SAN FRANCISCO (AP) — The latest update to OpenAI's artificial intelligence model can mimic the rhythm of human verbal responses and even try to detect people's moods.
This effect is reminiscent of Spike Jonze's 2013 film Her. There, the (human) protagonist falls in love with an artificially intelligent operating system, leading to some complications.
The new model will appeal to few, but OpenAI says it runs faster than previous versions and can reason about text, audio, and video in real time.
GPT-4o, which stands for “Omni,” powers OpenAI's popular ChatGPT chatbot and will be available to users in the coming weeks, including those using the free version, the company said in a short The announcement was made during a live streaming update. CEO Sam Altman, who was not a presenter at the event, simply posted the word “she” on her social media site, X.
During a demonstration with Chief Technology Officer Mira Murati and other executives, the AI bot chatted in real time, adding emotion, especially “more drama,” to its voice on request. It also helped him take the steps necessary to solve simple mathematical equations without spitting out the answers first, and the more complex software on the computer screen also helped him with his coding problems.
It also looks at a selfie video of a person's face and infers their emotional state (we think they're happy because they're smiling), and translates between English and Italian to help inform conversations between people who speak different languages. I've shown you how useful it is.
Gartner analyst Chirag Dekate said the update took less than 30 minutes but gave the impression that OpenAI was trying to catch up with its bigger rivals.
“Many of the demos and features that OpenAI showed off seemed familiar to me as I had seen advanced versions of these demos that Google showed off at the Gemini 1.5 pro launch,” Dekate said. Masu. “Last year, Open AI had a first-mover advantage with its ChatGPT and GPT3, but compared to its competitors, especially Google, there is now a capability gap.”
Google is holding an I/O developer conference on Tuesday and Wednesday, where it will announce updates to Gemini, its AI model.
