I was so excited about the new voice feature, and I understand that it is new and there will be bugs, but the bug that I’m running into is so frustrating and honestly defeats the purpose and my use case for using the feature.
Before, on the commute home, I used to be able to have long conversations with the old voice feature. As a non-coder, developer-wannabe hack, I flushed out some great ideas going back and forth with voice chat over long drives. Now, it can’t handle more than a two-minute conversation. I’ll start a conversation with it, I’ll respond, and once it responds to me, it will get maybe 100 words in and just drop off. It’s not just the voice that’s dropping off—it’s the text generation as well. I understand that the new feature’s functionality, and what makes it so good, is that it doesn’t have to generate the text and then read it. It can do them simultaneously. But when it can’t handle doing both simultaneously and just stops generating text completely, and the conversation just drops off, there’s no point.
When I ask it to resume, its response, unless I specifically stipulate to go back a paragraph, a lot of times will just see my ask as a new prompt and not continue on the same chain of thought as it was going down when it cut off.
I love the responsiveness of the voice feature, and I think it’s going to be amazing, but as of right now, it’s almost not worth my time.
Is anybody else running into the same bug? Any recommendations or fixes?