TMTPOST--OpenAI has introduced a new AI product it claims can reason and solve complex problems in fields like mathematics, coding, and science, marking a significant step toward humanlike cognition in machines. These AI models, called "o1," are being hailed as a testament to the rapid technological advancements of recent years, as tech companies like Google DeepMind, OpenAI, and Anthropic compete to develop more sophisticated AI systems. The race is now focused on creating autonomous "agents" — personalized bots designed to assist users in work, creativity, communication, and navigating the digital landscape. OpenAI announced that the o1 models will be available in ChatGPT Plus starting Thursday, primarily targeting scientists and developers rather than the general public. The company also revealed that the o1 models significantly outperformed existing models like GPT-4o in the International Mathematics Olympiad qualifying exam, achieving an impressive 83% score compared to GPT-4o's 13%.