ChatGPT developer OpenAI announced the launch of its "Strawberry" series of AI models designed to excel in complex problem-solving. These models, codenamed "o1" and "o1-mini," are capable of reasoning through intricate tasks and surpassing previous models in fields like science, coding, and mathematics.

The models, first reported by Reuters, leverage a technique called "chain-of-thought" reasoning, breaking down complex problems into smaller, logical steps. This approach, previously used as a prompting technique, has been automated within the models, allowing them to tackle challenges autonomously.

OpenAI emphasizes the models' enhanced thinking process, noting they "spend more time thinking through problems before they respond, much like a person would." This training enables them to refine their reasoning, explore different strategies, and recognize errors.

The o1 model achieved a remarkable 83% score on the International Mathematics Olympiad qualifying exam, a significant improvement over its predecessor, GPT-4o, which scored only 13%. Furthermore, the model demonstrated superior performance on competitive programming questions and surpassed human PhD-level accuracy on a benchmark of science problems.

The o1 model is now available in ChatGPT and its API, marking a significant step forward in OpenAI's pursuit of advanced reasoning capabilities within AI.