Garden of Machines
17 April 2015 - 22 August 2015
The Stack and the Post-human User: an Interview with Benjamin Bratton
Benjamin Bratton is a theorist whose work spans Philosophy, Art and Design. He is Associate Professor of Visual Arts and Director of The Center for Design and Geopolitics at the University of California, San Diego. He is also Professor of Digital Design at The European Graduate School in Saas-Fee, Switzerland. His research is situated at the intersections of political & social theory, emerging computational media & infrastructure, and interdisciplinary design methodologies. In his book _The Stack_, due to be published in December 2015 by MIT Press, Benjamin Bratton develops a political perspective on computation on the planetary scale. He proposes that we view the vast array of recent and upcoming digital and digitally supported technologies as aspects of a single phenomenon that he names 'The Stack', referring to the layered architecture of certain software-hardware systems. The Stack, he argues, has the characteristics of a platform. Bratton visited the Netherlands to deliver a keynote speech at the 2015 Sonic Acts festival, the theme of which was the Anthropocene, the current geological age in which humans have become the primary source of geological and biological change. At Sonic Acts he focussed on one of the constituting layers of the Stack: the Earth layer. Bratton was invited to the Netherlands with the support of Het Nieuwe Instituut's International Visitors Programme. Klaas Kuitenbrouwer, e-culture expert at the Institute, interviewed Bratton during his visit, as part of the research for the Garden of Machines exhibition. Bratton will be delivering this year's Premsela Lecture for Het Nieuwe Instituut.
Benjamin Bratton interviewed by Klaas Kuitenbrouwer, curator of Garden of Machines, 29 February 2015
What is the Stack? Does it already exist or is it a speculative notion?
It is both. It is the name for the accidental megastructure of planetary scale computation and it is a way of seeing that megastructure as a single meta-technology. Instead of seeing various kinds of computation, like smart grids, cloud computation, next generation interfaces, addressing systems, robotic artificial intelligence as a bunch of different species of computation, all moving on their own, we can understand them as forming part of a large composite armature that is in form and content not so unlike the software-hardware stacks (like OSI or TCP-IP). There are modular layers in the Stack, each of which has a specific role to play, and each is both independent and interdependent in relation to the other layers. I think there is value in thinking from this totality, and that this would allow for the possibility of designing and re-designing this infrastructure.
This is not only an issue of engineering, because computation has become the form and the content through which we organise our ideas, economies and cultures. It is not only a way in which governance operates, but it has become governance itself.
If the Stack is not designed, but has emerged, and it is - as you claim- governance itself, how is it 'designable'?
The fact that is has arrived unintentionally, doesn't mean the way we have to respond to it has to be equally accidental. Also the Stack didn't just fall from Mars, it followed from a set of initiatives that appear maybe a bit more coordinated than we anticipated that they were. Approaching it as a whole allows us to be a bit more deliberate about its future.
If it is this vast, enormously complex entity, and if it is governance itself, does this suggest that it is un-overseeable? This seems to imply that we are inside it, rather than looking at it from the outside. How does that allow for a design or engineering approach?
The way we can go into this, the way in which design can work in relationship to it from its inside (because we are inside of it) is by conceptualizing it through an image of totality, but not necessarily designing the entire thing at once. You don't have to boil down the ocean in order to make a move. Part of the rationale of stack systems in general is that anything that operates on one layer can be replaced by a completely different mechanism as long as it communicates protocollogically with the layer above and below it. Stack systems are in and of themselves designed to be re-designed. It is like the Ship of Theseus paradox. Every piece of wood or metal is replaced over time on this ship, and yet the ship remains the same ship, it occupies the same space, metaphysically. This is also the way stack systems work. Over the next decades almost every component of this mechanism may be replaced, and yet the diagram remains.
Part of the original impetus for the project of my book is to look at the way planetary-scale computation has deformed the political geography of the West-Phalian state system, the coherency of the modern nation state. So instead of understanding the plane upon which the subdivision of the earth would work as a kind of horizontal layer, subdivided into loops on land, in fact what we see is that the real contestations of sovereign claims is one on top of the another. It is an issue of multiple sovereignties, some private, some public, some state, some platform, some market. All claiming to be able to address something according to its own sovereign logic, superimposed over each other. And this super-imposition implies certain verticality. We need to find a conceptual model through which that verticality and that layering can be understood. Not as the exception to the rule, not a kind of pirate-space, but rather as the new norm. We need it to be able to deal with it as the new norm, without being forever scandalized by the way in which it works. We need to energize the frictions that these layers on top of one another actually provide. The Stack functions not only as a piece of technology, but also as a piece of geography: as a way of mapping jurisdictions, sovereignties.
