Back up your NFTs and discover
new artists at

Join Now
July 10, 2023

The Interview | A. Michael Noll

One of the fathers of computer art discusses his career at Bell Labs and beyond with Douglas Dodds
Credit: A. Michael Noll, Hypercube, 1965. Courtesy of the artist
Now Reading:  
The Interview | A. Michael Noll

I first met A. Michael Noll in person a decade ago, though I’d been familiar with his work for much longer. After exchanging emails for several years, I included him in a V&A exhibition I co-curated in 2009, entitled “Digital Pioneers.” We then attended a seminar at the School of Visual Arts in New York in 2013 in connection with an exhibition entitled, “The American Algorists: Linear Sublime.” We spoke at the event and got on well, thankfully! 

A few months later, in March 2014, I visited Michael at his home in New Jersey, passing close to Bell Labs on my way, where he had undertaken groundbreaking research in the 1960s. In 1965, iconic works such as Gaussian-Quadratic (1963), Computer Composition with Lines (1964), Vertical-Horizontal No. 3 (1964), and Ninety Parallel Sinusoids with Linearly Increasing Period (1964) were exhibited alongside those of Béla Julesz as part of “Computer-Generated Pictures” at the Howard Wise Gallery in New York. The exhibition is thought to be the first public display of computer art in a commercial gallery in North America, and one of the first in the world. His works have since been exhibited in numerous major shows, from “Cybernetic Serendipity” at London’s Institute of Contemporary Arts (ICA) in 1968, to “Chance and Control: Art in the Age of Computers” (2018) at the V&A, to “Coded: Art Enters the Computer Age, 1952-1982” (2023) at LACMA.

Sitting in his home office back in 2014, we discussed his time at Bell Labs, his work for the Executive Office of the President, as well as the future of artificial intelligence and digital art. Sadly, I didn’t manage to record our conversation at the time, so it was a real pleasure to have the chance to speak to him again earlier this year for Right Click Save. The following transcript represents just a few highlights from our wide-ranging conversation, which lasted for several hours and could have gone on for much longer!

A. Michael Noll, Ninety Parallel Sinusoids with Linearly Increasing Period, 1964. Courtesy of the artist and Victoria and Albert Museum, London 

Douglas Dodds: So Michael, how did it all begin?

Michael Noll: It all began in 1961 when I was working at Bell Telephone Laboratories Incorporated, also known as Bell Labs in Murray Hill, New Jersey. During the summer of 1962, I had an internship in the research division where I was working mostly on speech processing, pitch detection, and technical things involving the microfilm plotter — plotting out data, and using their big IBM 7090 digital computer. 

On one occasion, somebody used the plotter and something went haywire, and it came out with lines all over the place. A person named Elwyn Berlekamp comically called it “computer art,” but I thought to myself: “gee, people were making computer music, why don’t we do computer art seriously, combining random algorithms with mathematical equations?” So that’s what I did.

DD: Had anybody actually used the term “computer art” before 1962?

MN: Not that I know of, but it’s possible. In August 1962, I wrote a technical memorandum documenting what I had been doing that was distributed to probably 40 or 50 people at Bell Labs. Computer music was [already] going on there, and now there was digital computer art. I titled the technical memorandum, “Patterns by 7090,” since my management didn’t want me to use the term “computer art” at all. What I didn’t know at the time was that there was a front page of Life Magazine showing computer art by Maughan MasonLissajous patterns that he had done with an analog computer. 

Back then, if you used the term “computer,” you had to put the words “analog” or “digital” in front of it. Most computers back then were analog.

In the early 1960s I started doing three-dimensional stereoscopic presentations of data, so that we could look at the data in real 3D, not just X and Y, but X, Y, and Z. People were doing computer animation back then, so I started doing 3D computer animations. Stereoscopy required polarized glasses and projectors in order to see it. I went to a ballet performance in New York and realized that we could do a stereoscopic, 3D, computer-generated ballet with stick figures running around on the stage.

