Neill Blomkamp’s Underrated Sci-Fi Flop Sneakily Provided The Solution To Our AI Malaise

We live in the world that sci-fi pioneers like Isaac Asimov warned us about. Inequality heightens while political authoritarianism is on the rise, while companies pursue the development of AI-powered military robots and job-ending AI. Most science fiction, at least what makes it to screen, appropriately highlights these dangers, but other than “don’t do it!”, film history doesn’t always offer advice on what we should do about these future technologies. Chappie, Neill Blomkamp’s exploration of near-future AI, suggests a novel solution: make AI-powered bots more human, not less.

Chappie takes place in a world where Johannesburg gets the world’s first robotic police force, the “Scouts,” thanks to inventor Deon Wilson (Dev Patel) and the company Tetravaal. A set of local criminals, led by Yolandi (Yolandi Visser) and Ninja (Ninja), inherit the world’s first truly autonomous, sentient robot when they kidnap Deon and his latest project. The catch: its consciousness starts as a child, and it must be raised like one. Between Ninja’s desire to use Chappie for crime, Yolandi’s sensitivity to it as a sort of child, and Deon’s desire to test the limits of his experimental consciousness, Chappie goes through a complex array of developmental stages. Meanwhile, at Tetravaal, Deon’s work rival, Vincent (Hugh Jackman), pursues the development of his walking tank of a robot, the Moose. And as he “grows,” the atypical Chappie becomes part of the resistance movement against the growing horde of military robots.

There’s an inherent in-world controversy about Chappie and how his functions should be used. On the one hand, Ninja wants Chappie to be a murderous tool, the function that its base model was intended for, before Deon’s experimental cognitive update. He simply wants Chappie to be his tool, for theft and crime and so on. On the other hand, Deon and Yolandi recognize Chappie’s capabilities and its implicit personhood. Rather than being a mere object to program, Chappie’s advanced software learns like a child does, therefore going through true stages of development on the way to personhood. Deon teaches the growing robot to love books and paint, and when Ninja reacts violently, Yolandi tells Ninja, “Maybe he’s more than just a stupid robot who shoots people, just let him do his stuff, he is a kid!”

Chappie implies a lot about psychology and society: children, and therefore people, often grow to reflect the contexts and situations in which they’re raised. Train them to be violent, authoritarian, and oppressive, and they may become so. Nurture creativity, kindness, and empathy, and they’ll internalize it. But Chappie explores the notion that if AI was true intelligence, an actual mind instead of an ordinary processor connected to a hard drive full of data, it might well grow (and therefore need to be trained, or rather, raised) the same way.

Chappie highlights certain transparent dangers of AI development run amok: powerful murder machines serving the interests of the wealthy in highly unequal societies. Those threats are at their most dangerous at the Scout’s traditional level of AI, where they’re given dictates but have just enough intellect to find the best ways to oppress people in the guise of “law and order.” Chappie, by contrast, has empathy. It has emotional attachments to Yolandi, to its Maker. He experiences fear and therefore wants peace and safety. His emotional attachments make him hesitant to harm humans once he understands what that means.

There’s considerable scholarly and scientific debate over what true AI would entail, how to define personhood, and what consciousness even is. But if an AI, like Chappie, utilizes chain-of-thought reasoning (breaking down complex tasks into logical sub-steps before answering), has self-awareness, the ability to discern truth from falsehood, develops complex emotions, and so on, does that make it true intelligence? Chappie’s journey suggests if true artificial intelligence is treated like intelligence instead of an advanced murderous implement, it could be the answer to the question of how to develop it responsibly.

Of course, that’s not why it’s being developed. AI is being funded to replace labor, not to replicate humans with their agency, needs, and issues. And maybe, in that case, the answer is to reject its development entirely until society changes. Nevertheless, while Chappie wasn’t welcomed with entirely open critical arms when it first premiered 10 years ago today, it provokes a thoughtful alternative path that speaks directly to crises we’re facing a decade later. Perhaps, in the long run, the key to doing AI right to allow them full, complex humanity in all its messy and emotional detail.