OpenAI unveils newest AI model, GPT-4o
By Clare Duffy, CNN
New York (CNN) — ChatGPT is about to become a lot more useful.
OpenAI on Monday announced its latest artificial intelligence large language model that it says will make ChatGPT smarter and easier to use.
The new model, called GPT-4o, is an update from the company’s previous GPT-4 model, which launched just over a year ago. The model will be available to unpaid customers, meaning anyone will have access to OpenAI’s most advanced technology through ChatGPT.
Based on the company’s Monday demonstration, GPT-4o will effectively turn ChatGPT into a digital personal assistant that can engage in real-time, spoken conversations. It will also be able to interact using text and “vision,” meaning it can view screenshots, photos, documents or charts uploaded by users and have a conversation about them.
OpenAI Chief Technology Officer Mira Murati said the updated version of ChatGPT will now also have memory capabilities, meaning it can learn from previous conversations with users, and can do real-time translation.
“This is the first time that we are really making a huge step forward when it comes to the ease of use,” Murati said during the live demo from the company’s San Francisco headquarters. “This interaction becomes much more natural and far, far easier.”
The new release comes as OpenAI seeks to stay ahead of the growing competition in the AI arms race. Rivals including Google and Meta have been working to build increasingly powerful large language models that power chatbots and can be used to bring AI technology to various other products.
The OpenAI event came one day ahead of Google’s annual I/O developer conference, at which it’s expected to announce updates to its Gemini AI model. Like the new GPT-4o, Google’s Gemini is also multimodal, meaning it can interpret and generate text, images and audio. OpenAI’s update also comes ahead of expected AI announcements from Apple at its Worldwide Developers Conference next month, which could include new ways of incorporating AI into the next iPhone or iOS releases.
Meanwhile, the latest GPT release could be a boon to Microsoft, which has invested billions of dollars into OpenAI to embed its AI technology into Microsoft’s own products.
OpenAI executives demonstrated a spoken conversation with ChatGPT to get real-time instructions for solving a math problem, to tell a bedtime story and to get coding advice. ChatGPT was able to speak in a natural, human-sounding voice, as well as a robot voice — and even sang part of one response. The tool was also able to look at an image of a chart and discuss it.
They also showed the model detecting users’ emotions; in one instance, it listened to an executive’s breathing and encouraged him to calm down.
“You’re not a vacuum cleaner!” the female voice of ChatGPT (which sounds remarkably similar to the Scarlett Johansson-voiced digital companion from the 2013 film “Her”) jokingly told the staff member.
ChatGPT was also able to have a conversation in multiple languages by translating and responding automatically. The tool now supports more than 50 languages, according to OpenAI.
“The new voice (and video) mode is the best computer interface I’ve ever used,” OpenAI CEO Sam Altman said in a blog post following the announcement. “It feels like AI from the movies; and it’s still a bit surprising to me that it’s real. Getting to human-level response times and expressiveness turns out to be a big change.”
Murati said that OpenAI will launch a ChatGPT desktop app with the GPT-4o capabilities, giving users another platform to interact with the company’s technology. GPT-4o will also be available to developers looking to build their own custom chatbots from OpenAI’s GPT store, a feature that will now also be available to non-paying users.
The updated technology and features are set to roll out to ChatGPT in the coming months. Free ChatGPT users will have a limited number of interactions with the new GPT-4o model before the tool automatically reverts to relying on the old GPT-3.5 model; paid users will have access to a greater number of messages with the latest model.
OpenAI said more than 100 million people already are using ChatGPT. But an updated ChatGPT experience — and the ability to interact with it on desktop and through improved voice conversations — could give even more people reason to use its technology. The moves comes at a time when integrations of AI into more widely-used consumer products by Google and Meta, like Instagram and Google Assistant, may make those companies’ technology more widely and easily accessible.
This story has been updated with additional developments and context.
The-CNN-Wire
™ & © 2024 Cable News Network, Inc., a Warner Bros. Discovery Company. All rights reserved.
