‍

Matt Hall and John Watkinson on Beginning Movements

Contemporary artists and creative technologists Matt Hall and John Watkinson describe their theory and practice to Peter Bauman (Monk Antony). The pair explains how projects like CryptoPunks and Autoglyphs continue decades-old traditions of web-based creative interventions while revealing the potential of blockchains in art.

‍Peter Bauman: You’ve talked a bit about your core philosophy before—its grounding in curiosity, technological experimentation and creative hacking. But how would you describe it now? How has it developed, especially with the success of your projects?‍

‍Matt Hall: It took us a little while to figure out and we did through working on things, determining what we found interesting and where we found we were at our best.

The realization we came to was that when working with new or newish technology—or at least something that feels like that to us—its potential hasn't really been revealed yet. One of our strengths is that we can figure out a way to illustrate how it's interesting. We felt that when we first started working together. It was the early days of the web and there was lots to figure out there. Then we did the same for mobile.
‍

For blockchain, it took us a little while as well before we realized that there's something really cool here. There's something fundamentally new and interesting here. 

Everybody's using it for money. We know that. But there's something else here. It was like CryptoPunks was the experiment that illustrated to us what was interesting about that new thing.
‍

‍
‍
‍John Watkinson: Where there's some new tech—some new capability that's interesting and has promise, but maybe hasn't quite crossed over into that mainstream use case yet—is the most comfortable place for us to operate.

‍Peter Bauman: Had you been aware of any of those earlier experiments with the blockchain before 2017—the work of the McCoys, Rhea Myers, NILIcoins or anything like that?‍

‍John Watkinson: I think the only thing we had heard about is Rare Pepes. When we were looking into whether it would be possible to do a collectible on here. The only thing that seemed close was Rare Pepes, using Bitcoin and some other tokens. That was still a little different in that most of the Rare Pepes or all of them were basically just a coin. There might be 200 of them or a thousand, or there might be one. But each one was its own coin category. That was the closest thing we could find. And so we were looking into how they were doing that but we weren't too satisfied. It was very complicated.

That's what brought us to, “Oh, wait, Ethereum; it seems like the perfect place to do this thing.” But we weren't aware of any previous efforts at that or anyone who was encoding artworks into a transaction.

‍Matt Hall: We weren't even really thinking of it as an art project at that time either. So we weren't looking into art world precedence. It crossed over by itself. And then people were telling us, “This is artwork.” And you're like, “Oh, all right, cool." [laughs]

‍Peter Bauman: Does AI now fall into that category of new technologies whose “potential hasn’t been revealed yet,” as you said?‍

‍Matt Hall: That is in that category. Obviously, when something's received $100 billion in funding, it's probably pretty well known. Perhaps we haven't seen what's really interesting about it yet. It's similar to what we were talking about before. There's something clearly interesting and new here.
‍

What is actually really great about it? What's the essence of what's really good about it? Is it writing work emails? There's something else that feels possible with it that would be really, truly special.

‍
‍John Watkinson: Even though it's very mainstream and there's huge investment by all the big companies, there is a feeling of, “Are we really doing cool stuff with it yet?” Or is it all very up the middle—similar to what you could say in the early days of blockchain? Everything was just money, the most obvious use case. We might be in that phase with AI as well.

‍Peter Bauman: What are the limits of your interventions with new technology? I spoke recently to net art pioneer Joan Heemskerk about her work with quantum computing and networks, for example.‍

‍John Watkinson: We're not going to be the types to try to get access to a quantum computer and do some weird stuff with it.
‍

We're more about what's that cheap, accessible thing that anyone could do but maybe no one's doing much with yet because they don't know what to do. 

Even when we were working with smartphones in the pre-iPhone era, there were a few that had really cool features. But they were big and bulky with really terrible battery life and bad CPU performance. Yet there was something really cool there where it's like, “Oh, having this in your pocket is really transformative. There's something going on here.” That's where we operate rather than the more deep science stuff that isn’t a consumer thing. Even though it's cool and interesting, that's not quite as much for us.

‍Matt Hall: We want to make something that's interactive in some way, something that people can use and experience. It should be affordable or on something they already have—ideally, a phone. That's why we weren't the first people on stuff because it was too early. It's got to clearly have some adoption and user base for us to get interested in it, probably.

‍Peter Bauman: It's aligned with that punk ethos, which Simon Denny spoke to me about, saying he’s excited when “good ideas can trump technical knowledge,” like in punk rock. Part of the punk ethos is hacking the current system and tools to fit one’s own artistic aims. How do you subvert and harness the power of systems thinking in your practice?
‍

‍‍
‍
‍John Watkinson: We came at it a little bit sideways in that we were experimenting with the tech first and then we were educated later. It was pitched back to us. We launched CryptoPunks and then it was a success—not a big financial success yet—but people were doing it. They were feeling ownership; they were buying and selling.

