The New Voice-to-Text feature is a Step Back from the Original

ChatGPT Feature requests
1

I just wanted to share some feedback about the recent changes to the voice-to-text feature. I really preferred the previous version, where you could tap the microphone, see the little timer running, and speak freely for as long as needed. After that, it would add your words to the text box, and you had a chance to review or edit before sending. It felt natural and user-friendly.

With the new setup, it’s a one-shot deal. There’s no way to pause, add on, or edit after speaking, and it feels more rushed and limiting. I also noticed you can’t use voice-to-text when captioning an image, which used to be one of the most helpful moments for it.

I really appreciate all the improvements that have been made overall, but I hope this specific feature might be reconsidered or made optional. The old style made longer or more thoughtful responses much easier to manage.

Please tell me I’m not the only one who thinks this.

2 Likes
2

Bro I agree the push to talk was way better :+1:

1 Like
3

I agree but for me personally, it was never as long as needed if you spoke for over 5 minutes there were continuous errors (e.g., network error, and transcription error) I reported this and suggested a change in the feature since I like the blue visual and waveform but miss the timer and ability to pause and edit. They responded and said that they took notes and were looking into it.

2 Likes