Writing versus speaking with superwhisper
superwhisper has been making the rounds recently as an app for writing with your voice. It, along with OpenAI, seems to eschew the idea that a word should begin capitalized.
Nevertheless, I’ve been thinking about this app a lot lately. It claims to take your voice and turn it into legible writing. I tried it, found it frustrating, and went back to typing. It just doesn’t work for me. Namely that it seems to get confused when I’m speaking versus not speaking, a vital feature I was hoping would just work.
I’ve tried voice to text for years now, tested every app available and they all fail at the basic idea of just capturing what you say and nothing more or less. The better solutions are ones that take already recorded audio and transfer sound to text. But it’s not quite the same as something that could work live. Live just doesn’t cut it though, at least not with any apps I’ve tried. Training myself to speak differently is an option, but defeats the purpose of hoping that hardware and software can just get out of the way.
The allure is there, though. I’d love to freeform speak, have my thoughts captured, and turn that into writing.
However, as Jason Snell mentioned on Upgrade this week, speaking isn’t the same as writing. When I sit down to put words to page, even with my somewhat undetermined style, I’m still thinking and formatting. When I get stuck on a sentence I go back to the beginning, re-read my words, insert a sentence or two here and there, and then return to my most forward sentence and continue writing.
This process is different than talking to a friend, or even podcasting. In the latter example you generally have an outline, a buddy to riff off of, and an understanding from your audience that the conversation is free flowing. That works fine with audio, or video, where the inflection of your voice, pauses, and even smiling while speaking impact the words you say.
In writing you have none of that. You only have the words. So a little more structure is crucial to bring some semblance of legitibility.
So, while apps like this are exciting, I don’t see them replacing the way I like to write. I’m also not trying to write perfectly. I want legibility, and hope for my intent to come through. But perfection isn’t the point. Pulling an idea through, conveying that idea, and hopefully resonating with someone, these are the reasons I write. I want to figure something out, so I write. I’m curious about something, so I write.