Cinema Multiverse Lounge: Enhancing Film Appreciation via Multi-Agent Conversations

Ryu, J.*, Kim, K.*, Heo, D., Song, H., Oh, C., & Suh, B.

* denotes equal contribution.

Proceedings of the 2025 CHI Conference on Human Factors in Computing Systems

Multi-AgentConversational AIFilm AppreciationHuman-AI Collaboration

What if a film discussion had several AI voices?

A memorable film often leaves us with more than one possible interpretation. Cinema Multiverse Lounge explores whether AI can enrich that experience by bringing different perspectives into a conversation, from a character’s point of view to a filmmaker’s or another audience member’s.

Let viewers choose their conversation partners

We built a system with AI agents representing different film-related personas. Thirty participants used it across three discussion sessions and could freely choose who joined the conversation. We studied how they assembled these groups and how the resulting discussions shaped their engagement and interpretation.

An illustration of a viewer discussing a film with AI personas representing a director, actor, character, and audience members. — The project's teaser illustrates a discussion across several film-related perspectives. These are AI personas, not the actual people depicted by those roles. Cinema Multiverse Lounge.

Differences made the discussion interesting

Participants explored alternative readings and developed richer interpretations of the films. Unexpected disagreements between personas could also make the experience enjoyable. The value came from encountering different viewpoints, rather than receiving a single authoritative explanation of what a film meant.

Design for exploration

The study points toward AI companions that help people develop their own interpretations. Giving viewers control over the mix of perspectives is part of that design. These findings concern film appreciation in the studied sessions; they do not establish that adding more agents improves every conversation or makes their claims more accurate.

Read the original abstract

Advancements in large language models (LLMs) enable the development of interactive systems that enhance user engagement with cinematic content. We introduce Cinema Multiverse Lounge, a multi-agent conversational system where users interact with LLM-based agents embodying diverse film-related personas. We investigate how user interactions with these agents influence their film appreciation. Thirty participants engaged in three discussion sessions, freely selecting persona agents such as film characters, filmmakers, or anonymous audiences. We explored how users composed different combinations of personas, the factors affecting their engagement and interpretation, and how diverse perspectives influenced film appreciation. Results indicate that interactions with varied agents enhanced participants’ appreciation by enabling the exploration of multiple viewpoints and fostering deeper narrative engagement. Moreover, the unexpected clashes between different worldviews added a fresh and enjoyable layer to the interactions. Our findings provide empirical insights and design implications for developing multi-agent systems that support enriched media consumption experiences.