JOHN F SIMON, JR: "CREATING AN AI TO INNOVATE LIKE ME"

John F Simon, Jr: My improvisational drawing practice started because I had a goal to design a creative artificial intelligence, a general AI that could innovate like me. This was in 1999, and at the time I was busy coding finite automata for my art appliances. Those generative software pieces produced more variation than I would ever see in my lifetime, but the variation was always within the code structure. I began dreaming of software that could exceed those boundaries in a coherent way.

To discover how to write such software, I asked myself the question, "How do I create?" I thought if I could find the hidden rules guiding my drawing, I could code them, and that would be the basis for the AI. I sat quietly and began making marks on paper, allowing my hand to move freely, and at each turn, I tried to look inside and see in myself what was guiding that decision.

For example, if I made a line and then made one perpendicular to it, I would wonder, "Why did I do that? Why is it square? Why not an angle? Or a curve?" Needless to say, I never found a concrete set of rules because creative decisions depend on countless factors that are difficult to quantify, like "I did it because it felt right." The code never got anywhere, but that style of introspective drawing brought me so many benefits that I have continued to do the practice every day since then.

JFSjr: A number of people are subscribed to the email distribution of the daily drawings on iclock.com, and I see that others regularly check the site. I often receive comments via email when a particular drawing resonates with someone. It does help me to know that there is a small audience enjoying the flow of drawings. I hope that my steady output will encourage others to create. I also feel like I can bring a moment of art into people’s lives who may not encounter art very frequently. These things help my discipline and effort, so, yes, these external factors influence the existence and duration of the practice.

JFSjr: Daily drawing has had a profound effect on me and my life. In the late 1980s, I coded digital drawing tools and spent late nights in the studio with a big screen and Wacom tablet, creating drawing after drawing and modifying the tools as I went.

The drawings from that time showed me how my non-verbal thoughts found expression in a drawing. It was very therapeutic and helped me work through a lot of grief. When I started my Creative AI project in 1999, I was fortunate from the beginning to shape my practice around introspection. I like to say, "wrong question, right answer." The question was wrong because there is no single programmable set of rules for making drawings. The answer was right because the research method was identical to a method of insight meditation, and I soon began to reap the benefits.

This ultimately led, about 15 years later, to my book, DRAWING YOUR OWN PATH: 33 PRACTICES AT THE CROSSROADS OF ART AND MEDITATION, where I attempt to describe the benefits of daily drawing. I am part of a small community of 'creative contemplatives' that formed around the book, and I never expected that to happen, but this practice is now a part of my daily life.

JFSjr: The funny thing is that not only did I not draw as a kid, but I actively disliked drawing because I wasn’t good at it. I mistakenly thought that being good at drawing was a talent you were born with and didn’t appreciate that drawing is a skill that can always be developed. However, I was still creative, and I chose to do photography because it was a kind of 'instant realistic drawing' for much less work.

This sense of using technology to amplify skill led me into computers, and ironically, automated drawing tools were my first creative software efforts. I wrote about these programmed drawing tools in MOBILITY AGENTS (2005).

JFSjr: My love of photography led me to learning code. My undergraduate degree combined computers and photography. I produced a series of photographic prints chosen from Viking Orbiter images of Mars. I approached the selection of digital satellite images with the sensibility of a photographer, often looking for Weston-like compositions in the Martian sand dunes. This was five years before Photoshop was released, and I could already see the wave coming.

JFSjr: I don’t know if I ever wanted to be an artist, but I realized one day that I was one.

JFSjr: It was the day I looked back and realized that I had easily abandoned promising career paths in scientific research, university teaching, and computer systems consulting to spend all my time in the studio coding and drawing. I started to accept that I needed to be doing creative work full-time and should figure out how to get paid for it.

JFSjr: Living in NYC in the early 1990s, I became determined to sell my art in SoHo, the hot neighborhood for galleries at the time. I studied the market and spent a lot of time looking at what art was selling. The positive feedback from this approach was that I got to spend a lot of time looking at art!

In the mid-1990s, Sandra Gering, my longtime art dealer and friend, gave me her 'seductive object' speech, outlining a stereotype for collecting. That shook me out of being a purely conceptual artist and encouraged me to think about materials, surfaces, scales, etc., of the art I was creating and how important these aspects were to someone buying it.

