Generative art

Generative art uses some kind of system. The system removes some control over what it makes from the artist. In other words, it has autonomy. Once the system starts running, the artist’s involvement in it generally stops.

Most generative art involves a set of instructions. Generally, the artist creates them. The instructions are often digital, but they don’t have to be. They could be a list of actions for other people to make something, for example, such as the work of Sol Lewitt.

Generative art often features elements of chance. For instance, the artist could list a number of options and let the system decide which one to use. Linking a few different decisions together can quickly give plenty of scope for complex outcomes. Allowing the system to work this way is often what gives it autonomy.

What makes it art or not is harder to define. The creator of the system decides how much autonomy it has. At one end of the scale is minimal input from the artist. This would give completely random behaviour. At the other end is complete control from the artist. This would not technically be generative. The thing that may make it art is fine-tuning the system.

Philip Galanter suggests the idea of a sweet spot between chaos and order. Somewhere between the two is a balance that the artist can decide is the correct one. The quality of that decision might be what makes it art. If the generative output has the reference points used to discuss art in general applied to it, maybe that’s also part of it