Somebody suggested it might be interesting to look at a painting and see if a computer could make a similar version. That got me into the work of Piet Mondrian, and I started looking at his paintings. He did a series with horizontal, vertical, and circular bars, so I made a computer version of it. Because of my interest in psychology, I got the idea to show reproductions of both works to a hundred people and see which they preferred, as well as which they thought was done with a computer. That experiment was published in a peer-reviewed journal, The Psychological Record.

A. Michael Noll, Computer Composition with Lines, 1964. Courtesy of the artist and Victoria and Albert Museum, London

DD: Do you know of anybody else who attempted to do that, to codify the process of making art?

MN: No, nobody else had tried to do it. 

DD: It’s rather like what’s going on now with artificial intelligence and people using data sets of paintings to generate their own images. You were doing it in the 1960s?

MN: Yes, everything I’m seeing that people are doing today is very much a newer version of what went on in the ’60s.

DD: But basically, you did it first?

MN: Well, me and others. Martin Krampen and others were doing a conference in Canada and invited me to speak. I wrote a very broad paper for it called “Computers and the Visual Arts.” I believe that, even today, it is a fine introduction to computer art and should be more broadly available. But it is challenging to know who was the first, and I try to avoid such claims. Periods of time are usually ripe for specific innovations and a number of people often “stumble” separately upon something new to them. All deserve credit. 

For my part, I combined randomness with mathematics to create Gaussian-Quadratic (1963). That was my favorite piece of all the things I ever did. I liked it because it had a cubist block, which reminded me subconsciously of Picasso’s Ma Jolie (1911-12), which I’d always enjoyed seeing at MoMA along with Van Gogh’s The Starry Night (1889). 

My work wasn’t done with the Picasso in mind but, gee, that particular version just popped out randomly. The result was just: “wow.” 

I also made a computer version of Bridget Riley’s work, which is Ninety-Parallel Sinusoids (1964). If I were doing an exhibit someday, I would love to have a large reproduction of Gaussian-Quadratic next to Picasso’s work for people to view together. I used to take paper versions of Gaussian-Quadratic and color the lines with a magic marker. [...] I then decided to register the copyright for the work, so I sent off the form to Washington, listing the work as “computer art.” What followed was a cute exchange with the Copyright Office about whether or not you can copyright creativity

The Copyright Office said that they couldn’t register something that was done by a computer — it had to be done by a human. I then told them that a human had written a program, which involved some random elements. They replied that they couldn’t register the copyright for anything done randomly. I explained that it would only appear as random numbers to humans, but that they were actually calculated by an algorithm. In the end they accepted it.

A. Michael Noll, Gaussian-Quadratic, 1963. Courtesy of the artist and Victoria and Albert Museum, London 

At the time of Gaussian-Quadratic, I also had the idea of using a computer to generate sculptures out of lines in advance of their construction, so you could look at them in 3D from different directions. Then someone else suggested doing a 3D animation of a rotating 4D hypercube projected in 3D. So I did that too. These were all done in the first half of the 1960s in a fairly short space of time. 

I then had the idea of putting words in 4D space, and projecting them in three and then two dimensions. That was used to do the title sequence for the famous AT&T documentary, Incredible Machine (1968). I had a projector with a polaroid adapter with both left and right images — that was true 3D. I would take it out and show it to students in my lectures and things would jump right off the screen. It was an exciting and fun period of time.

Then Howard Wise got involved. He saw that Béla Julesz was doing dot stereograms at Bell Labs with random patterns where you could actually see some shape. He was trying to draw conclusions about where the brain did its processing, but those patterns looked very pretty on their own. Howard Wise invited Béla to do a show at his gallery on West 57th Street in New York, and Béla invited me to join. 

DD: That show was called “Computer-Generated Pictures,” wasn’t it? But there was no mention of “art”?

MN: Well, Béla didn’t consider his stuff as art — it was science, not art. I considered my work as art, not science.

