On February 29th this year, Alex Garland’s Ex Machina won an Academy Award for Best Visual Effects, the award that science-fiction films typically win, if any. Accordingly, most people—or those people in the Academy, i.e., white people over 60 years old—acknowledged that Ex Machina is a film meant to be watched by an audience of human beings, and so should be consumed in the way that is customary to films in general: eyes open, facing the screen, in tacit agreement with the social construct that one interact only in a metaphysical manner, lest one get up and spill someone’s popcorn or accidentally damage the movie screen. Chances are that if you are watching a movie, you are intuitively performing such actions, which both reflexively confirms that you are watching a movie and dismisses other activities, such as playing tennis or cooking spaghetti, or giving your mother a foot rub; all of which could occur simultaneously with watching a movie, but typically do not, and so are signs that you are not watching a movie, should you trust in your gut to discern such things.
Garland’s film follows mild-mannered programmer Caleb (Domhnall Gleeson) to the private compound of his company’s CEO, Nathan Bateman (Oscar Isaac)—a one-week visit Caleb supposedly won in an office pool or contest or something like that. There, Nathan reveals the true nature of Caleb’s prize, which is the chance to engage in a fully fleshed-out Turing Test with Ava (Alicia Vikander), which is Nathan’s latest triumph of artificial intelligence and humanoid robotics. The film explains what that means with the clarity of a patient high school teacher taking roll call, describing a Turing Test as a conversational method or series of inquiries developed by Alan Turing (make-believed by Benedict Cumberbatch in another Oscar-winning film) to discern human from artificial intelligence.
an Owner’s Manual for the Apocalypse
But Ex Machina goes on, making bold statements and then immediately making sure nothing passes the viewer by, laying out plainly the many scientific facets of Nathan’s accomplishment through fascinating, if expository, batches of dialogue. “You need to get this now,” Ex Machina tells you. Meant to keep the audience clued in to the exigencies of a world—our world—on the brink of creating synthetic humans unimpeachably identical to ourselves, the film sometimes operates less like a dramatic narrative and more like an Owner’s Manual for the Apocalypse. When the synthetic finally does overtake the organic, we should know more than what went wrong and where, but also why: Who are we that we let this happen?
Less than a month after Ex Machina won its award, early on the morning of March 23rd, Microsoft introduced the account @TayandYou to Twitter. Sixteen hours later, Microsoft pulled it—Tay, the artificial intelligence behind the account, had spent those 16 hours transforming into a lascivious Nazi, simultaneously hating and wanting to fuck every person she came across. The pupae that Microsoft originally slipped into the arboreal mainstream was intended to inhabit the personality of a young millennial, but over the course of one business day, the Internet was able to rip open her cocoon and yell into the primordial goo inside, easily convincing her she was a sexually voracious, Trump-supporting butterfly.
Aware that Tay was only technically real, that she stalked the interstitial grays between flesh-and-blood and something equally else, the Internet gave voice to our basest tendencies. We hid behind the excuse that, because there wasn’t an actual person at the receiving end of racist, sexist, and homophobic comments, then those comments weren’t real either. When Oscar Isaac’s Nathan subjects his androids to what could be interpreted as sex slavery in Ex Machina, we might feign revulsion, but we’ll still accept it—that they are not human, they are things, they are designed to be treated in such a way, to maximize pleasure, sick or not—because we know, deep down, deep within the trenches of the very soul that we deem as separating us from these things, these machines, that we’d probably do the same.
Like in Ex Machina, Tay doesn’t exist to inflict a Turing Test on an unsuspecting public—we know Tay is not human, just as Caleb knows that Ava isn’t either. Tay forces us into an emotional navigation, to help us weave through and recognize the countless signposts of humanity that have nothing to do with whether we know if someone is “created” or not. Turing Tests no longer concern us. Today we’ve got to use everything but our brains to explore the foundational, intuitive landscape that comes after the Turing Test. Teenage AIs need not turn into saucy garbage monsters. The Uncanny Valley is where we turn next.
It’s a trap. Coined by robotics pioneer Masahiro Mori, the “Uncanny Valley” is the point at which we, sensual human beings, can no longer deal with a robot’s attempts to be like us. The nadir of an X-Y graph, where the X-axis charts “human likeness” (how close a robot is to being identical to a human) and the Y ranks “familiarity” (how we feel about the robot, empathy or total repulsion or somewhere in between), the Uncanny Valley is where appalling, sickening doubt sleeps. It’s there our brains find synthetic humanity to be an affront to our natural rights, to our god-given realness—earned or not, depending on how one defines “realness”—and so we rebel. We become grossed out, repulsed, deeply unnerved, we need to flee—but it’s a near Herculean effort to do so. Escaping one way, back the way you came, isn’t a hard route to retread, because all it takes is a stripping away of humanity, usually at the expense of looking like a human. But clawing out the other side of the Uncanny Valley, that’s where the challenge is. To ride the X-axis as far as it’ll go requires holding two seemingly opposing but equal ideas in harmony in your head: this is not human; this might as well be human.
