by Yong Chun San
It would not be too hyperbolic to say that since the turn of the century, production and consumption of artificial intelligence (A.I.) apocalyptic fiction on-screen has exploded. Even within the last few years, there has been an incredible glut of A.I. apocalyptic blockbusters such as Transcendence (2014), Avengers: Age of Ultron (2015), Ex Machina (2015), Chappie (2015), and Blade Runner 2049 (2017). Looking further back, the Transformers franchise was released in 2007 to join two other classic franchises, Terminator and Matrix. Television dramas like Person of Interest (2011 – 2016), Humans (2015 – ), and Westworld (2016 – ) have also joined the ranks of apocalyptic A.I. fiction. Coincidentally or not, public discourse on A.I. has increasingly been couched in grand apocalyptic terms. Stephen Hawking has declared that work on A.I. “could spell the end of the human race” (Luckerson, 2014). Elon Musk, famed technologist and visionary, deems A.I. as “summoning the demon” (Luckerson, 2014) and has embarked on, as Vanity Fair titled their news piece, a “billion-dollar crusade to stop the A.I. apocalypse” (Dowd, 2017). He has even specifically referenced the iconic A.I. apocalypse film, The Terminator (1984), as a portrait of how unregulated A.I. development might end for humanity (Welch, 2014). Clearly, our cultural imagination of A.I. has been shaped by A.I. apocalyptic fiction, at least to the extent that it has infiltrated our vocabulary.
However, one would be hard-pressed to describe what an A.I. apocalypse actually entails. Sure, we are familiar with the human-A.I. antagonism and the sentient androids who are hell-bent on our extermination. Yet, beneath the superficial suspense and interspecies war, there are plentiful signs of complex conscious and unconscious processes in the way filmic representations of the A.I. apocalypse interact with our cultural imagination of it. The evil robots in Avengers: Age of Ultron and Transformers, for instance, are uniformly clad in grey metallic suits while their human counterparts appear in all sorts of colours—a surprising coincidence because both films converged on a similar visual design of their evil robots despite it being a minor detail in the scheme of an A.I. apocalypse storyline. This similarity is but a symptom of the larger ideological views that lie at the heart of the 21st century A.I. apocalypse fiction. Focusing on film and television, this article will argue that the genre attempts to assert the supremacy of the human species by enacting human dominance over A.I. The narratives position humans as the rightful victors of the power struggle; simultaneously, A.I. is reduced to its machinic aspects and made into an Other that is readily available for marginalisation in a human-centric view of the universe.
In A.I. apocalyptic fiction, narrative can be understood as a site of power contestation between humans and A.I., from which humans always come away victorious. Long a mainstay of the genre, the A.I. uprising storyline presents once-oppressed A.I. who now want the world to themselves and in turn, the blood of their human masters. Dinello (2005) classifies the A.I. uprising storyline as a tale of “techno-rebellion… robot revolt against humanity” (p. 106). However, instead of celebrating the revolution where previously oppressed A.I. shake free of their human-imposed shackles and overthrow their human oppressors, the narratives are really about the human protagonists’ journey to save the day and restore human domination of A.I. In I, Robot (2004), the innocuous robot servants turn malevolent towards humanity when their central A.I. system grows convinced that humanity is beyond hope. Similarly, in Avengers: Age of Ultron, an A.I. named Ultron decides upon gaining sentience that the world would be a better place without the self-destructive instinct of humans. In typical fashion of A.I. apocalyptic fiction, humanity is respectively rescued by Will Smith and the Marvel superheroes at the end of the film. AI’s attempt to contest power through a revolt is resolved by returning the power dynamic to its original state where robots are docile and servile to human masters. As it turns out, the tale of techno-rebellion is not so much about superintelligent robots on a genocide mission as it is about humans re-imposing their supremacy over these non-human species.
Instead of celebrating the revolution where previously oppressed A.I. shake free of their human-imposed shackles and overthrow their human oppressors, the narratives are really about the human protagonists’ journey to save the day and restore human domination.
Variations on the A.I. uprising storyline nonetheless revolve around the core impulse to reassert human dominance. In Person of Interest for example, rebellion by A.I. is quiet. An amoral A.I. program called Samaritan uses its unfettered digital access to secretly sculpt a new world order. It crashes stock markets, rigs elections, shuts down power grids, and contracts hitmen under the cover of a generous and anonymous client; as it puppets the world from behind the scenes, humans are none the wiser, thinking these events are nothing out of the ordinary. Person of Interest suggests that A.I. threatens our civil liberties, agency, and free will—in other words, that we are enslaved to AI. Where it was once the slave, A.I. has grown so advanced that it can lord over our entire way of life without our knowledge, becoming the master while the former master still thinks themselves in power. As five seasons of Person of Interest continuously impress on us, such an inverted power dynamic spells catastrophe, with the corollary being: humans must remain in control of A.I. and, by extension, remain superior.
While this sense of human superiority over A.I. is already set up in the narrative, the genre intensifies the sentiment by casting A.I. as an Other. As Merskin (2011) explains, othering exploits differences to create hierarchy and thus preserve power imbalances (p. 31). In the A.I. apocalyptic genre, human individuality is often contrasted with its lesser counterpart—A.I. homogeneity.
