OpenAI has announced a new flagship model of generative artificial intelligence called GPT-4o, where “o” stands for omni. The developers claim that the model can process text, speech and video and will be available to users within a few weeks.
According to the technical director of OpenAI, GPT-4o provides “GPT-4 level intelligence” (the company’s previous language model), but improves its capabilities in terms of interaction with the user and the world around him. He called it “the future of human-machine interaction.”
GPT-4o significantly improves the experience of working with the Open air chatbot based on artificial intelligence, ChatGPT. The platform has long offered a voice mode that converts chatbot responses from text to speech, but GPT-4o allows people to communicate with it as with a real interlocutor.
For example, users can ask a GPT-4o-based ChatGPT question and interrupt ChatGPT while answering. Judging by the demonstrations, the model provides responsiveness in almost real time and can even pick up nuances in the user’s voice, in response to generating voices in “a number of different emotional styles” (including singing).
In addition, ChatGPT based on GPT-4o will be able to literally “see” the world around through a mobile phone camera, describe and analyze what is happening around. According to Murat, these functions will be developed in the future. While GPT-4o can now look at the menu image in another language and translate it, in the future it will be able to allow ChatGPT, for example, to “watch” a broadcast of a sports game and explain its rules to the user on the fly.
Earlier, Bloomberg wrote that Apple has entered into an agreement with OpenAI in the field of introducing artificial intelligence (AI) technologies into iPhone smartphones. The company will announce AI features that will be available in the new Apple device model in June at the annual Worldwide Developers Conference.
Leave a Reply