By developing organic transistors that mimic neural synapses, the team has identified how molecular interfaces can be engineered to create "neuromorphic" computers that learn and store data simultaneously, potentially reducing AI energy consumption to the level of a light bulb.
