The premise of extracting meaning from noise is flawed. It doesn't matter what process you use to arrive at the meaningful conclusions if the source material is completely meaningless. It's analogous to a fallacy of false premise: you can literally arrive at any conclusion given a false assumption.
"random" encompasses an infinite array of variable parameters -- linear blur, Gaussian blur; white noise, brown noise, pink noise -- some of these can tend to make animals appear in dot patterns, or melodic phrases arise from static, or whisky taste of butterscotch and leather