DD: Even though you were working in a scientific lab?

MN: Yes. A lot of people got interested in programmed art at that exhibition. Nam June Paik saw the show and then visited Bell Labs a few years later and tried to do some computer art; he succeeded too. He actually came to Bell Labs a few times and was involved with me, but he went off and learned programming on his own. Gregory Zinman also did a computer animation at the time. Then, in around 1965-66, we hired a graphic designer, Aaron Marcus, to come to Bell Labs to work on page design and layout, but he also knew programming, so he did some early computer art too.

In the early 1970s, Max Mathews was doing a lot of computer music. Peter Denes had put together an interactive computer system, and Max was using it for music as did Emmanuel Ghent and Laurie Spiegel. Laurie wrote a paint program that produced lots of interactive stuff with color graphics. I recently wrote a little piece about the contribution of women at Bell Labs as there were a fair number of female programmers but many names have been lost to history. 

A number of women were involved in creating computer art in the 1960s and ’70s — Vera Molnar comes to mind immediately, as does Darcy Gerbarg. What I always look for is creativity and curiosity. I don’t care if the creator is male or female. Some of the female programmers at Bell Labs finally got promoted to full status and others went off to get doctorate degrees, including Joan E. Miller.  

Howard Wise Gallery, NYC Exhibition Announcements, 1965. Courtesy of The Anne and Michael Spalter Digital Art Collection

DD: In your paper about the Mondrian experiment you divided the cohort into different groups of people — the scientists, the experts, as well as non-technical people.

MN: I wanted to see if there was a preferred amount of disorder in patterns, and if people with artistic training have a different sense of aesthetics than those who do not. This was a complicated experiment and we had a hard time figuring out how to analyze the data. In the end, the conclusion was that there was a middle level of disorder, where people preferred having not too much mathematics and not too much structure. They liked something in between. I was also trying to get people in the field of aesthetics interested in using computer-generated patterns as stimuli to investigate human preferences in art.

DD: Although you worked in a scientific lab, you were actually really interested in art, weren’t you? 

MN: Yes. Back in high school I had taken art classes. I was very familiar with museums and visited them frequently in New York. Even in engineering school, I studied engineering drawing, which is very structured and certainly an art form in itself.

DD: Right Click Save readers are familiar with pen plotters, but they might not be so familiar with microfilm plotters. Could you explain a bit about the software and the hardware that you used to create some of your early images?

MN: The microfilm plotter was a very high precision cathode ray tube that had a mask with alphanumeric characters in it. They would take the beam, deflected through whatever character you chose, and then position it on a screen. While on that screen, the 35mm camera lens would be open — in essence, it was a high-speed printer. But, using a dot, the beam could draw and also be used as a graphics plotter. 

I used the Stromberg-Carlson SC4020 microfilm plotter for five years. That machine got replaced with a newer model, but, by 1968, I wasn’t doing this anymore because my intent was to generate interest for artists to start programming. The Howard Wise exhibition was very educational for people but, of course, nothing sold. Regardless, I don’t think Howard cared. He was independently wealthy and wasn’t trying to make money. He was educating the public. The question is, why didn’t it take off?

DD: What happened?

MN: Billy Klüver got involved with Experiments in Art and Technology (E.A.T.), and he believed that engineers had to collaborate and work with artists. They couldn’t do both — there had to be two people. All the shows he did attracted a considerable amount of attention in the New York area. And then Ken Knowlton and Leon Harmon did the Computer Nude (Studies in Perception I) (1967) and that attracted even more attention. It was hard to compete with that Nude, which exhausted the interest in computer art and, I think, slowed the whole thing down. It’s also unquestionably sexist. 

In my mind, the Experiments in Art and Technology was a disaster. Most of the technology never worked right, but because it involved Klüver and Robert Rauschenberg it grabbed the attention of the big money in New York. 
A. Michael Noll, Computer-animated 3D stereoscopic movies made at Bell Labs: random line object (kinetic sculpture); computer-generated ballet; 4D hypercube, hypersphere, and other 4D objects; and molecular motion. 1965-66. Courtesy of the artist

DD: So you moved on to other things?

