Researchers at OpenAI have proposed a new approach to training artificial intelligence models. The method, called 'model-based control', involves planning online and learning offline. This allows the model to efficiently explore its environment and adapt to changing conditions. According to the authors, this approach can improve performance in tasks such as robotics and game playing. The technique is based on recent advances in reinforcement learning and control theory. --- Why it matters: This matters because it could lead to more efficient and effective AI systems that can learn from experience and adapt to new situations. Source: https://openai.com/index/plan-online-learn-offline