First-Hand:Generative Art Memoir

Submitted by A. Michael Noll

August 3, 2025

Copyright © 2025 AMN

What Is Art?

“What is art?” is a topic that makes for great discussion amongst art scholars. If an elephant takes a brush in its trunk and applies paint to a canvas – is that art? Do other elephants look at the work and appreciate it as art? Is computer-art art made by a computer and enjoyed by other computers?

At one time, photography was not considered art. Then Ansel Adams and other photographers elevated photography to be accepted as an art form, and museums today have departments of photography. Almost all art forms and movements initially had challenges in being accepted as “art” – impressionism, cubism, and abstract expressionism. Even today it is hard to accept a large canvas painted in one solid color as really “art.”

The digital computer is a medium – but a medium that can be programmed to follow rules in making an artistic pattern. When can a pattern that results be considered as “art”?

Generative Art

The popular term today is “generative” art. It is art that was created following algorithms, or rules, by a digital computer, written by a human programmer-artist. Most artists follow rules of some kind in making their art. Artist Piet Mondrian clearly followed rules in his “Composition With Lines.”

Generative art usually combines rules with the disorder of randomness. But nothing done by a computer is truly “random.” Everything is by the rules of a program, although a program can be made to create numbers and images that might appear “random” to humans. The computer with its element of pseudo randomness can perhaps “surprise” the programmer-artist and thus appear “creative.”

The question of “randomness” became an issue when I attempted in the mid 1960s to register the 1965 copyright for my computer-generated art piece “Gaussian-Quadratic” at the US Copyright Office. However, only the ultimate work of a human could be registered – not anything truly random. I explained that I was the ultimate human who wrote the program that told the computer what to create, and the work was registered.

Ultimately, a human has written the program that controls the digital computer. Computers do not think -- they are not “intelligent,” although they might sometimes seem to surprise and produce the unexpected. Is that creativity? They can also be programmed to make decisions that might be unexpected. Human qualities of intelligence should not be attributed to machines. Today’s “artificial intelligence,” or “AI,” simply seems a new overly-hyped name for what used to be a digital computer – and is decades old. John R. Pierce liked to contrast the “artificial intelligence” of computers with the “natural stupidity” of humans.

Patterns by 7090

I started employment as a Member of Technical Staff at Bell Telephone Laboratories, Incorporated (aka: Bell Labs) in 1961 at its Murray Hill, New Jersey facility, after graduating from Newark College of Engineering. I was hired to perform research on the human factors of telephone distortions and other parameters. During the summer of 1962, I was assigned to the research division to perform research on pitch detection.

I was using the IBM 7090 mainframe computer and the Stromberg-Carlson SC-4020 microfilm plotter for my research. An error in a program caused a graph to look like some form of abstract art. I decided to program such abstraction deliberately – I had stumbled upon what today is called “generative” or “algorithmic” art. I described the results of my “computer art” in a Technical Memorandum (dated August 28, 1962) that was circulated widely within Bell Labs. My management cautioned me from calling it “art” and instead I called the results “patterns.” One pattern that I particularly liked was “Gaussian-Quadratic,” probably because it looked cubist and reminded me of Picasso’s painting Ma Jolie.

The patterns involved mathematical equations with pseudo randomness calculated by the computer. Once a program was written, I could then run it again and again on the computer, creating a series of similar patterns, yet each unique, and I could pick which one I liked best. I had the patterns printed on paper from the 35 mm microfilm and would color over the lines with color markers and give them as “art” to my colleagues to hang in their offices. It was all great fun, and most of the programming was done in evenings and weekends – this was not my prime work at Bell Labs.

This was 1962, and I was not aware of anyone else doing digital computer art. Analogue plotters were being used to create Lissajous patterns, some by Vaughan Mason. In the Fall of 1965, his art and mine were shown at the Fall Joint Computer Conference in Las Vegas.

I always had an interest in art. I had an art class in high school and studied engineering drawing in college. I visited local museums, particularly the Museum of Modern Art. My father was a machinist and carpenter, who liked to build and make things. My mother was a stenographer, typist, and a fine letter writer. I guess I learned from them both. Computer programming, art, and writing were their legacy to me.

