Rachel Falconer: The origin of your practice is a deep fascination with the natural world and its logics of evolutionary patterning. How would you describe your relationship with the ever-diminishing binaries between human and more-than-human entities, including the collaborative dynamics of generative art?
William Latham: I became fascinated by evolutionary processes and patterns while studying in London at the Royal College of Art (RCA) in the 1980s. During that period, I spent countless hours at the Natural History Museum nearby and the feeling of wonder at nature that I experienced there has never left me. You mention the logics of evolutionary patterning but alongside the awe-inspiring grand design(s) of evolution there is this tremendous variety of individual forms. There’s also a sense of abundance that is overpowering at times — masses of writhing worms or wiggly maggots feature strongly in my current work.
The fecundity of nature has always fascinated me as well as the ephemeral nature of individual forms. From the outset of my artistic career, I was already looking for a way of capturing forms in the process of becoming, in an endless succession.
I read and was deeply influenced by the mathematician Benoit Mandelbrot, who coined the term “fractal” to describe natural phenomena such as coastlines, snowflakes, mountains, and trees, whose patterns repeat themselves at smaller and smaller scales. Against that background, I began my FormSynth rule-based evolutionary hand drawings, in which a limited list of seven primitive geometric 3D forms were subjected to shape-changing operations that deformed the forms in specific ways. These drawings showed multiple generations of forms, with each generation showing the impact of the operation selected. They were vast, intricate and, of course, very laborious.
.jpg)
Toward the end of my time at the RCA I began experimenting with the BBC microcomputer, programming in BBC BASIC, whereby primitive shapes could be input and a series of operations output. The computer randomly selected the FormSynth operation to be used, and I would imagine the effect of the 3D transformation and draw the outcome by hand. This was a hybrid between a hand-drawn system and a computer-programmed system that I suppose would fall, albeit only just, into the category of “more than human.”
Then, in 1987, I went to IBM as a research fellow, gaining access to mainframe computers and support from world-class programmers and mathematicians. That was where the work I’m best known for really got underway. It was also where this question of “more than human” — or perhaps we could call it “human plus” — became central to my artistic practice.
Suddenly, I had a computer spawning images at a rate that would have been inconceivable by hand, yielding an abundance found in nature which an individual human mind could never emulate and a complexity in the mutations generated that was also beyond the human imagination.
At the same time, although the computer was producing images in a sort of simulation of Darwinian evolution, only those that fit with my very specific aesthetic made it past each stage. While this was a kind of “evolution by aesthetics,” it’s worth emphasizing that the positioning of the human remains central and essential to this category of “more than human.”
The best example of my collaboration with the computer — in addition to my human collaborator, Stephen Todd — can be seen by juxtaposing my early hand drawings with my recent evolutionary black-and-white Gothic series, whose level of detail and mirroring effects require “more-than-human” effort. This does not mean that humans are less busy at all, as Stephen can confirm. The work of rotating and refining the images sometimes feels endless; it requires one powerful PC plus two humans toiling.
RF: How does your distinct aesthetic resonate within the apparatus of VR (virtual reality)? The imagery that you deploy often has loose associations with bio-horror as a genre while VR as a regime of oversight suggests a particular hierarchy of guided experience. How do you approach the agency of VR as an apparatus and as a means of control over the spectator?
WL: VR has stunning potential as a creative medium because the viewer is put inside the artwork. In a sense, they are the artwork. There is this fundamental idea of the “artwork” and “the viewer/recipient,” who is often regarded at a distance. In VR they are fused together. In Mutator VR, the user’s body movements respawn the organic forms, which can be seen close up with absolute clarity. Serpent-like forms surround you and respond to your moves, which is a surreal experience that is by turns amusing, awe-inspiring, and downright strange.
The temptation with the medium of VR, given the level of realism that can be achieved, is to use distortions and tricks that play on human paranoia and phobias such as vertigo.
In Mutator VR, particularly the version we presented at the Centre Pompidou in 2021, I reference the sense of enclosure expressed in Magritte’s Listening Room paintings. Indeed, there is a point when viewers can find themselves trapped in a tiny mirrored cube. Many people enjoy this distortion of space, which has been described as the “Wonderland effect.” For others, it may trigger bad feelings of claustrophobia, which is obviously not our aim.