MN: I was trying out an interactive computer environment with 3D input and stereoscopic 3D output. I was now doing this interactively: a 3D drawing program with a 3D joystick and 3D input. Ultimately, I automated the 3D input device so it became a tactile machine that allowed you to feel things that were virtual that didn’t really exist. That haptic [technology] was patented in the late 1960s, early ’70s, but those patents have now expired. 

In 1971, I was invited to Washington to work at the Executive Office of the President at the Office of the Science Advisor. This was a completely different career direction, which involved dealing with computer security and privacy issues — very top secret, sensitive topics. 

DD: Were you headhunted, or did you just up sticks and go?

MN: The person at Bell Labs who was in charge of research, Bill Baker, was the informal advisor on science technology to a number of presidents, including Richard Nixon. At the time, Nixon needed a new Science Advisor, and Bill suggested Edward E. David Jr, who used to be at Bell Labs. Bill decided that I might do well working for Ed, so I got sent down to Washington but I had no idea what I was doing. 

I became heavily involved in science technology policy at the Executive Office of the President — [I was] in over my head, really. 

After a few years, that whole thing collapsed. I wasn’t smart enough to feather my nest in Washington and make contacts with people in order to set up a higher paying, more impressive job. I was too stupid to do that, so I had to go back to the lab.

I subsequently got involved with the Picturephone, [trying to address] why people weren’t buying it. That got me involved with the marketing people at AT&T. I ended up transferring over there and, around 1980, I started to see the market for what I call “digitext,” which was the ability to send text messages to people and get answers back in real time, instead of talking on the telephone. 

A. Michael Noll, Vertical-horizontal number three, 1962-63. Courtesy of the artist and Victoria and Albert Museum, London

DD: A lot of your images have copyright dates on them, why is that? 

MN: Back then, the copyright date was the date on which a work was first exhibited publicly, not the day it was created. The reason I copyrighted it was because of the Howard Wise exhibition in 1965. AT&T owned half of Bell Labs and they didn’t like the idea of people from the labs doing computer art or computer music, so they tried to stop the show from going ahead. That only failed because Howard Wise had already announced it. Both Béla Julesz and I were told: “no publicity.” The way to stop publicity was to copyright the works in our name and not let anybody make photographs of them or else publicize the work. Most of the works were actually copyrighted a year or two earlier.

DD: Did Bell Labs ever attempt to take them back when they realized that computer art was a big thing? 

MN: No, it was too late.

DD: Leon Harmon and Ken Knowlton had a similar experience with the Computer Nude (Studies in Perception I), didn’t they? Bell Labs didn’t want to be associated with it at first?

MN: Yeah, that was a little bit different. I don’t know exactly what happened there, but the Nude still gets a lot more attention than anything else. It was incredible because when newspapers reproduced a photograph, they used dots of various sizes, depending on how they wanted it to look. With darker things, for instance, the dots are bigger. But Ken and Leon used the dot itself as a little picture element with different amounts of grayscale, so if you got real close, you saw all these little pictures. When you backed off, it all blurred and you saw the bigger picture.

The original was very large, maybe six-feet long and, one day, it was hung in Ed David’s office as a sophomore prank when he was away on vacation. If you were really close to it, you couldn’t tell it was a nude. But when you backed away, you could see it. 

Ed was pretty high-level management at Bell and when he returned from his holiday, he asked that it be taken away. He believed that the idea of hanging a nude was not acceptable. I think it might have ended up with Rauschenberg for a while, and then it ended up in The New York Times. It’s had tremendous publicity ever since.

Leon Harmon and Ken Knowlton, Studies in Perception I, 1997 (orig. 1967). Courtesy Victoria and Albert Museum, London

DD: I wanted to ask about some of the other exhibitions that you participated in, such as “Cybernetic Serendipity” in 1968? Did you actually get to see the show in person?

