From the moment of birth, our brains begin rewiring themselves to decode the raw sensory input entering from our eyes and ears. A baby takes about three months to learn to track objects, five months to develop color vision, and seven to recognize objects. Is it possible, using the "impossible images" coming out of machine learning, to reverse or damage this process? Will looking at such images too long or too often begin to alter my perception of the world around me? Pic related.