OpenAI on Monday announced significant improvements to its flagship chatbot, ChatGPT. This includes improvements to voice, text, and vision functionality, as well as a new desktop app that will give all users a “faster” service.
Mira Murati, Chief Technology Officer at OpenAI. (Very briefly) Former CEOsaid in a video briefing that its latest large-scale language model, GPT-4o, makes real-time conversational voice, text, video, and audio available to developers and other users. Murati said the company's enterprise products will still have usage restrictions, but all users will have access to the new model's features.
“This is the first time we've really taken a big step forward in terms of ease of use,” Murati said.
in blog post, CEO Sam Altman said the new model advances the company's mission to provide advanced AI tools to everyone. “We are a business, so there will be a lot to charge,” he wrote. “…and it will help bring free and better AI services to (hopefully) billions of people.”
The announcement may have come as a bit of a disappointment to those who were excited by rumors of a potential partnership with Apple and search engine capabilities to rival Google. In his post about X, Altman ruled out any of these developments being part of today's announcement.
Is OpenAI just playing catch-up?
Chirag Dikeit, vice president and analyst at research firm Gartner, told InformationWeek that he was shocked by OpenAI's announcement and that the product's advances are catching up with Google's latest Gemini AI product. “I know that’s a bit of a harsh view,” he said in an interview. “This is a model…basically what Gemini was able to do months ago…now OpenAI's competitors are unleashing the full force of their data, infrastructure, research, innovation and other resources. And we're starting to see the gap in OpenAI being shown in that context.”
Still, these new features represent the Microsoft-backed GenAI company's biggest product announcement since last year's announcement of ChatGPT Enterprise.
“…the new audio (and video) mode is the best computing interface I've ever used,” Altman wrote. “It feels like an AI from a movie. I'm still a little surprised that it's real.”
Murati said the new model is faster in up to 50 different languages and will be available to developers via OpenAI's API starting today. “That means a developer will be able to start building 50% cheaper at twice the speed with a rate limit that is five times higher,” Murati said.
Members of the OpenAI team also sat down to demonstrate the new model's capabilities, showing off the model's ability to determine a user's emotional state and its ability to pause to clarify a question. OpenAI researcher Mark Chen shows how the model can be used to calm the nerves with detailed instructions on breathing techniques, and corrects when Chen intentionally hyperventilates. I was also able to do that. “Hey, slow down!” said ChatGPT's voice.
The ability of the model to support code and simple calculations was also demonstrated. Interestingly, the coding demonstration was performed on a MacBook. Bloomberg previously reported, citing sources, that the company was close to a deal with Apple to bring ChatGPT functionality to its next operating system.
The AI arms race is far from over
According to Bloomberg Intelligence, GenAI $1.3 trillion market by 2032 That's because OpenAI, Meta, Google and its parent company Alphabet, and Microsoft are all vying for position to create increasingly sophisticated language models at scale.according to StatistaThe global market for AI is expected to reach $184 billion in 2024.
For Gartner's Dekate, today's announcement shows that the ultimate winner of the AI arms race is far from certain, with Google, Microsoft, and Amazon pouring billions of dollars and research into the game. “What has changed is that OpenAI no longer defines the leadership moment for generative AI,” he says. “We're not sure if OpenAI will drive the future of generative AI. Perhaps it could come from elsewhere.”
