← New search

Other meanings of Generative art

Art & Technology

Generative art

Generative art is art created through autonomous systems, algorithms, or rules, where the artist defines a process and the system produces the final work, often with an element of unpredictability. It spans digital and physical media, from computer graphics to robotic painting, and has roots in both early computer art and conceptual art practices.

1960s
Origins in computer art
Decade
1965
First computer art exhibitions
Year
2018
First AI-generated artwork sold at Christie's
Year
1

Definition and core principles

Generative art is defined by the use of an autonomous system—such as a computer program, a set of rules, a mathematical formula, or a biological process—that makes decisions within the artist's parameters, producing a work that is not fully predetermined. The artist's role shifts from crafting the final image to designing the generative system and selecting or curating its outputs. This approach emphasizes process over product, and often embraces randomness, emergence, and iteration.

Key principles include the use of algorithms (e.g., fractals, cellular automata, genetic algorithms), the incorporation of chance (as in John Cage's aleatoric music, which influenced visual artists), and the possibility of infinite variation from a single set of rules. The term was popularized by Philip Galanter in his 2003 essay "What is Generative Art?", where he defined it as "any art practice where the artist uses a system, such as a set of natural language rules, a computer program, a machine, or other procedural invention, which is set into motion with some degree of autonomy contributing to or resulting in a completed work of art."1

2

Historical development

The roots of generative art trace back to the 1960s, when artists and engineers began using computers to create visual works. Pioneers include Frieder Nake, Georg Nees, and A. Michael Noll, who produced algorithmic drawings on plotters and exhibited them in 1965 in Stuttgart and New York. These early works were often geometric and abstract, reflecting the computational constraints of the time.

In the 1970s and 1980s, artists like Harold Cohen developed autonomous drawing machines (e.g., AARON) that could create original artworks based on rules about form and color. The 1990s saw the rise of software-based generative art with tools like Processing (created by Casey Reas and Ben Fry in 2001), which made algorithmic creation accessible to a broader audience. The 2010s brought generative adversarial networks (GANs) and other AI techniques, leading to works like Edmond de Belamy, created by the Paris-based collective Obvious, which sold at Christie's in 2018 for $432,500.2

3

Techniques and media

Generative art employs a wide range of techniques, from simple random number generation to complex machine learning models. Common methods include fractal generation (e.g., Mandelbrot sets), L-systems for simulating plant growth, cellular automata like Conway's Game of Life, and evolutionary algorithms that iteratively select and mutate designs. Artists also use physical systems, such as pendulums, wind, or chemical reactions, to generate patterns, as seen in the work of artists like Sol LeWitt, whose wall drawings are instructions for others to execute.

Media span digital images, animations, sound, sculpture, and interactive installations. For instance, the work of Refik Anadol uses data as a material, creating immersive installations that visualize large datasets through generative algorithms. In the physical realm, robotic arms and plotters can translate digital generative designs into tangible objects, as demonstrated by the work of artists like Jürg Lehni.

4

Lesser-known aspects

Beyond the well-known pioneers, generative art has a rich history of overlooked contributions. For example, the British artist Desmond Paul Henry created mechanical drawing machines in the 1960s using modified analogue computers, producing intricate line drawings that predate many digital works. Similarly, the Japanese artist Hiroshi Kawano, a philosopher and computer scientist, began creating computer art in the 1960s, often using early programming languages to generate abstract compositions.

Another niche area is generative music, where algorithms compose sound, as in the works of Brian Eno and the software SSEYO Koan. In the realm of literature, generative poetry and prose have been explored since the 1950s, with the Oulipo group using constrained writing techniques. More recently, generative art has intersected with blockchain technology through NFTs (non-fungible tokens), enabling artists to sell unique digital works, as seen in the Art Blocks platform. These examples highlight the breadth of generative art beyond visual media.

Glossary

Algorithm
A step-by-step procedure or formula for solving a problem, often used in generative art to define rules for creating works.
Autonomous system
A system that operates independently of direct human control, making decisions based on its programming or environment.
Cellular automaton
A discrete model studied in computability theory, mathematics, and theoretical biology, consisting of a grid of cells that evolve according to rules based on the states of neighboring cells.
GAN (Generative Adversarial Network)
A class of machine learning frameworks where two neural networks contest with each other to generate new data with the same statistics as the training set.
Plotter
A computer-controlled device that draws images on paper using pens or other tools, often used in early generative art.

This article focuses on generative art as art created through autonomous systems, algorithms, or rules, excluding other uses of the term in music or literature.