IT Home June 8th news, ChatGPT has upgraded its advanced voice mode for its paid users. This update has significantly improved its voice in terms of voice intonation, nature and emotional expression, making the interactive experience smoother and more "human".
According to OpenAI Introduced, this advanced voice mode upgrade further optimizes the naturalness of the voice, adding more delicate tone changes, closer to the real speech speed (including pauses and emphasis), and more accurate emotional expression, covering a variety of emotions such as sympathy and irony.
In addition, advanced voice mode also adds intuitive and efficient multilingual translation capabilities. Users only need to request speech for language translation, and they will continue to provide translation services throughout the conversation until the user requests to stop or switch languages.
IT Home noted that the update was based on ChatGPT's improvements to the voice mode earlier this year, aiming to reduce voice interruptions and optimize voice accents.
During testing, OpenAI found that this update could occasionally cause a slight drop in audio quality, including unexpected changes in tone and pitch. These issues are more evident in some voice options, but audio consistency is expected to gradually improve over time. In addition, despite the upgrade, there are still a very small number of “illusions” phenomena in the voice mode that may produce unexpected sounds similar to advertisements, nonsense or background music. Currently, the development team is actively investigating these issues and is committed to finding solutions as soon as possible.