We had people from the art community say, “Oh, this is actually a great way to edition digital art.” They were teaching us about that history and saying that this could be an important next step for digital art.

After that, we started researching and learning about generative art, which we knew very little about then. That led to Autoglyphs, where we were thinking, “Now that we have hacked the system a bit to have this way of making digital art have this almost physical feel, how can we take that to its most extreme place—its most pure form? The idea of Autoglyphs was that it wasn’t just residing on-chain but created there as well.

The entire genesis of the piece and everything that follows is on chain—on this public, world-accessible computer.

‍Matt Hall: It's specifically Sol LeWitt's work, which I had encountered before but had never really grasped the idea fully. When we started thinking about Autoglyphs, we re-encountered LeWitt at a Whitney show. It was almost like a solution to a problem we were having: How do we make this as pure as possible on the blockchain?
‍

The execution feels like a limitation of the concept if it becomes the focus, whereas we want the algorithm to be the focus. It was like, “Oh, this is the solution. The instructions are the art.”

That was the final piece to make it feel like a really coherent and—perfect is the wrong word but—a self-contained, independent project that felt pure. What, in retrospect, we feel has become interesting to us and ended up being one of the powerful concepts was the idea of limiting—using the blockchain to limit something that's infinite.

If you have a generator that can produce infinite outputs, that can make each output almost not feel special. It's like, “Okay, I bought a poster from the guy, and it's one of infinite outputs. Fine.” Editioning allows it to be a market asset but also algorithmically limits and defines the outputs so they feel special. You can also still see the variation possibilities. You know there's something at work behind it but each one is now a finite thing, not an infinite thing anymore. We did that without thinking about it expressly in CryptoPunks. It was imperfect in the way that the algorithm and the generator were separate. Then Autoglyphs was the refinement of that, the perfection of that.

‍John Watkinson: There's this tuning that I feel goes on, where every generator has the right number of outputs that you should generate from it. You want it to be a rich enough set where you see what's possible; you see all the possibilities. But you don't want it to feel redundant where you can't tell one apart from another.
‍

Every generator has a sweet spot where it just feels right, where you experience the full expressive capability of the generator but you don't feel like it's a ridiculous amount.

512 Autoglyphs felt like the right amount of expression for that algorithm, which was a very limited algorithm because it had to run within a single transaction block.
‍

‍
‍
‍Peter Bauman: These two projects opened the door for what’s commonly referred to today as long-form generative art. Many subsequent artists have made similar editioning calculations, yet you were some of the first to really think about it. Did you realize at the time that we hadn't even begun to scratch the surface of what algorithms, infinite systems and blockchains could achieve? Obviously—and you’ve said previously—you did not see the unfathomable success coming. But did the creative potential of what you'd uncovered hit you?‍

‍John Watkinson: I don't think it did.

‍Matt Hall: No.

‍John Watkinson: Especially because Autoglyphs came at a time of maybe the very lowest interest. There was some excitement, and all that, of course, was tied to crypto having a lot of interest in general leading up to the winter of 2018. Then it all went underground again. It seemed like there was not a lot of interest in crypto in general, certainly not in any innovation with it.

‍Autoglyphs was almost a project that we wanted to do for ourselves at that time. We had the idea not long after CryptoPunks. We thought about it and even took a few cracks in 2018 but gave up because we hit some technical limitations. At some point, we made that Sol LeWitt connection where we realized, “It can be this representational thing. It doesn't have to be about the final output.” We took steps towards it. But it really just felt like a labor of love for ourselves—something that felt cool and interesting.

We had an opportunity to put them in a little art show that Georg Bak was curating so that pushed us to get it done. [Note: That show was Automat und Mensch at Kate Vass Galerie with Jason Bailey and we interviewed Bak, Vass and Bailey for the five-year anniversary of the exhibition].
‍

Then we thought, "That's it. All right, that might be the end of our art career." There was not any sense of, “Oh, this is the beginning of a movement.”

We've done this thing. I don't know if anyone cares, really. It's a little bit esoteric. It's not easy to explain what's significant about it to someone like our parents. It is a lot easier to explain to people who have a tech background or grew up with tech. We didn't anticipate it being a mechanism that a lot of people—a lot of established artists—would use soon after. It felt like a last act for us in this space at the time.

‍Matt Hall: This all becomes obvious in hindsight. You're feeling around in the dark and you feel something interesting but you don't really know what it is. You just go ahead and only after time can you look back and realize, “Oh, that's what was cool about it.” Even we weren't sure exactly how interesting what we were doing at the time was. It seemed cool to us. But it definitely didn't seem like this was going to be what it has become now.

