Dictate into ChatGPT for as long as you want now, and remember that every word of it still has to go somewhere. I’m still a bit of a traditionalist when it comes to this functionality. I would almost rather record things (like a lecture or meeting) using just a recording app or device, then take the recording and create a transcript from it. Then I can run it through whatever AI I choose. Yes, it takes a bit longer, but I control where it goes and how it is used.
Also, many companies have built hardware and software that can do what ChatGPT can now do (but ChatGPT is a bit more limited… for now). I wonder if, with the release, those hardware and software companies are concerned about their business model. They do have “first mover” as an advantage; however, OpenAI just has to release some hardware that does this and fold their services into the integration. The writing has been on the wall for a while.
I still don’t like doing mass dictation to my AIs, but I have a feeling this may change as these audio features become more baked in.