Wise Gallery Show

Bela Julesz was a colleague at Bell Labs who was using the computer and plotter to make “random” patterns that he used for his experiments of human perception of three dimensions (random-dot stereograms). His patterns were shown on the cover of an issue of Scientific American magazine and were seen by Howard Wise, who had an art gallery on West 57th Street in New York City. Wise saw “art” in Julesz’s patterns, and asked Julesz if they could be exhibited at the Howard Wise Gallery. Julesz asked me to join him. The show “Computer-Generated Pictures” was in April 1965.

Julesz did not consider his patterns as “art” – I indeed considered mine as indeed “art” since they were made for pure enjoyment – not for scientific experimentation.

Howard Wise was a visionary and championed many new arts forms, such as Gerry Oster’s op art, Nam June Paik’s video art, and Per Biorn’s interactive art. Paik saw my computer art, and asked whether he could visit Bell Labs to learn programming to make art. He came starting around 1967, and indeed made some art, but then continued his video work.

Mondrian Experiment

A colleague at Bell Labs asked me whether I could program the computer to mimic an existing work of art. I needed a work in black-and-white that consisted mostly of lines, since that was what the computer could be programmed to make most easily. Piet Mondrian made a series of works about dock pilings in water that culminated with his “Composition With Lines.” It consisted of short lines in varying widths placed vertically and horizontally, mostly within a circle. I programmed the computer to make its version, utilizing “randomness” for lengths, widths, and placement, but following Mondrain overall.

With my background in the psychology of human factors, I decided to show reproductions of the two patterns to people at Bell Labs, asking which they preferred and which they thought was done by the human Mondrian. The majority preferred the computer pattern and believed a human made it! The experiment (“Human or Machine”) was published in the journal The Psychological Record. It is an early Turing test of machine versus human.

Other Work at Bell Labs

Much work was going on at Bell Labs in electronic and computer music. Max Mathews and John R. Pierce encouraged this work, which they justified as an extension of research into speech and sound. Musicians were visiting Bell Labs to use the computers there (Laurie Spiegel was a composer who also programmed an early interactive paint program to make her computer art). The Swiss conductor Herman Scherchen visited Bell Labs but was not impressed by the computer music and instead applauded our efforts in computer animation. Leopold Stokowski visited one day to see the research in speech and music. Roy Disney visited but saw no relevance of computer animation to Disney at that time.

Physicist Billy Kluver left Bell Labs in 1966 to form Experiments in Art and Technology (E.A.T.) in New York City. This movement was involved with Robert Rauschenberg and received much publicity. It had nothing to do with computer art, and possibly absorbed so much attention that computer art was ignored. In many ways, in my opinion, E.A.T. combined bad art with bad technology.

However, one use of the computer did obtain much publicity and attention when Leon Harmon and Kenneth Knowlton of Bell Labs used it to digitize a photograph of a nude woman (“The Nude”). These were the days of Playboy magazine, and nudes were “in.” But Gaussian-Quadratic” was not sexy like a nude woman and could not compete. Oh well. Knowlton considered the whole episode a “sophomoric prank.” It certainly was not research, and should have been an embarrassment to Bell Labs.

Stereographics

Stereoscopic 3D viewers were “the toy” in my childhood. It was natural then that I programmed the computer to calculate stereographic projections and plot the results as stereo pairs to display scientific information – and art. I also programmed 3D stereographic animation in which films were created by the plotter, initially in 35 mm and later in 16 mm. In addition to plotting scientific 3D data, I used the technique to create a computer ballet in 3D with little stick figures moving randomly about a stage. I wrote a paper describing this for Dance Magazine. The BBC came to Bell Labs and made a short documentary about my computer choreography around 1968.

A colleague at Bell Labs suggested I use my 3D animation to display a four-dimensional rotating hypercube, which I then programmed. Another colleague Man Mohan Sondhi explained the projection mathematics to me. The hypercube looked like a 3D cube turning inside out, and became a classic animation. I programmed words rotating in 4D, which were then projected to 2D. This technique was used to create the title sequence for the AT&T documentary “Incredible Machine” and also for Arthur C. Clarke’s TV special “The Unexplained.” I guess I should have left Bell Labs to start a computer title business in Hollywood, but I always avoided personal risk.