As an artist — rather than a producer of entertaining or horrifying experiences where participants know what they’re letting themselves in for — you need to take responsibility for your viewers’ reactions and shield them from having a bad experience. However, this is also very different from the role of a traditional artist, who might be looking for a particular reaction but cannot reasonably foresee every viewer’s reactions nor be held accountable for them. Our software is designed so that a viewer who becomes claustrophobic can easily back out of a tight space — choice is part of their participation in the creative process.
.jpg)
RF: Throughout your career you have often worked closely and explicitly with individual programmers or a programming team. Indeed, your collaboration with Stephen Todd goes back 35 years. How do you feel about artists becoming more open about acknowledging their human and nonhuman collaborators? What are the emerging hierarchies of collaboration in the context of human-machine co-creation?
WL: To me, collaboration has been key. I took my first steps working on the computer together with the programmer Mike King. Then, when I joined IBM, I worked initially with Peter Quarendon and then Stephen Todd — all great programmers. For an artist to be able to work with expert mathematicians and software engineers, with the cross-fertilization of ideas that it involves, opens up all sorts of possibilities. I have developed a special relationship with Stephen over the years, who as well as being a brilliant mathematician is also someone with a high level of artistic sensibility. He seems to understand my artistic quirks and aesthetic so well that he sometimes gives me what I’m looking for even before I’ve asked for it.
We have always regarded the computer as a creative partner. While I’ve often described myself as a gardener, selecting and culling the images generated by the computer, that is not intended to downplay the computer’s “creativity.” The images it generates take me beyond the limits of my imagination, so it really is a joint endeavor.
RF: Your historical evolution work of the late 1980s and ’90s is often cited alongside that of Richard Dawkins and Karl Sims. What was it like working during this period?
WL: When Richard Dawkins published his first academic paper on The Blind Watchmaker, I had already been working on my FormSynth drawings for a couple of years. These were enormous, painstakingly elaborated evolutionary trees of organic forms, and in the era before my work transitioned to computers, they looked incredibly detailed. One key aspect common to the work of Dawkins, Sims and my own was that there was a grammar that would be mutated.
Karl had a Lisp-style grammar; Dawkins had a branching-type grammar; and I had what I called my horn-based FormGrow Grammar. I got to know Dawkins when he visited the Arnolfini in Bristol to give a lecture during my exhibition “The Conquest of Form,” while I got to know Karl at the SIGGRAPH conferences, where he discussed his geometry and unique style of rendering. That was where I also met the Japanese artist Yoichiro Kawaguchi, who was making great art with a branching-type grammar though without a specific mutation interface.
Following its initial run at the Arnolfini, “The Conquest of Form” toured across the UK, including the Natural History Museum, and then to Germany, Japan, and Australia. As the tour continued, the types of venues switched from art to science, with often big attendee figures. However, this switch resulted in an increased dislike of my work from the art world (with a few exceptions, such as the art critic Martin Kemp).
In their minds I had crossed the art-science divide. Rehashing the readymade originated by Duchamp was fine but using a computer was not! The term “artificial intelligence” was first coined in 1956 but in the art world of the 1980s and ’90s it would have been a dirty word.
.jpg)
RF: How do you reflect on the histories of SIGGRAPH, which required supercomputers or mainframe computers and was largely rejected by the art world, and the recent developments in generative art that have found a market thanks to the NFT?
WL: In the 1980s and ’90s, the art world was very hostile to computer art. This experience was shared by all the computer pioneers, from Vera Molnar onwards.
The reason for that suspicion — one might almost call it a hatred — may be that the computer could produce millions of variations, which was seen as undermining the authenticity of the image.
Although events in the 1960s such as Experiments in Art and Technology (E.A.T.) along with Pop artists and Bell Lab programmers, not to mention the wonderful Jasia Reichardt’s exhibition “Cybernetic Serendipity” at the ICA showed huge potential for mixing computing with art, they were very much one-off events.
Given the lack of interest from the art world, the annual SIGGRAPH Art Show became the main forum for computer artists to meet and show work together in an atmosphere that was incredibly stimulating and supportive. SIGGRAPH’S golden age was probably from 1985 to 1990. Of course, it continued far beyond that and is still a major conference today, but from around 1990 it was no longer the sole arena for computer art. Other events such as the eclectic IMAGINA conference held in Monte Carlo and ISEA International, hosted in the Netherlands, were starting up and I presented my work at both.
Another challenge for artists was that, to do any interesting work using a computer, you needed to get access to mainframe or supercomputers to obtain the necessary rendering speeds. Karl Sims joined Thinking Machines, Kawaguchi joined Nippon Electronics, and I joined IBM. It was only in the 1990s that PCs emerged with the famous Silicon Graphics machines with modeling packages such as Wavefront.
