*And this makes sense only after introducing an additional fact. Thanks to random, probabilistic hiccups with the ion channels now and then neurons will occasionally have a random, spontaneous action potential from out of nowhere (which is looked at in depth in chapter 10 when considering what quantum indeterminacy has to do with brain function [psst—not much]). So neuron A intentionally fires off ten action potentials, followed soon after by two random ones. That might make it hard to tell if neuron A meant to yell ten, eleven, or twelve times. By calibrating the circuit so that the inhibitory feedback signal shows up right after the tenth action potential, the two random ones afterward are prevented, and it is easier to tell what neuron A meant. The signal has been sharpened by damping the noise.