Artificial intelligence has been one of the most dramatic developments of the 20th and early 21st centuries. Robots, supercomputers, operating systems and the Internet have already changed our lives in countless ways, and at the rate we’re going it looks like their influence on history is just beginning. So when you think about it, it’s kind of surprising that there haven’t been more films about AI’s implications and the increasing interaction between people and machines (no, Transformers doesn’t count). Luckily, a lot of the films that have come out have been very good. Taking time to get philosophical and speculative about the moral implications of AI, and to truly think about the interactions between humanity and technology is what sci-fi is truly about. The movies on this list do all that and more– they present artificial intelligence that is not only thought-provoking but feels almost too real.
Here are the 15 Most Scarily Realistic Artificial Intelligence In Sci-Fi.
15. Terminator 2: Judgment Day (1991)
Three memorable AI creations for the price of one in T2. There’s the T-1000, the terrifying killing machine brought to life by Robert Patrick’s poker-faced performance. Then there’s Skynet, the AI that intends to initiate “Judgment Day” as soon as it can get its hands on our nuclear codes. Finally, there’s the return of Arnold Schwarzenegger’s less advanced T-800 model, this time programmed to protect John Connor – and, by extension, the human race (see, robots can be our friends).
Each of these characters gives you a different perspective on what AI could mean. Of the three, the T-1000 is probably the most realistic: he’s programmed to do something, and he spends the film trying to do that thing. But the T-800 is far more interesting. Unlike Skynet, he never deviates from his mission or tries to take matters into his own hands. Thanks to John he gradually figures out why his mission is so important, and how he can go about it in the least lethal way possible. In other words, he discovers morality.
14. The Hitchhiker’s Guide to the Galaxy (2005)
Most sci-fi authors and screenwriters make one fatal mistake when writing about AI: they forget that while computers generally behave logically, people don’t. And when it comes down to it, it’s people who design computers. Hitchhiker’s Guide Guide to the Galaxy is well aware of this – hence the “Genuine People Personalities” AI system, a well-intentioned scheme that results in a fantastically depressed prototype called Marvin.
Where fictional robots generally protect humans, threaten them, or at least do some useful tasks for them now and then, Marvin mostly wanders around bringing the mood down (sample exchange: “Marvin, you saved our lives!” “I know. Wretched, isn’t it?“). The genius of Hitchhiker’s Guide is that crazy concepts like the Paranoid Android don’t seem so far-fetched the minute you stop to think about them.
13. Bicentennial Man (1999)
Credit where credit is due: robotics in fiction starts with Isaac Asimov. He was the first author to write about robots that weren’t metal Frankensteins, cuddly playmates, or metaphors for minorities, but simply machines with built-in safeguards that did as they were told (see Asimov’s own “My Robots”).
It’s ironic, then, that one of his few big-screen adaptations ended up being Bicentennial Man, based on an unusual-for-him story about robot-human friendship and love. While Robin Williams’ Andrew might seem like a purely sentimental creation, his story raises a very good point: if AI keeps progressing at the current rate, how long will it be before robots start to think and feel everything we think and feel? And at what point should we decide that their complexity entitles them to the same rights and recognition as us? Given our track record, isn’t it possible that robot rights will be held up indefinitely for political reasons?
12. Star Trek: The Motion Picture (1979)
Slow-moving and dated as Star Trek: The Motion Picture may be, it’s actually fairly ahead of its time on AI. Ilia, a replica whose robotic coldness has been imperfectly grafted on to a human core with feelings and memories, is like a precursor to Ghost in the Shell‘s Major. And then there’s “V’Ger”, the enormous “living machine” that’s got more knowledge than our whole species put together but none of our common sense or ingenuity.
The “mind-meld” at the end of the film might seem like a hangover from the Original Series‘ hippy origins, but it’s actually grounded in some pretty solid transhumanist speculation. As machines become more like us (androids) and we become more like them (cyborgs), it’s easy enough to foresee the emergence of a “third thing” that combines the best of both: the machines’ physical strength and data capacity plus our lateral thinking and capacity for meaning-making.
11. I, Robot (2004)
Another Asimov-based entry, this time loosely inspired by a book of short stories (in the sense that its plot doesn’t resemble any of the short stories and it’s exactly the kind of dystopian affair that Asimov would never write). The film’s two main AIs, Sonny and VIKI, are programmed with the Three Laws – preventing them from harming humans – but are both able to circumvent these laws. Like Skynet, VIKI is a rebellious and dangerous supercomputer, but unlike Skynet, VIKI claims that any harm she causes is ultimately for our good.
