Play Engines

Left to our own devices forever, We watch the sun rust at the end of its days Alicia E. Stallings, ”

The Machines Mourn the Passing of People

I want to write now about machines. Not as enemies, not as future fictional technologies that will supplant us, but as devices that, as the poem goes, “[…] were kicked like dogs when [we] were broken”; devices not all too human, nor too alien either. Machines as things we play with, and by doing so we give them a different status. Elevated or not, it depends on your view on humanity – but a different status, nonetheless.

I want to write not about kicking the machines, but about being complicit with them. I don’t want to write about exploits as dangerous practices, but as exploits as forms of misunderstood collaborations between machines and people, combined forms of resistance. Exploits as aesthetics, as perhaps a most desirable form of aesthetic collaboration.

I want to write about play engines, not game engines. That means I don’t want to write about “game design”, but about making people play, about a world more playful. I won’t write about the kind of objects we can create with and for computers, to play. I write about the computer itself as a machine of play.

When we think about making people play with computers, our instinct is to look at how to turn anything into a game. I am not going to write now about how this is a complicated issue – the best approach to it was formulated by Sebastian Deterding. What I am interested in doing is following the logic, just a bit: if we want playful interactions with computers, it is only logical to use, or at least be inspired by, the technologies that facilitate the creation of videogames.

And so we look at game engines, and how they enforce certain aspects of what we consider “playing with computers”: game engines help with rendering things onscreen (and that includes VR); game engines run the game loops, managing the assets and calculating a game state on regular cycles; and game engines facilitate certain types of interaction, via dedicated controllers or the appropriate use of keyboard and mouse. Or it can get worse: I fear the future of a successful Facebook running only on Oculus and powered by Unity: 3D VR gameful social networking. Global warming cannot save us fast enough.

We think too narrowly. We think about what can we do with the computer, not what we can make the computer do. We must think beyond game engines; we must think about computers as potential play engines.

There are many things modern computers can effortlessly do, many more now that they are all part of centralized, privately owned “cloud” networks. All of these things are apparently at the service of our leisure and work: we can play games, alone or with many others; we can work, alone or with many others; and we can be alone, with many others. Computers are social, work, and leisure engines: much like game engines streamline and facilitate the creation of games by providing a support for the most common structural requirements of games, computers also structure work (unlimited undos, automatic backups, online sharing), socialising (liking, disliking, retweeting), and leisure (winning, losing, participating). All of these are, of course, re-ontologization processes.

So how are computers play engines? There are many different ways – most of which rely on re-appropriating the computer and its context, with mischievous effects. But I want to look at other forms of expression, ones that actually use the computer as a collaborator, and as a material. For these forms of expression, the computer is a gateway device for the re-ontologization process. Before the world is changed to be processed and computed, it needs to go through (but also made for) the computer. Any expression needs to be translated for the computer machine to be able to act on it. That translation is the moment for play. And a key process for appropriating the potential of computers as play engines is that of piping.(*)

Think about the computer and its networks as basically a series of tubes, if you wish (this line of thinking will take you far). These tubes “transport” data from one place of the computer to the other, where algorithms process that data into information. But with a little bit of knowledge, we can do our own plumbing. For instance, piping the data from the mouse, the memory, or the fantastic dev/urandom to the soundcard is not only possible, but also fun and playful. If you are on linux box, just try cat /dev/urandom | aplay. Or try the same with any picture: cat summer_holidays.jpg | aplay. This is not hacking. It’s simply taking one set of data, and literally piping it somewhere where it does not belong, but has an effect. It is playing – appropriating the machine, taking over the re-ontologization process.

Most of the systems we work with, most of the systems we make -even those we want people to play with- are subject to the rational myth of efficiency and functionalism. This is of course not a new reflection. I want to think about computing machines not as data churning devices, but as playthings. Computers are things we can play with. If we want to, computers become play engines – perhaps the most powerful, most radically transforming play engines we have ever witnessed.

