When I read that the topic was “randomness,” I instantly thought of Pisarro’s paintings that use pointillism. Although he placed every dot of paint in each pointillism painting, they weren’t calculated to their own rigid, specific positions. The placement of those dots are also random to an extent, because when people “randomly” generate things, it is not wholly random. So then I wondered about computer-generated “fine art” and I came across a project called “Grow Your Own Picture” by Chris Cummins, that generates the Mona Lisa painting using polygons of different shapes and colors, which can be manipulated by the website user. Chris gives credit for the idea of genetic algorithms to Roger Johansson, who was able to regenerate the painting using polygons of different colors, transparencies, and shapes. I was really intrigued that he was able to take so many sharp shapes and blend them to mimic the original painting so convincingly. I have also never heard of the term genetic programming or genetic algorithms before. It was so interesting to hear about programming that generates a “gene pool” of shapes and then does its best to sift out the “most fit” shapes to the Mona Lisa image. Cummin’s and Johansson’s approaches to this replication of fine art vary in level of randomness, in that with Cummin’s project, someone can interact with it and manipulate the content, so it has more opportunities to be different or random. They are also random in a similar sense, in the code that generates the art.
Chris Cummin’s interactive Grow Your Own Picture https://chriscummins.cc/s/genetics/#
Roger Johansson’s Mona Lisa replication via code