The Precariousness of Chaos and Order: An Introduction to Generative Art

The generative art group exhibition, In Our Code, arrives at Unit London on 13th September 2022. This article takes readers on a journey through the conceptual origins of generative art, examining how this genre explores the distinction between humans and machines, the balance of freewill and chance, and the notion of artistic authorship.

What is Generative Art?

The year is 1792, and a small manuscript of 192 single-bar fragments of music appears in Berlin. Its author is Wolfgang Mozart, and its title declares that it will make composers out of those who know nothing of composition. All that is needed is a pair of dice. Each fragment has been enumerated and assigned a position in a 12 x 16 grid where the x and y axis correspond to the number the dice roll, and the order they are rolled in. This system, known as a Musikalisches Würfelspiel (lit. music dice game), can generate over 46 quadrillion randomised compositions.

Mozart, Musikalisches Würfelspiel, 1792

It’s now 1971, and Sol LeWitt, the enigmatic pioneer of conceptual art, hangs a set of instructions on a wall in the Boston Museum urging museum-goers to take a hard pencil to the panel in front of them, place 50 random points across it, and connect these points with straight lines. The result is a soft grey matrix of crisscrossing lines, dark constellations appearing where these lines insect. In the 50 years that follow, LeWitt’s instructions, named Wall Drawing no 118, are repeatedly executed by humans and computers alike. Each time, a new variation of constellations and geometric angles emerges.

Unemployed veteran, Wall Drawing #118 (After Sol LeWitt), 2016

The decades come and go. 2021 dawns and Tyler Hobbs releases the algorithm Fidenza on Art Blocks, Erick Calderon (Snowfro)’s marketplace for code-generated NFTs. Hobbs has programmed Fidenza to output a collection of 999 digital images of curves. While the outputs are capped, the variations are not. Instead, an infinite number of curve combinations are possible within a system of probabilities, assigned by Hobbs, to features such as length, angle and colour. It is only when the future collectors of Fidenza’s outputs chose to mint (aka create or buy) an NFT, that the algorithm will randomly generate a new and unique image. Not even Hobbs knows exactly what form the curves in these images will take, but in 24 minutes, the collection sells out, and in some, gentle strings of colours wind around each other. In others, heavy blocks of wonky black and white line the screen.

Tyler Hobbs, Fidenza #942, #943 and #944, 2021. Courtesy to the artist.

These works of Mozart, LeWitt, and Hobbs are markedly disparate – from music, contemporary conceptual art, to computer generated art. Their eras are distinct, their mediums are contrasting, and yet, as systems, they share a single overarching feature that defines them and their outputs as generative art: randomness.

Computer Generated Art

Despite the rich history of artists submitting to randomness, it wasn’t until the 1960’s that generative art was recognised as a movement. This is no coincidence. The 1960s were when artists – usually those with the charm or connections to wangle their way into university science departments – gained access to computers.  

The computers of this era were slow and hulking mainframes, but were nevertheless, as the Godfather of computer-based generative art Herbert W. Franke puts it ‘the universal machine par excellence’ for generating randomness. No human brain can produce permutations with the speed that a computer can. Artists interested in the overlap of mathematics and art quickly recognised this and began, as Franke did, to see the computer as ‘not just as a tool, but as a kind of partner who could do some things better than I could.’

While computers allowed generative art to reach its full permutational potential, NFTs have allowed it to reach its full market potential, enabling artists to sell digital works in a way that wasn’t previously possible using blockchain technology. 

Herbert W. Franke, Oscillogram from “Experimental Aesthetic”, 1959. Courtesy to Museum für Angewandte Kunst.

Long-Form and Short-Form Generative Art

Blockchain has also facilitated new evolutions of computer-generated art which heighten viewer participation and dilute artistic control. As an algorithm whose outputs are unknown to the artist or collector before the moment of minting, Fidenza is an example of ‘long-form’ generative art, a new sub-genre named and pioneered by Hobbs. In contrast, its predecessor, ‘short-form’ generative art allows artists to privately generate outputs and select the ones an audience encounters. Unattractive or repetitive variations can remain unseen. By requiring algorithms to reliably generate outputs that are both beautiful and varied, long-form computer-generated art forces artists to calibrate the balance of chaos and order with a new-found level of precision.

Even before the arrival of long-form computer-generated art, no other movement has so decisively called the question of authorship into question. To what extent is the artist the creator, and to what extent is the machine? Distinguishing between the two, in turn, cracks open another set of questions; to what extent can we even separate artists from tools or man from his machines?

“The Idea Becomes a Machine that Makes the Art” 

Moreover, while there may be little doubt that both a machine and an artist exist, what exactly is the art? This question hints at generative arts’ overlap with conceptual art. As a pioneer of both, Sol LeWitt established the notion that it is the idea and not the output which is the art piece. ‘The idea,’ he wrote, ‘becomes a machine that makes the art,’ but in the generative world this is by no means dogma, and it is often the output that the market values.

Perhaps the most unanswerable queries that generative art provokes are those inherent to randomness. Contemplate the topic for too long and questions of free will, of chaos theory, and the tyranny of chance will start to surface.

For many, it is these questions that underscore the richness of generative art, and in an age where computers and algorithms increasingly shape our lives, few mediums are better suited for reflecting on the ways that the digital and human have merged.

Vera Molnar, (Des)Orders, plotter drawing, ink on paper, 1974. Courtesy to artist and Digital Art Museum.

By Kitty Horlick