The world of artificial intelligence is reaching a new milestone. OpenAI presented o1 on Thursday, an innovative generative model. It can analyze before responding, particularly in areas such as mathematics, with speed surpassing that of humans.
OpenAIthe company behind ChatGPT, has just unveiled the latest model of its artificial intelligence. Baptized o1this model can “reason” like a human. No more guesswork! o1 is designed to provide accurate and relevant information, even on complex topics.
A model designed for reasoning
Unlike previous versions, o1 stands out for its ability to analyze before providing an answer. The objective displayed by OpenAI is clear: develop a artificial intelligence said “general”that is, capable of thinking like humans, while limiting the risks of errors or hallucinations, commonly encountered in previous models.
A revolutionary AI
The general manager ofOpenAISam Altman, praised the launch on X, calling o1 a true “paradigm shift”. He stressed, however, that this technology is “still imperfect, still limited.”
At first glance, o1 looks impressive, but after prolonged use, you realize its limitations.
Progressive availability
For now, o1 is accessible in beta version for paying ChatGPT users. They will then be able to preview this new feature IAThis testing phase is crucial to gather user feedback before a possible wider deployment.
