I have been experimenting with random art since 2014.
This page provides a brief introduction and provides context for the galleries on this site.
Let's start things off with a motivating example. Suppose you flip a coin 100 times, drawing "|" or "–" depending on the outcome, and then arrange these in a 10x10 grid. That might look something like what we see here.
Instead of allowing just two possible outcomes, we can allow for more. We can also change our predetermined actions beyond vertical and horizontal lines. Under this general framework, the art takes on many different characteristics. Here are some examples of what that can look like.
Something really different.
There is a lot more to say about how exactly these are made, but these too are just several rules followed using random numbers as inputs.
Instead of lines, we can draw circles, or other shapes. We can imagine rolling a die and drawing one of six possible sizes of circles. Set some rules, pick which attributes to fix and which to control, and follow your rules a predetermined number of times and see what comes out. The random numbers used can come from any distribution we like.
These choices along with the design of assigning parameters to visual attributes is my contribution to the artwork. Each piece is unique, even with the same sets of rules followed the same number of times. I am a curator as well as a creator. I group images into collections based on the relationships among the pieces. Some galleries contain images that followed the same rules. Others may share most attributes but one rule is changing from piece to piece.
I invite you to observe these images and pick apart how they were made – what rules were set here? What is the same? What is different? What is changing from image to image?
A screenshot of the original GUI-version of the software from June 2014, as it was approaching its final design state. Eventually I'd restart from scratch.
The software I have written to create these artworks has undergone several evolutions. Part of my original motivation for this project was to help visualize the interactions among probability distributions and learn how to program a graphical user interface. A snapshot of it in action can be seen here.
low number of choices
The visualizations of the underlying distributions really helped build intuition for how to control the general appearance of the outcome.
a slightly older version
As mentioned, the software kept changing (and still is), constantly adding new features. Notice the absence of the [ C | D ] button, which at some point toggled sorting.
It was important to me that the software was flexible and could create a wide variety of random images, from dots to lines to squiggles and shapes. After a semester of work, I stopped adding features and turned my attention towards exploration (and towards beginning graduate school). In 2016 I began to I overhaul my process in favor of a script-based approach using Python, which made adding new artistic styles much less time consuming than the GUI seen here but required a significant effort .