Imagine you want to enjoy the rich, life-like quality of ChatGPT’s spoken responses but don’t wish to speak your prompts aloud. Perhaps it’s late and you’re trying not to wake your spouse, or you’re on a crowded train heading to work, earbuds in, and you want to keep your ChatGPT conversation to yourself. Sure - you can click the button to manually read aloud each ChatGPT response, one after another, but that gets tedious and doesn’t promote the natural flow of conversation. That’s where mixed input comes in!
Type your prompt using the usual on-screen keyboard but receive a spoken response, just as you would in normal voice mode. I’m imagining this setting being easy to access, perhaps by long-pressing the voice mode button to reveal a context menu where mixed mode could be chosen (don’t forget to make it screen reader accessible ;)).
This would give users greater flexibility in how they interact with ChatGPT and integrate it into the rhythm of their daily activities.