Tool Innovation

Tool innovation research explores how agents, both artificial and biological, discover and create tools to solve problems, focusing on improving efficiency and adaptability. Current research investigates the use of large language models (LLMs) augmented with symbolic solvers to enhance logical reasoning and tool utilization, as well as the development of self-supervised learning frameworks for robots to design and use tools from readily available materials like paper. This work is significant for advancing artificial intelligence, particularly in robotics and problem-solving, and has implications for automating complex tasks in various fields, such as agriculture.