Self-Assembly Gets Automated in Reverse of ‘Game of Life’

In cellular automata, researchers can now reverse-engineer simple rules creating complex structures. Google's Alexander Mordvintsev developed neural cellular automata (NCAs), working backward from desired patterns to discover generating rules. NCAs' most impressive feature is self-repair—damaged patterns regenerate, mimicking biological organisms. This breakthrough could impact understanding biological regeneration and creating distributed, energy-efficient computers operating without central coordination, potentially surpassing traditional neural networks.