• Amazon companies will invest up to 4,000 million dollars in Anthropic, the great rival of the creator of Chat GPT
  • AI OpenAI already has ready the third version of DALL-E, its most creative artificial intelligence

ChatGPT has become in less than a year the most well-known artificial intelligence tool in the world. With hundreds of millions of registered users, this chatbot, developed by OpenAI, can answer any question asked in natural language or invent texts from a few guidelines using any style imaginable.

But, from now on, he will also be able to speak, listen and see. The latest version of the tool, both in its web version and in apps for mobile devices, adds the possibility of using images and voice to ask the different questions.

ChatGPT, in addition, will now be able to give answers not only as a block of text but aloud, using five voices created, of course, with artificial intelligence. This will allow, for the first time, to maintain a natural conversation with this long model of language – which is how the technology that makes it possible for ChatGPT to create answers that seem to be written by a person is known.

It is, of course, a much slower experience than talking to a real person. The answer to a question can come after several seconds of awkward silence, similar to text answers.

Thanks to these new abilities, however, from now on it will be possible to use CHatGPT in a greater variety of situations. The app, for example, will be able to identify the different parts of a bike in an image, and help the owner adjust or fix something that has broken. You can also comment on the different points of interest of a landscape or recommend different recipes just by analyzing a photo of what is inside the refrigerator.

Find out more

Technology.

Speaking to Ernie, the censored Chinese ChatGPT: "Freedom is relative"

  • Editor: LUCAS DE LA CAL (Correspondent)Shanghai

Speaking to Ernie, the censored Chinese ChatGPT: "Freedom is relative"

The world wide.

John Maeda, Microsoft's design and AI genius: "Artists understand this technological age better than anyone"

  • Writing: ISRAEL ZABALLA Madrid

John Maeda, Microsoft's design and AI genius: "Artists understand this technological age better than anyone"

The new features are currently restricted to a small group of users and are exclusive to those who pay for the ChatGPT Plus service or the enterprise version of the application. According to the company, access will be expanded in the coming weeks among subscribers, although tools such as synthetic voices will initially be available only in English.

These new features will also allow third parties to offer new services. Spotify, for example, has just begun piloting an instant podcast translation service using presenters' own voice.

From the company, however, they point out that these new forms of interaction do not necessarily improve the reliability of the answers, one of the problems of this type of tools, since they can, sometimes, invent the data that they do not know to give coherence to the final text. "We are transparent about the limitations of the model and advise against the use of higher risk, such as specialized information, without additional verification," OpenAI explains.

  • ChatGPT
  • Artificial intelligence