Why trust Greenbot
TestingCatalog has just discovered eight new voices that could eventually be added to ChatGPT’s advanced voice mode through reverse engineering. The new voices, named Fathom, Glimmer, Harp, Maple, Orbit, Rainbow, Reef, Ridge, and Vale, come with significant improvement to the current four voices: Juniper, Ember, Cove, and Breeze.
Tibor Blaho, an engineer from TestingCatalog, posted samples of the new voices on X. He noted that these voices not only vary in gender and accents but also possess the ability to mimic animal sounds.
The introduction of these voices comes after the initial rollout of ChatGPT’s Advanced Voice Mode in July to a select group of ChatGPT Plus users. OpenAI’s Voice Mode, part of the GPT-4o model, integrates text, vision, and audio modalities into one cohesive system. This model aims to offer a more natural and human-like interaction, setting it apart from other AI platforms like Google Gemini and Meta AI. That is, if you ignore the controversies regarding similarities between one of its demo voices and actress Scarlett Johansson.
The new voices are expected to enhance the user experience by providing a more varied and realistic interaction. For instance, Fathom and Vale are described as having posh British accents, while Reef has an Australian tone.
Additionally, the ability of these voices to replicate animal sounds adds a unique layer to the interaction. According to Blaho’s demonstrations, some voices attempt to mimic the actual sounds of dogs and crows, while others simply say “woof” and “caw.” This feature could be particularly appealing for users looking to build their own virtual pet using the system.
Another significant aspect of these new voices is their impact on content creation. They offer creators in fields like storytelling, podcasting, and other creative domains the ability to use AI as a voice actor or narrator, with new possibilities for creators to experiment with different vocal tones and styles.
These are, however, not yet available to the general public. TestingCatalog has noted that the current dropdown menu does not have space to accommodate eight additional voices.
OpenAI has not provided a specific timeline for when these voices will be widely accessible, but the ongoing development and testing suggest that a broader rollout may be on the horizon.
Moreso, user feedback and safe ethical considerations should play a crucial role in driving future improvements and innovations in OpenAI’s voice technology in order to meet user needs and avoid possible backlash from the public media.
