Screensavers, Generative Images and Entangled Algorithms
The emergence of the screensaver a few years before the explosive growth of the world wide web brought a large audience into contact for the first time with computer-generated images and the algorithms (the processes computers use to carry out calculations) that produced these images. The artist Rafaël Rozendaal is fascinated by the aesthetic of the early screensaver, which is the subject of the exhibition Sleep Mode. Klaas Kuitenbrouwer spoke with Jan Robert Leegte, Rosa Menkman and Matthew Plummer-Fernandez about the development from early generative images to today’s complex algorithmic culture. This article has been compiled from two separate interviews with the artists.
Jan Robert Leegte (NL) has been active as an internet and browser artist since the early days of the world wide web. Rosa Menkman (NL) is a glitch artist and theorist who achieved fame around 2010 for her work around image compression in the context of digital visual culture. Matthew Plummer-Fernandez is a ‘bot’ artist and blogger on, among other sites, Algopop, where he writes about the role of algorithms in daily life. Works by Jan Robert Leegte and Rosa Menkman were acquired along with works by Rafaël Rozendaal and others in December 2016 by the Stedelijk Museum in association with the Museum of the Image (MOTI) in Breda. Matthew Plummer-Fernandez is co-curator and speaker at the Bot Club, the Thursday Night Live! programme about current developments in algorithmic culture, the first edition of which will take place on 2 March 2017.
Using the content of Sleep Mode and their own experiences as a starting point, all three reflect upon the increasing entanglement of human culture and computational processes, about various notions of digital simulation and about the development of different audiences in relation to digital culture.
Since the early days of screensavers, the ways that human authors, code, hardware and the public affect each other has changed radically. At that time, generative images were still an exotic phenomenon that frequently referred to pure mathematical laws that allowed the computer to simulate natural processes. Today large parts of the media, the economy and other sectors of society are hopelessly entangled in indistinct interactions between human psychology, biased computer models, filter bubbles, statistical forecasting, misplaced suspicion and misplaced trust in technology.
KK: Alongside yourselves as makers, various kinds of software such as browsers, digital platforms, hardware and the wider environment of the internet all have a voice in your work. How do you see the interaction between yourself as author and the roles of all the other software and hardware elements?
JR: For me that is a very material question, almost like asking: ‘What sort of paint do you use?’ But it was a real process of realisation. When I began in the late 1990s there was barely any history for my sort of work. At that time I always showed it in Windows ’98 with Windows Explorer. For me, the formal idiom of that interface referred to modernism and minimalism, more so than that of the Mac from that time. That specific form was important. See, for example, Scrollbar Composition from 2000.
I didn’t make any more online work for a while after that. When I looked at my old work again in 2007 I saw that it was totally transformed. I viewed Scrollbar Composition on a Mac with the new aqua scroll bars. It had become something totally different in visual terms. At first that created a distance and the work didn’t interest me so much any more but at a certain moment I saw that that transformation was essential. I began to understand the digital ecology of the development of the web.
I now make a lot of works that use APIs and here the transformation and vulnerability can actually be a point of departure. You don’t know precisely what will happen in the black box of an API. For example, when I showed Random Selection Random Image in Mexico, very different images emerged than in the Netherlands. The Flickr database is highly localised. If you log on in Mexico, the image results for your searches come from the region around Mexico. Then you see the tactics of a company such as Flickr unexpectedly expressed in your work. Nowadays I really enjoy giving in to this.
In January I was in a show, Electronic Superhighway, at the Whitechapel Gallery in London. Together with the musician Dragan Espenschied I had the idea of showing a work in three different epochs. Then you suddenly feel the performative characteristics of such a work, that it has developed over time.
RM: I’m always interested in pulling apart all those voices or actors to see which ‘affordances’ and specific qualities they possess. I think it’s important to analyse this and to show it in order to get a better notion of how the technology actually works and how it influences my perception and perspective. Sometimes, for example, a compression with certain qualities plays an important role.
"Sometimes there are enormous political choices and issues behind the choice for a certain piece of software. I try to capture and show motivations and consequences, either in images or sometimes in sound."
I always focus on a small part of the enormous chaos of elements that play a role in this and try to elucidate it from different perspectives. In short, this means that I often attempt to anthropomorphise various elements. I did this, for example, in DCT, a work from 2016 in which I attempt to offer a perspective on how JPEG compression works by following the compression from compression to compression.
MPF: I begin to see my work as reconfiguration, of code repositories, platforms, Google searches and such. For me it is interesting to bring those voices into the process as part of a generative process. I try to reconfigure or recycle those elements into something humorous or satirical. I kind of think of them as entangled systems that are by their own nature quite open-ended. I use APIs or other tools of which the end result is also open-ended to turn pre-conceived ideas about what you should do with those tools on their head.
