If you want your AI to be more human-like, you need a good voice. AI voice technology has exploded in realism and one of the companies making the most realistic voices is OpenAI. According to a new report, OpenAI could triple the number of voices for ChatGPT.
We’ve heard a few scarily realistic voices from OpenAI. One of them even closely resembled actress Scarlett Johansson. At the moment, there are four voices that you can choose from in the mobile app. These voices are available to all users including free users. However, there are some advanced voice features in production on their way to ChatGPT Plus users.
To hear the voices, you’ll need to tap on the little headphones icon in the app. You’ll be able to choose the voice that you want to hear. Just speak to ChatGPT just like you were speaking to a person, and you’ll get an answer promptly. When you start a conversation with a voice, it will still be saved.
OpenAI could triple the number of available voices in ChatGPT
This isn’t official just yet, so you’ll want to take it with a grain of salt. Tibor Blaho was able to reverse engineer the ChatGPT app and extract code for an additional eight voices that might be coming to the platform. He posted a video to X showcasing these new voices. We don’t know what these voices will be called, but the video shows the possible names. We don’t know if these will be the official names or if they’re just code names.
The names in the video are Fathom, Glimmer, Harp, Maple, Orbit, Rainbow, Reef, Ridge, and Vale. They’re a mix of male and female voices.
Along with having different tones and vocal characteristics, these new voices might also be better at emulating animal sounds. In the video, we hear the voices saying “woof, woof” and cawing like crows. The believability fluctuates a bit between the voices, but that could just be because the company is still working on them.
Not only that but the voices could also be more expressive. When reading back text that’s bold or italicized, the voice could add more emphasis. For the time being, we don’t know when these voices will hit the market or if they’ll be available to free users. Only time will tell.