How Might We?
How might mixed reality, volumetric media, and AI reshape storytelling, interaction, and presence?
This page collects the questions, systems, and experiments guiding my work.
How might perception become interface?
Spatial Synesthesia: a Vision Pro build where gaze drives orchestral composition in real time.
Built with visionOS + RealityKit.
Selected for demonstration at ASU Emerging Media.
How might volumetric cinema respond to space in real time?
ASU MA capstone: a mixed reality film in 4D Gaussian Splatting, with actors captured and steered by spatial agency in real time. Pipeline work continues at ASU Los Angeles.
Now:
- capture → compression → playback pipeline
- browser-capable delivery for WebXR stages
Pushing from recorded performance toward spatially responsive cinema.
How might AI reshape creative production without flattening authorship?
- 50%+ faster iteration: AI-assisted workflows, human review gates
- SIGGRAPH 2025, Vancouver — Immersive Pavilion Subcommittee; AI prototyping for live production
- Agentic direction in MR film: where it holds, where it fails
How might volumetric media be delivered efficiently across real-time environments?
Compress, ship, and run splats in real time:
- WebXR delivery pipelines
- mesh compression + texture baking
- load-time vs. fidelity tradeoffs
At Structural Color Gallery, cut WebGL exhibition load times by 50% across 20 artworks; 20+ usability sessions.
Cross-cutting threads
Problems that sit outside a single project—carried forward from lab, set, and exhibition work.
- What makes a volumetric actor feel present—not just visible—in MR?
- Where is implicit input (gaze, vision) robust enough to ship?
- How do spatial stories adapt when the room is part of the scene?
- When does AI direction deepen a take, and when does it erode it?