‍Peter Bauman: The concept of the instructions being the art, like you mentioned, was that arrival point just intuitive?‍

‍John Watkinson: No, before we thought about it from the Sol LeWitt point of view, we were wrestling with that. We were thinking about representation because we had these very extreme limitations on Autoglyphs. We thought, “How do we make something that we think looks compelling—surpasses our bar for aesthetically achieving something—and that people are going to want to look at and return to, while still managing these really tight computation and representation constraints?”
‍

When we saw Sol LeWitt fitting instructions on a single page that then turns into this enormous wall art, the light went on. 

“Oh, it can be a representation." And what we did is not exactly like Sol LeWitt, because we leave it a lot more up to chance. Sol LeWitt was extremely prescriptive with his stuff.

‍Matt Hall: If I remember back to that time, we were just starting to talk to the art world about CryptoPunks. Our sense was there's still some confusion here. They're not quite getting it. There were a lot of questions around, “Well, where did these CryptoPunks come from?” “Where is the image?” “Am I buying the image?” And we're like, “No, not quite. It's this thing.”

One of the questions was, “Is this on the blockchain?” And we wanted to answer “yes.” It was a nuanced answer for CryptoPunks but we wanted to say just “yes.”

Part of the solution was to have it represented by instructions rather than to be literal. That led to all the decisions you see there—that the generator was going to live on-chain and that the output had to be written on-chain, where you could go see the output on Etherscan. That felt important to us.

The idea and importance of the blockchain were not easy to understand. We had to say, “Don't believe us. Go look there to verify that the contract is secure. Go see. It's there. You don't have to trust us.”
‍

‍
‍
‍Peter Bauman: Instruction-based work with more randomness, like you mentioned, John, reminds me of Fluxus. That could be a fun museum show. Speaking of art historical connections, your collaboration goes back to Toronto in the net art days of the '90s. How did that time inform the cyberpunk vibe of CryptoPunks? Do you see a connection between those early net days, net art and CryptoPunks?‍

‍John Watkinson: Absolutely. It's easy to forget how much the early internet was an entirely counterculture phenomenon. Regular people weren't really accessing it much other than email. There were definitely a lot of people finding each other who had alternative lifestyles—that real punk vibe. Since then, it's become the largest industry in the world—basically, the tech industry. It's completely devoid of that essentially now.

When the crypto stuff came around, we detected a lot of the similar culture. Crypto became a new magnet for that. Plus, there was this economic cyberpunk ethos, where people were trying to escape existing financial and power systems. It definitely reminded us of that. It's like, “Oh, we're back in this era now.”
‍

That's a good indicator: If the weird people are into something, it's worth taking a look at it. There's something going on there if that same crew that we knew from the '90s is into something else now.

‍
‍When we started, we were working on the idea of collectibles already. A lot of the stuff we were focused on was even for younger kids—how they could collect something and have it feel real. But once we realized this was going to be a crypto phenomenon, and not for kids, then somehow it very quickly became CryptoPunks. It felt right that it was this punk thing. That was just our association with it.
‍

Matt Hall: I can't remember CryptoPunks not being that name. We got that name super early. We were both just using it. I don't know.

‍John Watkinson: It's true. And naming a project can be really hard. But that one, I don't even remember it being named. It was immediately that.

‍
Before we had even worked out anything, we had a little bit of the artwork and we were researching how it could be done. We didn't even know yet that we could do it. I remember Matt was just starting to dig into the Ethereum documentation—and it was already CryptoPunks.
‍

‍

‍Matt Hall: Exactly, and the early Internet did feel really like West Coast punk in a weird way—cyberpunk, not in that Blade Runner-y way but that '90s Wired magazine way where you're hooking up stuff to your PC and trying to get on the Internet. I can remember also thinking, “No one's ever going to make any money on this. Everything's free. This place is crazy.”

It felt amazingly permissionless in that way, too, where you could put a website up and then somebody on the other side of the world could just be there immediately for free. Having grown up with long-distance charges to the town next door—once I accidentally ran up a $160 phone bill calling some piece of junk BBS one town over—then to suddenly be completely international, it felt pretty incredible. We've always been looking for that feeling again.

We've seen the cycle a couple of times now where it gets taken over by big companies. Sometimes those companies become big because they do something well that we need. But the end result is you can't do wacky stuff anymore—easily, at least.

‍John Watkinson: I grew up in a fairly small town and I had no Internet access. But you could get Wired magazine. I remember reading maybe in '93 or '94 an article in Wired about some website, some Internet thing. You read the article and then you just put the magazine down, thinking, “That sounds like a fun thing.” [laughs]
‍
‍Matt Hall: [Laughing] Those books, too. Those huge books, like the thousand best websites.

‍John Watkinson: When I went to University of Toronto, I was studying computer science. The first day I was in the computer lab, the person sitting beside me was using a web browser. So I fired up Mosaic or whatever, and I was starting to hit a few websites. Then they were like, “Oh, you got to download Netscape. It's way better.”

