OpenAI announced on Monday (25) a new and important update to ChatGPT. Now, the artificial intelligence tool is able to receive commands with images and by voice, as well as respond by talking to the user – but only if access occurs via smartphones. The news will arrive first for users of the paid services ChatGPT Plus and ChatGPT Enterprise.
The company said the update will arrive for all users over the next two weeks — with the first customers getting the news starting today. OpenAI also reveals that, after this period, it will release the use of these novelties to developers. However, it did not reveal when the new features will arrive for non-paying users.
ChatGPT now listens, sees, and speaks—at least on mobile phones
The new ways to send commands to ChatGPT, in addition to showing the evolution of artificial intelligence, should deliver more convenience to users – especially the feature of image prompts. Last week, OpenAI reported that Dall-E 3 has been integrated into ChatGPT.
As the OpenAI account on X/Twitter and for example on its website shows, ChatGPT will be able to identify a photo and assist the user after analyzing the content of the image. On the social network, the company posted a video of the AI explaining how to adjust the seat of a bicycle.
ChatGPT asks for a more detailed picture of the seat attachment point, tools the user has, and the bike manual. In the video, the artificial intelligence identifies that the bank needs an Allen key, also known as a hexagon, and that this tool is in the box — analyzing the image uploaded by the user.
OpenAI also showed off the voice prompt and spoken response feature. In the example, ChatGPT has a female voice and, according to a user’s voice command, tells a children’s story. The story is not completed by the AI in a single prompt, but the user makes one more command to follow up the children’s tale.
If you subscribe to ChatGPT Plus or Enterprise, you can enable the voice feature (if it has already been released to you) through the app’s settings.