RF: How do you view the recent growth in popularity of art produced with generative AI as well as the rise and fall of public interest in NFTs? How do you think your work sits alongside that currently being produced by artists working with generative AI?
WL: The NFT phenomenon is very interesting and I am starting to work in this space, coming at it very late and feeling some trepidation.
Given the dependency on having huge Twitter and Instagram followings, it feels a bit like a race with maybe not enough time for artistic reflection. However, the ability to register a digital artwork, ruling out unrestricted reproduction and unacknowledged derivation, is a game changer as it authenticates a digital image and makes it definitive.
The explosion of AI art is a fascinating development, representing a paradigm shift which I believe is as radical as when cubism took the art scene by storm. I applaud the democracy it entails whereby anyone can make art (provided they belong to the 66 percent of the global population with access to the internet). However, while some of the Midjourney images I’ve seen have been stunning, sometimes they feel a bit rushed or else that the artist has made do with the images selected. It’s almost as if the minting process kicks in too rapidly, without enough time spent on the earlier stage of selection. That said, it seems clear that artists as opposed to casual users often persist longer to get better results.
When I pick and breed images, the computer can spew out wonderful 3D forms that I could never have imagined myself. But generally the process doesn’t stop there. The mutated form gives me an idea of what I want, but I have to keep chipping away, refining and refining until I have that “eureka!” moment when I recognize I’ve got there. While the computer is my partner, I never let it have the final say artistically, at least not yet!
RF: During the height of ’90s cyberculture, you worked in the rave scene creating visuals with different bands and then in computer games for over a decade with a team of 70 people. How does that experience resonate with you today at a moment when digital artists of all kinds are finally getting their due?
WL: ’90s cyberculture has a lot of parallels with today. Back then, fractals had captured the wider public imagination in the same way that AI is now doing. There was a massive cultural drive in art, graphics, digital sampling, music, and books by authors such as William Gibson and Bruce Sterling. The World Wide Web had just been created and the huge growth and increased sophistication of video games fueled a creative explosion. Despite being cold-shouldered by the art world, I was at the center of it!
When I left IBM in 1993, I set up a small studio in an attic where we produced album covers and rave music videos. That was where I was approached by Warner Bros. to produce Organic Art for PC. This was a dynamic screensaver that generated in real time, continually creating new mutations for users to discover, which went on to sell millions of copies. Suddenly my images were everywhere in the cyberculture movement. I was working with the band The Shamen to produce stage visuals and album covers, while DJ Danny Rampling often featured my art at Shoom in London, where Acid house emerged.
My Mutator Art was a good fit with experiences around the new drug, Ecstasy, which meant that my artworks were being shown in large warehouse raves. There are parallels with the current preoccupation with “hallucinations,” which at the time tended to be described more as “psychedelic.” I’m not fond of either term, which I regard as ways of labeling certain images as “weird.”
The ’90s were a bit more anarchic, experimental, and rough and ready than today. No one was interested in selfie moments, which seem to be one of the main drivers for people to attend events and private views. There wasn’t the same VIP culture. Raves were held in warehouses, not swanky venues.
RF: The Mutator technology has also been adapted and extended for scientific visualization projects, covering real-time virus modeling, protein folding, and medical visualization work, in collaboration with Goldsmiths Professor Frederic Fol Leymarie and teams from multiple universities. Why?
WL: The Head of Bioinformatics at Imperial College London, Professor Michael Sternberg, knew my art well and told me that there was a similarity between protein Beta-sheet secondary structures and my horn structures. Our scientific collaboration started there, with Stephen and Peter Todd as well as Frederic Fol Leymarie becoming part of that conversation, which ultimately led to the interactive visualization platform called FoldSynth. We have since worked with other institutions such as the MRC Weatherall Institute of Molecular Medicine in Oxford and the Wellcome Trust. In a way, this was a homecoming for Mutator because my very first form, which was mutated at IBM in 1987, was a protein structure that Stephen Todd and I had adapted for art.
I am fascinated to learn about structural biology and applying ideas back into art, in particular the FormGrow grammar. My current black-and-white “Mutator Infinity” drawings are highly influenced by electron microscopy, while the 3D physics horn “whiplash effect” that I’m currently using is derived from protein modeling. Although there is a clear separation between the scientific visualization work I do and my art, one constantly feeds into the other. My recent “Virtual Virus” collaboration with York University built on the fact that many of my FormGrow forms share with virus structures the classic features of a capsid shell with spikes as well as coiled elements that are similar to RNA and DNA.