In 1999, when I premiered the first Art Appliance, I tried to balance the design of the housing with the style of the software. By bringing software into a sculptural object, John Hanhardt told me I was 'acknowledging the market'.

Selling your art is not required to be a successful artist. Art and the art market are separate. But if your goal is to sell art, it’s easier to sell if you make something people want to buy.

JFSjr: Working with Björk on BIOPHILIA was fun, stressful, and exciting.

My interactions with Björk were memorable. She radiates creativity, and brainstorming app design with her was enjoyable. All of her collaborators were given creative freedom for their app designs, allowing me to work at my best level. I aimed for the piece to be memorable. Her ability to inspire others enabled her to trust a large group of people to create a singular product. We worked hard on tight deadlines and pioneered a new concept. I'm not aware of anyone else creating another app album. Being part of something so unique was thrilling.

I vividly recall attending an elaborate release party with my wife in a loft in Chelsea, featuring a live performance with a large choir. We were informed that the entire project, including the source code, was acquired by MOMA and exhibited there in 2023.

Upon the project's completion, I received an exquisitely crafted wooden box containing a copy of the album and artwork from the project. Björk had them specially made as gifts for all the collaborators. Truly considerate and wonderful, isn't it?

JFSjr: So much to say about EVERY ICON. I wrote the first version 27 years ago (1997) as a Java Applet and a Palm Pilot app. I sold copies as editions and created a licensing system similar to the NFT contracts used today. The work was highly successful in the art world, being featured in the 2000 Whitney Biennial and showcased in numerous articles and exhibitions.

In 2021, witnessing the explosion of NFTs, I realized that the blockchain was the natural home for EVERY ICON and began exploring the best way to launch it. Fortunately, my friend and patron John Borthwick contacted me and stated, "EVERY ICON needs to be an NFT."

John assembled a talented team that included Harri and Arran from @divergence_art, Sam Spike curating for @FingerprintsDAO, and Danika Laszuk from @betaworks. We launched in December 2021 during the peak of ETH and quickly sold out. It was indeed a thrilling ride.

JFSjr: The original COMPLEXCITY was created in 2000, during the peak of the dot-com boom, as a joyful tribute to New York City. At that time, my art appliances were selling well and gaining recognition, prompting me to develop a work that paid homage to the city, particularly midtown NYC, where my studio overlooked the skyline. The piece consisted of nine separate programs running simultaneously, each exploring endless variations of New York themes. For instance, one section symbolized elevators with colored boxes moving up and down shafts, creating ever-changing compositions. Another section was a traffic simulation inspired by Mondrian’s Broadway Boogie Woogie.

In 2023, I began discussions with Art Wrld about releasing an updated version of ComplexCity that would be archived on the blockchain, offering me the opportunity to reimagine the piece following my move out of Manhattan. The NFT version reproduces the 'classic' nine modules of COMPLEXCITY and then remixes them into five new neighborhoods. The transition from my life in NYC in 2000 to residing outside the city in 2023 represented the dissolution of boundaries and the merging of different aspects of my life.

It was both enjoyable and enlightening to revisit and revise a software artwork after such a long period. Technological advancements allowed me to incorporate elements that were previously unattainable in terms of computing and display speeds. I expanded on existing ideas and introduced new ones, resulting in more intricate interactions—as expected in COMPLEXCITY.

JFSjr: I love both realms and have spent decades working on this boundary, attempting at certain points to create pieces that dissolve or merge the two. The art appliances were constructed to showcase the dynamic, ever-changing nature of software, housed within an object designed for visual and textural appeal.

I aimed to create pieces where screens were naturally integrated into the composition. To achieve this, I built up surfaces next to screens, covered screens with laser-cut plexiglass, reflected screens on mirrors inside boxes, and projected the light from screens onto thin materials.

At other times, I created pen plotter drawings, laser engravings, and CNC carvings to explore the digital control of materials. Similarly, I have published software pieces that can be viewed on any screen to take advantage of the malleability of code and its continuously variable nature.

JFSjr: Within the broader range of explorations of AI-assisted imaging that I am conducting in this new set of work at EXPANDED.ART, the BIRD piece represents the deepest venture into machine hallucinations. I am utilizing a control net, LoRAs, and prompts to modify some of my existing drawings that were originally created on paper. For the most part, I tone down the creativity, but sometimes there is a detail in a drawing that the AI just can't let alone.