Does this mean that the West-Phalian state is no longer the norm? Or is it part of the new norm?
This is not a story about the disappearing state. Quite the contrary. Just as cloud platforms (Google, Facebook and others) begin to take on more and more services and functions that were previously provided by the state: cartography, currency, legal identity, even leading to brand patriotism, etc., as the cloud platforms begin to take on roles of the state, the states themselves are evolving into cloud-platforms.
Where is the friction then?
These transpositions are not always symmetrical. They give rise to multiple forms of friction. We can see this for instance in the Google-China conflict. We can see it in the way we don't know how to make sense of the new surveillance capacities of states, and us being scandalized by such things. The argument can be made that the history of states is a history of what it is that states are technologically capable of seeing at any point in historical time. The cloud allows the state to see things it had previously not been able to see. And it shouldn't be unexpected that it went ahead and filled that space. So when we look at the role of the cloud in China or in the United States or in any other place, far from the situation in which the state is somehow being vaporized, it is in many ways refortified, but in ways that make it something quite different than what it was before.
So, parts of the Stack can be redesigned, they can be made to do new things, to do things they didn't do before. But somehow the system is stable: no matter what you replace it with it will always remain the Stack?
Yes, you don't have to do a whole sale replacement in order for it to remain in motion. What we see is that at any point in time lots of different things occupy any one layer. We can see the way different platforms could each have not only a different business model, but also a different political model for what it is the cloud polis might be. Facebook models the cloud polis on the reification of the self-display of the subject. Google models the polis on the rationalisation of information structures and algorithmic reason. Apple models the polis on this benevolent dictatorship of closed, smooth surfaces. Amazon models the cloud polis on the amplification and acceleration of molecular flux - it is about object as the object of political knowledge, not people. Each of these are not only different business models, they are different political models, and they all occupy the cloud layer simultaneously. Each of these cloud platforms are optimizing against one another, but they each make use of this larger apparatus, to access city layer, address layer, user layer, that in and of themselves are not their domain. It is not the things they have control over. But the reason they can work on their own layer is because the other layers are already provided for them.
You've elaborated on the cloud layer and the various actors that occupy that layer. On 28 February 2015 at Sonic Acts you talked mainly about the Earth layer. Could you briefly expose the layers of the Stack again?
Earth layer, cloud layer, city layer, address layer, interface layer, user layer . You could have seven, you could have nine, this is a schematic. The Stack is an enormously hungry machine, its carbon appetite, its energy appetite; its mineral appetite is vast. Understood as a single massively distributed machine, it is the hungriest machine we have ever made. The earth layer is where this extraction, a terraforming of the earth into the needs of this machine takes place. The whole thing is operating perched upon it and fed by its relation the earth. This is not only the source in terms of what it is composed at the chemical level, it is also an object of its primary objects of knowledge. The earth itself is something measured and modelled and rationalized and surveyed and visualized and governed. This is one of the first tasks of the Stack, which introduces a whole lot of paradoxical and ironic situations. Not the least of which is climate change: on the one hand, taken as a whole, this huge technological transformation of the earth over the last one hundred and fifty years has put us on the anthropocenic precipice of a sixth great extinction. At the same time it is through this apparatus of satellite systems, networked sensors, oceanic measurements, that we even know this is happening in the first place. The capacity to measure and model and to even have a sense of the reality of climate transformation is because we have the instruments through which this is articulated. It's a snake eating its own tail. That's one of the paradoxes of the earth layer.
One other thing I deal with in this chapter is some of the ways in which this produces contested jurisdictions and contested sovereignties. Historically you can argue that new jurisdictions arrive in response to emergency situations, or non-normal, exceptional situations at least. If you look at something like an association of island states, all member countries are near the equator. Not all of them are islands, but their common predicament in relation to climate change is rising sea levels. The fact that they have this common relationship to the climatic emergency makes them a common political actor in this form. In the same way you have coalitions between Texas, Russia and Norway, the oil producing countries, which are essentially jurisdictions of the causal agents. Around these geologic transformations that happen at the earth layer, there are these re-alignments of interest, re-orientations of different kinds of federations.
