A new study from MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) demonstrates that large language models (LLMs) can be used to automatically generate and refine robot task plans, significantly reducing the need for manual programming. The system, named Code as Policies, allows robots to interpret high-level instructions written in natural language and autonomously write the code needed to execute complex, multi-step tasks. This approach bypasses traditional, time-consuming programming methods and enables more flexible and adaptive robotic behavior. The research shows particular promise in domains like manufacturing and domestic assistance, where robots must respond to dynamic environments. While effective, the method currently works best in predictable settings and requires safeguards against errors in the generated code. Read the full article at https://technologyreview.com/2023/09/12/1079602/robots-ai-language-models-generate-code/.