MN: No, I didn’t. Back in the 1960s, traveling from the US to any place by air over the ocean was a big deal. The first time I did it was to attend a conference in Israel organized by Vladimir Bonačić. On the way back, I stopped in London and visited a bunch of museums including the Tate Gallery, British Museum, and Imperial War Museum, as well as Harrods of course.

For years before she curated “Cybernetic Serendipity,” Jasia Reichardt travelled the planet looking at what was being done involving computers and technology and art. When you look at what she covered in that exhibition, it was an incredible documentation of the state of computers and art in 1968. When you compare the catalog with that of “The Responsive Eye” (1965) at MoMA, you realize that Reichardt was far ahead at the time. There were dozens of people in London: Maughan Mason with his message patterns and pen plotter, as well as the so-called “three Ns” [Frieder Nake, Georg Nees, and Noll]. 

DD: But obviously you’re familiar with the work of Nees and Nake in Germany?

MN: At the time I did my work, I was not. At the time of the show at the Howard Wise Gallery, I was not familiar with them and what was going on in Germany. [...] I had heard of what they said they were doing. But, even so, they don’t go back to the time frame of 1962.

DD: So you were creating images, but not exhibiting them for a few years?

MN: I was certainly documenting [my work] at Bell Labs. I’m not claiming that I was the first to do anything, but there’s documentary evidence [for the 1962 date]. I got permission from Bell Labs a couple of years ago to reproduce the technical memorandum and post it at my website. In theory it’s proprietary to Bell Labs even today, but I got permission from the legal department to do it. 

Back then, it was mostly analog computers that had been hooked up to a pen plotter. A lot of people were doing those kinds of patterns, many of which weren’t really programmed. The things I was doing used a digital computer, which was deliberately programmed with elements of so-called randomness — calculated randomness — and mathematical algorithms, which nowadays people call “algorithmic art.”

DD: Or even “generative art,” which seems to be the more favored term these days.

MN: My objective was to educate. At Bell Labs, there was an environment [committed to] educating the public about science and technology, computers, sound, and electronics. 

I was trying to educate artists about the prospects of the computer as a new medium and, incidentally perhaps, a programmed, intelligent creative medium that could surprise you.

I envisaged a human observer being wired up so that their emotional state could be determined by an artist. The sounds and patterns you would see would elicit from you the various emotions specified by the art. Nobody’s ever done that, but I wrote it up at the time. 

Documentary featuring A. Michael Noll, Computer-Generated Ballet (1964), programmed at Bell Labs, 1968. Courtesy computerartist

DD: Are there things that you predicted or hoped would happen that perhaps haven’t arrived yet?

MN: I saw the use of the computer in choreography. In 1967, I wrote a paper in Dance Magazine called “Choreography and Computers.” I was trying to excite choreographers to use the computer to help them create a dance work by moving around little stick figures in order to produce a score, which dancers could then interpret. If I had continued in that direction, I would have studied human movement. My idea would be for people wearing black to have little white spots on them, which could easily be tracked by the computer. I didn’t get to do it, but I think others finally did.

DD: What was it really like working at Bell Labs? Did everybody really dress in suits,  the way they seem to in the films? Was it anything like today’s startup culture?

MN: If you were an engineer, you were expected to wear a white shirt with a tie and a jacket. During the summer, many of the buildings were not air-conditioned but the computer center was. That’s partly why I got involved with the computer, so I could hang around in a nice, air-conditioned environment. People would stay after work and play ping pong in the hallway. They even spent time in the labs on weekends and holidays. You couldn’t keep them away. 

The things that went on at Bell Labs in the 1960s were nobody’s full-time job. These were just creative people having fun. 

DD: Was Bell Labs the Silicon Valley of its day, then?

