OpenAI’s ChatGPT is now more complex than ever with the introduction of voiced conversations. Unlike previous versions of the generative chatbot, users will now be able to voice queries to the AI program instead.
Revealed in an announcement on OpenAI’s blog, ChatGPT voice functionality is rolling out to the service now. Those who still use the popular internet tool can pretend they have a friend with unlimited thoughts.
Alongside voice capabilities, all versions of ChatGPT will also include AI image functionality. Users will now be able to take a picture of any object and have a fully voiced discussion with the chatbot about the image.
“Voice and image give you more ways to use ChatGPT in your life,” OpenAI writes in the features’ announcement. “Snap a picture of a landmark while traveling and have a live conversation about what’s interesting about it. When you’re home, snap pictures of your fridge and pantry to figure out what’s for dinner (and ask follow up questions for a step by step recipe). After dinner, help your child with a math problem by taking a photo, circling the problem set, and having it share hints with both of you.”
However, this new advanced feature won’t be available to all ChatGPT users. Instead, OpenAI is locking the advanced features behind the ChatGPT Plus and ChatGPT Enterprise subscription.
ChatGPT voice will have five AI generated voices to pick from, three feminine and two masculine. Dubbed Breeze, Sly, Cove, Ember and Breeze, these realistic AI voices will read out responses to users such as generated recipes, jokes or conversations.
ChatGPT image recognition has been built from a prior OpenAI model designed to help blind people. The Be My Eyes mobile app was a service designed to help blind people recognise objects with AI, and that information and training has been pumped into ChatGPT.
While ChatGPT voice and images will be locked behind a paywall, more users are expected to have access sometime in the near future. The software is available on iOS, Android and PC.