Virtual realities, the simulacrum and electric sheep

Where is the divide between reality and representation?

As Gilles Deleuze notes in his 1983 work Plato and the Simulacrum, the power of the simulation is not in its opposition or otherness, but in its almost imperceptible similarity to reality.


A radical French thinker, Deleuze’s philosophy centres upon concepts of multiplicity and difference. Writing with partner and psychoanalyst Felix Guattari, Deleuze championed a rhizomatic approach to knowledge and understanding; allowing for multiple, non hierarchal ideas to coexist and feed into one another. His writing rejects a dualistic or binary understanding of the world, claiming that the “rhizome has no beginning or end; it is always in the middle, between things, interbeing, intermezzo”— it is perhaps this ambiguity of origin which endows much of popular culture’s rendering of AI, VR and the cyborg body with an innate sense of dystopia.

Rhizome, or an “image of thought”, is a term Deleuze and Guattari use to describe theory and research that allows for multiple, non-hierarchical entry and exit points in data representation and interpretation. It presents history and culture as a wide array of attractions and influences with no specific origin.

Ridley Scott’s Blade Runner (1982) follows the story of Rick Deckard, a bounty hunter, or blade runner, hired to retire (a dehumanised translation for kill) replicant cyborg beings. These cyborgs are designed to be “more human than human”, albeit lacking in emotional depth and sense of empathy, and in theory distinct from humans. However, when he encounters an experimental replicant named Rachael — who has been implanted with false memories, and thus believes herself to be human — the border between human and machine becomes hazy. The film never clarifies whether Deckard is himself a replicant, an ambivalence which problematises this grey area between appearances and truth.

Take Philip K. Dick’s Do Androids Dream of Electric Sheep? (1968), a loose basis for Scott’s film; the opening of the novel documents a conversation between Rick Deckard and his wife Iran surrounding their Penfield mood organ, a device used to regulate the emotions of users. Iran tells Rick that her schedule lists “a six hour self-accusatory depression,” much to her husband’s despair. She informs him that she was tinkering with the machine one day when she stumbled across the setting for despair, and has scheduled it for twice a month — “I think that’s a reasonable amount of time to feel hopeless about everything… don’t you think?” There is almost a tongue in cheek approach to these slippages between reality and simulation — moments of dark humour can be found throughout the text, in which a point is made of such uncanny moments in which our characters break a fourth wall, and discuss a collision between reality and its simulation.

What draws my attention is an awareness of the constructed nature around them, and how the characters choose to inhabit the world in light of this knowledge. Deckard yearns for a real animal to replace his mechanical sheep, rather than tending to a “fake” he finds demoralising; he discloses to his neighbour of how he used to own a real sheep which died suddenly, and had a replica of him made without arousing suspicion; not the same, but “almost”. I have come to think of the novel’s title as an extension into this territory of the “almost,” a world in which androids count electric sheep to fall asleep, almost perfectly emulating human characteristics but slightly off kilter — a technology which parodies and simulates.

We’re told that we must fear the ability of these technologies to replicate, and to position them as the uncanny doppelgänger. A positioning of the virtual as a false image (distorted to the point of seeming real), attempts to strip the copy or the simulacrum of its power for productivity. But why are we conditioned to view that which is, arguably, made in our image with such hesitation and trepidation?

§The artist Terence Broad engages with such themes of replication, simulation and representation within his work Autoencoded (2016), a technologically manufactured replication of Scott’s Blade Runner. Broad writes extensively about his thinking and practice, sighting the mind/body split of philosopher René Decartes as a focal point of his artistic motivation.

Descartes assertion of ‘cogito ergo sum’/‘I think therefore I am’, hoped to reconcile a unification between a thinking mind and existence.

Using an autoencoder, Broad constructed the film by teaching a neural network to distinguish between data fed to it, to memorise and then relay from memory a string of images which present a hazy and dream like impression of Scott’s work. This method of feeding the network compressed data, and allowing it to reconstruct what it has “seen” perhaps attempts to endow the technology with a sense of agency. The artist presents the ability for such a virtual system which ‘perceives images but is not embodied within the environment that the images represent’ to exacerbate the assertion of the thinking mind as the root of reality. The film picks up on a disparity between the human and its representations, and both Broad’s work and the materialisation of A.I. represented within sci-fi such as Blade Runner become part of a process of humanising technology and giving rise to the possibility for a subjective machine.

This draws an interesting parallel with Warner Brother’s response to the piece, issuing a DMCA takedown notice for Broad’s work on Vimeo, on the grounds of copyright infringement —in other words: ‘Warner had just DMCA’d an artificial reconstruction of a film about artificial intelligence being indistinguishable from humans, because it couldn’t distinguish between the simulation and the real thing’. A key aspect of Scott’s film, at least philosophically, is the attempt to distinguish between the representations of the real and “reality” itself — and whether this is can ever be clear cut. The importance of distinguishing between real and fake, origin (primary) and offspring (secondary/copy), original and implanted resonate not only throughout science fiction dedicated to the future of humanity and cyborg bodies, but more broadly to discussions of the nature of the human and of ontology. Creating machines with a sense of human instability and irrationality is potentially the most challenging and yet exciting aspect of the android and of A.I.

Before taking his own life in November 1995, Deleuze penned an essay entitled The Actual and The Virtual — defining the virtual as a kind of potentiality, and of continuous multiplicities, we can begin to think of the virtual space perhaps not as an extension of the physical world, but as a arena for potentiality, and a space to work through, to confront and to challenge. And maybe the fear of the doppelgänger or the android stems from precisely this — that it forces us to confront ourselves, to rethink that which we believe we fully grasp, and the basis of all our understanding.


Perhaps the virtual challenges our grasp upon our own perception. Do our attempts to control simulation and mimicry enable a potentially more objective view of our domain?