I’d love to have an audio only feature where I can talk and listen to responses given the context of my focus window.
For example, adding an voice/audio channel for interaction would add a new async layer to the experience. I’d want it to be aware of where my cursor is and what i’m highlighting and also the main file that has focus, etc.. But the main thing would be to just enable my mic and have it continually listening and respond to my questions (when i obviously authorize it; i’d of course want to turn it off when co-workers come around).
For example,
ME: “hey I’m looking at this new component I didn’t write. How many buttons are in the code?”
C: I see three buttons, “save”, “cancel” and “delete”.
ME: “okay go to the delete. Are these html or imported components? Also, go find the test file, and give me a report on how the tests are looking do you have any suggestions. Jump to the delete button i want to see it.”
C: …[etc]
Basically, just let me enable a mode to talk to and listen to responses from the AI.