personas into a set of posts, replies, and anti-social behaviors; shifts these behaviors appropriately in response to design changes; and enables exploration of "what if?" scenarios where community members or moderators intervene. We contribute techniques for prompting a large language model to generate such social interactions, drawing on the observation that large language models have consumed a wide variety of these behaviors on the public web. In evaluations, we show that participants were often unable to distinguish social simulacra from actual community behavior, and that social computing designers could use them to iterate on their designs.