Centuries ago—as far back as the twentieth century A.D.—human researchers dealing with similar neural networks comprised of pre-AI silicon intelligences discovered that the best way to make a neural network creative was to kill it. In those dying seconds—even in the last nanoseconds of a sentient or near-sentient conscience’s existence—the linear, essentially binary processes of neural net computing jumped barriers, became wildly creative in the near-death liberation from off-on, binary-based processing.
