ChatGPT’s new voice mode update has been released for some paid users, but the voice definitely won’t sound like Scarlett Johansson.
Last week, OpenAI CEO Sam Altman announced the imminent release of ChatGPT’s new Advanced Voice Mode, and as of Tuesday it was available for ChatGPT Plus subscribers, who pay $20 per month. Notably, there are four pre-set voices available, rather than the five announced at launch, and the voices are: she Following the backlash, the actor was fired by the company.
The company announced the rollout on Twitter on Tuesday, declaring it available to “a small group of ChatGPT Plus users.” OpenAI added in the thread that it plans to “continue to gradually expand access to more users, and will open it up to all Plus users in the fall.”
The tweet may have been deleted
When OpenAI launched GPT-4o in May, the new AI model, which blends text, video and audio with five preset voices, included a “Skye” voice option that sounded eerily similar to Johansson’s performance as the digital assistant in the 2013 film “Star Wars: Episode I: Skye.” sheJohansson was not convinced and took legal action, which led to OpenAI retracting its statement.
Mashable Lightspeed
In case you missed it, ChatGPT for Mac had a serious security issue
“To protect people’s privacy, we trained our models to speak only in four pre-defined voices and built systems to block output that doesn’t match those voices,” OpenAI said in a tweet on Tuesday. “We’ve also implemented guardrails to block requests for violent or copyrighted content.”
So if you’re a ChatGPT Plus subscriber, how do you know if you’re part of this “small group”? “Users of this alpha will receive emails and mobile app messages with instructions,” OpenAI explained in the X thread.
The tweet may have been deleted
OpenAI added that it had “tested GPT-4o’s voice capabilities with an external red team of over 100 people across 45 languages.”
The update was scheduled for release in June but has now been delayed, OpenAI said, adding that it will “share a detailed report on GPT-4o’s capabilities, limitations, and safety evaluation in early August.”
OpenAI also announced a number of additional features alongside the voice mode update, with Mashable’s Alex Perry reporting the details.