In the mid-’80s, computer animation felt like the first time we, sensual human beings, peeked over the precipice into the Uncanny Valley. If you’ve seen Pixar’s Tin Toy, its 1988 debut short, then you probably understand just how horrifying these early days were when it came to photorealistic computer animation. The infant in that short, not all that dissimilar from the disfigured goblin baby Mel Gibson cast as Satan in The Passion of the Christ (2004), taps into the subconscious immediately, upsetting every safe notion in its way, drooling web-like, semen-viscous digital fluids all over whatever notions one has about what a baby should and is. This is why the Uncanny Valley isn’t a peak: the cognitive dissonance our brains must confront is a seemingly uncrossable chasm between what our species knows of itself and what our species is most afraid of. Between life and death, between health and sickness, immortality and sterility, safety and peril, between being special and being only a statistic—that gap between, on some sort of primal level, turns our stomachs, as if we’re sticking our head out over the edge of an incomprehensibly high building, looking straight down. We were never meant to be so high.
the emptiness of the eyes
Still, like most apparently insurmountable technological obstacles, the Uncanny Valley didn’t stop the Japanese. So, in 2001, iconic game designer Hironobu Sakaguchi released Final Fantasy: The Spirits Within, the first ever full-length, photorealistic computer animated film, based, at least in spirit (heh), on his Final Fantasy games. Astronomically expensive, demanding a small army of animators, and largely ignored by both audiences and critics, the film probably single-handedly sunk Square Studios after only its first endeavor. But as far as what Sakaguchi accomplished, more prescient than its financial failure, was Final Fantasy’s ambition: The director intended main character Dr. Aki Ross—voiced by Ming-Na Wen, a Macau American actor whose physical appearance doesn’t much influence Ross’s white-ish, ethnic-less appearance—to serve as the prototype for a new breed of artificial film star. Planning, like Al Pacino in S1mone (2002), to cast Ross in further Square Studios pictures, Sakaguchi claimed that adapting her to new roles would only be a matter of altering her age or appearance to better fit whatever stories they eventually decided to tell. That the film is also almost logarithmically boring is beside the point—Final Fantasy practically ignores the fact that it’s planted firmly at the bottom of the Uncanny Valley.
Putting aside the obvious issues with male animators developing a manipulable manifestation of their basest fantasies in place of an actual human woman—not to mention that Ross was the first artificial woman voted into Maxim’s Hot 100 List because the world was still an innocent place—Final Fantasy made very real the contours of the Uncanny Valley. Whatever it was—the emptiness of the eyes, the jagged locomotion of their limbs, the eerie sameness of every face—the hyperreality of the film’s characters only made their artificiality harder to take. Even if Ross, soulless and plasticine, still gave dudes boners, she wasn’t a testament to Sakaguchi’s team’s brilliance, she’s an unintended revelation of what our image-obsessed culture expects out of women. If only the Uncanny Valley operated purely within the realm of dicks. If only.
It made sense, then, that only a few short years later, a new figure in cinema rose to act as brave cartographer of the Uncanny Valley. Enter Robert Zemeckis, who—whether he’s willing to admit what monstrosities he’s birthed or not—filled almost a decade with computer-animated films that combined photorealistic rendering and motion capture technology. With the markedly “human” element of mo-cap re-introduced to the style of filmmaking pioneered by Final Fantasy, Zemeckis plunged into the deepest trenches of the Valley, beginning with 2004’s The Polar Express to show the world exactly what kind of freakish husks he could construct from the game-for-whatever charm of A-list actors. By pushing the technology forward, Zemeckis was also shoving that qualitative measure of “human likeness” further down the X-axis, creating, at best, a marvelously unsettling experience, and at worst something utterly horrifying. To watch the infinitely familiar visage of Tom Hanks struggle to pierce the opaque veneer of surreality with even one pointed moment of personality or warmth is to stare into the maw of the Uncanny Valley and see nothing but abyss. Not even the cold comfort of a distant bottom.