I do not see this as direct criticism. It is more like performing a strange, mischievous social role, exposing the norms of the system of the technical construct that programmers, users, everybody seems to obey. By turning everything around, by twisting the plot slightly, that set of rules and norms is breached.
KK: Generative processes play a role in your work. But does that also mean that it relates to generative artistic practices?
JRL: I strongly associate the generative with pieces of software that make things entirely by themselves. I couldn’t connect with that whole movement of formalist, algorithmic art in the 1990s. But the networked, working with mining is a generative process.
The fact that you can now work with more intelligent systems and with queries and APIs makes everything much richer. Since the emergence of the internet, the term ‘generative’ has a much broader scope. Generative processes are now ubiquitous.
RM: No matter what technology I employ, it’s always tied up in all sorts of complex interactions. I am also part of that. I think that the word ‘generative’ is too stuck in a certain tradition and time. You could use the word to describe aspects of my work but it doesn’t describe the most interesting parts of my work. For example, you could describe the work A Vernacular of File Formats, which has been acquired jointly by the MOTI and Stedelijk Museum in Amsterdam, as a self-portrait, but then you haven’t grasped its essence. The essence of my work resides more in exposing the rules, the protocols that lie hidden in the technology and that dictate our use of it.
At the University of Amsterdam I studied with Geert Lovink, one of the protagonists of the Tactical Media movement since the 1990s. I see myself as a part of second generation of tactical media thinkers and artists. Though of course tradition belongs to an older generation and the issues and media have developed since then and can’t necessarily be compared.
MPF: I recognise something in the ethos of the tactical media people. But they worked through disruption, by doing something profoundly dissident or antagonistic in short bursts of activity. ‘Bot people’ (like me) tend to make things that will run for much longer, and instead of disrupting it is the opposite: they kind of ‘fix’ it, but in a wrong way. Have you seen the film Brazil ? The Robert de Niro character, the renegade plumber, comes in and fixes it, but not according to procedure.
Within art practices that are developing, you see an interesting move from closed systems to entangled systems, from generative art in software like Processing [link] to much more complex systems that use APIs, or that are outputting to social media.
These kinds of systems that you can construct are also themselves entangled with audiences. They use all kinds of inputs and other sources. That’s what makes such open-ended systems interesting, these more external voices.
I see my work as more in line with a conceptual art practice, where the procedures are the work, and the material output is the trace, the documentation. I want to see what the material output is of a system that for instance utilizes Google search. ‘Bots’ is just one way of framing that type of practice. There are many more interesting outcomes that can be generated with such configurations.
Much of our daily existence could be seen as the physical output of a generative process. That widening of generative processes is also what makes it relevant to a wider audience. People can identify with their own use of Google, Facebook or Amazon, and they can see for themselves how those systems construct their behaviours, their norms, their lifestyles.
JRL: A friend of mine is a UX (user experience) designer and works on solutions to problems in digital systems in the dairy industry. This sector is totally beyond the difference between the digital and the physical; it’s totally governed by algorithmic processes. A dairy farmer keeps track of the developments in the data and can manipulate various sliders on his screen to determine how much milk he gets from a particular cow at a certain time. The computer also determines how much feed the cow needs. It’s a black box for the farmer.
The cow’s entire life is governed by algorithms. There is some room for manoeuvre but it is entirely robotised. There are probably other unexpected sectors within which these developments are much more advanced than we might assume.
KK: More and more realities are being algorithmically modelled and optimised. We are to Facebook what those cows are to the milk-production system. There is often a mathematical model behind those sorts of systems that functions as a normalising standard for the algorithms’ interventions. In the first screensavers and in early generative art, simulations – mathematical models of reality – often played in the foreground: the idea that an aspect of reality could be captured in an algorithm and imitated in the computer. With the emergence of ‘machine learning’ and digital platforms, dynamic models of parts of reality play a role in more parts of society. Does the idea of simulation play a role in your work?
JRL: A classic idea of simulation plays an important role in my Minecraft work, for example in Spiral Jetty, which is literally a participation in a digital, simulated environment in order to reflect upon the artistic tradition of Land Art. Robert Smithson (the maker of the original Spiral Jetty in 1970) subjected himself physically to an ambitious project in an inhospitable area. He attempted to succumb to the landscape. That sort of idea is also important in Minecraft. For that reason, I made my Spiral Jetty in Survival Mode.  That was a very conscious choice for the idea of simulation. Otherwise I would just have been playing with Lego. First I had to find a good place, like Smithson had.
"I had to travel, search and find, while beating off monsters. I had to build a hiding place and make tools."