I imagine this is the machine equivalent of when someone points out a face in a painting to you, and you can't unsee it. The marks in this particular base drawing must have strongly resembled a bird. I experimented with various settings and obtained numerous bird variations. I found the way this bird was rendered to be truly striking.

During a conversation with my son, he mentioned that the right side looked like my typical work, but the bird did not. I felt that this aspect made this image a successful collaboration with the AI.

JFSjr: Artists have always engaged in dialogue with one another. I regularly visit museums and galleries in NYC and have done so for years. I am aware that what I see influences me. The impressions I gather ultimately impact my work. AI systems automate and accelerate this assimilation process.

Think about all the photographers, illustrators, cartoonists, and painters whose art has been utilized to train these large image generators. If I am collaborating with anyone, it is with this creative collective. I have some aggregate of impressions of many artists to thank for the way that BIRD was rendered.

JFSjr: A productive collaboration with AI would be one where the novelty that arises feels like a natural extension of my creative process. A similar feeling to the outcomes in my improvisational drawings: I am surprised and delighted but can still call them my own.

JFSjr: I read a lot of science fiction as a young adult so Asimov’s robots and HAL from the film 2001 may have put the idea of AI in my head.

The first time I tried AI as a programmer was when I studied LISP in 1988 and wrote a natural language parser to move graphical objects around a screen.

JFSjr: I thought it would take a long time and much more powerful machines to make AI work well.

JFSjr: There was a long period, maybe 25 years, where the AI breakthroughs were always 'coming soon'. The combination of big data, high speed bandwidth, and GPU clusters seems to have crossed a threshold. I am amazed at what is happening now and am energized to use the new tools. 

When I wrote EVERY ICON (1997), the irony was that brute force computing would never scratch the surface of discovering coherent images in noise. Now stable diffusion gives us that power. We have a vehicle to fly through possibility space, a tool to find coherence and push beyond. The promise was fulfilled in an unexpected way.

JFSjr: Yes and no. Every system has boundaries. A software that exceeded its structure would have to be self-modifying and the path of modifications would also have to be varied. That being said, AI models can evolve, weights can be changed through use, images added, and other open collective actions can make the software feel more like a living thing than a machine. That open ended quality makes AI exciting to interact with because the code structure is only part of the game.

JFSjr: So complicated. A question I get asked a lot when I teach improvisational drawing is, "How do I know when the drawing is done?"

I often have students practice a technique that is similar to what Shinzen Young calls 'noting gone':

Sit quietly and look at your work. As your eye passes over the drawing, a detail may interrupt the flow, or an imbalance in the composition irritates you, or some color distracts your eye. If that happens, work to correct it and then return to looking. When you can sit quietly and note the absence of any thoughts about changing what you see, the drawing is done.

For exhibitions I’m comfortable choosing any work I’ve finished like this, but selecting work for exhibition usually depends as much on the exhibition space, the budget, the vision of the curator, etc. as the content of the images.

JFSjr: If I sit with a finished work, I never think about how to change it. What I notice is how I’ve changed since it was made. Seeing older work often triggers memories from the work’s production process and reminds me of how I saw myself at the time as an artist.

When I see Color Panel v1.0 from 1999, I remember how limited the processing power of the PowerBook 280c was and how I had to craft the code to get the motion and timing perfect. If it was perfect then, it is perfect now; and it also carries the style and capabilities of 1999.

JFSjr: In general I am excited by how NFTs broke into the public awareness and it seems to me that we have only begun to explore the possibilities of digital ownership and smart contracts.

I think collectors need more elegant ways to buy and archive NFT collections. Wallets could be streamlined and made safer. There could be dedicated online spaces to organizing, browsing and showing off collections. It would be better for sales if crypto was more stable or NFTs were universally priced in fiat or stablecoins.

JFSjr: Yes! Absolutely. NFTs happened at a perfect time for me. NFTs made my life easier in many ways. Sales of EVERY ICON brought me a much needed infusion of revenue just after the pandemic shut down my gallery shows and killed sales.

A younger generation, many of whom were not oriented toward fine art, saw and collected my work–that is a goal for any creator. The blockchain turned out to be perfectly suited to make code based art accessible and functioning. That in itself makes my life better. I am constantly maintaining older software artworks and being able to release code art that runs on all devices makes me sleep much easier.