ChatGPT’s new Advanced Voice Mode was released to a few lucky users. It lets you talk to the AI with real voices, not just text.
This feature aims to make conversations more natural. You can even interrupt it, and it will respond to your emotions.
However, there are limits. It uses one of four preset voices and won’t mimic real people. Also, it can’t produce copyrighted audio or music.
But users are already having fun with it, making it beatbox and create sound effects for stories.
A user named Ethan shared a clip where ChatGPT made a short rap and beatboxed. Another user, Kesku, had it add sound effects to a sci-fi story.
The AI’s ability to create characters and add emotions makes it seem almost human. Ethan showed how it can recreate famous movie scenes with emotion.
The AI can also speak multiple languages and make realistic animal sounds. One user had it make cat noises, while another used it to get advice on caring for a new kitten.
It even works as a real-time translator, using the device’s camera to translate Japanese text on a GameBoy screen.
Though it’s in alpha now, OpenAI plans to expand this feature to more users soon. It’s a promising step towards making AI interactions feel more like talking to a real person.