Technology media outlet testingcatalog published a blog post yesterday, September 2, reporting that by reverse engineering the ChatGPT application, it was found that OpenAI Soon to be expanded and addedcolloquial (rather than literary) pronunciation of a Chinese character, making reading aloud sound more natural and expressive.
There are indications that OpenAI may introduce an additional eight new voices in the future, each with a unique codename, which may be rolled out incrementally in the future.
Another interesting feature of these new voices is that theThey are able to express sounds more naturally, such as animal calls or other non-verbal sounds.
In addition, they can convey or emphasize specific emotions when reading aloud bolded or italicized text. However, it's important to remember that these are still TTS (text-to-speech) voices, and are likely not related to the advanced speech patterns that are currently in alpha.