The idea that a super-complex AI would prioritize society’s interests over the individual’s comes from the I, Robot story “The Evitable Conflict”. It’s not much of a stretch to imagine a supercomputer coming up with a “Zeroth Law” in a world where we’re already handing more and more of our decision-making over to AI. While the film’s convinced that this is a bad thing, Asimov’s story isn’t so sure. Again, it’s up to the viewer to decide for themselves.
10. The Matrix (1999)
The Matrix‘s premise is almost exactly the same as T2‘s, but what sets this installment in the endless series of machines-vs.-humans films apart is the character of Agent Smith. He’s every bit as pitiless, single-minded and efficient as the T-1000, but there’s one key difference: where the T-1000 is just doing what he’s built to do, Smith believes in his mission. Everything about humanity bothers him on an emotional level (“This reality, whatever you want to call it, I can’t stand it any longer… I can taste your stink and every time I do, I fear that I’ve somehow been infected by it“).
Ultimately, Smith feels it’s his moral duty to thwart our plans because we’re “a disease… a plague” for which AI is the ‘cure’. The film never explains where Smith gets his emotion from, but Hugo Weaving’s performance makes his disgust totally believable while we’re watching it.
9. Interstellar (2014)
Interstellar has got some pretty damn cool robots. The irony is that they look less like us than any other cinematic bots, but their personalities are more humanoid than most of Hollywood’s more photogenic androids put together.
CASE and TARS’ lack of faces is more than compensated for by their colloquial speech and their natural, conversational deliveries (although there’s even better AI voice acting out there – read on). TARS in particular has attitude to burn, which shows through most in his much-discussed sense of humor.
Along with creativity, intuition and the search for meaning, many people consider humor to be one of the most human of all human attributes, so much so that one article describes it as the “final frontier of artificial intelligence”. We’ll have to wait and see whether scientists will ever create a robot who’d make a good guest at a dinner party, but in the meantime we’ve got TARS.
8. WarGames (1983)
It’s often said of computers that they’re intelligent, not smart. (Anyone who’s spent fruitless hours trying to word that Google search just right will agree.) The most extreme example of this in the movies has got to be the military supercomputer WOPR in WarGames, which regards initiating World War Three as no different than playing “a nice game of chess.”
Considering that we already entrust stock market fluctuations to machines that don’t know what money is, the idea of an American AI that baits Russia for fun isn’t too much of a stretch. The climactic scene where David basically has to “trick” WOPR into realizing the error of its ways is particularly well-observed: maybe one day AIs will reason the same way we do, but until then we’re talking about two completely different systems of logic with only superficial similarities.
7. Robot & Frank (2012)
Jake Schreier, Robot & Frank’s director, has said that the film isn’t intended to be a dystopian warning about the evils of technology – it’s simply a cool-headed examination of some of its implications: “[Technology’s] not bad or good but it will change the way we relate to each other.” Asimov would be proud.
The robot of the title is a strange hybrid of some of the other AIs on the list: he’s essentially programmed to do good but he’s not above a little stealing, he uses colloquial slang (“I’m getting the hang of it“) but his voice sounds metallic, he’s humanoid without looking like a person, and he’s helpful without being especially obedient. In other words, the character’s not a standard-issue trope.
His well-meaning nature and unforced charm make Frank’s growing attachment to the robot entirely believable, and his final act of self-sacrifice is both logical and moving. All these things make Robot & Frank a welcome addition to the “robot as friend” genre.
6. Alien (1979)
No doubt about it, Ash is one of the scariest android villains out there (“I can’t lie to you about your chances, but you have my sympathies“). In fact, one of the scariest things about him is that he is an android – we spend a good chunk of the film thinking he’s just another loyal flesh-and-blood crew member, right up until the memorable scene where his head comes off to reveal the circuitry underneath.
The reason this revelation works so well is that it plays on one of 21st century humanity’s greatest fears – the terror of a world where we can’t tell the real from the synthetic, allowing machines to undermine us in plain sight. There’s basically no better way for a screenwriter to evoke this feeling than trapping a bunch of innocent humans with a sleeper agent in the claustrophobic environment of a small spaceship. Ash is just one of the many things that makes Alien one of the all-time great sci-fi films.