This is again a question about the extent to which we should read _The Stack_ as a metaphor, or a diagram or model, or an actual machine. You seem to be describing something akin to an informational ecology, not just as a metaphor, but in the sense of the way ecology works in biology, with an interdependence of various systems all occupying different niches, processing inputs into outputs, maintaining their own stability in relation to the stabilities of other systems. I don't think an ecology has ever been described as a Stack, but the notion of the food chain seems to bear some resemblance.
I think ecologies are generally far more complicated than stack systems. Part of the way stacks systems work is in this highly reductive, diagrammatic format, and it is the reductiveness of this that allows it to be very flexible in terms of what we would design. The nested homeostasis inside a flux that you see in ecological systems is a different model and a different diagram.
So the Stack is not a cybernetic idea?
Not primarily, but there are ways in which you can read some of the questions that arise from the Stack idea with cybernetic vocabularies that may be useful. There are ways in which technologies that have begun as technologies for observation or measurement or surveillance, have become technologies of design or transformation or governance. Some things that were sensors have become effectors, in cybernetic vocabulary. We see this happening all the time: the transposition of something that was a sensor into an effector or a map into an instrument. This is one of the key logics of the interface layer. One way to read the history of the graphical user interface with terms from the history of the diagram. With diagrams there is this reductive synthesis of information into a set of spatial and conditional relationships. It is a way in which an event or system or process is rendered and synthesized into an image. This chain of signification works one way; the diagram is a kind of sensor. A diagram is a way in which something that is happening in the world is captured in a model. 'Interface' inverts the chain of signification: in manipulating the interfacial diagram, you can affect the thing the diagram is representing. It may be possible to identify the circumstances in which this could give rise to a nested homeostasis. This is not something I deal with in the book, but it could be done.
However, I would caution against reading The Stack as a kind of systems theory. The best way to understand the Stack is as a kind of platform. Not all platforms are stacks, but all stacks are platforms. Platforms we need to understand not only as a technical model but as an institutional model as well. There is some literature on platforms in organizational literature and computer science literature, but there is no precise working definition of what a platform is. The are provisional definitions, but nothing that really explains why it is that Google is a platform and an urban street grid is a platform too, and the tetrahedral body plan is a platform? What reins this all together? In the book I try to draw out some of the generic properties of platforms. One is that a platform is a set of mechanisms put in place that allows for the undirected use of these mechanisms to produce emergent effects, which are not intended or planned in advance with the initiation of the mechanisms. A platform is a set of strategic potentialities. The way one is set up in advance will contain the phase space of possibilities in a certain way. In the same way that Google will organise the logic of the Internet around information rationalisation, it doesn't make the content of the Internet, it makes the mechanism by which everybody else makes the content of the Internet. And in doing so I think there are two forms of platform value that are produced by this. There is a platform user value, by which the information that I put on the platform is more useful to me because it is on this platform in such a way that it can be circulated. Thereby there is an increase in the opportunity costs of using any other platform and a decrease in costs in continuing to use the one platform. And there is also a platform surplus value and that is that the platform extracts more value of the information that is put there as well. There is much more to say about platforms, but one of the side effects, I hope, of using this model is a more specific and empirical theory of platforms as such. There is a lot of confusion about how we might govern planetary scale computation that is unnecessarily brought about by understanding it through a binary matrix of states versus markets, or public versus private. Like in the example: 'How is it that that the state should govern these private actors?' I think by understanding the way states and these markets function both as platforms in a certain way, the real questions (which are not easier) of what the political status of the user is in relation to the platform can at least be configured. States have citizens, markets have consumers, and platforms have users. And what the ethical responsibility to and from users and platforms is, is something we don't have a good language for.
Would you say that the market and the state operate as platforms?
They can, but not necessarily. There are ways in which markets operate in different kinds of things that are not platforms, states as well What I'm trying to point at it that they do in certain ways, and this deserves our attention. It would seem that the interface is very much the place where the politics of a platform are shaped. The interface structures structuring the potentiality of the platform, towards the platform itself, and towards the user.