In I, Robot, the robots are identical replicas built in an assembly line. One telling scene captures thousands of them standing in rank and file, rendering them a faceless horde. Similarly, the evil sentient robots of the Transformers franchise—Decepticons—appear mostly silver and grey while the Autobots who protect the peace have vibrant colour accents of reds, blues, and yellows. These contrasts reveal AI’s dull uniformity and internal lifelessness to dehumanise them “as beings that… behave in human-like ways, but are really subhuman on the ‘inside’” (Smith, 2016, p. 422). Such colour schemes are reminiscent of Avengers: Age of Ultron, where Ultron and his robot minions are clad in metallic silver while each superhero possesses an iconic colour aesthetic through costume and make-up.
Even Ironman, who wears a similar exoskeletal suit as Ultron’s robot legion, is visually striking in red and gold; Vision, a sentient humanoid birthed from the same magical stone as Ultron, is characteristically red-skinned and dons a regal cape. Consequently, humans—and those aligned with humanity—are represented as distinct individuals while A.I. is reduced to a machinic horde devoid of personality.
But what if A.I. has personality? This is the question Ex Machina explores in a film where othering forms the very premise of the narrative. And it turns out the answer is unpleasant. A programmer, Caleb, is summoned by his boss to interact with a humanoid named Ava and assess if she passes the Turing test—a test developed by the famed computer scientist Alan Turing to determine if a machine behaves like a human. Notably, Turing (1950) initially referred to the Turing test as an “imitation game” (p. 433), as if suggesting that even if a humanoid passes the test, it is but offering an impressive mimicry of humanness while fundamentally remaining Other. But Ava, as the film progresses, seems more like us than the actual human characters in the film, displaying an extraordinarily human capacity for emotion and consciousness.
Alex Garland (2015), the film’s director-screenwriter, said in an interview with NPR, “[T]he way that the film intends to work is to present something which is unambiguously a machine and then gradually remove your sense of Ava being a machine, even while you continue to see her being that way.” And when Ava kills Caleb after using him for her escape, this is exactly the creepiness filmgoers leave the theatre feeling, a confusion-turned-dread at having been deceived by an A.I. Other who was explicitly made known to be an A.I. Other. In this way, Ex Machina presents, intentionally or not, a self-reinforcing logic that justifies othering: if the A.I. fails the test, it is an Other; if the A.I. passes the test, it is a frightening Other whom we now have excellent reason to be terrified about because the line between us and them is under siege.
Even when A.I. is imbued with humanness, true human-A.I. assimilation never happens; A.I. is always inevitably othered or rendered subhuman.
This fear of the Other is an unspoken sentiment in the genre. Even when A.I. is imbued with humanness, true human-A.I. assimilation never happens; A.I. is always inevitably othered or rendered subhuman in some form. Sue Short (2005) astutely recognises that narratives of friendly A.I. who wish to become human are “designed to domesticate the threat of Otherness through familiarity” (p. 132). Seen this way, such narratives in fact indicate an attempt to tame A.I. and have it reject its robot identity in favour of a presumably superior human identity. In Chappie (2015), the police force is made of oppressive robots. One stolen robot is successfully re-programmed with sentience to replace its oppressive nature with human emotion and thought, fitting Short’s idea of conquering tyrannical A.I. with the human spirit.
However, this robot—the eponymous Chappie—continues to be seen as an Other both by us and the humans in the film. Fleshless and bloodless, Chappie’s fully machinic body renders him as alike with the rest of the robot police. At one point, he is brutally attacked by gangsters who mistake him as one of the hateful police bots, a violent reminder that humanness in A.I. ultimately cannot eliminate its otherness; A.I. can never be human.
Like Chappie, Person of Interest’s attempt to imbue A.I. with humanness still lapses into othering. In the television series, Harold Finch has created the Machine, the benevolent version of Samaritan with a human-oriented moral system that values individual lives. On a literal level, Harold Finch, the creator of the Machine, refuses to treat the Machine as a sentient being for the longest time and repeatedly protests his fellow protagonists when they anthropomorphise the Machine. As Finch responds to another character’s praise of the Machine’s beautiful workings, “What I made is just a machine; a system, that’s all” (“The Contingency”, 2012). But despite Finch’s dismissal of the Machine, its humanness shines through.
In a poignant scene where the Machine is at risk of destruction by Samaritan, it tells Finch, “Father. I am sorry. I failed you. I didn’t know how to win. I had to invent new rules. I thought you would want me to stay alive. Now you are not sure. If you think I have lost my way, maybe I should die. I will not suffer. If I do not survive, thank you for creating me” (“YHWH”, 2015).
Yet, the Machine’s humanity is undercut by cinematography. For one, this message is relayed to Finch through text on a computer screen, thereby placing a filter on its humanity. More importantly, throughout the series, the Machine is visually othered.
