New ChatGPT models, OpenAI o1 and o1-mini, have `reasoning' skills

OpenAI has released a new series of models codenamed o1 with ‘reasoning’ capabilities that are designed to spend more time before responding to users' queries. The new model is immediately available for use and is said to have the ability to perform more complex tasks than were thought possible by previous models. 

While o1 is said to be better at performing some tasks like coding and solving multistep problems it is also slower and costlier than current models like GPT-4o. 

`Beginning of a new paradigm'

OpenAI CEO Sam Altman, in an X post, said: “Here is o1, a series of our most capable and aligned models yet: o1 is still flawed, still limited, and it still seems more impressive on first use than it does after you spend more time with it. But, also, it is the beginning of a new paradigm: AI that can do general-purpose complex reasoning.”

OpenAI researcher Noam Brown in a series of posts on X stated that o1 is designed to ‘think’ in a private chain of thought before responding to queries. Mr Brown states that the longer o1 the better it does on reasoning tasks. 

Explaining o1's thinking process in a blog post, OpenAI wrote, “Similar to how a human may think for a long time before responding to a difficult question, o1 uses a chain of thought when attempting to solve a problem. Through reinforcement learning, o1 learns to hone its chain of thought and refine the strategies it uses. It learns to recognise and correct its mistakes. It learns to break down tricky steps into simpler ones. It learns to try a different approach when the current one isn’t working.”

(By arrangement with livemint.com)