OpenAI today announced the most capable upgraded version of its artificial intelligence The model to date—a model that takes even more time to deliberate on questions—just one day later Google announced its first model of this type.
OpenAI's new model, called o3, replaces the company's O1 Introduced in SeptemberLike O1, the new model spends time considering a problem to better answer questions that require step-by-step logical reasoning. (OpenAI chose to drop the “o2” moniker because it is already the name of a mobile carrier in the UK.)
“We see this as the beginning of the next phase of AI,” OpenAI CEO Sam Altman said on a livestream Friday. “Where you can use these models to perform increasingly complex tasks that require a lot of reasoning.”
OpenAI says the O3 model scores much higher on a number of measurements than its predecessor, including models measuring complex coding-related skills and advanced math and science aptitude. It is three times better than o1 in answering the questions asked ARC-AGIA benchmark designed to test the ability of AI models to reason on extremely difficult mathematical and logical problems that they are encountering for the first time.
Google is doing similar research work. Noam Shazir, a Google researcher, yesterday Revealed in a post on The company has developed its own reasoning model, called Gemini 2.0 Flash Thinking. Google CEO Sundar Pichai called it “our most thoughtful model yet.” own post,
The two duel models show the competition between OpenAI and Google to be more fierce than ever. It is important for OpenAI to demonstrate that it can continue to progress as it looks to attract more investment and build a profitable business. Google, meanwhile, is desperate to show it remains at the forefront of AI research.
The new models also show how AI companies are moving beyond enhancing AI models to achieving greater intelligence.
OpenAI says there are two versions of the new model, the O3 and O3-Mini. The company is not yet making the models publicly available, but says it will invite outsiders to apply to test them. OpenAI also provided more details today about the techniques used to align o1. This involves the model reasoning about the nature of a given request to ask whether it might violate its guardrails.