ChatGPT now has eyes, ears and mouth

OpenAI has equipped its text AI ChatGPT with voices. The company said that paying users of the service will be able to interact with the chatbot via voice in the future.

The user can choose from five voices – Juniper, Sky, Cove, Ember and Breeze – and receive real-time answers to questions in a voice chat. The voices were created in collaboration with voice actors, writes OpenAI.

Unlike voice assistants like Amazon’s Alexa or Apple’s Siri, the voices actually sound less artificial. OpenAI has published a few examples of text on its blog, which are read by the ChatGPT voices.

” class=””>In the blog post, OpenAI warns against misuse of the technology. While the feature offers the opportunity to create barrier-free and creative applications, it also poses risks if criminals use it for fraudulent activities – for example to settle claims claiming to impersonate other people, which is why the technology is only designed for this one specific use: voice chat.

ChatGPT learns to see to help with problems

With the new software update, OpenAI also brings a function that allows the AI ​​to recognize image content. ChatGPT teaches you to see, so to speak, writes OpenAI.

How does this work? Here’s an example: the user takes a photo of a landmark while traveling. He then uploads the image via the ChatGPT app. The AI ​​recognizes the motif and can tell you something about its origin

OpenAI sees another potential application in problem solving. For example, you can take a photo of a grill and ask the chatbot why it is not working. The AI ​​could help find the error.

OpenAI plans to make the new features available to all paying members of its service in the next two weeks. The voice feature is only available for the app on iOS and Android, it is said. However, image recognition will be available on all platforms.

Used sources: