I wish I could have the text output instead of getting a vocal answer. We all can read way faster than we can listen to speech, and I don't need a lot of the fluff I get from LLMs. But I guess that's what most users want? Something that's closer to a conversation with a human? But I don't care about being polite with an LLM, it feels like an anti-pattern because it's too much overhead.
Ideally I would:
* query by voice
* get the response in text as soon as possible, maybe even as I'm still talking, with the LLMs correcting the response as I give more information, and without needing me to touch anything
* follow up by voice, even before the response was finished, and without needing to press anything
> We all can read way faster than we can listen to speech
I can’t. At least not with reading every word. I can finish an audiobook much faster than reading the book on my own.
And I also can’t pick up as much reading by myself. My preferred way of consuming is now reading while a screen reader or audiobook is played at 1.5x or faster speed. Much better retention because
That's the crux of the issue though, isn't it? I don't read words by reading single letters, and I don't think I read sentences by reading every single word either. That'd be the main reason why listening is slower than reading: the forced linearity.
To be fair, my "all" was clearly an exaggeration, I'm sure there are people for whom this doesn't apply, but I'd still expect it to apply to 50%+ of the populations of countries where literacy is not a significant issue anymore.
Interesting, it's exactly the opposite for me. My understanding of voice is much lower than of reading, and much slower. I think the biggest factors are that reading is faster than words are typically spoken, yet I can also pause and dwell on sections I don't get the first time, so I really get it and then proceed.
Obviously, brains work very differently, so there really is no one-size-fits-all solution; we need the varied methods.
Ideally I would:
* query by voice
* get the response in text as soon as possible, maybe even as I'm still talking, with the LLMs correcting the response as I give more information, and without needing me to touch anything
* follow up by voice, even before the response was finished, and without needing to press anything
I hope we'll get there eventually!