The Porous Wall of Consequences - 2025

This is a core component of my MFA thesis and an exploration of how generative AI can operate as a situated performance design system. The work integrates LLM tools for TouchDesigner developed by Lyell Hintz (Dot Simulate) to generate multiple versions of a narrative alongside real-time visualization.The audience speaks directly to the system through Assembly AI’s speech-to-text API. Responses are produced by a custom Gemini agent inside TouchDesigner and rendered aloud via ElevenLabs’ text-to-speech, which is simultaneously visualized. These responses also drive live imagery through Google Gemini 2.0 Flash’s experimental image-generation API. The Installation ends with a physical fabric tableau becomes a mutable “painting” that viewers can transform by moving and repositioning textiles, processed locally through Stream Diffusion. The project engages the ongoing struggle of Iranian people under a theocratic regime by asking participants to confront questions about media and perception, the use of violence, systems of power, and the ethical responsibility of both citizens and spectators. (Presented in The University of Texas at Austin’s Theatre and Dance department on 1/24/2026)