Researchers are now harnessing the creative potential of generative AI to design robots with enhanced physical capabilities. In a recent breakthrough, a team at MIT employed a diffusion model, a type of generative AI, to reimagine the design of a jumping robot. By providing the AI with the basic parameters and the goal of maximizing jump height, the model generated a novel design with curved linkages that allowed it to store and release more energy. The resulting 3D-printed robot could jump significantly higher than its human-designed counterpart, showcasing how generative AI can not only produce text and images but also contribute to tangible engineering innovations by exploring a vast design space beyond human intuition.
