Waiting To Be Signed · interviews on generative art, on-chain
← Index
Interview // JUN 2023

Andreas Gysin

Title: Digital. Mechanical. Kinetic. Materialized
Role: Generative artist
Platform: Verse
Duration: 1h 9m
Hosts: Will & Trinity
Listen on Spotify Guest on X Guest website Download MP3
#037 · Digital. Mechanical. Kinetic. Materialized
Self-hosted audio // press play
1h 9m
MP3 ↓

Spot an error? Highlight the words or lines you want to fix — an “edit” button appears, and the panel opens with your selection ready to edit.

Will: All right, hello and welcome everyone to another episode of Waiting to Be Signed, a special interview episode. We're joined today by Andreas Gysin, who is with us in advance of an exhibit coming up, the first Verse solo exhibit, called Materialized. We're really excited to jump into that with him and talk all about it. Of course, Trinity here as well. Before we get into it, let's all just say hi. How's it going, everyone?

Trinity: I'm doing very well. Thank you for that intro, Will. I can't wait to talk to Andreas.

Andreas Gysin: Same here.

Trinity: Andreas, welcome as well. We're so excited to have you here. Oh my gosh.

Andreas Gysin: Hi. Hello both, and hello to everyone.

Will: I'm also super happy to be here — it's a real treat to have you on. We first encountered your work through fx(hash), which I think most of our listeners will know you from, but you've got a much longer history in your art practice that we'd love to get into. And of course, we'll talk about the Verse work you're releasing soon too. But before we get into all of that, for anyone who might not be familiar with you, can you give us a bit of your background in art and coding as a way of introduction?

Andreas Gysin: Sure. I don't originate from art and coding — I originate more from graphic design. While studying and working on graphic design, I discovered the computer and code, and I realized there was great potential in using this kind of tool for the job, and to use code to generate images I wouldn't be able to generate otherwise. So that's the background. The art part came much later.

Will: How did you transition into what we generally know as generative art? Your work on fx(hash) is long-form generative, but you've done a lot of projects prior to that using things like signage donated to you, or bought from places like train stations decommissioning their analog signs. So what bridged you from the graphic design world into making these installation pieces and this more conceptual work?

Andreas Gysin: I'll go back to graphic design quickly, because there was a series of posters commissioned to me and my friend Sidi. We were working together — we don't have a studio, but we work together, and we'll come back to him later. We got commissioned a series of four posters for an electronic music festival, and we could do whatever we wanted. Since it was electronic and experimental music, we thought this kind of poster needed an electronic treatment too. That was the first time we used a generative system — a piece of code that would generate the images for this series of posters. We built custom software and generated a lot of variations, a lot of possibilities. So this long-form idea was already present in that approach, even if I'd now call it a parametric program, where you change parameters and the image follows.

That series of posters was the beginning of this generative, image-making-through-code approach. A few other projects followed, all in print. These posters or images were made for printing, and we considered the printing techniques — offset printing, for example, where you have four channels, or silkscreen printing, where you have only one channel. So the notion of the final printing technique was embedded in the thinking behind these pieces.

At a certain point we transitioned, quite naturally, into projects without a commission or an outside brief — no client asking us to explore these techniques or experiments. We started doing our own experiments, our own research. That's an interesting aspect that still holds for me today: the research that comes from graphic design projects with a client, a specific audience, a specific context, very often shifts and becomes a new starting point for an art piece — and also the other way around. A lot of research from artistic, personal context flows back into commissioned work in a different domain. So I don't think there's really a distinction between these two fields. Of course, a commissioned work has certain constraints you need to follow, but the research in both fields nourishes each other, and the methodology of the work is basically the same. Internally, we don't see a big distinction in how we approach these projects.

Trinity: I love these early influences, especially that origin story through graphic design — we can really see that in a lot of your physical work. Since you've already started talking about Sidi Vanetti, maybe we can briefly talk about that collaboration. It would be wonderful to know how it began — you've been longtime collaborators, right?

Andreas Gysin: Yes. I met Sidi — we were both studying graphic design at a school with very few students, less than 20, maybe even less than 15. For logistical reasons, the class was split into two classrooms, left and right. I was on the left side, he was on the right. For the three years we studied together, we didn't really communicate much. He was also far more advanced — the more advanced students were on his side, and on the left we were the beginners. But we started collaborating on the same topic for the final project. We had the same interest, so we started exchanging thoughts around it, and that was our first collaboration.

Then we started working almost by chance at a movie theater — at night, at the entrance, giving away tickets and popcorn. Then we worked for the same graphic design studio as collaborators, and started working together on projects, and kind of never stopped. I managed to move away from Switzerland several times in my life, but I always came back, and even when I was away we never really stopped working together. We still don't have a studio or a legal entity keeping us together on paper — we just work and get things done.

Trinity: When you're doing collaborative work, what is that process like? What does each of you bring to the table, and how do you navigate that creative process when making new work?

Will: In particular, what happens when there's a disagreement? How do you resolve it?

