 |
Phone screen displays GPT-4o |
ChatGPT-4o will be rolled out to testers in the coming weeks.
A new version of ChatGPT can read facial expressions, mimic human speech patterns, and conduct near-real-time conversations, its creators have revealed.
OpenAI demonstrated an upcoming version of its artificial intelligence (AI) chatbot, called GPT-4o, in a real-time presentation. The chatbot, which spoke to the presenter over the phone, appeared to have an uncanny command of human conversation and its subtle emotional cues — switching between robotic and vocal voices on command, adapting to interruptions, and intuitively processing facial expressions and surroundings.
During the demonstration, the AI voice assistant showed off its skills by completing tasks such as translating languages in real time, solving math equations written on a piece of paper, and guiding a blind person around the streets of London.
Sam Altman, CEO of OpenAI, wrote in a one-word post on the social media platform X after the presentation ended. The post was a reference to the 2013 film of the same name, in which a lonely man falls in love with an AI assistant.
To demonstrate its ability to read visual cues, the chatbot used the phone's camera lens to read an OpenAI engineer's facial expressions and describe their emotions.
These new capabilities are a huge improvement over the limited voice features in previous models—which were unable to handle interruptions or respond to visual information.
“We’re looking at the future of human-machine interaction,” Mira Murati, OpenAI’s chief technology officer, said at the press conference. “We think GPT-4o is really changing that paradigm.”
The new voice assistant is expected to be released in limited form to alpha testers in the coming weeks, followed by a wider rollout that will begin with paying ChatGPT Plus subscribers.
The announcement also follows a Bloomberg report that the company is close to reaching a deal with Apple to integrate ChatGPT on the iPhone — opening up the possibility that GPT-4o could be used to upgrade Siri, the iPhone's voice assistant.
However, this new technology also raises concerns about safety and making fraudulent phone calls.
Comment (0)