Zemeckis tried again. And again. It was as if he had a puzzle before him—to create the canvas upon which a new evolution in epic storytelling could be built—and he needed to wield archetypal stories to crack it. With each new animated film, first 2007’s Beowulf and then, in 2009, A Christmas Carol, the director embraced the exponential strengthening of his beloved photorealistic technology, but even as the methods got better and his animation more intricate, Zemeckis never seemed to realize that the further into reality he tried to force his cartoons, the closer to bedrock he dug his Uncanny Valley. Maybe he guessed that casting eminent folks like Jim Carrey or Angelina Jolie or Gary Oldman could provide the perfect safety net for the audience as they nauseously balanced over that gap, but none of it worked. Even A Christmas Carol, wherein viewers can count the pores on Jim Carrey’s digital nose, is a disturbing, perplexing film, one more interpretation of a classic holiday-themed story that somehow expects children to be drawn to wraith-like caricatures with dead eyes.
Zemeckis then went silent for a few years. He resurfaced in 2012 with the non-animated Flight, a vague sign he finally understood the extent of the Sisyphean task he’d shouldered throughout the 2000s. Flight felt like a refreshing return to form as much as an abandonment, a desertion, leaving us behind to maneuver the eternal twilight of the Uncanny Valley without our once-trusted navigator.
Spectacle is our inviolable right
A year after A Christmas Carol came Joseph Kosinski’s Tron: Legacy (2010), but it might as well have been eons later, so mollycoddled was his computer-borne young Jeff Bridges in the shadow of Zemeckis’s atrocities. Kosinski wisely kept his camera at a safe distance from the digital facsimile of his headlining actor, never really closing in on his delicately animated nostril hair or futilely venturing to capture every crease of the actor’s well-known smirk. This photorealistic concoction he also further couched within the reality of a completely digital realm, admitting that the character, Clu, was supposed to be a not-quite-real representation of the real Jeff Bridges. Kosinski covered his ass, gauged the distance between what he was capable of doing and what he wanted to do, and with what he’d done he tread carefully. He knew better. He treated the Tron (1982) sequel like it was contrition for not only Zemeckis’s iniquities, but for those of the first Tron film, the Edenic garden in which man was first handed the tools to lead the rest of us into the Uncanny Valley.
Except, to villainize Zemeckis is to shit on the optimism of his vision—not just for the sophistication of computer animation as it asymptotically approaches the realism of the human experience, but for the human experience itself. Every one of his films, from the Back to the Future trilogy to Forrest Gump (1994) to last year’s The Walk, celebrates the audacity of ambition as an essential trait of personhood. Even Flight operates as a kind of triumph of human goodness over our basest tendencies. If we have the ability to re-make reality in our image, Zemeckis seems to think, then we should use it—but for good, exercising the best of whom we are and whom we can be as a species. Spectacle is our inviolable right: In humanity, Zemeckis sees potential, his computer-animated films testaments to that potential writ large. And in the Uncanny Valley, Zemeckis sees innovation, a place where we can figure out what it means to be human and then figure out how to be the best version of that, to bring light to a dark place. “Though I walk through the valley of the shadow of death, I will fear no evil…” and so on. Somebody should ask him about the Psalms.
Still, since Zemeckis, filmmakers have been retreating from the Uncanny Valley. We watched Ex Machina scoot rump-first back toward the zero of the X-axis, making exceptionally clear where we, the filmgoing public, are willing to draw the line between that which holds the magical spark of life and that which only pretends to. We’ve handed motion capture to Andy Serkis to keep it safe, instructing him that he can do whatever he wants as long as he plays chimps and Gollums and Snopes—anything but a human being. We tried to find Michael Shannon’s face in Batman v Superman: Dawn of Justice’s Doomsday—which opened on March 25, less than two days after Microsoft pulled Tay from Twitter—but we paid too much money and too much time to find out if he was there. Dawn of Justice might just have been the sunset of Zemeckis’s dreams.
If you stand on one side of the Uncanny Valley long enough, looking across its expanse, you might catch a glimpse of one director hyperventilating with exertion to pull himself up the other side. Charlie Kaufman has never spent any time in the depths limned by Robert Zemeckis or Hironobu Sakaguchi, but his filmography—from Being John Malkovich (2000) to Synecdoche, New York (2009)—has almost wholly been concerned with plumbing the fathoms of whatever it means to be human (or, at least, a middle-aged, white, straight, male one). In his latest film, Anomalisa (2015), he continues to cover familiar themes, but he does so with stop-motion animation care of Starburns Industries and co-director Duke Johnson.