Andreas Gysin: Disagreement is always interesting, and we always manage to resolve it. The simpler part of the question is that we have certain roles. I'm the only one who codes, so the software part is usually on me. That doesn't mean Sidi doesn't understand code — he understands it very well, he understands the mechanism, and when I talk to him about certain aspects of a program he knows exactly what I mean. He just doesn't write the code himself. He, on the other hand, comes from print — all the parts that require printing knowledge are his. He does a lot of book layout, poster design, and exhibition design, so he has huge expertise in that pure graphic design work, and that knowledge often comes into our projects.

Will: Building on that — what's the difference between something you make solo, under your own name, versus something that comes out as a collaboration? The projects we know you for, like Towers and Device_1 from fx(hash) — were those entirely solo? Or are they collaborations in some way that people just don't know about?

Towers — ertdfgcvb

Andreas Gysin: No, those are solo pieces. Sidi is completely unaware of crypto art and NFTs — he's not on any social media, he lives in a very small village of 300 people with his family and his dog. That doesn't mean he's an outsider — he knows how everything works, and quite well. But these more ephemeral aspects don't interest him, and he's not very aware of them. I explained a few things to him when I first heard about the crypto art movement, the NFT scene. I started publishing things online, dropping a few pieces as an experiment — I was curious. I'd seen these things happening a few years ago and wanted to try. Before that I had a website where I published interactive or generative works that would run in a browser, just experiments I could throw online. So it felt natural to continue that kind of experimentation in this new context, with the difference that eventually some pieces could be sold. That was a real novelty for me, and for many of us. I just continued that attitude on my own. At a certain point I told Sidi, "Hey, look, I'm doing this thing." We have several things going on that aren't part of a collaboration, even professional projects — we have our own lives.

Will: Yeah.

Andreas Gysin: Our own clients, our own artistic projects. He has his own artistic collaborations too. So we're kind of independent in a certain way.

Trinity: Let's talk more about NFTs. You've described discovering the scene and releasing things, but when did it hit you that the blockchain was a legitimate way to release, distribute, and sell your work?

Andreas Gysin: I saw things happening, and for me it's really just a channel to reach an audience, a mechanism to sell and deliver pieces and sign them digitally — that's where my interest in the blockchain ends, to be honest. It's a fascinating piece of technology, medium complexity I'd say — there are pieces of software that are much more interesting and complex that would also merit this kind of discussion and attention, and unfortunately they don't receive it. For me it's really just a distribution mechanism. I'm rarely interested in the other things that float around it.

To be honest, there's an aspect I don't find interesting at all: unfortunately, this way of distributing software or images or clips via the blockchain also ties into a much broader scene of crypto money, and I really don't like that speculative aspect. The NFT part, the crypto art part, is just a small piece of it. Older digital artists were waiting for some mechanism that would let us sell or monetize a piece of software, or something completely intangible — and then it happened through this mechanism. I think most of us weren't so happy that it came bundled with this huge weight you have to drag along if you want to sell a piece of software.

Towers — ertdfgcvb

Will: I think we agree — it's a topic we explore a lot on the show, and we wrestle with our own distaste for the broader crypto community while enjoying collecting art, learning about art, and being part of this generative art scene on the blockchain. And you're right, NFTs have been the thing that catalyzed artists' ability to make a living off this work in a much more accessible way than ever. Maybe that brings some bad things with it, but also a lot of good. Before we started recording, you mentioned you'd just been to a conference with a bunch of older-generation generative artists — I imagine a lot of them haven't embraced this technology, and some are pretty against it. So what pushed you to give it a shot? So many people say, "I'll never do crypto, I'll never do NFTs, it's all a scam, bad for this, bad for that." What convinced you personally to give it a try?

Andreas Gysin: I didn't have a strong opinion. I just wanted to try it, and I saw it happening around me. When I saw a few people I admired at the time getting into it, I thought, okay, maybe I follow these people I respect and see what happens with them. The promise of huge gains never really materialized for me, but it still felt like something worth trying. Before this, we had limited ways to distribute this sort of work — software artists especially were trying to find outlets, whether screensavers or apps you could buy from an app store. Then this appeared, and I told myself, maybe it's time to give it a shot.

Trinity: And it looks like it's worked out pretty well, financialization and the crypto market aside. The NFT scene and the entire crypto economy do bring a lot more eyeballs and interest. Will and I were sideline art appreciators who'd go to museums on occasion before we really found out what fx(hash) is or does. Especially since COVID, people are appreciating things more digitally and virtually, in online communities rather than real-life ones. It's really changed who collects art, who appreciates art, and who has the ability to buy and sell it. How has this impacted you, beyond just having a new audience — and how do you think it's impacted the legacy or traditional art world?

Andreas Gysin: I think one of the most important aspects is that regional limitations were completely zeroed out. There's no limit anymore. Being able to reach a global audience is incredible — not only for me, but for all the artists I met in this space. Artists who were under the rock, or simply invisible even to me, people I wouldn't even know existed. This thing brought us together, and there's so much happening now. That's the total, absolute upside of this whole phenomenon — I think it's even more important than the financial aspect.

Trinity: Agreed. The financialization is a bonus.

Andreas Gysin: A total bonus, of course.

Towers — ertdfgcvb

Trinity: But ultimately, a world with more art appreciators and more awareness is always going to be a better world. I don't have any doubt about that.