Minecraft has its own process that you go through as a player, also with disappointments. That’s also very funny. Then I climbed into a high tree, but that didn’t give me a good shot. I had to climb up another tree in a different place or stack stones to make a screenshot. This work also plays with the idea of simulation in the receptive sense. Most people know Spiral Jetty only from photos and documentation. Very few people have been there because it’s in the middle of nowhere. It is also now underwater. People also know my Spiral Jetty in Minecraft only from the documentation.
I was very intrigued by the qualities of landscape in the digital domain. Not only in Minecraft, but also, for example, in Flickr. That’s what Mountains & Dropshadows is about. This work should of course be compared with what Eva en Franco Mattes have done: the re-enactments of famous performances by Marina Abramovic and Gilbert & George in Second Life. I think of Second Life more as a culture simulator than a nature simulator like Minecraft. In my opinion, those re-enactments of performances wouldn’t make any sense in Minecraft. This sort of work is more about what is simulated than about the more abstract idea of the nature of simulation.
"Simulations are not only models or reproductions but are speculative machines in which something new can always happen and which can make predictions."
RM: The concept of ‘simulation’ has a limiting factor: many simulations take on the visual form of the process they simulate. There are so many possibilities in more fluid translations of actions that do not always need to be observed or do not directly resemble the things they simulate. Take cancer research, for example: by simulating or modelling the growth of cancer not in imagery but in sound, the growth and development of a tumour can be observed via a rhythm. Some forms of cancer grow at a different ‘rhythm’: for people, development is easier to understand and assess via rhythm than through an image. I think there’s an enormous wealth and future within the field of more experimental forms of simulation and translation.
MPF: That early notion of simulation triggers curiosity, almost in a historical perspective. It goes back to the weird dichotomy of the virtual and the real, and the idea that these digital processes were there to compute or mirror physical processes. My work called Peak Simulator was really about simulation. It was looking at the history of simulation and generative art, at the first modelling of 3D terrains. The first computer-generated landscape was made by Loren Carpenter and was used for Star Trek 2.
When I was asked to do something for the Lake District, a beautiful, mountainous area in the UK, I decided to take his original algorithm and then build a mountain ridge and put it side by side with a real natural landscape. I wanted to take it from its simulated nature and put it back in the physical world, as a form of physical trace of an algorithmical process. It is easy to suggest that algorithms are part of the digital domain and not of the physical, but it is interesting to see such processes materialise. The Amazon warehouse is a good example, a huge physical space that is entirely arranged by algorithms.
Simulation started with the question of how to model the outside world, but we have moved beyond this to almost a reversal, in which physical processes have now become the traces of primary digital processes. I am not particularly interested in algorithmical processes that emulate natural phenomena. That is too romantic.
"The element I want to bring in from the physical world is the messiness of it. The human element."
In the early days of generative art, media art had a highly speculative dimension, the exploration of technology. But the computational has become banal, rather than sublime or futuristic. My work doesn’t have to find strength in newness, it is about the now, not about the future at all. It is also not about technology as such, it just takes technology as a context.
For me technology becomes more interesting to play with when it becomes banal. I would rather make a piece on Facebook than on virtual reality. VR is not banal enough yet. It is still exoticised. Whereas Facebook is as boring as a shopping centre. Everyone knows that. So that is where you can do your software art equivalent of The Dawn of the Dead.
KK: Perhaps the screensaver had a banal aspect?
MPF: The screensaver did not have a place in media art that relied on the technological sublime. Screensavers were rather simple pieces of software, and were not designed to run on supercomputers. Maybe you can regard it as a kind of folk art. The screensaver became this cultural artefact that was overlooked at the time because it was so ordinary.
Reception - the role of the audience
KK: What has been developed in the reception of algorithmical artefacts? Do you notice a development in the public?
RM: The public is more literate. Technology and interfaces have become part of our visual culture. The visual language of digital culture has permeated into the smallest niches of life. I see this development not only within the public but also in the questions I am asked in interviews. People ask more complex questions and reflect more. They accept that the work is not a gimmick, but they are prepared to see more layers. The days when my work was seen just as a beautiful or ugly picture are thankfully in the past. But of course you can’t talk of just a single public. Many people who visit exhibitions are informed and literate, whereas it’s probably different when I step outside of my ‘discourse’.
JR: The visual language of digital culture has been assimilated. I’m now in a commercial gallery! After fifteen years the collectors are coming! And they understand it!
You can distinguish several layers in the group of people around a gallery like that. There are the collectors: their art-historical understanding is better that it used to be: ‘Do I read this in the context of performance art or visual minimalism?’ The post-digital discourse has had an enormous effect: the understanding that computational processes can refer to reality outside the computer and vice versa. That has grown enormously in the past few years. Almost no one sees them now as divided worlds; that is a broad cultural shift.
