Black Squares

Kazimir Malevich, Black Square, 1915

Around the time I started making painting machines I started seriously asking myself whether painting was compelling at all. But like anything you find in the world, if you spend enough time studying it you will be rewarded with endless fascinations.

My starting point was what I was working on: machine-produced drawings and paintings. I quickly ran into the artist Vera Molnár, who was very early (1959 on) in seizing on the use of computers for creating art.

As I was looking through her work and that of other early pioneers in the space I made the observation that it was often visually similar to that of the artists who were a part of the Suprematist movement starting around 1913. Those artists were after "the supremacy of pure artistic feeling." And they found it in fundamental geometric forms. Malevich's Black Square (at the top of the page) is sometimes called the zero point of painting. After the Suprematists broke down painting to its absolute essentials they strove to "go beyond zero" and build it back up to accomplish whatever goals they found interesting.

El Lissitzky, 1o Kestnermappe Proun, 1923

I found it fascinating that the limitations of early computers meant that the practical place to start with them was near the same zero point where the Suprematists ended up when they tried to break down painting until there was nothing left. From the perspective of the development of artificial intelligence and robotics, it's sort of like the Suprematists were reverse engineering their own visual and motor systems, but assuming they would find something like the logical machines that were changing their world at the time. So they ended up accidentally predicting where computers eventually started their artistic development.

I suspect that the actual zero point for human painting has yet to be found. And that it will be found by a machine through a process that mirrors that which the Suprematist painters set about. Machines are uniquely advantaged in the analysis of complex, dynamic systems like those we humans are.

Curiosity about what the hypothetical real zero point for painting might look like led me to consider what it would take to build the machines that would find it. Looking for the data I would feed to the machines I came back to the paintings themselves. I saw that paintings are imprints of the complex dynamics of the painters minds and bodies working together based on learning borne of long experience. I felt it would be interesting to make a machine that could x-ray those dynamics for any particular painting, purposefully cutting away some of the surface area within which the viewer might lazily stash meaning.

The first experiment I did was to simply test my implementation of AR with paintings. I was using the wonderful AR.js so I could have the AR run in an accessible web app. After writing a bit of code in JavaScript to generate AR web apps with embedded 3D assets I made a test painting and took it into different environments to try to visualize a world with mixed-reality artifacts.

The second experiment I did was to convey the dynamics of a painter whose mind I had full access to - a machine. Specifically an AxiDraw plotter.

I built a system to generate a plotted drawing and a corresponding augmented reality application which could show the "x-ray" of the drawing process directly on top of the work, by visually recognizing and tracking the square black border of the graphic and rendering 3D graphics on top.

With that proven out I repeated the same exercise but with me as the painter. I tracked my motions while painting using an HTC Vive VR system and used the same software system to generate an augmented reality application visualizing the topology of the brush motion that went into the painting.

I failed to learn anything about the actual zero point for human painters, but the experiments were an excellent petri dish for thought.