All visuals are generated live and controlled by text prompts and audio-reactive AI.