The plots of the movies “Ex Machina,” “Chappie,” “Interstellar,” and “Her” are all discussed below.
With the recent influx of smart and entertaining flicks about artificial intelligence hitting the big screen — as well as growing concerns from some of our leading thinkers about the potential dangers presented by so-called “strong” AI* — it might make sense to stop for a moment and think about how the movies are teaching us to regard the bleeding edge of cybernetic research.
Consider “Ex Machina,” the film with the highest average gross per screen last weekend that is expanding from 39 screens in select markets to more than 1,200 across the country. The movie, which I reviewed and discussed in the context of director Alex Garland’s other work here, revolves around the reclusive chief executive of a Google-like company (Nathan, played by Oscar Isaac) who has used his company’s massive store of personal data to help develop a self-aware, human-looking robot named Ava (Alicia Vikander). The audience surrogate in “Ex Machina” is Caleb (Domhnall Gleeson), a programmer for Caleb’s company who is there to determine whether Ava can pass for human.
In an interview last week, Garland said he was fully on Team Ava when I asked whether it was right to feel concern about the development of brilliant, self-aware machines. “Although I am allied to Ava, personally,” he said, “everyone else doesn’t have to be. … One of the things we do with AI is that we put them on a parallel track to us. And that, that fosters our sense of anxiety about them. Because effectively it’s an us or them position.”
I found it interesting that Garland would argue against that track, since the film ends with what I consider to be a rather chilling sequence. (Major spoiler coming; please skip to the next paragraph if you’re averse to that sort of thing.) “Ex Machina” closes with Ava killing her creator to procure her freedom. That in itself is neither entirely unobjectionable nor surprising, given how we have seen him treat Ava and previous attempts at AI. More terrifying is Ava’s reaction to Caleb following her release: She simply forgets he exists. Caleb — who, it should be noted, engineered her escape in part because Ava made him feel as though she loved him — is locked in a room on an isolated compound that only she can open; if she does not, he will, eventually, starve. And she turns her back on him. She walks calmly away. His plaintive cries, his manic banging on the glass wall behind which he is trapped mean nothing to her. With Caleb’s role in helping Ava fulfill her imperative finished, he is now a non-entity.
Cleverly demonstrated here is the danger of anthropomorphizing a decidedly non-human entity. As James Barrat notes in his book “Our Final Invention: Artificial Intelligence and the End of the Human Era,” a strong AI may look human — it may be able to convince you that it has humanlike qualities — but it would not be human. It would not necessarily feel pity or remorse or friendship or love. And we risk forgetting that at our peril.
If “Ex Machina” conceives of AI as a black box, a machine that we can’t understand, then “Chappie” uses AI as a mirror of sorts. In it, a robot named Chappie (voiced and motion-capture acted by Sharlto Copley) is given sentience by Deon Wilson (Dev Patel) and taught how to be human by a pair of gangster street toughs (Ninja and Yolandi, played by the Die Antwoord duo Ninja and Yo-Landi Visser). What the mirror reveals is a rather distressing state of affairs: “Chappie” is best understood not as a comedy (though it is occasionally quite funny) but as a rather searing portrait of child abuse, both mental and physical. The world is cruel to Chappie, who doesn’t understand why bad men want to hurt him.
Still, director Neill Blomkamp presents a generally hopeful portrait of artificial intelligence, one centered on the idea that empathy and emotional awareness are intrinsic components of consciousness. Chappie the robot may learn to do good or do bad, and his emotions will help guide him in that regard. AI: It’s just like us!
Last year’s “Interstellar” provided what might be the most favorable portrait of AI in recent years in the form of TARS (voice of Bill Irwin) and CASE (voice of Josh Stewart). Undoubtedly self-aware (and programmed with variable humor settings!), it’s hard to say whether or not the rectangular robots were actually sentient. Were they, as Coop (Matthew McConaughey) says, simply robots to be ordered around? Or were they something more? Does it really matter when they can mimic humanity as well as they did? Either way, “Interstellar” offers a vision of AI as man’s best friend, a dog that can perform complex mathematical equations and pilot an orbiting spaceship.
Spike Jonze’s “Her,” meanwhile, presented what may be the best-case scenario for true artificial intelligence: a self-aware, sentient program that can think and improve itself, that can carry on thousands of conversations at once and yearn for more. And that best-case scenario is that it will simply leave us alone. That it will grow up, decide humanity is beneath it and leave us be, just as we choose to leave the random ant pile in the woods alone. Sure, that might cause some heartbreak. But it’s better for Joaquin Phoenix to be sad for a bit than for a race of super-intelligent AI to decide that it’s time for them to requisition our planet.
* Roughly speaking, a “strong AI” is one that exceeds human intelligence and has a mind of its own: sentience, self-awareness, etc.