Let's take the GUI [graphical user interface] as a generic starting point. In order for any interface to be useful, to be operational, it has to be reduced into a set of things that allow participation to be at the scale of the gesture, as opposed to conceptualising the whole. This is inevitably a kind of ideological reduction of what those possibilities are. The extent to which we become 'culturated' into a particular space of that reduction is what I call an interfacial regime. There may be multiple interfacial regimes that we come in contact with over the course of the day, each of which is describing the rest of the whole to us in a particular way. They have a narrative logic, they have a value proposition, but they are also tools by which those logics are instantiated. It's a value system and when you use it, it materially reinforces itself. But the interface is also the way in which the rest of the Stack sees the user.
In your Stack model, the user can be human, but doesn't have to be. What is the user in the Stack?
When we say user-centred design, we focus on human users. Now, anything that can initiate what I call a column, that can activate all the layers of the Stack, can be a user. It is worth thinking about the layers as actually working sequentially, when one message is sent from one user to another user. The user sends a message to the interface layer, the address layer, all the way down to the earth layer, and then to the other user back up through all the layers of the Stack. It is important to keep in mind that in the simple movement of one message of a user to a user, the entire apparatus is invoked, the whole at once. But anything that initiates these columns is a user. That could be a person, but it could also be a high-speed trading algorithm, it could be an animal, vegetable, mineral, driverless car, whatever you want it to be. We see most of the traffic on the Internet is already non-human. This co-participation within the space of the Stack, with other machinic, animalian, algorithmic co-inhabitants will be one of the more difficult philosophical challenges for us to deal with. This is not radical cosmopolitanism, or the deeply universal suffrage in the Latourian sense of a Parliament of Things, because it's not a parliament, it's a machine. It is not a philosophical recognition; it's a mechanical co-participation. It's not about transference of sovereign will through mechanisms of representation by which some sort of transparent majority outcome is what steers it as a whole - that is not the mechanism. But to deal with the highly contingent status of the human user towards other users will be a big problem. I actually foresee a whole range of different varieties of humanist fundamentalism over the course of the next decade, pushing back against this.
This is one way in which understanding the platform as a third kind of organisational principle next to the market and next to the state becomes very urgent because otherwise we can only see users as either consumers or citizens and they could be very different things as well.
Exactly. And the kind of discourses we have about the status of the user, from Wikileaks etc., are trying to counter-weaponise the atomic, anomic individual in relationship to the state apparatus. But I think the potentiality for the user as a subject position is not about the individual withdrawal from this thing and a kind of Second Amendment weaponisation through encryption and privacy, but rather through the multiplication and pluralisation of user positions.
One of the real problems of interface design is that is supposes that an interface is used by one user at a time. It individuates and interpolates people as individual users, but in fact we are collaborating through these mechanisms in different ways and if we could understand how it is that we would design these interfaces for these distributed collaborative user positions we would have something going there. The example I use is proxy-users. When you have a proxy user system, the user is in one location, but as far as the interface is concerned it thinks it is coming from somewhere else. The user is not identical to the person. You could also have two users that are actually one person. It almost doesn't matter at one point whether you know which composite user you might be participating in at a particular time, any more than it is necessary for the bacteria in your body to know what your driver's license number is.
The question of how we will even define the parameters and delineate the forms of the different kinds of user positions that any of us move in and out of over the course of the day, is interesting. In other words: when we can conceptually separate the idea of the agency and political rights responsibilities of the user from those of the individual human organism, and no longer understand these as isomorphically mapping on one another, we can have a much better conversation. Not because that would be 'good', but because it would help us to get a handle on where we are right now, at the very least.
To a lot of people human agency is still the only real source of agency in existence, although many begin to get used to the idea of machinic agency: there is artificial intelligence, there are robots.
But those are still projected as virtual types of agency versus real (human) kinds of agency.
The user layer in the Stack is open to many different kinds of agency. Could you elaborate on non-human, and non-machinic kinds of agency in the user layer of the Stack? Is it conceivable that for instance icebergs, clouds, animals or plants develop a kind of user agency beyond just being 'read'?
Yes, I think it is possible to imagine this in a number of different ways. Just as there is no clear difference between a human user and a machinic user, or a physical machinic user and an algorithmical machinic user, in the same way there is not necessarily a clear-cut separation between the vegetable phylum and other types of users. Those kinds of cyborgian composites can operate with plants and bees; rivers and icebergs just as well as they can with us. There are a couple ways in which you can think about this. One is that the forms of communication that take place in a complex localized environment are already much more dense and rich than those we have the ability to model right now. Everything is talking to everything else constantly.
