Abstract
Communication is a fundamental human activity that has been continually reshaped by technological advancements. As digital and blended environments become central to daily life, people increasingly meet co-workers, attend conferences, and socialize with friends in virtual spaces. Yet real-world communication remains richer and messier than most virtual tools can support: people interrupt, overlap, whisper, form subgroups, and drift in and out of talk, continually renegotiating who is included, attended to, or heard. Recent advances in generative AI further expand the design space of communication systems, as conversational agents can participate in conversations, support the authoring of multi-party interactions, and help configure the spatial and temporal dynamics through which people communicate. Together, these developments create opportunities to rethink communication as a dynamic process shaped by space, time, composition, and interpretation.
This dissertation proposes conversational dynamics (the spatial, temporal, authoring, and revisiting structure of a multi-party conversation) as a first-class design material for human-human and human-AI communication. I design and implement four interactive systems that explore how these dynamics can be authored, adapted, and revisited across different communication settings. These systems examine how spatial configurations affect group formation, awareness, and presence; how temporal configurations support turn-taking and participation; how multi-party human-AI conversations can be scripted, configured, and tested; and how users can later make sense of conversations. Together, these contributions advance the design of interactive systems and the empirical understanding of multi-party communication, laying the groundwork for flexible, expressive, and human-centered conversational technologies.
