11:41 AM
Tuesday, May 14, 2024
OpenAI, the developer of the Chat GPT robot, announced yesterday the launch of its new artificial intelligence model, GPT-40, and said that it can conduct a realistic voice conversation and deal with texts and images.
The new voice capabilities allow users to speak to GBT Chat and get real-time responses without any delay, as well as interrupting them while speaking, both features of real-life conversations that have eluded previous AI voice assistant services.
“It feels like artificial intelligence is a fantasy,” said Sam Altman, CEO of OpenAI. “Talking to a computer never came natural to me, and now it does.”
In a live conference, OpenAI researchers demonstrated the new voice assistant capabilities of GPT Chat. In one demonstration, GBT Chat used its visual and audio capabilities to talk to a researcher to solve a mathematical equation on a piece of paper.
In another demonstration, the researchers demonstrated the GPT-40 model’s ability to translate in real time.
OpenAI’s demos were science fiction-esque, with GPT Chat and his interlocutor at one point engaging in banter. The OpenAI researcher told the chatbot that he was in a great mood because he was showing “how helpful and amazing you are,” and ChatGPT responded, “Stop that! You’re making me blush!”
Mira Moratti, chief technology officer at OpenAI, said at the event that the new model will be offered for free because it is more cost-effective than the company’s previous models. However, it clarified that the paid version of GPT-40 will be broader in scope than the free one.
The company said that GPT-40 will be available in the next few weeks.
Read also:
Selfies of the Pharaohs… 12 strange images of artificial intelligence from ancient times