I find myself switching between ChatGPT voice mode on my phone, and Codex a lot to copy and paste requirements that I’ve brainstormed via voice mode. It’d be nice if I didn’t have to switch back and forth and that voice mode feature was available to me now that we have a mobile Codex. It’s kind of like I’m needing to switch between two apps (which I guess I am) but it would be nice if it was integrated together so that in plan mode I can brainstorm with my voice alongside Codex and then when I’m done we can see the requirements based on what I’ve brainstormed with Codex.
By now, I think you can improve it by connecting your ChatGPT thread to your repository with the Github app. ChatGPT is able to read your repository and give you more accurate information about your project during the brainstorm.
Another approach is the speech-to-text in the Codex App.
Or you can go further and build your own rich chat with Codex App Server and some real-time voice api or some application that automates it for you.
But if you get some better approach it would be nice to know.
Pretty cool ideas, I’ll experiment around with this and definitely suggest better approaches when I come across them!
This is pretty cool and similar to what I’m talking about. Thank you for sharing!