MN: At the time, New Jersey was the innovation state. The Edison laboratory in West Orange was the forerunner of today’s industrial research labs, and the David Sarnoff Laboratories in Princeton were responsible for color television. Bell Labs originated the transistor and a host of other innovations. At the same time, there were little electronics firms, workshops, machine shops, drafting people, and machinists scattered all over New Jersey. Bell Labs is now part of Nokia, so it’s Nokia Bell Labs, and there’s still an archive of the laboratory’s history and a little museum of some of the innovations. In 2015, I produced a short history of Bell Labs

A. Michael Noll, Computer Composition with Lines, 1964. Courtesy of the artist and Fleet Library at Rhode Island School of Design, Visual + Material Resource Center. 

DD: You moved to California for a while, didn’t you?

MN: I was working at AT&T consumer products at a time when marketing was falling apart. AT&T was a very sick company, so I got out of there. I had been teaching the basics of technology part-time at New York University, but was invited to join the faculty at the Annenberg School at the University of Southern California. I ended up spending 25 years out there and was Dean of the school for a few years. But then it came time to retire, and since I was born in New Jersey, I figured I’d die in New Jersey. 

California was an exciting period. I always hated Los Angeles but I love classical music and the quality of the musicians and concerts available there was incredible. It’s also a great museum town: LACMA, the Norton Simon Museum, and The Huntington Library are all wonderful. Southern California is probably even more exciting nowadays than New York. It certainly is for classical music.

DD: Your work featured prominently in LACMA’s recent exhibition, “Coded,” curated by Leslie Jones. Have you been able to follow recent developments in digital art as well as AI research?

MN: Yes, I was very pleased to see that. Everything you hear today about artificial intelligence all goes back to the 1950s and ’60s, and to artists like Harold Cohen. The father of Telstar, John Pierce, was one of the real geniuses at Bell Labs. He asked me to be the co-author on his book, Signals: The Science of Telecommunications (1990), which was one of the greatest honors I’ve ever received. John wasn’t too impressed with the artificial intelligence of computers but thought it would be better than the natural stupidity of humans.

As for cryptocurrency and NFTs, I’m still trying to figure them out. I just chuckle at the idea of people spending millions of dollars to have the right to look at something — not even to own it! It leaves me stunned.
Protect your NFT collection and discover new artists with ClubNFT

A. Michael Noll has had a varied career in communications as a researcher at Bell Labs, a pioneer in computer art and animation, staff member to the White House Science Advisor, AT&T manager and planner, academic professor and administrator, author, columnist, classical music critic, archivist, and biographer. He continues to write about telecommunication and other issues. Dr. Noll is professor emeritus at the Annenberg School for Communication and Journalism at the University of Southern California. He was dean of the Annenberg School for an interim period from 1992 to 1994. Dr. Noll spent nearly fifteen years performing basic research at Bell Labs and is one of the earliest pioneers in the use of digital computers in the visual arts and stereoscopic animation for art and science. 

In the early 1970s, Dr. Noll was on the staff of the President’s Science Advisor at the White House and later worked at AT&T identifying opportunities for new products and services. He is a Senior Affiliated Research Fellow at the Columbia Institute for Tele-Information at Columbia University’s Business School and was a member of the adjunct faculty of the Interactive Telecommunications Program at New York University’s Tisch School of the Arts. He has published over ninety professional papers, was granted six patents for his inventions at Bell Labs, and is the author of twelve books on various aspects of communications. Since 1990, he has published over 130 op-eds and column pieces in various national newspapers and trade publications and also wrote over 30 reviews of classical music for the Classical New Jersey Society. He archived the papers of Dr. William O. Baker of Bell Labs and is also responsible for the William O. Baker website.

Douglas Dodds is an independent curator and researcher. He was previously a senior curator at the Victoria and Albert Museum, which now holds an extensive collection of computer-generated art from the early 1960s onwards. V&A exhibitions include “Digital Pioneers” (2009-10) and “Chance and Control: Art in the Age of Computers” (2018).