(CLO) On Friday, OpenAI said it is testing two new "artificial super intelligence (AI)" models capable of reasoning: o3 and o3 mini.
OpenAI will release a stripped-down version of the o3 mini at the end of January, followed by the full version later, according to CEO Sam Altman. He believes these models will outperform existing models and attract new investments and users.
With Microsoft's support, OpenAI's o1 AI model is designed to spend more time analyzing and providing answers to complex questions.
Illustration: AI
o1 models are capable of reasoning and solving complex tasks in the fields of science , coding, and mathematics.
OpenAI is internally testing the O3 and O3 mini models, with the goal of creating more powerful versions of the O1.
OpenAI is currently inviting researchers to participate in an exclusive testing program for o3 models, which will end on January 10.
The release of ChatGPT in November 2022 has sparked an arms race in the AI industry. OpenAI’s rapid growth and new products have helped the company attract attention and close a $6.6 billion funding round in October of this year.
Google is also constantly improving its AI models, with the launch of the second generation of its Gemini AI model in early December. Google aims to regain its lead in the AI technology race.
This shows the increasingly fierce competition among technology companies in developing AI models capable of handling complex problems.
Ha Trang (according to OpenAI, Axios, Reuters)
Source: https://www.congluan.vn/openai-thu-nghiem-mo-hinh-sieu-tri-tue-nhan-tao-co-kha-nang-suy-luan-post326870.html
Comment (0)