The Simple Beauty of Generative Design

In our our current cultural conversation, "algorithm" has become a four letter word. Intertwined with the fates of platforms like Facebook and Youtube, a once-benign computer science term is now often said with suspicion or contempt, evoking fears of misinformation, surveillance capitalism, and civic rot.

Most people view code as, at best, abstract, utilitarian, and coldly technical—a tool for building digital infrastructure, more similar to a hammer than a paintbrush. At worst, they see it as the root of some of our biggest problems.

Code is, at the end of the day, though, just language—one that we happen to use to communicate with computers, not all that different than the languages we use to talk to one another. One of the defining features of good languages is that they are malleable—we can use English to write both songs and dissertations, love letters and hateful screeds. Similarly, javascript and other coding languages are not exclusively well-suited for building apps. They can also be tools for storytelling and self-expression. They can, in a word, be generative.

In her essay on the subject, artist Amy Goodchild identifies autonomy as the key element of generative artwork, and defines autonomy as "a process not under direct human control." Her definition is purposefully expansive and academic—both Jackson Pollack and Tyler Hobbes are, in this conception, generative artists—so much so that I wonder if she's sacrificing clarity for comprehensiveness.

For the purposes of this post, I'm going to narrow the field a bit and say that generative design is:

A code-based solution for creating artwork based on a combination of explicit rules and planned spontaneity

Not bad, but let's break that down a step further:

• code-based: a simple way of saying that the work in question is produced by a set of instructions executed by a computer
• artwork: the output of a generative creative process—most often either images or animations, though you can certainly create generative work in any medium
• explicit rules: we're talking about art that is created indirectly through rule-based algorithms (not the scary ones). In a sense, generative art is more like knocking over dominoes than it is like drawing a picture.
• planned spontaneity: generative works are almost always probabilistic in nature—which is to say that they rely on randomness and embrace elements that are intentionally beyond the control of the creator

Borrowing an analogy from another art form, if others forms of art are like oil painting, where the creator has a high degree of control over the final output, generative design is more like water color. You purposefully sacrifice some level of direct control to the influence chance and happenstance—fully embracing so-called "happy accidents." The imprecision of the medium is often times the point.

To demonstrate the point, let's look at a simple, elegant example of generative design: the 10 Print pattern.

References 10 Print (Book) The Coding Train