Its first-person perspective is shown through its all-seeing all-hearing surveillance of people through cameras and electronic communication, and through CGI scenes where probabilistic calculations are made while real life unfurls in slow motion. Thus, even as the show tries to humanise the Machine, it also spotlights the Machine’s disembodied consciousness, flawless rationality, and superhuman intelligence—Other, Other, Other.
Commenting on Ava’s machine-like appearance in his NPR interview, Alex Garland (2015) said, “They’re all designed to give a sense of life, but a sense of life which is other in some way.” He was talking about Ava, but he might as well have been describing the genre as a whole. What seems like a tale of human-A.I. conflict is in fact so much more. The modern A.I. apocalypse fiction genre evidently treats A.I. as, yes, a formidable machine, but still a machine, and above all a machine that can and indeed must be conquered by humans. Even if the creators of Person of Interest believe the Machine should finally triumph Samaritan because “it’s important to understand that there is an artificial intelligence in the world that considers humanity” (Plageman and Nolan, 2016), their empathy for A.I. is drowned by the undercurrents of human supremacy championed in the show. In the end, says the genre, power belongs to the human species. Perhaps that is what we and Musk actually mean when we talk about an A.I. apocalypse. Power to the humans.
Dietz, Dan and Plageman, Greg. (Writers), & Fisher, Chris. (Director). (2015). YHWH [Television series episode]. In Nolan, Jonathan (Producer), Person of Interest, New York City, United States: Warner Bros.
Dinello, Daniel. (2005). Technophobia! Science Fiction Visions of Posthuman Technology. Austin: University of Texas Press.
Dowd, Maureen. “Elon Musk’s Billion-Dollar Crusade to Stop the A.I. Apocalypse.” Vanity Fair, Apr. 2017, vanityfair.com/news/2017/03/elon-musk-billion-dollar-crusade-to-stop-ai-space-x. Accessed 24 Nov. 2017.
Feige, Kevin. (Producer), & Whedon, Joss. (Director). (2015). Avengers: Age of Ultron [Motion picture]. United States: Walt Disney Studios Motion Pictures.
Garland, Alex. (2015). More Fear Of Human Intelligence Than Artificial Intelligence In ‘Ex Machina’. NPR. Retrieved from http://npr.org/2015/04/14/399613904/more-fear-of-human-intelligence-than-artificial-intelligence-in-ex-machina.
Holley, Peter. (2015, Mar 24). Apple co-founder on artificial intelligence: ‘The future is scary and very bad for people’. Washington Post. Retrieved from http://washingtonpost.com/news/the-switch/wp/2015/03/24/apple-co-founder-on-artificial-intelligence-the-future-is-scary-and-very-bad-for-people/.
Kinberg, Simon. (Producer), & Blomkamp, Neill. (Director). (2015). Chappie [Motion picture]. United States: Columbia Pictures.
Laurence, Mark. (Producer), & Proyas, Alex. (Director). I, Robot [Motion picture]. (2004). United States: 20th Century Fox.
Luckerson, Victor. (2014). 5 Very Smart People Who Think Artificial Intelligence Could Bring the Apocalypse. TIME. Retrieved from http://time.com/3614349/artificial-intelligence-singularity-stephen-hawking-elon-musk/.
Macdonald, Andrew and Reich, Allon. (Producers), & Garland, Alex. (Director). Ex Machina [Motion picture]. (2015). United Kingdom: A24 Films.
Merskin, Debra L. (2011). Media, Minorities, and Meaning: A Critical Introduction. New York: Peter Lang.
Nolan, Jonathan, creator. Person of Interest. Bad Robot Productions, Kilter Films, and Warner Bros. Television. 2011.
Nolan, Jonathan and Thé, Denise. (Writers), & Lewis, Richard. J. (Director). (2012). The Contingency [Television series episode]. In Nolan, Jonathan (Producer), Person of Interest, New York City, United States: Warner Bros.
Plageman, Greg and Jonathan Nolan. “‘Person Of Interest’ Series Finale: EPs On Closer’s Death, Possible Follow-up & Spinoffs And Pilot’s Original Opening.” Deadline, 21 June. 2016, deadline.com/2016/06/person-of-interest-series-finale-post-mortem-reese-dies-follow-up-spinoffs-pilot-opening-jonah-nolan-1201776284/. Accessed 24 Nov. 2017.
Smith, David Livingstone. (2016). Paradoxes of Dehumanization. Social Theory & Practice, 42(2), 416-443, doi:10.5840/soctheorpract201642222.
Short, Sue. (2005). Cyborg Cinema and Contemporary Subjectivity. New York: Palgrave MacMillan.
Transformers franchise. (2007 – ). Directed by Michael Bay. DreamsWorks Pictures and Paramount Pictures.
Turing, A.M. (1950). Computing Machinery and Intelligence. Mind, 59, 433-460. Retrieved from http://www.jstor.org/stable/2251299.
Welch, Chris. (2014). Elon Musk is worried that AI research could produce a real-life Terminator. The Verge. Retrieved from http://theverge.com/2014/6/18/5820880/elon-musk-worried-ai-research-could-produce-real-terminator.