"The post-digital has now truly arrived."
But then there’s the very material question that collectors have difficulty with: how can it be commodified? A non-collecting audience is less concerned with the question of commodification and can accept the work more easily. Take the question of authorship. People are used to the idea of a maker producing something, but I ‘subcontract’ an important part of the work to APIs or other computer-generated processes. I can explain that and we also know that from Warhol: appropriation and suchlike. Process-based things also have a long history I can refer to. With a little bit of explanation, the public is on board nowadays.
RM: I’m not a big fan of the term ‘post-digital’. Many professional sectors have become post-digital, but many others haven’t. For me it’s interesting at this moment to make translations not within art or visual culture but between totally different discourses.
The technology and the language within which we work has now been developed to such an extent that they also have a form in other professional sectors and for me that represents an enormous new wealth of possibilities. People from the legal world or the medical sphere, for example, are also having a discussion about what digital technology means for their material and practice.
MPF: I think of the audience as existing in layers, depending on how distant they are to the performance of the work. If you’re very close, you don’t see the bigger picture. These include the co-users of Thingiverse, where Shiv Integer is operating. Then as you step away, there is a secondary audience of people that read about the project and then see what is happening, and see how these immediate audiences are reacting. Then you zoom out even further, to the documentation of the project in a gallery space, for a third, even more reflective audience, but they don’t have the thrill of being directly involved. These layers help me to not get too universal about the idea of the audience. There is no one single audience. And all audiences are constructed.
The first layer of audience is the most important, because they are really part of the project. They are interesting because their reactions can be so genuine. However, the other two are also very important. The project only resolves itself when it reaches the third audience. That reflective moment is also what I’m going for. Werner Herzog said in an interview recently that for him a film is not complete until an audience has seen it. They are the final ingredient, when they bring their own ideas, their own experience of the cinema. Only then is the movie making complete.
KK: In terms of the content of your work or the actual processes at play in your work, is there anything in the development of the reception of those that you find particularly important or fascinating?
MPF: The reactions of the Thingiverse participants to Shiv Integer point to the kind of implicit social norms that appear to be violated. They are not the technical rules, but the human-made rules, although the rules are also generated by the particular platform in play.
One example was that many users were angry that they were getting kicked off the front page of Thingiverse, because Thingiverse is set up so that everyone sees the same front page, which is updated as soon as there is new work. For many people this is their two minutes of fame, when everyone in Thingiverse is going to see what they have just made. If there is a bot that continuously puts up new work, those two minutes are reduced to one minute. That is a big thing for the users of Thingiverse.
So the first layer of audience is angry. The second layer of audience then says: this is a fault of the platform, they have created a system where everyone is looking at the same front page. They see this social routine of getting yourself on the front page, that is also technically constructed, through the way the system is set up. And then the third layer of audience looks at the whole thing as a comment on the reputation economy, on the platforms and such like. Each layer is creating something for the layer above. The first reading of the project is authoring the reading of the next layer of the audience. The work is not finished at the outset, and then there is an audience. It is unfolding. In this quite literal way, the layers of audience are all performing a part of the generative process. The work is not about computation but about the entanglement of culture and algorithms.
KK: To what extent are you consciously working with different audiences?
MPF: It is in development. The more success I have with such projects, the more I want to do more of them. For Disarming Corruptor I was designing the software and then people were downloading the software and then supposedly using it, and then there were onlookers who were commenting on that use. I started to see layers emerge. Shiv Integer was putting that into practice. I am enjoying developing that idea.
RM: I’ve noticed that the way my public and I relate to each other has changed. Or perhaps it’s just my role that has changed. In the past I looked mainly at art but today I try to step outside my filter bubble and to be the ‘audience’, the observer within other professional sectors or discourses, such as law or medicine. I enter their field and look and try to understand how they are using digital media. In the meantime, we have matured within our own professional field. We are better equipped to develop the conversations further when we transfer or broaden them.
MPF: The context for such work has developed. So much of our lives are now mediated by computational systems. This also means that these works are not merely technical pursuits (as early generative art often was). They have become more discursive and much more reflective about the world we live in.
1. Application Programming Interface is a means by which a piece of software can ‘talk to’ another piece of software or a database and extract information from it.
2. Mining, the retrieval of unstructured data from the web, is a metaphor and makes an analogy between the mining of raw materials and the mining of ‘raw’ data.
3. Minecraft can be played in Creative Mode or in Survival Mode. In Creative Mode there are unlimited building materials and no monsters. It is simply a matter of building with digital blocks in a simulated world. In Survival Mode players are beset by zombies, spiders and other monsters and must mine the materials themselves and fabricate tools in order to build.