I became intrigued by interactive 3D and the interactive laboratory computer system of Peter Denes at Bell Labs; I designed 3D input devices and displays for this system. I also devised in the late 1960s a scan display system in color that utilized a buffer memory (now called a frame buffer). Charlie Matke, John Dubnowski, and Jack MacLean assisted me in these projects at Bell Labs.

Eventually, I invented an input/output device that could give tactile feedback. This research was the topic of my doctoral dissertation at the Polytechnic Institute of Brooklyn and resulted in a patent issued to Bell Labs with me as the inventor. My tactile system allowed users to see an object in 3D, and also at the same time, to feel the shape of that object.

I was doing research in other areas during these years. I was on a team at Bell Labs that was attempting to unscramble the audiotape from the Apollo 1 disaster. That led to a project to determine the effects of peak clipping on shouted speech, which I did with the assistance of Jack MacLean.

Bell Telephone Laboratories, Incorporated

My work in computer art was possible because I was employed as a Member of Technical Staff at Bell Telephone Laboratories, Incorporated (aka: Bell Labs) in Murray Hill, New Jersey. Bell Labs not only had the very best computers and equipment, but also the very best scientists and engineers, who were always eager to offer their opinions, criticism, and assistance. This environment resulted in a legacy of early digital art and animation done at Bell Labs in the 1960s.

Bell Labs welcomed outsiders: students, interns, academics, and even visiting musicians and artists. As stated earlier, video artist Nam June Paik came to learn programming with me, and created computer art. Artist Stan VanDerBeek came and collaborated with Ken Knowlton to create many computer-animated art films.

Innovation

Innovation occurs because the time is ripe, just as roses bloom in the spring. There is a “first” rose, but that does not matter – the ripe time occurred. So too with digital computer art. I had the computer and the cathode ray tube (CRT) plotter for my use. The time was ripe. I did it, and so too did others, although at the time we probably did not know of each other.

There are visionaries who see the coming of new innovation. Jasia Reichardt was one and organized the exhibit “Cybernetic Serendipity” at the Institute for Contemporary Art in London in 1968. She also later edited a book Cybernetics, Art, and Ideas in which she nicely edited together two of my published papers.

Some digital art pioneers seem forgotten. Leslie Mezei in Canada was a true pioneer in digital computer art. Martin Krampen and Maurice Constant were other Canadian champions of digital art and animation. Krampen organized a conference in Canada for which I described the possibilities for digital computers in the visual arts. Vladimir Bonacic in Yugoslavia was another champion, who later went to Israel to direct the Bezalel Academy and organized a conference on digital art, which I was invited to attend – my first flight across the Atlantic. I showed my computer animations and art works. Abraham Moles and John Whitney also attended.

Art historians, such as Grant Taylor in Pennsylvania and Margit Rosen in Germany, are now discovering the early work I did at Bell Labs. The collector Michael Spalter has my early work and considers some of the images as “iconic.”

My papers and works are in such institutions as the Los Angeles County Museum of Art, the Victoria & Albert in London, the Buffalo AKG Museum, the Huntington Library, the Museum of Modern Art, the NYU Dibner Library, and elsewhere. But some major museums still show no interest in my early works, much to my dismay.

Career Twists

I become curious about a topic or doing something new, and once I have done it, I become bored and move on to something else. By 1965, I had done my computer art, written about it, and had also created stereoscopic computer art and animation. I had designed the tactile system and also the scanned display system (frame buffer). After obtaining my Ph.D. in 1971, I left Bell Labs to work on computer security and privacy issues at the Executive Office of the President in Washington, DC.

After Washington, I worked in marketing at AT&T, investigating new services, such as videotex and digital text. We also investigated the market for video teleconferencing. I was an adjunct professor at Martin Elton’s Interactive Telecommunication Program at the NYU Tisch School, where I learned how to teach basic technological concepts to non-techies. That led me ultimately to a full professorship at the Annenberg School at the University of Southern California. Along the way, I wrote many op-eds and column pieces, and also was a music critic, and even dean for two years. I am now fully retired, lost in the past, and preparing for my final exit and journey. Sigh.