So, it wouldn't be a matter of introducing information flows as such into the situation. But you could prosthethisize some of these information flows, so that something operating in one position could sense or activate or participate with information in some way that is different than its normal capacity might be. This could change the nested layers of homeostasis that are happening in a situation. So yes, the prosthetization, cyborgisation of plants, flowers and bees is where we could see such a thing. Geo engineering at a biotechnological level. For another way to think of the design space related to this it is worthwhile to go to the address layer. Anything that has the capacity to participate in this system has to be identifiable, has to have a place on the map and a position from which to speak. Now, with the Internet this largely works through the IPv4 addressing system. This has worked pretty well, but had to build a couple of work-arounds on it, because its total theoretical address space is about 4 billion addresses, less then one per human on the globe. So in a real, robust, ubiquitous computing environment it can't really scale. IPv6 is the new iteration, which has a 128 bit address space, so if you do the math and divide that by 7 billion people, it comes down to roughly 10 to the power 23 addresses per person. That is Avogadro's number in addresses per person.
So, if you wanted to go through your entire life and distribute an IP address to every single thing you ever have been in contact with, it would never exhaust. You're nearly down to the level of molecules in the things you come in contact with. But if you subdivide that number by a larger number of users, not 7 billion humans, but when you're down to individual daffodils and bees and insects having an allocation of addresses to a certain degree - then you would get a conception of the site condition by which such a scenario (with non-human users) could be designed.
The internet-of-things scenario at this moment is still very much about how the toothbrush talks to the refrigerator. It is very much about things that within our natural sense of what an object scale would be. But when you have this deep addressability, it is possible to imagine things at very different physical scales being able to exchange real information with one another, in ways they haven't been able to before. It's not just that there would be a conceptual relation between these objects; there would also be an actual exchange of information possible.
Something in the scale of 10 to the minus 9 meter can exchange information with something in the scale of 10 to the 9th meters. What do they say to each other? We don't really know. It's this idea of a chain of information exchange between things at scales that previously were highly mediated at the very least or otherwise totally incommunicable to one another, that allows for this kind of alien communication that is enormously exciting.
There are ways in which you can understand all of these natural processes as somewhat algorithmic, and in this way you might come to think of the Stack as a secondary planetary computer, on top of the first one. Maybe this relates to more of an interlacing of this synthetic computation and natural computation that is already at work.
The Stack as a way to communicate across scales doesn't happen in organic ecology. If this happens it will happen through the interface layer. At the moment interfaces are mainly human, partly machinic artefacts, intended for humans and machines. In this scenario, what would be the interfaces by which non-human organic users would access it?
If the interface is first of all a map of the space of possibilities, instrumentalised in such a way that when it is manipulated, it can essentially do whatever it is representing. What does that mean for a bee or a tree or a forest or an aardvark? Each of these creatures is already sensing the world around them one way or another - even if it is just the tick on a branch that is sensing heat - there is some way in which the world - part of the world - is made sensible to them.
This is exactly the right question, the question that follows from the implication of the Stack. What would be the vernacular in interfaces that we might employ on behalf of these non-human users to introduce them into this system, and what would be the conditions by which they would be able to create, propose, optimize another regime of interfaces for one another that humans don't necessarily need to participate in?
The next thing to do would be to develop an archaeology of what would be the likely precedents for them - if we're able to recognise them, because we might not even be able to do that.
There is a parallel in which the non-human user is discussed here and the way it is discussed in Object Oriented Ontology and Speculative Realism, against the backdrop of the Anthropocene and post-humanism. But I would want to make an important differentiation here: what OOO and SR can contribute to describing this situation is as true now as it was a century ago or 2015 BC - theirs is an ontological position.
But I'm not making an ontological argument here. I'm making an argument about a technological mechanism and infrastructure that would produce a condition for instrumental relationships in which these things are co-participating in a very specific way which would be impossible 20 years ago, let alone 2000 years ago.
So let's not conflate the functioning of a particular apparatus with an ontological argument that may be as idiosyncratic or universal as you may find it to be.
At Sonic Acts and on other occasions, you came up with the term 'the post-Anthropocene'. It seems to refer to the user-layer in the Stack?
Could be. To the condition under which a post-Anthropocene might emerge&
Do you mean a period in which humans are extinct?