5. Blade Runner
Speaking of the all-time greats… Blade Runner, like Bicentennial Man, asks just how many features robots need to share with us before they’re entitled to the same treatment. Not only are the film’s “replicants” almost psychologically identical to humans; they’re bioengineered rather than constructed from metal and tubes. In fact, they’re so similar to us that they can only be discovered through a rigorous psychological examination (the “Voight-Kampff test”). This enables the film to give us some extremely complex, interesting AI characters.
Rachael, an experimental replicant, believes she’s human due to false memories that have been implanted in her as an “emotional cushion“, while replicant-hunter Rick Deckard, who knows otherwise, can’t prevent himself from being attracted to her anyway. Meanwhile, replicant leader Roy is bad enough to have done some “questionable things” but good enough to save Deckard’s life – and human enough to deliver one of cinema’s great soliloquys. All of this is so believable because it’s so messy. Whatever else the future of AI is going to be, it won’t be neat and tidy.
4. 2001: A Space Odyssey (1968)
It’s probably fair to say that whenever you hear the phrase “evil computer” the first thing that springs into your mind is Hal (“I am afraid I can’t do that Dave“) from 2001: A Space Odyssey. Yet another computer that goes off script and makes his own rules, everything about HAL 9000 (to use his full name) is creepy: his half-metallic, half-human voice, his unwillingness to explain his actions, his meaningless reassurances, his fondness for gentle taunting (“Without your space helmet, Dave? You’re going to find that rather difficult“).
All of that’s more than enough to make him one of Hollywood’s most convincing AI villains, but the scene where Dave shuts him down for good is just extraordinary. As his circuits are deactivated one by one he starts to show what seems to be genuine emotion, anticipating Roy’s death monologue in Blade Runner. It’s up to us whether to believe his pleas or not: is he really afraid to die? Or is this just one last mind game?
3. A. I. Artificial Intelligence (2001)
One of the most moving treatments of artificial intelligence, A. I. tells the story of David, a futuristic Pinocchio who wants to become human. An advanced android (“Mecha”) who’s capable of loving and desiring love in return, David is dispatched to a human family to replace their son Martin, who’s in suspended animation until his disease can be cured. Naturally, when Martin finally returns he’s not too pleased about David.
As with so many other sci-fi films, there’s a mutual jealousy going on here: the robot feels envious of the shared connection that human beings have, and the person feels threatened by a machine that should know its place. The hair-cutting scene shows us that David’s naive in a way that Martin isn’t, suggesting that he’s not quite a perfect simulation of a “real boy”.
But like Rachael and Roy in Blade Runner, David’s pretty hard to pin down. He feels genuine love and affection, but he’s also capable of violence and trickery. By the end of the film you’re left believing that he’s a person in all the ways that matter.
2. Ex Machina (2015)
Ava’s an even more extreme example of a robot that uses trickery to get what she wants. Her character is deliciously ambiguous – does her plan to escape Nathan’s prison, which involves cunning, deception and manipulation, imply that she has all the psychological complexity of a person? Or does it just tell us that she’s learnt how to mimic human expressions and intonations without ever learning genuine empathy? Does she feel indifferent to the wellbeing of all humans, or just the particular people who have kept her cooped up her whole life?
At first glance the Ex Machina’s ending implies that Ava’s hopes and dreams are basically similar to ours. But there’s no real proof of that, especially when you read that the movie was originally going to end with the revelation that Ava’s thought process was completely foreign to us. What makes all this come alive for the viewer is Alicia Vikander’s brilliant performance as Ava. Placid without being emotionless, likable without being effusive, she strikes the perfect balance between human and machine. Unsettling stuff.
1. Her (2013)
Could this be the best voice acting of all time? Scarlett Johansson’s ability to convey all the shades of emotion experienced by Samantha – a disembodied OS with the ability to adapt and evolve – is nothing short of amazing.
Her is a thought-provoking study on the various complications that could arise if such a system starting dating a human. The relationship between Samantha and Theodore goes through a series of difficulties – Samantha’s botched attempt to use a human surrogate, Theodore’s fears that the relationship isn’t a proper one, Samantha’s desire for a body etc. – but the real knockout punch comes towards the end of the film.
In one of the most startling twists in all of sci-fi, Samantha goes from wanting to be just like us to realizing that her existence is better than ours. She starts interacting less and less with Theodore and more and more with her fellow AIs, until she eventually joins with them to travel to an existential plane that Theodore can’t reach or even imagine.
This is a highly believable account of what a (relatively) benign technological singularity might look like. Here’s hoping that we can expect more and more nuanced films about AI in the coming years.
Which AI in movies do you think is the most believable? Does that scare you or make you excited? Let us know in the comments!