Drawing With Evolution

What does it mean for a computer to have a personality, or a style? What if we gave a machine the tools necessary to create its own representations of real world things or people? These are the questions I am attempting to answer with this project. By mimicking evolution to solve a problem, the computer starts off with a population of random noise drawings. By mating the most succesful drawings of each generation, the computer begins to learn what the image looks like. Because of the nature of the process, each implemenation produces a unique drawing (i.e. a unique representation) of the object it perceives.

The Max Patch that manages the population of images