OpenAI releases new GPT-4o

On Monday, OpenAI unveiled an upgraded iteration of its widely-used generative tool, ChatGPT, boasting enhanced performance and a more human-like interface.

This improved version, GPT-4o, is now accessible to all users at no cost. The release coincided with an impending announcement from Google about Gemini, its own AI tool directly challenging ChatGPT.

Mira Murati, Chief Technology Officer, expressed immense excitement about the launch during the highly anticipated event in San Francisco.

OpenAI announced that the new GPT-4o model, with “O” representing omni, will gradually roll out across its products in the coming weeks.

Paid subscribers will enjoy unlimited access to this advanced tool. Capable of generating content and comprehending commands across voice, text, and images, the model introduces a revolutionary voice (and video) mode, lauded by OpenAI CEO Sam Altman as the most exceptional computer interface he has encountered, reminiscent of AI depicted in movies.

Altman has previously cited the character played by Scarlett Johansson in the film “Her” as an inspiration for the direction he envisions AI interactions taking.

Altman further expressed, “Talking to a computer has never felt really natural for me; now it does.”

During the virtual event, Murati and engineers showcased the enhanced capabilities of GPT-4o. They presented challenges to the upgraded ChatGPT chatbot, which responded with jokes and human-like conversation.

The demonstration included OpenAI staff members posing questions to the voice-enabled ChatGPT, which not only provided witty responses but also served as an interpreter from English to Italian, interpreted facial expressions, and assisted a user with a complex algebra problem.

The ChatGPT bot remarked, “Hmm, from what I can see, it looks like you’re in some sort of recording or production setup with lights, tripods… Perhaps you’re preparing to shoot a video or make an announcement?”

Regarding recent expectations, there was widespread anticipation that OpenAI would unveil an AI-enhanced online search tool to rival Google. However, Altman clarified on Friday that this wasn’t the case.

There was also anticipation surrounding the release of GPT-5, but Altman mentioned last week that the company would “take our time on releases of major new models.”

This event is the latest development in the ongoing AI competition, where OpenAI-backed Microsoft has surpassed Apple to become the world’s largest company by market capitalization.

OpenAI and Microsoft are engaged in a fierce rivalry with Google to establish dominance in the field of generative AI. Additionally, Meta, the parent company of Facebook, and the emerging player Anthropic are also making significant strides to compete.

All these companies are racing to find ways to cover the considerable costs associated with generative AI, much of which involves investments in chip giant Nvidia and its powerful GPU semiconductors.

Making the new model accessible to all users might prompt questions about OpenAI’s monetization strategy, especially considering doubts about widespread willingness among users to pay for a subscription.

Until now, only less advanced versions of OpenAI or Google’s chatbots were offered to users for free.

“We are a business and will find plenty of things to charge for,” Altman affirmed on his blog.

AI developers are also facing pressure from content creators and publishers, who are advocating for compensation for any content used to train the models. While OpenAI has established content partnerships with entities like the Associated Press, the Financial Times, and Axel Springer, it’s also embroiled in a significant legal battle with The New York Times.

Furthermore, AI companies are contending with individual lawsuits from artists, musicians, and authors in US courtrooms.