OpenAI is gradually introducing its new advanced voice mode for ChatGPT, available to a limited number of ChatGPT Plus subscribers. The feature, unveiled at the GPT-4o launch event in May, was initially criticized for sounding similar to Scarlett Johansson’s voice and was later delayed due to security concerns.
The new voice mode promises to overcome the limitations of the previous version, offering a more fluid and natural conversational experience. During the demonstration in May, users were able to interrupt the chatbot and ask it to tell a story in different ways, receiving appropriate and coherent responses.
The enhanced voice mode was scheduled to launch in June, but OpenAI decided to delay it by a month to “get to our launch standards.” During that time, the company said it worked to improve the model’s ability to detect and reject certain content, bringing in more than 100 outside experts to test the model’s safety. OpenAI also added new filters to block requests to generate music or other copyrighted audio.
One of the main criticisms of the new voice mode was that the voice of “Sky” sounded similar to that of Scarlett Johansson, who played an AI in the movie “Her.” OpenAI decided to remove the voice shortly before the actress sent a letter to the company demanding an explanation for its creation. The new mode will only use four preset voices created with professional actors, and will be able to block the imitation of other voices, both individuals and public figures.
OpenAI plans to roll out Enhanced Voice to all ChatGPT Plus subscribers this fall, ushering in a new era of AI interactions. What do you think of this new feature? Let us know in the comments below.
