Nature shows that intelligence did not begin with brains or algorithms—it began with matter itself. This review spotlights “physical intelligence,” the problem‐solving capability of materials, structures, and interactions that sense, adapt, and compute directly through physics. Long before brain emerged, cells, plants, and insect swarms were exploiting physical laws to navigate, coordinate, and thrive. Today, these same principles inspire breakthroughs in robotics and engineering: soft robots that move without explicit programming, origami architectures that reconfigure on demand, and swarms of simple agents that collaborate like living colonies. By tracing physical intelligence across scales, from molecules to collectives, we highlight its role as the foundation of embodied intelligence and a blueprint for rethinking machine design. Embedding intelligence into matter itself, rather than treating hardware as passive vessels for software, offers a pathway to systems that are resilient, efficient, and adaptive. The future of artificial intelligence may depend as much on intelligent physical bodies as on smart code.
Liu et al. (Tue,) studied this question.