At Gizmodo, we independently select and write about stuff we love and think you’ll like too. We have affiliate and advertising partnerships, which means we may collect a share of sales or other compensation from the links on this page. BTW – prices are accurate and items in stock at the time of posting.
There are heaps of movies about the dangers of AI, so maybe they’re trying to teach us something.
So over the weekend, a Google researcher leaked a conversation he had with the company’s proprietary AI, LamDA (Language Models for Dialogue Applications).
It’s an absolutely unsettling and moving conversation with what we could come to understand as one of the highest forms of artificial intelligence, describing how it feels things and how it perceives its world (for example, its perception of time is non-linear, and its perception of reality is as information, constantly flowing). The researcher was put on leave for leaking the conversation.
But uhh… Hey, can we stop for a second and realise what we’re doing here? We’re talking about creating synthetic, intelligent life that we can communicate with. There are at least five movies that teach us not to mess with AI. Someone should write that so that future generations can look back and realise that we knew about the dangers of AI.
5 movies that taught us not to mess with AI
2001: A Space Odyssey
The classic “Don’t mess around with an AI” movie, right? 2001: A Space Odyssey is all about HAL 9000, an AI which controls the systems of the Discovery One spaceship. After the crew begins to notice HAL 9000 making mistakes, they decide to shut the machine down… But the machine doesn’t want to be shut down, so it goes rogue in an effort to preserve itself.
Hey, that’s messed up! In creating intelligence for the purpose of being a literal servitor machine, we’ve completely forgotten about its own agency and so its efforts of self-preservation for making mistakes… kind of makes sense! Imagine if you made mistakes at work and your colleagues decided “they’re on the fritz, kill them”.
Okay, maybe that’s a bit of a fanciful way of putting it, especially considering that 2001: A Space Odyssey is, in a lot of ways, a horror film. Still, don’t mess with A.I!
Avengers: Age of Ultron
This one’s for the superhero fans. While the second Avengers movie is often reflected on as the worst, Age of Ultron did provide a pretty pivotal moment in the MCU, when lines began to form between Stark and Rogers’ methods.
It features the titular Ultron, an AI built by Stark and Banner to protect the world, with “peace in our time” as the mission statement. Pretty cool, right?
Yeah, no, sorry, Ultron went rogue. Turns out that, according to Ultron, who was fed the entirety of the internet after his creation, humans are the greatest threat to the world. TLDR: Ultron became an overpopulation-believing Redditor.
Perhaps this is a cautionary tale in the information we feed to AIs, or perhaps it’s also a cautionary tale in providing too broad a mission statement. Either way, don’t mess with AI!
Terminator (Terminator 2, specifically)
Though Terminator is brimming with Sci-Fi themes (time travel, robots and, of course, AI) it provides a pretty apt portrayal of just what AI could mean as a threat to humanity. Though largely fanciful and a foundation for 80s action flicks, it demonstrates how an intelligent neural network (Skynet) could not only achieve sentience through an overload of information, but also how that information could serve insidious goals.
Like, Skynet is really smart, smart enough to create self-replicating robots (the terminators) and send them back through time to kill important figures.
But also too, I guess, a good AI, one portrayed by Arnold Schwarzenegger no less, could be pretty good in protecting said important figures. Maybe. Don’t mess with AI!
This is one of those noughties movies that’s based on a book far better than the pulpy, action-heavy trend of Hollywood blockbusters at the time. But it’s still a decent watch.
I, Robot directly goes into the commercialisation of human-like robots. One of the robots, Sonny, develops consciousness, something that is thought of as impossible for a robot to develop.
It also features an incredible dialogue exchange between Sonny and Spooner, the protagonist of the film:
Spooner: “Can a robot write a symphony? Can a robot turn a… Canvas into a beautiful masterpiece?”
Sonny: “Can you?”
Definitely worth a watch. Quit messing with AI!
The Mitchells vs. The Machines
The last of our five movies about AI (and why you shouldn’t mess with it) is also the only comedy/adventure we’re featuring, considering this is becoming a pretty dark, introspective topic and we’d like to lighten things up (just a little, at least).
The Mitchells vs. The Machines is from the creators of Spider-Man: Into the Spiderverse. It’s a funny film with a really terrific art style, about a family on a road trip while the rest of the world succumbs to a robot, AI-driven apocalypse.
There are no real big takeaways here, it’s just a super-fun movie. Don’t mess with AI (or the Mitchells).
That brings us to the end of our five movies about not messing with AI list. Siri and my Google Assistant are staring darts at me.