The First Computer Artists
Georg Nees & Frieder Nake
Two mathematicians who hung plotter drawings in Stuttgart galleries in 1965 and called them art.
Stuttgart, February 1965
In February 1965, in the Studiengalerie of the Technische Hochschule Stuttgart, Georg Nees hung drawings made by a program on a digital computer and drawn by a pen plotter, and called them art. The exhibition, titled Computer-Grafik, opened on 4 February and ran from the 5th to the 19th, held under the Ästhetisches Colloquium of Max Bense's Institute of Philosophy and Theory of Knowledge. By the conventional count in the field's histories, it was the first exhibition of computer art anywhere in the world.
Nees worked outside the art world, as an industry mathematician at Siemens's Schuckertwerke in Erlangen, where he had been since 1959. He made these drawings on his own time.
Two months later, on 6 April 1965, A. Michael Noll and Béla Julesz opened Computer-Generated Pictures at the Howard Wise Gallery in New York, running through 24 April. A second Stuttgart show, of work by Nake and Nees, followed in November. Histories group the three pioneers as the "three big N's" (Noll, Nees, Nake), scientists who found independently and almost at the same time that a machine built for calculation could be made to draw. Nees's show is simply the earliest dated.
Bense, generative aesthetics, and "artificial art"
Nees reportedly showed his graphics to a fellow train passenger, Helmar Frank, one of Bense's doctoral students, who told Bense; Bense, intrigued, arranged the February show almost at once. Nake, who knew both men, later said that Bense did not invent their method but "prepared the frame of mind" in which such image-making became thinkable.
Bense's larger project was generative aesthetics: deriving aesthetic value from measurable, programmable rules instead of from taste or genius. It reached print the same year in a slim pamphlet, rot 19, which paired six of Nees's earliest drawings and his own commentary with Bense's essay "Projekte generativer Ästhetik" (Projects of Generative Aesthetics). rot 19 is often called the first publication anywhere specifically about computer art.
The show also gave the work a name. As the story is often told, at or around the February opening a visiting artist-professor from Stuttgart's art academy asked Nees whether the machine could be made to draw the way the artist himself drew. Nees is reported to have answered yes, of course, on one condition: that the artist first explain exactly how he drew. In the irritation that followed, Bense is said to have coined künstliche Kunst, artificial art, on the spot. The term separated art as purely human activity from art produced by a machine, while still counting the machine's output as art of a different kind. The exchange survives only in secondary retellings, with no named eyewitness or primary transcript, so treat it as reported speech rather than a quotation. Nake himself later took up the term in his essay "Art in the Time of the Artificial."
(A 2024 technical reconstruction of Nees's 23-Ecke reports that Bense, reprinting it in rot 19, cropped a row from it and dropped Nees's attribution. It is the only source that says so.)
The plotters they used
Both artists used the same model of machine, the Zuse Graphomat Z64, an early flatbed pen plotter, at different institutions. Nees used the one at Siemens in Erlangen, having been instrumental in the company's 1964 acquisition of it, and programmed largely in ALGOL. Nake used the Graphomat at the TH Stuttgart computing center. As a student he had written the control software that let it be driven by algorithmic instruction, and his first test graphics were written in machine language because the ALGOL tooling was not yet finished. ALGOL 60 control came later and is documented for Walk-Through-Raster in 1966. Nake drove his plotter partly through an SEL ER56 computer.
A plotter of that era could move a pen, put it down, draw a straight segment, lift it, and move again, one pen at a time. It could not fill shapes, it drew curves only as many short straight chords, and color meant changing the pen. Any disorder in the drawings came from pseudo-random numbers computed in the program. The simulated plotter in this lesson's tool has the same limits, and it runs reconstructions of four of their algorithms with parameters you can change.
Four works and their algorithms
Each of these works is described precisely enough to be reconstructed as an algorithm.
Nees, 23-Ecke (twenty-three corners, 1964). A grid of closed polygons, each built from exactly twenty-three line segments, traced by a random walk Nees called "ELIRR" that alternates axis-parallel and diagonal steps drawn from a large modulo random-number generator. The commonly reconstructed grid runs fourteen columns by roughly nineteen or twenty rows, depending on which of two known print variants you are looking at. It was shown at the February 1965 exhibition and reprinted in rot 19.
Nees, Schotter (Gravel, 1968). A field of identical squares in rows, with random offsets in position and rotation that grow from the top of the sheet to the bottom, so the grid breaks up into a heap toward the foot of the page. The Victoria and Albert Museum holds a print of it, inscribed "Computergrafik mit Siemens-System 4004" and dated to the 1968–1970 window. Commonly reproduced versions show roughly a dozen columns across roughly twenty rows; the museum record gives no grid count, so treat that figure as an approximation.
Nake, Hommage à Paul Klee (1965). A horizontal-and-vertical proportional framework derived from Klee's 1929 painting "Hauptweg und Nebenwege" (Highways and Byways). Nake wrote random variables into the parameters, so the machine made its own formal choices within limits set by the inherited structure. It was realized as a screenprint, 49.2 by 49.2 centimeters, in an edition of forty, produced with an SEL ER56 driving the Zuse Graphomat Z64, and is held by the V&A. It is one of the most reproduced images in the histories of early computer art.
Nake, Walk-Through-Raster (1966). A small repertoire of signs placed on a grid. Written in ALGOL 60, it used one function to choose a sign and a second, probability-weighted function to choose each following sign from the one before it.
The tool runs all four, listed under Program as 23-Ecke, Schotter, Klee, and Raster. Set Columns and Rows, move the Disorder slider, and press New seed to get other drawings from the same program.
