OpenAI improves ChatGPT’s voice features for a select group
AI company OpenAI is starting to roll out new voice features for its ChatGPT chatbot to a small number of ChatGPT Plus subscribers in an early alpha test. said on X on Tuesday.
The startup gave a sneak peek of the advanced voice mode during its Spring Update in May, where it also introduced the GPT-4o model.
Users with access have taken to social media to share their initial experiences, including getting help with French sayings, imitate an airline pilot speaking from the cockpit and imitating seven Regional dialects of the USThe New York and Midwestern accent could use some improvement, but the chatbot knows that New Yorkers fold their pizza.
OpenAI isn’t alone in its ambitions for voice chatbot functionality for subscribers who pay $20 a month for perks like early access. Google also shared plans for a more conversational Gemini chatbot via its Gemini Live feature for Gemini Advanced subscribers, who also pay $20 a month. Meta’s Meta AI chatbot can also chat with users wearing its Ray-Ban glasses.
This is an example of how tech companies continue to roll out new models and features in an appeal to users, which is also a constant game of one-upmanship. The prize? The largest share of the generative AI market, which expected to be worth $1.3 trillion in 2023.
Hello, ChatGPT
According to OpenAI, the advanced voice mode lets you have more natural real-time conversations with ChatGPT. It also senses and responds to your emotions — and you can interrupt them if you want.
You can open ChatGPT with the familiar phrase: “Hey, ChatGPT.”
Further, details about what exactly this advanced functionality entails are unclear. A spokesperson did not respond to a request for comment.
Subscribers in the alpha test will receive a notification in the ChatGPT app, along with an email with instructions on how to use it. The goal of the early trial is to monitor usage and improve the model’s capabilities and security ahead of a broader rollout, a spokesperson said in an earlier email.
OpenAI will expand access to additional subscribers in the coming weeks, and plans to offer advanced voice functionality to all Plus members in the fall. In addition to early access to new features, Plus members also get an always-on connection and unlimited access to GPT-4o. (If you’re on the free version, you’ll be reverted to the previous GPT-3.5 model if you ask too many questions or experience heavy traffic.)
ChatGPT first introduced voice functionality in September 2023.
The Advanced Voice mode will include four preset voices, Breeze, Cove, Ember and Juniper, that OpenAI developed with voice actors in 2023. There was originally a fifth voice, Sky, but this was discontinued after actor Scarlett Johansson, who voiced virtual assistant Samantha in the 2013 film Her, complained about the similarities to her own voice.
CEO Sam Altman apologized to Johansson in a statement, but said the voice was not meant to resemble hers.
In a related blog postOpenAI said it selected the voice actors for its voices based on finding talent from diverse backgrounds, and on voices that feel timeless, voices that are approachable and trustworthy, voices that are warm, engaging and charismatic, and voices that are natural and easy to listen to.
OpenAI said ChatGPT cannot imitate voices and added filters that block requests to generate copyrighted audio.