I think that this project kept me sane during lockdown. In the middle of the epidemic I was evolving aesthetically pleasing viruses!
RF: What are you trying do with the Creative Machine and where will it go next?
WL: The Creative Machine Symposium is a biannual event that brings together AI and generative artists, curators, scientists, and computer engineers. It has grown beyond expectations from its humble beginnings in 2014 when my colleague, the mathematician Frederic Leymarie, and I put on an exhibition of works by 25 digital artists in Goldsmiths’ St. James’ Hatcham Church exhibition space accompanied by a half-day conference.
The Creative Machine 2014 was one of the first events to engage seriously with AI in terms of both theory and artistic practice. What differentiates Creative Machine is that we mix artists and scientists with a spread from veterans and highly experienced creatives through to bright young individuals with novel work they want to show. Over the years, the artists exhibited have used a range of novel technological approaches, including machine learning, cellular growth simulation, fuzzy logic, organic structure mutation, and automated aesthetic selection to lead them into new, uncharted creative domains, all of it accompanied by lively discussions. This year, Creative Machine is off on a big adventure, traveling abroad for the first time to be hosted in China in November. We’ll be announcing the museum venue and dates in the coming weeks.
With thanks to Miriam Frendo.
William Latham is well-known as an early pioneer of generative art through his Mutator Evolutionary Art, created at IBM in the late 1980s and early ’90s. His work, which exhibits strange organic, often serpentine, forms, is produced using his “alternative evolutionary software system,” developed with Stephen Todd and team, which Latham uses to pick and breed 3D forms freed from the limits of the human imagination. His work was shown widely in museums and touring shows in the UK, Germany, Australia, and Japan at that time. Latham’s work is focused on using evolutionary processes to create art centered on the idea of the “artist as gardener,” originating from his time as Henry Moore Scholar at the Royal College of Art, where he created his FormSynth drawings and etchings — a blueprint for his later Mutator work. On leaving IBM in the mid-nineties Latham set up a studio in Soho, London and moved into Rave Music where his organic art had built up a following in the emerging Rave and Cyberculture scene. After three years, he and his team of around 70 then began developing games with Warner Bros. and Universal Studios. In 2002, they created the hit game, The Thing, based on the John Carpenter movie.
In 2007, he moved away from entertainment, becoming Professor in Computer Art at Goldsmiths, University of London, where he restarted his creative collaboration with IBM mathematician and programmer Stephen Todd. They resurrected and extended their old Mutator code and pushed the technology into VR, creating highly novel organic immersive experiences for the public. This VR work has been shown in many touring exhibitions in China, Japan, Peru, Belgium, and the UK. Mutator VR uses Latham’s core evolutionary ideas with influence from Rave and interactivity from games. His recent “Fantasy Virus Mutator” and “Infinity Mirror” series are influenced by electron microscopy as well as the engravings of Albrecht Dürer. Latham’s work is in the permanent art collections of the Centre Pompidou, Paris; The V&A and Calouste Gulbenkian Foundation, London; and the Henry Moore Institute, Leeds.
Rachel Falconer is an independent digital art curator, academic, and founder of the experimental curatorial collective Mutable Prototype Syndicate. She currently holds the position of Head of Digital Arts Computing BSc and is a lecturer across undergraduate and postgraduate studies in Computational Arts at Goldsmiths, University of London. As an independent curator, researcher and writer she operates at the critical intersection of contemporary art practice, emergent technology ecosystems, feminist technoscience, and networked culture. Drawing on her experience in the film, advertising, and gaming industries to inform her systems-based practice, Falconer has conceived and delivered multi-scale interdisciplinary curatorial platforms and public programming, alternative exhibition models, residencies, and innovative public research platforms in an international context spanning a diversity of critical modes of engagement with technology. She is regularly invited to speak at public events and has participated in public programs at institutions including Transmediale, Berlin; Tate; Barbican Centre; ICA; V&A; Somerset House Studios; The Photographers Gallery; The Lumen Prize; Arebyte; Gazelli Art House; and Whitechapel Gallery, London; Rhizome; the New Museum, New York; and SONAR +D, Barcelona. Her writing and research has been published across a wide range of platforms including: Routledge, Sternberg Press, BOMB Magazine, Dazed, Frieze, The Guardian, the British Journal of Photography, and The White Review.
