A new study from MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) demonstrates that large language models (LLMs) can significantly accelerate the process of robot motion planning, a traditionally slow and computationally intensive task. The research introduces a method where an LLM, such as GPT-4, breaks down complex navigation instructions into smaller, manageable sub-tasks described in natural language. A separate, smaller AI model then translates these sub-task descriptions into code that a robot can execute. This hierarchical approach, tested in simulated environments, allowed robots to navigate complex, multi-step routes more efficiently than with traditional planning methods alone. The work suggests a promising path toward more fluid and responsive human-robot interaction by leveraging the high-level reasoning of LLMs. For the full details, read the complete article at https://technologyreview.com/2024/05/06/1092520/llms-robot-navigation-planning/.