Nake's 1971 essay, "There should be no Computer Art"
Both artists showed at Jasia Reichardt's Cybernetic Serendipity at the ICA in London in 1968. Three years later Nake turned against computer art. In October 1971 he published a short, blunt essay, "There should be no Computer Art," in the Bulletin of the Computer Arts Society (pages 18 and 19).
His argument was political. He did not believe computers should be used to manufacture another art fashion, and he worried that computer art was becoming a market commodity instead of a question about signs, communication, and society. He wrote:
"aesthetic information as such is interesting only for the rich and the ruling" — Frieder Nake, "There should be no Computer Art," Bulletin of the Computer Arts Society, October 1971
He proposed that computer scientists study the alienation between artists and technology, analyze artistic sign-repertoires rigorously, and ask what aesthetic information might mean for society at large, rather than produce salable prints.
Nake stopped making and showing computer art but stayed in academic computing. He became a professor of interactive computer graphics at the University of Bremen in 1972, returned to showing his early work publicly from around the mid-1980s, and in 1999 co-founded the compArt database, one of the sources that make this history reconstructable today.
After Stuttgart
Nees died in Erlangen in 2016. His dissertation, "Generative Computergraphik," completed under Bense in 1969, is widely cited as the first doctoral dissertation on computer art, though the sources hedge that with "arguably."
Manfred Mohr came to a similar rational, generative approach on his own in Paris, after reading Bense's information aesthetics, and showed alongside Nees and Nake at Cybernetic Serendipity in 1968. Vera Molnár pursued the same questions in her own way.
Three of these artists later received the ACM SIGGRAPH Distinguished Artist Award for Lifetime Achievement in Digital Art: Mohr in 2013, Molnár in 2022, and Nake in 2025, decades after the two Stuttgart shows. The Manfred Mohr lesson and the Vera Molnár lesson cover their work.
The Studio
Adjust the controls and save the studies you want to keep.
Exercises
1. Find where Schotter's grid breaks
~8mLoad Schotter and set Disorder to 0, which gives a perfect grid. Raise it slowly. The jitter increases down the sheet, so the top rows keep their order while the bottom rows scatter. Find the Disorder value where the grid first stops reading as a grid.
Goal: Find the Disorder value at which Nees's grid turns into a heap.
2. Two versions of 23-Ecke
~7mLoad 23-Ecke and set Disorder very low: each polygon is a tight knot. Set it high and the same twenty-three segments spread across the cell. The only thing you changed is the step size of the random walk.
Goal: Produce two distinct versions of Nees's 23-corner polygon by changing only the step size of the walk.
3. What did the framework contribute?
~8mLoad Klee and set Disorder to 0. What remains is the bare horizontal and vertical framework Nake took from Klee's painting. Raise Disorder high and the boundaries start to wander. Decide which parts of the image come from Klee's structure and which from the randomness Nake added.
Goal: Separate the contribution of the inherited framework from the contribution of the random parameters.
4. Recover a drawing from its seed
~5mMake a drawing you like with any program and note its Seed. Press New seed a few times, then type the old number back into Seed. The same drawing returns, because every random choice the plotter makes comes from the seed. The program, its settings, and the seed are enough to redraw the work exactly.
Goal: Recover a specific drawing by re-entering its seed, and see that the program and its settings, not the print, are what reproduce the work.
Sources & further reading
Primary sources
Recorded conversation between Wolf Lieser and Frieder Nake discussing Georg Nees. Digital Art Museum.
The train anecdote (Nees showing Helmar Frank his graphics) and Nake's characterization of Bense as having prepared the frame of mind rather than initiating the work.
Frieder Nake, "There should be no Computer Art," Bulletin of the Computer Arts Society, October 1971, pp. 18-19.
Nake's renunciation essay and the confirmed phrase about aesthetic information serving only the rich and the ruling.
Frieder Nake, "Art in the Time of the Artificial," Leonardo Electronic Almanac.
Confirms Nake's own continued engagement with Bense's künstliche Kunst (artificial art) framing.
Museum and foundation records
Digital Art Museum, artist page for Georg Nees.
Nees's dates (1926-2016), Siemens Schuckertwerke career, the 1964 Zuse Graphomat Z64 acquisition, and the 1969 dissertation.
Digital Art Museum, artist page for Frieder Nake.
Nake's dates, first 1964 drawings on the TH Stuttgart Graphomat, key works, the 1971 renunciation, and the Bremen career.
Victoria and Albert Museum, collection record for Schotter, Georg Nees.
Medium, 1968-1970 dating, and the inscription Computergrafik mit Siemens-System 4004. Grounds the hedge on Schotter's unverified grid count.
Victoria and Albert Museum, collection record for Hommage à Paul Klee, 13/9/65 Nr. 2, Frieder Nake.
Screenprint dimensions, edition of forty, SEL ER56 and Zuse Graphomat Z64 hardware, and the Klee framework derivation.
Scholarly sources
Zellyn, "23 ecke," technical reconstruction of Georg Nees's algorithm, 2024.
The ELIRR random walk, the 23-segment polygon structure, the 14x19/14x20 grid ambiguity, and the single-sourced rot 19 cropping claim.
A. Michael Noll, "The Howard Wise Gallery Show: Computer-Generated Pictures (1965)." Leonardo, vol. 49, no. 3.
Confirms the April 1965 New York show's dates, used for the roughly simultaneous three-pioneers framing.
compArt (University of Bremen), exhibition record for Computer-Grafik, Studiengalerie der TH Stuttgart, 1965.
The February 1965 show's exact dates and venue, under Bense's Ästhetisches Colloquium.
Related lessons
Exhaustive Variation
Manfred Mohr (b. 1938)
From 1973 Mohr built his work on the cube's twelve edges, and later on higher-dimensional hypercubes.
Order and Disorder
Vera Molnár (1924-2023)
She added a small, controlled amount of randomness, her 1% disorder, to strict geometric order.