Andreas Gysin: If you believe art is able to change the world, then yes.

Will: I want to ask a question on this same topic, but starting to diverge into your artistic practice. In prep for this interview, we watched a more recent talk you gave — I forget which conference — where you showed a lot of your work with signs. Near the end, you showed some of your code-only work, which looked like an early prototype for Towers before it was eventually released. You talked about working within the constraints of a given technology and letting those constraints force you into a more creative mode. Trinity and I play Magic: The Gathering, and the longtime head designer has this saying: restriction breeds creativity — you should look at a restraint as an opportunity, not a barrier. So how do you connect your work with these analog objects, where there's a lot of built-in constraint, to your code-based work? What constraints lead you to be more creative there? What boxes do you put yourself in, working with something like JavaScript or within the constraints of NFTs generally? Where do you find the creativity in that space?

Towers — ertdfgcvb

Andreas Gysin: This is a very important question for the way I work, both alone and with Sidi — this aspect of constraints. There's already a problem when I work on software-only projects based on screens, or inside a browser, because there you have to invent a sort of constraint yourself to force creativity.

But let's start with the physical work, where it's simpler: the objects themselves are limited, and the limit is built into the object. There aren't many things they can do, but the few things they're good at are usually what they were designed to do. Signage systems especially — electronic, electromechanical, or even purely physical ones — are designed to communicate a certain message in the most effective way possible. That's one of the things we try to harvest when working with these objects: we use their powerful existing language and try to find a project that pushes it, drives it to the maximum. We try to work with the constraints, not against them.

There's another attitude we bring to found objects: staying as conservative as possible. When we find a display or signage system, we try not to modify or alter it, not to add features that weren't there originally. Sometimes we slightly alter the software so it behaves a little differently; sometimes it's enough just to position it differently, or put it near something else, and the object starts to express itself in a powerful way. That's more or less the broader way we work with these displays.

Once you understand this way of thinking, you can see what happens when you suddenly work with a computer screen that has such high resolution there's no grid — you can have whatever color you want, display whatever information you want. Reasoning like that, you'd almost say the computer screen has no constraints, because whatever you imagine, you can display. That's why in some of my projects, I've put myself in the position of building artificial constraints. It's why I started mostly with text art projects when I began with NFTs — I told myself, I need some limit, some border, something that doesn't let me put whatever I want on the screen. So I said, okay, I'm just going to use text, and see if I can find the boundaries of this constraint system — how deep, how wide, how far I can move inside this artificial limitation.

Trinity: We definitely see that in both your work on fx(hash) and elsewhere. Towers comes to mind as one working within those constraints, but you also have that wonderful project — I think there are a couple of variations — where the code on the screen looks like a simple text editor, but the text is always transforming itself, undulating up and down and across the screen in beautiful ways. How do you come up with those constructs? What do you find inspiring about the way the text moves, and how do these ideas come to be?

Towers — ertdfgcvb

Andreas Gysin: For all these text-based projects, I need to give a quick flashback. Sidi and I started the text project when we were commissioned by an electronic festival to come up with a poster and a graphic system, several years ago. We designed a few prototypes and thought, this could work as something similar to ASCII art — so let's quickly build an engine that lets us experiment with it, and prepare a few prototypes for the poster and graphical systems to use at the festival. The project got rejected, but the seed of some of the projects that came after was already there.

We kept exploring this language on our own, but Sidi gets bored of an idea very quickly, so he stopped. I'm a bit more simple-minded, so I stayed continuously fascinated by this idea of text mode, ASCII art. It's an interesting limitation, and an interesting language too — there's a long history of ASCII art, text art, typewriter art, so I started studying the history of these graphical systems.

At a certain point, during COVID lockdown, I told myself: these ASCII projects almost always use the same kind of rendering engine — what changes is the algorithm itself. So let's build a playground where other people can write their own code, publish it on the platform, and distribute it, so we can remix each other's ideas and sketches. I built this thing with a preview on the left side and a code editor on the right, where you write code and it updates in real time as you type — if the line you're writing is correct, the image updates immediately on the left. So you always have the code and the result together.

I think the idea for the projects you mentioned came out of working on this. Towers, too — the idea came while I was working on this playground, actually because of a mistake. Instead of aligning all the text to the left, I made an error in a CSS declaration and centered it instead. All the lines of text formed this sort of tower, and the idea was born. So you could say Towers was a mistake.

Towers — ertdfgcvb

Will: Or a happy accident. As we often hear from artists in this space, bugs in code tend to produce some of the most delightful discoveries. Great story.

Andreas Gysin: Just to complete this part — the code that displays itself is a known concept in the more experimental side of programming, called a quine, if I'm pronouncing it correctly. A quine is a piece of code that displays itself. In JavaScript, the language of the browser, this isn't so difficult to achieve, because JavaScript allows something called introspection — the code can access itself. Not all languages allow this; writing a quine in a different language can become quite complex, but in JavaScript it's really simple. I always felt I was cheating a little bit comparing this project to a proper quine.

Will: Very cool.

Trinity: Amazing. How do you feel about fx(hash) and Tezos? That's something that's been on our minds, since we got into this world through fx(hash) ourselves. You've released a lot of this ASCII art on numerous platforms too, like Super Rare and Foundation, which are ETH-based. How did you come across the Tezos ecosystem and fx(hash) specifically, and what made you want to release there rather than, say, standing in line for Art Blocks?

Andreas Gysin: I started on these ETH platforms because those were the ones I could find, and some of them didn't ask any questions -- you could just create your wallet and start releasing pieces. There was no gatekeeping. But these platforms were optimized for images, movies, and GIFs, and I wanted to start with text, so I released a few ASCII projects there. I always created perfect loops. I already had this idea that a piece shouldn't really have a beginning or an end -- it should be something that continues to repeat itself, even if the loop is technically a GIF or a video that has to repeat at a certain point. The idea was this flowing thing.

Then Hic Et Nunc came along, which was even wilder -- they didn't ask any questions at all, and they allowed you to publish any kind of media: not just images or GIFs, but PDFs, text files, pieces of code. That opened up an interesting playground, because you could publish real-time pieces of software. I used the platform to publish SVG pieces -- scalable vector graphics you can display in the browser that also allow execution of JavaScript if embedded correctly. So at that point we were able to publish real-time work.

Towers — ertdfgcvb

fx(hash) came a bit later, and it was a great surprise -- well done, well documented, no questions asked, you could just release your piece. The fact that there was no curation, that the barrier was lowered, was a real invitation to participate.

And why not Art Blocks? I was contacted a few times about doing something there, but I couldn't really understand what was going on with the curation and some other aspects -- it wasn't clear to me. I looked at a few pieces on the platform and told myself they weren't really interesting. I was very ignorant -- I saw Chromie Squiggle and thought, this is too basic to be an interesting project. So I said, maybe I won't consider this platform for the moment. Probably a big mistake.

Will: The flip side now is that you're releasing in a curated fashion to some extent, with Verse. Around the time this episode drops, the auction or whatever mechanism you'll use to sell work in this Materialized solo exhibition will be ongoing. Can you give us the story of how Verse got in touch with you, and what convinced you to work with them this way? I imagine there was some level of curation and collaboration, where you had to put your trust in them to a degree.

Andreas Gysin: Layla from Verse saw a presentation I gave in Paris last year, at an event called NFT In. It ran from morning to evening -- short 15-minute presentations, one after another, with total focus on artists. There were a few more critical talks and roundtables, but mostly it was just artist after artist, each talking about a single project. I gave a small presentation. It was a really interesting moment, held in a beautiful old theater in the center of Paris -- we all had a great time.

Toward the end of the day, I presented, and Layla saw it. Afterward, she ran up to me and said, "You have to come to London, I want to do a show with you." I'd also presented several physical projects I'd realized with Sidi, going back a bit to our origins, not just NFT projects. So I said, if I come to London, I'll come with Sidi and we'll bring these physical projects too. I wasn't even fully aware of what Verse was about, but I trusted Layla completely -- she was convincing, she was great. So we had this agreement, and worked it out over the following months.

Layla put us in a tough spot almost immediately: she said, "We have this room for you," and sent us the floor plan -- and the room was huge. We'd never done an exhibition in such a big space before, and it was already crazy to imagine pulling it together in short order. Would we really fill this whole thing? She eventually found a smaller room, but it was still very big, and difficult to solve for with our projects.

Chromie Squiggle — Snowfro

So we worked out what we could bring. She insisted there be a long-form project of some sort, but we insisted that we came from the physical world and wanted to push that too. Especially for a space like this, we wanted to bring a bigger piece -- and for us, the centerpiece of the whole exhibition is a piece we've remixed over the years, always transforming. It's called Digits, a huge electromechanical piece that occupies almost the entire diagonal of the room. We even had to shorten it to make it fit properly. This will be the third iteration of the project, in a new form.

Trinity: Looking through the description of the exhibition, it looks like you're bringing six pieces, mostly physical, and then a seventh -- the long-form piece -- that will be sold as an NFT on Verse.

Andreas Gysin: There are also two physical pieces that come with a digital counterpart. This is a topic I've been thinking about for a while now, about a year and a half. These pieces exist as both physical and digital. That's not new in this space -- a lot of artists produce a digital piece that then gets printed, if it's a static image. Some pieces are print-first: the physical object exists before the NFT, and the NFT is almost just a way of selling it -- maybe the NFT display is a scan or image of the object, acting more or less as a contract.

I didn't want either of those solutions. I wanted something where the physical exists and the digital exists, and they can also live together. For the two pieces we're showing here, we constrained ourselves to a display made of two LED matrices connected together, giving a total surface of 64 by 64 pixels. If you look at 64 by 64 pixels on a computer display, or even a newer phone, that's probably a few millimeters wide. But turn those pixels into actual LEDs, and the surface becomes quite large -- it starts to have a presence in space, starts to exist as an object.

64 Pixels — Andreas Gysin & Sidi Vanetti

So this 64-by-64-pixel grid -- a bit more than 4,000 pixels total -- became another artificial playground we built for ourselves. We thought, let's imagine projects that work inside this ultra-reduced space of possibilities. That's how the physical pieces were born. The digital counterpart is a reflection of those same ideas. What's important to understand about the digital part is that the concepts behind this display are simple -- just a small story we want to tell. And that story can also be transported onto a different container -- not an LED matrix, but a computer screen.

Trinity: What's the process for creating something like this, where on one hand you're coding and designing for an LED display, but then recreating that representation digitally? Is it the same codebase running in different ways, or something separate entirely? Theoretically each pixel has to scale to fit the constraints of JavaScript and any responsive device -- I'm really curious about that.

Andreas Gysin: I'm so glad you asked, because there's a whole process behind this. We designed the hardware for the physical part too -- though not the LED panels themselves. Those are commercial, industrial components, and they're beautiful precisely because they lack any decoration or adornment. Just this industrial piece, naked, dry as hell -- and it already looks so good in that dryness, in just being that essential thing. We tried to keep everything around it as clean as possible. We built a custom aluminum backplate, folded and machined, which I'm quite proud of. But we also designed the hardware controller -- the LED panel doesn't run itself, it needs something to control it. Designing that was a way of signing the hardware, saying: this is a complete object, not just something you buy and then program.

We usually don't build hardware ourselves -- we work with engineers, and I gave this engineer some extreme constraints. That idea of limiting possibilities gets transferred to the people who have the misfortune of working with us. I asked him to make the controller as small as possible, and he nearly went crazy, but now we have this tiny square you can plug into the matrix and it lights up.

The code running on this hardware is C -- an old, fairly limited, low-level language with few instructions and a handful of types you have to work within. Translating that to JavaScript is pretty trivial; it would be much harder the other way around. There are even mechanisms for automatic transcription from C to JavaScript, and I considered using one, since it felt like the conceptually "correct" thing to do -- write the algorithm once, translate it automatically, no extra parameters or choices introduced in the process, guaranteeing the JavaScript is exactly what you wrote in C. But since these ideas are pretty small -- the algorithm is a small part of the whole setup once everything else is in place -- I just translated it by hand. It's really just a bunch of arrays and a few pointers, not the complex part.

I think the more interesting part of your question is: how do you treat a physical LED pixel, a few millimeters wide, on a computer screen? That bugged us for a long time, because it felt impure to represent this idea digitally. Do you draw small squares and call each one a pixel -- even though on a screen, a square is already made of several pixels? Do you imitate the roundness of the LED and draw small circles?

64 Pixels — Andreas Gysin & Sidi Vanetti

Our first answer was to just use physical pixels for the digital version. That would mean looking at the project and barely seeing anything -- just a few dots moving in the center of the screen, too small to recognize any composition. So I wrote a renderer that takes the pixel data and displays it on screen in a more arbitrary way. But I also built in -- maybe I shouldn't admit this -- an Easter egg that lets you display the pixels as 1x1 pixel data, so the on-screen version is, in fact, using only 64 physical screen pixels, even though it makes no practical sense.

Trinity: So if I have a viewport that's, say, 2000 by 2000, are you subdividing that and just lighting up a 64th of that pixel space on the screen?

Andreas Gysin: Yes. The piece is completely responsive, so it will run in whatever area is available of the browser, and it also runs as a web app. I introduced a few scaling factors, but at this point it almost becomes a kind of preview image. Sidi and I talked about these agreements at length, and after a very long discussion we agreed it would be okay to do a representation of the pixel on screen. So we represent a pixel by drawing rectangles, which isn't really a "proper" idea somehow, but we agreed to it because the algorithms are the interesting part for us. The stories these algorithms tell — and there are two algorithms across these two projects, 64 Pixels and Recursive Tiles — can also be told on a medium other than the physical piece they were originally made for.

Recursive Tiles — Andreas Gysin & Sidi Vanetti

Trinity: Conversely, if I bought 64 Pixels and owned it on the blockchain, but I'm handy with a soldering iron and have some LEDs lying around, is there a way I could pull the piece from the blockchain and recreate it physically in a 64-by-64 LED space? Or should I just email you on the side and ask how to do it?

Will: Or do we get physicals? As part of the sale, does the physical come with the digital? Is it both?

Andreas Gysin: We're still discussing this — maybe once the piece airs, the choices will have been made. The number of editions for the physical piece is extremely limited. We didn't want to produce a huge series of these things we built, but the software part could be much larger. So we're still thinking about whether there's a larger tier of just the software piece that you could own without access to the physical one — enough to enjoy the algorithm, try to understand it, or just watch it. For some people that might be good enough, if they don't want extra stuff in their homes. But the physical is the physical — it's what legitimizes the idea of these constrained algorithms. So yes, they are probably separate objects. As things stand now, they're still separate.

Will: If anyone's interested in other physical pieces from you, there's the LCD one you released at the end of last year — a slightly different take on a device that displays ASCII art. The floor is about 2.3 ETH, so go check those out. I missed out — I was looking at those when you released them and thought they were really cool. But let's talk about the long-form piece, which is probably going to be the most accessible piece in the exhibition. We don't know much about it beyond the description that's up there, and even in the notes you sent, you said the name is still a work in progress. We're two weeks out — is there a name for it yet? What can you tell us about this piece in general? The name we have written down right now is 256.

