raven wasp
We all know the previous model would occasionally say "Thanks for watching," likely because a lot of AI models were trained on publicly available videos where content creators often close with that phrase.
After using the new voice model since its release, I’m really eager to go back to the old, more functional voice model. The previous model actually worked and wasn’t just a conversational gimmick.
The new model interrupts itself at the slightest noise, like wind or a car in the background, and it constantly thinks I’ve "reaffirmed" what it said when I never said anything.
Please make the old model available again. This new one is practically useless. I only use it for casual conversations now, and I hate having to type out things to chatgpt just for it to have simple context via a search, which the old voice mode handled without a problem but was somehow downgraded and removed because of... voice?
The technical conversations I had with ChatGPT a month ago were in depth and complex and now, im just talking about The French Revolution because it has no idea that it was even updated because again, it cannot search in voice.
This new voice model is a MAJOR downgrade in every capacity. There's zero thats useful, if at all via voice.
uneven pike