Then I'm using Netscape and, crazily enough, they told me, “Oh, actually, your account on this thing, if you make a public HTML folder and set it to be world-readable in Unix, then it'll be a website. And this is the URL.”
‍

So the same day that I first browsed the web, I made a website. 

I didn't have much on it; it was just a “hello world” but that was astonishing. It was such a different time. And it felt so amazing. It was this weird counterculture that very few people had access to but then that very few people even wanted access to. It really was just a corner of culture that wasn't that important yet.

‍Peter Bauman: In terms of similarly emerging technologies today, like AI, is there a similar spirit? How do these early interventions with AI compare to those early interventions with the web?‍

‍John Watkinson: I think AI will always be different because it's not an organic, ground-up thing. There are extremely expensive upfront costs to make a model. It'll always be, to some extent, the domain of large companies. Because it arrived on the heels of massive investment, it came with massive hype as well. That doesn't mean there won't still be small communities that do cool, interesting things with it. But in terms of being a completely organic, from the ground up thing, I don't see that spirit.

I don't think we've lost something in general. You're never going to see the little pockets of interesting new culture and everything until they become larger. Even though we have the Internet now and everyone's hyper connected—you can find people so much easier—I still think that little pockets of culture and people doing interesting, cool things will eventually find a larger audience. I think all that is still intact.

‍Matt Hall: When John was telling the story of making a website the first day he used the Internet, it reminded me of our first computers, where you were programming a little bit to even load the game you wanted to play. You were accidentally typing in some commands. And then when you got bored of the one game you had, you started thinking, “Well, what's this other thing I’ve got?”

I always wondered, “What's the trade-off for people starting now?” There's so much more. Obviously, everything is much more incredible. Our computers were terrible but they were a little more approachable and understandable.

Now AI is amazing but mysterious. You can't really train your own model but you can go download one and do interesting things with it. You have the Internet; you can learn. There's so much more to see. You feel a little overwhelmed but you can learn things so much faster. I had one book about programming. That was it. So if I didn't get it from that book, I was stuck. It's such a trade-off.
‍

‍

‍Peter Bauman: AI is a loaded term—you’re right—running from those corporate generative latent diffusion models to GANs and other bespoke self-trained models. It can certainly feel overwhelming.‍

How do you see the trade offs or tension in the different ways your work is perceived? Looking back, you’ve achieved a remarkable amount. As artists, you’ve created lasting cultural work found in major museums. You’ve had successful professional careers. You were instrumental in solving the decades-long new media art editioning problem. You gave this space an identity with CryptoPunks and you demonstrated the potential of on-chain generative art with Autoglyphs. The art world was quick to embrace editioning. But it has been more critical of other unintentional elements, like the gamification and financialization of art.

‍John Watkinson: Talking about the unintended negative consequences—gamification and financialization of art—I'd say we're quite comfortable with CryptoPunks being almost this perfect embodiment of that.

We don't shy away from CryptoPunks being this very specific hyperfinancialization. Like you said, none of this was really intended. We've learned that this hyperfinancialization and gamification all previously existed in the art market. It's all present there but it's always a little bit secret, a little bit held apart. We like that CryptoPunks does the opposite—that it's all front and center.

If you go look at a CryptoPunks you can immediately see all the transaction history. There's something cool about that—the fact that that's inseparable from it. All this financialization is built right into CryptoPunks.

In terms of legacy, with Autoglyphs, it's pretty well understood. Whenever people talk about Autoglyphs, we're like, “Yes, you get it.” But with CryptoPunks, often the discourse is more about it being the first PFP and how it set up that whole movement.
‍

We say, “No, really, it's all about that market. It is this gamified, financialized thing. And it is pointing the finger right at the rest of the art market and saying, ‘This is what you all do.’” 

Not that we were in our ivory tower being like, “I know how to get these guys.” But certainly in hindsight, we're not going to blush and be like, “Oh, we don't want to talk about the financialization.” It’s been collected by museums and when they talk about displaying it, we're always pushing for displaying the financial part, too. We want to show it as this living financial entity.

So far, they've resisted that. No one wants to do that. But I think they will eventually, because it's a major part of this and inseparable from the rest. As time goes by, I think people will come around to that as an important part of the CryptoPunks' legacy.

‍Matt Hall: If you wanted to have a piece of work comment on the art market, there's no better way than to really mess up and be a part of the art market at the same time.



---



Matt Hall and John Watkinson are creative technologists who have worked broadly with software. They are best known for co-founding Larva Labs, where they created CryptoPunks and Autoglyphs. They also create large scale web infrastructure, genomics analysis software (Watkinson has a Ph.D.) and 8-bit roleplaying games.

Their work can be found in museums worldwide including Centre Pompidou, LACMA and ICA Miami.
‍
Peter Bauman (Monk Antony) is Le Random's Editor-in-Chief.