Andreas Gysin: 256 is right. The piece we just talked about is 64 Pixels, the other hardware piece is Recursive Tiles, and among the other physical pieces, one is called Digits, and another, which uses red-green LEDs, is called Red Green. We usually give titles that aren't really titles — they don't add any extra layer of information about the piece. The most obvious thing you could say about a piece usually becomes its title when Sidi and I work together. There's a certain honesty in that, and also a wish not to add extra complexity, extra layers of meaning for people to project onto the piece. We try to leave a little space around it rather than pollute it with meaning through the title.

64 Pixels — Andreas Gysin & Sidi Vanetti

In this case, we found ourselves facing an unconstrained screen space. We didn't know what to do with the screen, because you can do so many things — where do you even start? One topic we wanted to explore was gradients. It's less true of current screens, but in the usual model, you have 8 bits per color channel. In RGB, that's 8 bits, so 256 possibilities for red, green, and blue, and the same for grayscale values. So an average screen — again, less true with newer displays — can render a gradient of 256 steps. We told ourselves this would be our entry point, the constraint that would let us reason our way into something for this limitless surface.

It happened that a few years ago, maybe four years back, we did a design research project for a science fiction movie. The producers came to us to imagine signage systems for the future of society — some decades ahead, where part of society had slightly collapsed, not in a Mad Max way, not fully dystopian, but more complex than ours, more problems, a few robots and aliens around. That society needed a wayfinding system, a signage system that talks to machines, to humans, to extraterrestrial entities. It was a dream job as a designer — you can go completely crazy, imagine things that don't even need to function properly. We weren't preparing material that would end up in the actual movie, just ideas and imagery that the film's artists might later pick up. It was a pure research project.

Among the many things we imagined, we came up with the idea of taking a shape, an original message, and pushing it through a series of 256 grayscale layers, changing its shape as it passed through. The images we got were strange, yet somehow familiar — if you work with 3D imagery and look at a depth buffer, that's roughly how it would look. A black-and-white image, a bit like a CT scan of the brain, where things closer to you are lighter and things further back are darker.

We'd only explored this quickly for that short project — we were always in a hurry — and we wanted to pick it up again. I mention this because it's a case where a commissioned project flows into a freer, more artistic practice. That's what we're working on now. I built a renderer that works quite well in real time and lets you slightly modify the point of view. I think it's an interesting piece because, for me, it's not really a 3D piece — I very rarely work in 3D — it's completely two-dimensional, but the final result is this sort of stack. So I like to say it's 2.5-dimensional: not really 3D, not really 2D, something in between, with its own limits. Right now we're still exploring, with 3D, all the shapes we can model through this pipeline, and we're pretty happy with where it's at.

Of course, in a long-form project, if you remove color as a way of creating variation, you limit yourself a lot. We decided to go grayscale only, which complicates things. It's easy in long-form work if one of your parameters is a color palette — you draw a circle, change the palette, and you already have very different-looking outputs. Remove color, and you have to do all the work through form alone. And the last thing about this project — like all the projects in this exhibition, and basically all our work — is that it's in motion. These are moving pieces, not images.

Trinity: This is one of the pieces we haven't seen any previews of at all. Is it purely digital, or does it also have a physical component, or the potential for one?

64 Pixels — Andreas Gysin & Sidi Vanetti

Andreas Gysin: To contradict what I just said — in this case, we are presenting print output in the exhibition.

Trinity: Wow.

Andreas Gysin: Whoa.

Will: So are you curating stills that you enjoy out of the various algorithms? Since it's a 3D piece that moves, is that the approach?

Andreas Gysin: It moves — it's a continuously transforming shape — and you select one moment in that transformation and print it. Since it's black and white, Sidi's almost unlimited knowledge of different printing techniques already comes into play. But the print component we're using for the exhibition isn't really part of the project itself. We'll find a good solution, but probably not one of the more elaborate things he has in mind.

Will: You alluded to this a bit at the end of your description — that it might be difficult to produce a lot of outputs given it's a grayscale piece. Do you and Sidi have a number in mind? Obviously the most harmonious number would be 256.

64 Pixels — Andreas Gysin & Sidi Vanetti

Andreas Gysin: That's what we're thinking too, yes.

Will: And pricing — auction, flat price? Do you have a plan for how to release it?

Andreas Gysin: I try to leave those aspects, including the edition number, to them — they know what they're doing, and honestly it's not something I like to think about much, so I can concentrate on the piece itself. Of course Sidi and I need to have a say on the final number, because it depends on the quality and the range of variation the algorithm is capable of producing. If it's very limited, it doesn't make sense to make more than ten. But we're aiming for 256 — we're not quite there yet. We still have a few days to fine-tune it, and a few new ideas to experiment with.

Trinity: The entire thing sounds very exciting. Will, was there anything else you wanted to ask about 256?

Will: Yes, a little. Continuing on motion — a lot of your work has it, and a lot of your physical pieces also have a sound and rhythm component, coming from these analog technologies. Related to your digital work, and Device 1 in particular: did you ever consider adding an audio component? Watching the digits flow, there's so much rhythm to the way they move across the screen that you can almost hear it, if you're familiar with the sound of that type of technology. But the piece itself — unless there's an Easter egg I'm not aware of — has no audio. Is that something you've considered? Does audio add too much complexity, or is it just never satisfying in the way you'd want? It seems like such a natural extension.

