Artist Mario Klingemann, also known as Quasimondo, describes himself as an “obsessive orderer.” Over the past few years he’s been parsing and sorting millions of images from the British Library and the Internet Archive. During the process of rewriting and improving the tools he created to automatically classify the pictures, he stumbled onto new visual maps by forcing the images into a grid-like formation.
“What you see in the Emoji piece and the other ones is a bit of a side-effect of these tools and shows what my algorithms deems to be similar images,” Klingemann explains to The Creators Project. Forcing the images into a grid is a difficult task that might not be obvious to a “professional orderer,” he adds. In data visualizations, items are often organized into a superconnected relationship network that resembles a “can of worms,” with each element clustered on top of one another. “The reason that you see those a lot is that they are easy to do because you don't have think about the space that each element uses—you just stack them.” But bringing all of the images together into a single space is much more challenging.
Even if one only glanced at each image for a second 24 hours a day it would take them more than 40 days to browse the entire collection, estimates Klingemann. By adding tags and using algorithms to help in the monumental task, he hopes he can make the images sets more useful to the public. His work falls into his longterm goals of perfecting his “artificial creativity programs—bots that create autonomously.” Below, see some of the results of his algorithmic ordering experiments: