🐈 Github
https://github.com/ekkezia/listening_machines/tree/main/darwin_theatre
🖍️ Notes

I tried using Whisper to transcribe my commands ("up", "down", "left", "right") that will control the laptop cursor position using pyautogui.

Potential exploration to create a voice-narrated/commanded/triggered story that will take the computer screen as the show stage. Thinking of an existentialist computer that listens to an AI-generated voice command agent (e.g Siri, Alexa) that guide them step-by-step to ironically annihilate the computer process that's running the program to transcribe the voice agent command into computer events (e.g, mouse position, clicking).

Here it is! A simple narration to serenade the application that runs the code with the voice command (VS Code) to end itself. The actor is the cursor. And spoiler alert, it blindly (or more like... deafly) obeys the voice command that will eventually brings the program into its demise.

To put it simply, the program is triggered by saying "Darwin, take over" (an homage to Darwin's survival of the fittest theory and just making a little joke on how this voice command application will survive), and then it will concurrently optionally narrate the story with a machine generated voice, listen to the sound using Whisper and trigger the computer events.

Darwin happens to be the name of my computer OS too.

Blog image

Elizabeth Kezia Widjaja © 2026 🙂