What happens if you write software that generates random polygons and the software then feeds the results through facial recognition software, looping thousands of times until the generated image more and more resembles a face? Phil McCarthy’s Pareidoloop. Above, my results from running it for a few hours. Spooky.
prosthetic knowledge shows an image resolving
This is a cautionary tale of science, that you shouldn’t make a priori assumptions about results. If you assume that a noisy signal is a face, then you’re going to see a face.
This is reminiscent of some pseudoscience from so-called ghost hunters recording voices, in something called EVP (Electric voice phenomena). You record a tape and then filter the noise, keeping anything that sounds like a voice. Which is what you get, for the same reason as seeing the faces.