A Python program that listens for computer audio (not mic audio) when pressing CTRL, transcribes speech via Whisper API, sends the transcribed audio to ChatGPT, and generates a response.
Requires VB Audio Cable. Set default playback and recording device to VB Cable, enable "Listen to this device" for the recording cable so you can hear computer audio.
Add your API key where it says api_key = "your_api_key" Change the system message for ChatGPT to your liking.