A method of training a biological neural network using a controller, comprising: applying a cycle comprising: stimulating a neural network by said controller applying at least an input signal to the network; detecting an output response of the network by said controller; and modifying said stimulation by said controller for at least a period of time if said response matches a desired at least approximate response; and repeating said cycle of stimulation, detection and modification at least one more time until said neural network is trained to generate a desired output response for said input signal.