@[email protected] and to add to that, that is exactly how human Brian has been trained - well maybe over a longer time and a mix of hardware/software tweeking, but overall the same concept: produce the most probable sequence of signals based on a learnt distribution - likely the result of optimizing a fitness function of sort