Device 1 — ertdfgcvb

Andreas Gysin: It's very difficult for me, at least, to work with audio, to be honest. One of the first pieces I did with this split-flap idea, more or less, that you mentioned—these characters that don't just change to the proper one, but have to go through the whole list before they stop. This idea, which comes from my fascination with split-flap displays that we've used several times in our work, gives this whole movement and rhythm to Device 1 and most of my ASCII art pieces.

There's a precedent, though. This piece is about ten years old. I published it with a musician when I was living in Berlin—an electronic musician who contacted me wanting a design for his album cover. I designed the cover and a system of characters for it, and that system became an animated, real-time generative movie clip for one of his tracks. It worked well because the way he made music was through a generative system too—even though at a certain point he'd capture a result, freeze that endless composition, choose one version, and put it on vinyl. But the process of building the tracks involved a lot of randomization, a lot of systems. So I wanted to build a video clip for him that would also be driven by randomized factors. I prepared a way of placing cue points at certain moments in the track, and something slight would happen in the real-time clip at each one—not by analyzing the sound, but through those cues.

So I'm not so new to this kind of project with audio and video. I've made several for musicians, and I consider almost all of them a complete failure—except that one. I'm really proud of it, and I think it kickstarted most of my subsequent ASCII work. I still get contacted by musicians who want to do things with this kind of imagery, and I almost never have time. I once sent a few video loops to a DJ in LA I didn't know. Six months later he sent me back footage of himself performing in front of, I don't know, 200,000 people, with this ASCII art as the backdrop. I was blown away—okay, so this is what these loops were for. Somehow the rhythmic aspects of the project find their way back to music, but I've never managed a good combination of the two by myself.

I know Leander Herzog just released a piece with generative imagery and generative sound, and I deeply respect that—though the generative sound part was made by a musician. It's a very interesting piece, one to watch, and probably something without much commercial success, but that shouldn't stop an artist from doing it. Strangely enough, the most commercial pieces are static, and I really cannot understand that. This medium—the computer screen, at least—demands movement, not a still image. We've had still images for twenty thousand years already.

Will: That was the core of the episode we did with Leander. We had him on last year, and half of it is spent on that very observation. You should listen to that one.

Andreas Gysin: Ah, okay, I have to listen to it then.

Device 1 — ertdfgcvb

Trinity: Speaking of Lenny, his most recent piece, DOM1, uses generative sound working in conjunction with the visuals—we'll definitely be talking about it when we record our weekly episode. When I think about the work you've done, both in the NFT space and with existing signage and physicals, his project Very Large Array comes to mind as something analogous—it leverages the onboard sound card that exists in everybody's computer to create these super dissonant sounds that I can't stand, but Will loves. So thinking about how you might use the ASCII side of things in conjunction with the native hardware components available in these devices—there's something there. I don't know quite what it is.

Will: I just want Device 1 to click.

Trinity: I just want the click—because certain things on your sound card are being triggered directly, versus using Tone.js or whatever. Those are two very different things.

Will: Still great work without sound. I just love the exploration.

Trinity: I have another question before we wrap up, partly related to what you mentioned earlier about 256. At the start of the show you talked about using signage and symbols from culture to drive a lot of your work over the last few decades—things that are often regional, with specific meanings: here's what a train track sign looks like in France, or a slowdown sign in Germany. There's often a broader cultural meaning behind these signs. And obviously with ASCII characters, those are things we observe and interact with every day—a critical part of society. Is there other inspiration behind using them? And with 256, with its alien signage, putting yourself in the mind of what would be relevant in a futuristic social construct—I'd love to hear about the meanings and narratives that helps create.

Andreas Gysin: That's interesting—it's almost semiotics. What do these symbols drag with them, and do we want that meaning present in the final artwork? Do we want to push it, or try to reduce the existing meaning of these forms? I think one simple answer, especially in the ASCII pieces, is that each letter, each glyph—each symbol from that initial ASCII character set—I don't treat as a bearer of meaning anymore. They're not letters in these pieces. They become form. I'm interested in an L because it has a sharp corner. I'm interested in a Y because it points down, while an A points up. The meaning is cancelled.

Device 1 — ertdfgcvb

That's fairly simple when you're working with single glyphs. But in the train station piece, there's no single character anymore—each element of that signage already spells out a train station name. It's much harder to remove the meaning when you can only display a full word. I think that's the beauty of the project: it coexists on two planes. You're looking at a kind of choreography—enjoying the movement, the sound, the wind, the air moving a little—the shapes, the composition, the idea itself. But then you also start to read these place names being displayed, and there's an effect—something happens in the brain. You read "Paris" as a train destination, and even while you're watching the choreography, the word sticks somewhere in the back of your thoughts and enters that moment of looking at the composition.

So it's different—each project treats this a little differently, and controls the amount of meaning carried differently. Street signage is easier, in a way, because it's already so abstract: high contrast, full color, ultra-simple geometric shapes. As soon as you take them out of the context of the street, they almost take on a life of their own. They're already abstract; away from a street or a car, they transform quickly into something else.

