Session

DescriptionAs AI disrupts creative and economic models in arts and entertainment, most attention has focused on centralized, cloud-hosted capabilities and media generation. This special session explores a shift toward decentralizing AI models onto audience devices and engaging with spatial storytelling in the physical world, highlighting new experiential possibilities enabled by lower costs, enhanced privacy, reduced latency, and offline collaboration.

Jeff Burke (UCLA Theater, Film, and Television): Burke will open with research on future fan experiences powered by AI inference distributed across mobile, edge, and cloud devices—including media generation, vision, and language models on mobile hardware. He will use upcoming media and entertainment projects from UCLA REMAP to illustrate the technical tradeoffs and challenges of distributed processing versus cloud-first approaches, and to suggest key research opportunities.

David Beaudry (Beaudry Interactive): Expanding on Burke's use cases, Beaudry will discuss the link between systems innovation and audience experience in large-scale themed entertainment for clients including Disney and Universal Parks, Sesame Workshop, and Carnival Cruise Line. This domain brings together mobile, spatial, and embedded computing technologies in demanding real-world environments with unique requirements.

Achuta Kadambi (UCLA Electrical Engineering and Computer Science): Kadambi will conclude with a state-of-the-art technical perspective on AI approaches that construct, update, and act on models of changing worlds, covering work in his lab that illustrates how spatial intelligence requires grounded representation, persistent memory, active perception, and physically consistent reasoning. Within the talk, he will discuss limitations and opportunities of agentic processing with spatial AI that could enable new applications like those introduced in the preceding talks.