An interactive sound experience with a language model in Virtual Reality. It maps the inner workings of the AI, as well as the user's physical movements and behavior, onto sound.
Spoken words are transcribed and sent to a server, where the AI generates an answer. The hidden data behind the model's "thoughts", including its internal processing states, is sent back to the WebXR or Android interface. There, custom-built audio software transforms this data into sound in real-time.