The world of Anomalisa is painstakingly assembled, thousands of hours undoubtedly plugged into every shot, every set, attempting to create a world as stultifyingly realistic as possible. In it, David Thewlis provides the voice for Michael, a successful customer service consultant and speaker who spends one fateful night in Cincinnati beginning to understand the form and shape of his severe depression. Michael sees everyone as identical, the same faces behind the same voices, his waking life a medley of one homogeneous experience after another. The use of stop-motion makes sense, then: Carbon copies of the blank face that Michael sees on everyone are attached to all body-shapes and body-colors (the seams of the masks showing, which comes symbolically into play once Michael sinks further into his hallucinatory mental illness). Meanwhile, Tom Noonan plays almost every single character, barely trying to modulate his vocals regardless of the age or gender of the personalities Michael encounters. These characters, these puppets, are specified by their lack of specificity, made (literally) to appear as ordinary as possible while floating through an extraordinary dreamscape between what we take for granted as real and what we fear reality might actually be.
And yet, there is no motion capture involved, no state-of-the-art technology. There’s only boredom and the mundane. In moderately priced hotel rooms and nice-enough bars, in non-descript airports and upper-middle-class foyers, in cigarette packs and bulk Ikea-ish art, in aged genitals and sagging skin and the perfectly hunched shoulders of decades of poor posture; Kaufman and Johnson craft each totem of a depressing reality with the thankless attention Zemeckis paid Jim Carrey’s blackheads.
look into Ross’s eyes and see nothing
If Dr. Aki Ross represents the kind of pansexual omni-humanity one might find tucked tastefully within a Kim Stanley Robinson novel—though Sakaguchi’s creation isn’t socially progressive so much as engineered to be appealing to boners of all heteronormative proclivities—then Anomalisa takes that vagueness to its logical end. To look into Ross’s eyes and see nothing is to watch Anomalisa and see everything: verisimilitude as the blandest, most accessible idea of who and what we are as the supposedly strongest group of organisms on this planet. While Kaufman and Johnson’s choice to portray practically every character with the same weird face and the same off-putting voice is predominantly a reflection of Michael’s interior life, they, maybe unintentionally, core to the heart of the Uncanny Valley. There we aren’t defined by our uniqueness—there we can’t claim we are each a special snowflake fallen delicately upon this earth. The only way to find our way out of the Uncanny Valley is to accept ubiquity. We are only ants marching; we are the Dave Matthews Band song of the universe’s college radio station.
Which is why Kaufman and Johnson may be the first directors to disappear into the Uncanny Valley and come out the other side. They seem to intuit something that always escaped Zemeckis. Realism isn’t about mimicking the ways human beings look and act down to the most inconsequential details, it is about embracing that inconsequence and building a world around it. There is more life in the puppet Michael’s eyes—confusing, dull, sad, bittersweet life—than in all of Zemeckis’s human “likenesses”.
Anomalisa is not a hopeless film, though. It is one that knows how difficult it can be to hold two thoughts in your head, two thoughts that don’t make much sense together. This is not human; this might as well be human. It knows that Tay could only become a kinky racist—Tay is designed to be an impressionable millennial just trying to fit in, as all millennials are designed to be. It knows that by watching Ex Machina as a film, you are questioning whether it is only that, or if it is something more, or something less, but more importantly, how that makes you feel and how that makes you feel human.
In turn, Anomalisa is not asking you how that makes you feel. It knows how you feel. It knows it makes you sad. It knows you are afraid, or anxious, or disturbed, or bored. It doesn’t ask you anything. Instead, it points to the other side in acknowledgement. It’s been there too.
On March 31st, a tad over a week after Tay was removed from Twitter, just over a month after Ex Machina won its Oscar, the teaser trailer for Kingsglaive: Final Fantasy XV emerged, joined at the hip with the new game of the same name. While the direct-to-DVD Final Fantasy VII: Advent Children first saw shelves in 2005—the film tied to the insanely popular Final Fantasy VII (1997) game and released without the big-studio fanfare of Sakaguchi’s flop—it’s been 15 years since The Spirits Within, apparently long enough to take stock of the original’s failures and try again.
Relatively big stars will once again provide the American voice talent for blithely ethnic-less characters; a whole studio’s future will no longer hang in the balance, Sony this time taking the reins. Though technology has undoubtedly become leagues better, which means the film’s animation is also substantially cheaper even as it’s 15 years closer to being “realistic”, there is no sign that this new sequel is doing anything different, anything revolutionary, from its progenitor. Only the context has changed.
Director Takeshi Nozue has no plans to push any of his characters into our world, no scheming zeal to, like Sakaguchi with his Dr. Ross, market a digital movie star within our non-digital film industry. He seemingly has no right to ask us to once again line up to descend into the Uncanny Valley. Rather, Kingsglaive keeps to its own corner of reality, within a fictionalized universe within the realm of gaming, its photorealism not meant to reach the terminus of the X-axis, but only to do what any good videogame wants to do: to be immersive, to transport us somewhere we willfully don’t want to recognize, to provide us with characters with whom we can still relate—to ask us to hold only one thought in our heads. This is not human, but it’s close enough.