Will: That got deep. Trinity, was that the answer you were hoping for?

Trinity: We haven't seen 256, so we can't really talk about the science-fiction futuristic symbols specifically, but it sounds like the symbolism itself doesn't matter—it's about leveraging the geometries they create to convey not meaning, but a feeling, perhaps.

Andreas Gysin: I don't want to add anything after that—it was pretty beautiful.

Will: That feels like the moment to transition into ending the episode with a couple of quick questions. You already mentioned Lenny as someone whose work you enjoy and collect—is there anyone else you'd want to shout out, artists you admire and collect in the NFT space? Maybe some of the folks who inspired you to take the leap into NFTs in the first place. Who's out there that Andreas likes to collect, and what do you look for?

Device 1 — ertdfgcvb

Andreas Gysin: I like motion pieces, mostly. I've already talked about Lenny—he does a lot of things, and I think he's someone who has an idea and just goes for the project, which is super interesting because he doesn't really have a fixed style. For me that's usually a good sign. He explores different ideas—sometimes very simple ones, almost like a small joke or a bit of poetry, and sometimes more complex projects. Sometimes the language is very visual, sometimes it stays back and something else happens. But he does a lot of motion pieces, so I try to collect them all.

Another artist I like a lot who mostly works with motion is Kim Asendorf. His pieces, heavily motion-based, have a quality I also try to reproduce and seek out: they hypnotize me. They put me in a state where I forget about everything—I just look at the piece, and at a certain point I don't even look at it anymore, I'm dragged in, transported somewhere, and totally unconnected thoughts start appearing in my mind. Unlike Lenny, he has a bit more of a recognizable style, but he's still exploring an idea that needs exploring—his flip-flop, back-and-forth movement of pixels. I don't think he's squeezed all of it out yet, so I'm happy we get to continue that journey. That said, he's also shown he can go in different directions, that he's not limited to that one style or technique. Those two are very close to me.

Will: Two of our favorites as well.

Andreas Gysin: Kerem Safa is also very good—he generates repeating patterns. He's a musician, and you can totally see it in the work; there's a musical quality to it. He works by hand, so his pieces are drawn, but that repeating pattern, those oscillations, that back and forth—it's again an aspect I always look for in a project.

Will: Haven't heard of him.

Andreas Gysin: Go check him out.

Device 1 — ertdfgcvb

Will: Thank you. All right, Trinity, one rapid-fire question and then we'll wrap it up?

Trinity: What's the story behind your username? I feel like we should have started with this question.

Andreas Gysin: It's in front of your eyes if you lower your gaze.

Will: There's a sequence of keys that kind of makes sense, but how did you come to it?

Andreas Gysin: I needed a domain name to publish my projects online, and I was thinking about all these cool names involving some color—black, white, gray, dark—plus some cool word. I realized all these names were super bad, and I was pretty sure they'd work like a tattoo: I'd get super annoyed with one a year after choosing it. So I thought, well, I work with geometry—let's find the geometry on the keyboard. The problem is keyboards are international and they change: something that works on my keyboard doesn't work on an American one, or an Italian one, or a French one. So I overlapped many keyboard layouts—virtually, not physically—and found the area that doesn't change. Those central keys stay the same across almost all keyboards, and they build a square. So I kept it.

Trinity: If keyboard makers ever made one square, maybe that's where it would end up. A little crazy, but I appreciate it.

Device 1 — ertdfgcvb

Will: Trinity's a bit of a keyboard fanatic.

Trinity: I go through waves.

Andreas Gysin: Ah, you have a mechanical one?

Trinity: I have two or three, but I haven't used them since working from home. It's just not as ergonomic without a standing desk, but I love the clicky clackiness.

Andreas Gysin: We should have talked about this.

Trinity: All right, we'll do another interview, or you can guest host and we can have a side podcast about keyboards.

Device 1 — ertdfgcvb

Will: Waiting to click.

Andreas Gysin: Whoa.

Will: It's also so fitting hearing you describe the process for coming up with that name, having just heard you talk about deconstructing the meaning of letters, or ignoring their meaning entirely, and arriving at something based on form and function and universality rather than any meaning tied to the actual string of letters. It feels connected to your work in a way, hearing you speak. Andreas, do you have anything you want to say before we go? Any final words, any plugs, anything to preview from you after the Verse exhibition?

Andreas Gysin: I don't know anything about the future, but I really want to thank you.

Will: You're welcome, and thank you for coming on the show. It's always amazing for us to talk to an artist like yourself and get your perspective and learn more about you. Greatly appreciate it. And thanks to Layla at Verse for getting you on the show with us. That's it for this one. I hope you all enjoyed it. That was Andreas Gysin — check out his solo exhibition coming up on Verse right around the time this episode airs. Thank you everyone for listening.

Andreas Gysin: Bye.

Device 1 — ertdfgcvb

Trinity: Bye.

Andreas Gysin: The rail of the week. It's gonna be time. We're waiting. Always.

Will: We're waiting.

Andreas Gysin: To be signed.

Change log

  • Initial transcript — auto-transcribed (AssemblyAI) and readability-edited.