So what is a play engine? Play engines are machines that reconfigure the world so we can play, or be playful, in it. Toys are play engines. Playgrounds are play engines. Our bodies are play engines. Computers are play engines. Not because we can make or play games with them, but because they open up, and make worlds for us to play in, and with.

What’s at stake with this idea? Let’s face it, we have all stopped being humans, if we have ever been, and one of the things we are now is obedient data producers. The expansion of the infosphere, the revolution of cheap, networked, ubiquitous computation, has transformed users into not just users, but also (unwilling, unknowing) producers of data. Your phone produces data about you, and so does your smart tv, your console, and your sex toys. Now, what’s interesting is that we produce too much data (big data, they call it), so companies (specially marketing companies) are on an arms race developing the best algorithms to make sense of all these data.

Another way of dealing with this data deluge, historically entrenched in the paradigms of open computation but far from its idealism, is the proliferation of open APIs. Facebook has one, and so does Twitter, Google Maps, and even LinkedIn (it’s API probably wants to join your professional network). APIs are great because by tapping on the mix of hope and desperation of many that want to access the tech industry, lured by its (inhuman office hours, rampant sexism and agism, and gentrification capacities) bright future, they became instruments for free labor. Give people free access to (most of) our data, and someone will come up with a clever solution we can then cheaply copy.

But open APIs can also be forms of resistance. That’s why they are so heavily monitored, because the risk of rebellion, of misunderstanding, of purposefully breaking things is too high. We should take those risks. APIs are made to comfortably, easily transport data from user/producer to algorithm. But we can break that pipe. We can make our own pipes. Piping is a mode of resistance, driving data not to be processed at the algorithmic slaughterhouse, but in other places, for other purposes: for fun, for exposure, for pleasure. Piping allows us to own the stream of data, and turns the computers into something else than efficient processing machines.

You want examples? Look at things like Infinite Adam Curtis, Ad Nauseam, Sans Bullshit Sans, Mountains of Mouthness, Do Not Touch, or Antenna. These works, to some extent, reflect some of the qualities I write about – the will to bend computed data to a will different than that it was originally intended to please.

Computers are play engines when we use them as deviant tools, when we do not surrender to their corporate imprint of tempting, streamlined, pleasurable interpretations of the world. Computers are play engines we can play with, together. Much like videogames excel when they overcome the stretches of their engines and their inscribed rhetorics, computers are play engines when we see them as fooled instruments, when we feed them the wrong data to model more interesting worlds than the one they were expecting. And in their generosity, these play engines often return to us more interesting worlds, invisible or unthinkable until then. The computer play engine creates new worlds and joyfully presents them to us, twisted and beautiful in their own weirdness, drunken worlds dreamt up by data drunk machines.


This is part two of the posts dedicated to the research I am conducting at UCSC as part of my sabbatical. It’s very much a work in progress, if you haven’t noticed yet. Part three will be on play loops.

(*) I am appropriating the concept of piping here, and translating it from a purely technical term to whatever it is it becomes in this theory. It might be imprecise and technical readers will probably pull their hair and raise their fists, form a mob and DDoS me. However, my appropriation of the concept builds on that technical definition of piping, that is, it is not merely a descriptive concept but it can also be an actionable one: piping can be used to describe a process, but it should only describe those processes that we can effectively implement. My use of piping only works for understanding how to make changes/interventions in the flow of IO in a computer. Any other use will go beyond the boundaries of the use I am making of the concept.

[Note: I understand I am being weirdly optimistic, but read me right: I am not saying that thanks to APIs we can play in the world and make it ours. I’m saying that despite APIs, despite the ways in which we are being marketed and mined as just another Thing in the Internet of Things, we can resist. Resist by misinterpreting, by appropriating what’s given to us but also what we are made of. It would be easy to be glib, to complain about the rhetorical excesses of the new data era, to tweet and write ACM or Medium articles about it. But I want to propose alternative ways of thinking, forms of dissent I think can be productive, in the hands of the right people. Because, paraphrasing Brian Sutton-Smith, play is the opposite of conformity.]

Leave a Reply

Your email address will not be published.