Not necessarily, no. Someone said yesterday at the conference: 'There is no escape from the Anthropocene.' My response is: Well, geological eras come and go... remember your 4th grade map of the Devonian and the Cretaceous? Of course there is an escape from it. Just wait long enough. If we think of the predicament of the Anthropocene, as much as it is possible to do so, maybe stand back for a moment and think about the terms by which at that level of humans as a geologic actor, an alternative is possible. If anything the Anthropocene has proved that geo-engineering is possible.
In a short period of time we have geo-engineered the entire planet, very quickly. That it is possible to do so, doesn't mean it is possible to do so deliberately, but anyway, whether we want to or not or whether we take it on as a deliberate trajectory now, we will be geo-engineers at least for the foreseeable future. Whatever we do or don't do at this point will have effects at geo-engineering level. So it makes sense to begin to think about the site-conditions by which our design interventions work within this question. The only thing we can know about the post-Anthropocene is that it is defined by the condition that humans are no longer the dominant geologic actor. That means maybe we're extinct, maybe we become something else that is not recognizably human. Our DNA persists, but the human as a species diagram doesn't. We don't know. But I use the term in the sense of 'This is ultimately what we want to accomplish'. The post-anthropocene should be seen as the accomplishment. A sort of vague place over the mountain.
So, you don't mean the post-Anthropocene as either utopian, or dystopian, or in that specific way value-charged?
It is an anti-Apocalyptic notion. There is a similarity with the way in which the Anthopocene is invoked in some kinds of eschatological millenarian narratives. For some people that is what they like about the Anthropocene. The apocalyptic, end-of-the-world stories are part of the texture of their moral community. I find these essentially reactionary, deeply conservative dispositions. A way to counter the narrative of the apocalypse is to simply name the time that comes after it, as something that is going to happen whether or not you're there. I'm erasing their apocalypse, so they have to keep on going.
So the post-Anthropocene is a kind of rhetorical position.
As a concept it has more of a role to play in the book I'm working on now, on AI, robotics, synthetic biology and these sorts of things as well. Because all of those things put the status of the human in question, as a geologic, biotechnologic, bio-political actor, in ways in which The Stack isn't quite making that point. The Stack is not really making an argument at the level of millennia. It is not about human's relation to philosophy, it is a work of political philosophy.
I mentioned the post-Anthropocene also because certainly among Californians there is a big belief in The Singularity as another kind of eschatological story.
Equally conservative, equally reactionary.
But that's not at all the narrative that you work with. Could you elaborate on that?
I make a pretty strong distinction between transhumanism and post-humanism. It has to do with the way in which each employs the figure of the human. In the simplest sense, transhumanism is a project about a reification and amplification of the individual humans as self-encapsulated and self-mastered organisms that can innovate and expand themselves as individuals. The X-Men reading of Atlas Shrugged. Whereas I see post-humanism as the space of potential that becomes clear when the contemporary Copernican traumas of the anthropocenic predicament with the contemporary technologies like nanotech and biotech are understood as challenges to that conception of the human the transhumanists are fetishizing, and become pre-conditions for an alternative epistemology that would locate the specificity of human sentience in a different context. I see them really as projects working in totally opposite directions.
In short: transhumanism is excessively anthropocentric, and post humanism isn't.
Yes, post humanism would be deeply suspicious of anthropocentric starting points, whereas with transhumanism it is the whole point of it. While it is absolutely true that human thinking and one's own neuro-cognitive state are ultimately reducible to physics, that doesn't mean it is simple to put it onto a silicon substrate, as the transhumanists love to think. I think we can pre-suppose that the last thing to survive the transposition of the medium of embodiment of a neuro-identity network would be any coherent sense of ego and identity that existed in the biological body. It may be a very good project to try it, but because it would annihilate the previous logic of identity, not because it would amplify it.
How does the post-human map on the user in the Stack?
The position of the user in the Stack forces a kind of post-human epistemological framework to be deployed, because we have to make sense of our correspondence with these non-human actors that this mechanism we have produced has put us into. We're now cousins and neighbours with them in ways we previously weren't. The user is a technological relationship, one that requires the epistemological framework of the post-human. Non-human users don't necessarily map on the figure of the post-human, but both are cyborgian in nature. These are all interwoven; it is no longer animal versus vegetable versus machinic creatures.