A new study from MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) demonstrates that large language models (LLMs) can be used to generate detailed, step-by-step robot task plans from natural language instructions. The system, called Code as Policies, allows robots to autonomously perform complex, multi-step tasks by translating high-level commands into executable code. This approach bypasses the need for traditional, painstaking programming of each robot action, potentially making robots more adaptable to new instructions and environments. The research highlights both the potential for LLMs to bridge the gap between human instruction and machine execution and the ongoing challenges of ensuring safety and reliability in real-world applications. For the full details, read the article at https://technologyreview.com/2023/09/15/1080019/robots-ai-language-models-code-policies/.