What you need to know
- OpenAI used X to announce that it will delay the launch of enhanced speech mode, which was planned for this spring.
- The company is testing the new language model on a small group of people to gather feedback and insights before making it available to all Plus users.
- If everything goes according to plan, the remaining Plus users will have access to the extended language model in the fall.
Open AI was quick to tweet on Tuesday (June 26) that it would hold off on rolling out the enhanced speech mode it demoed last month. It had planned to roll out the feature widely in the software’s spring update, but it looks like people who wanted to experiment with it will have to wait a bit longer.
The new voice assistant is designed to have “human” conversations that users can interrupt, offering emotions and lifelike interactions like laughter, sarcasm, and humor that are a change from the usual robotic prompts we expect from an assistant. Although OpenAI promised a quick rollout for subscribers, the company has now announced a delay in the release of the updated “voice mode” (via Android Police).
We’re releasing an update to the enhanced voice mode that we demoed during our Spring Update and remain very excited about: we had planned to make this available as an alpha to a small group of ChatGPT Plus users in late June, but we still need a month to reach our launch level. …25 June 2024
While the new voice mode is still being tested, the company plans to give access to it to a small group of ChatGPT Plus users in late June. Some of them have already received an update on their phones. One of those users who got access shared an invitation she received with X. The invitation said: “The enhanced voice feature is in a limited alpha phase, bugs may occur and access to it is subject to change.”
Open AI said its systems will use the audio and video content the user shares through Chat GPT to better train the model. Users can also opt out of sharing their data and training the model by clicking “Improve model for everyone” in the settings.
The user also posted a tweet featuring the new voice of “Sky,” which appears to be different from the one we heard when Chat GPT-4o was announced. Initially, users on social media pointed out the similarities between the assistant’s voice and award-winning actress Scarlett Johansson. However, OpenAI denied that Sky’s voice was based on Johansson and even published a blog post explaining how the voices were chosen. The actress countered these claims with an explanation of how OpenAI contacted her about officially giving her the voice for GPT-4o, which she declined, and the company then quickly removed the Sky voice option from its platform.
The new voice has a “happier tone,” the user says, as the voice model describes the weather in San Francisco. Other users in the same tweet thread seemed unhappy with the new voice. One user said, “It’s horrifying; someone help him,” while others were particularly “cringe-worthy.” Listen below.
GPT-4o talks about summer in San Francisco, but in a very happy tone pic.twitter.com/6taoPeVLdh26 June 2024
OpenAI added in its tweet that based on the feedback they received from this small group of testers, they plan to roll out the model to Plus users in the fall.
“Exact timelines will depend on our ability to maintain our high standards of security and reliability. We are also working on rolling out the new video and screen sharing features we demonstrated separately and will keep you updated on that timeline.”
