Generative and algorithm art have become hot topics recently, particularly in the NFT art market where, along with much digital art, they have found an eager audience of buyers, however; whilst many headlines insist these “AI” generated images are a remarkable new phase in art, in truth it’s a resurgence of one with a history that arguably goes beyond computing.
Since the earliest days of computing, creative coders became interested in images and artworks generated by a computer, and explored a range of mathematical algorithms that created a set of rules for a computer to follow in creating an image. Most of the work was based in repeating geometric shapes to create a pattern, fractals, and even glitches which deliberately break the code to see what happened, or random factor aspects allowing for certain variations in the calculations.
Algorithms have existed for much longer than computers though; many of the geometric patterns found in Islamic tiling and decoration are, arguably, a form of algorithmic art growing from a culture with a fascination for mathematics and abstract patterns. Many Renaissance paintings make use of maths also, in composition, proportion, and perspective. These early examples of course are debatable, but with certainty the idea of using algorithms to create art emerged in the 1960s.
The first wave of algorithm art with the birth of computers
The 1950s had seen a wave of art similar to the Islamic art; using a range of strict rules and procedures to create patterns and images, and as computing became more widely available in the 1960s many artists, and computing experts, were attracted to experimenting with it.
An important early group of pioneers were the 3N artists, including Georg Nees, a maths, physics, and philosophy academic, Frieder Nake, a mathematician and computer scientist, and A. Michael Noll, an engineer and communication professor. Though they never worked as a unit, all three rapidly pushed forward ideas and boundaries in terms of applying computers, algorithms, and generative ideas into a range of artforms.
Nees played with random numbers and geometric shapes on a flatbed plotter, and sculptures created from computer controlled milling machines. Noll worked at Bell Labs which provided a home for many experiments in different kinds of computer art; Noll focused on the effects of media on communication, speech-signal processing, and early 3-dimensional graphics and animation. Nake focused on art produced with a precision plotter, and very much led by code designed to try and allow for the computer to be the biggest part of the creative process; perhaps most similar to the modern wave of “AI” generative and algorithm artworks.
AI, generative art, and the emerging NFT market
With the rapid emergence of the NFT art market, there has been a renewed interest in computer generated and algorithm art of various kinds; it never really went away, but as a format finally emerges where many artists can earn money from it, far more artists are willing to put time into experiments.
A term which often excites buyers is “AI”, the use of artificial intelligence; this is often very broadly applied such that many of the early 1960s experiments could have been argued as artificial intelligence, but as computing power and complexity has increased, so it becomes possible for computers to achieve increasingly interesting things.
A great example is making use of image databases and the Internet; there are several AI image mechanisms which can be fed words, and will create an image by running those words through an image database such as Google Images, and mixing the results together following an algorithmic method. The results are often fascinating; any given word will have many image-associations online, and as such what the AI chooses to represent can feel very much like an artificial intelligence creating an image.
And now that these images can earn money; AI generative NFTs have sold for hundreds of thousands of dollars, it’s certain things will continue to get more interesting rapidly, as artists seek ways to collaborate with AI’s, or build them to play as much of a creative role as possible, or simply seek unexpected behaviours and glitches amongst the carefully crafted code. For now, it is still artificial creativity rather than direct computer creativity, but as the technology continues to expand, so do the possibilities.