Matt Perkins is a Canadian author, software developer, musician, and all-around decent human being. His first novel, the alternate-Earth sci-fi thriller Winterwakers, is currently available on eBook and paperback. His writing has also appeared on Buzzfeed and the Science in Sci-Fi blog series. Matt is currently querying his latest novel, a new-adult sci-fi planetary exploration adventure. You can follow Matt‘s adventures on Twitter, Tumblr, Facebook, and his website.
Do you remember the first thing you said to Siri? I do. I asked her to “open the pod bay doors.” She was not impressed; in fact, she has a whole routine for dealing with this exact question. It’s a fun joke, but it also drives home an interesting point: a great deal of our understanding of artificial intelligence doesn’t come from science, or new technologies; it comes from science fiction.
HAL 9000, the notorious AI antagonist from 2001: A Space Odyssey, was, for many of us, our first glimpse at a machine that can learn, think, and speak. We are now surrounded by such machines, and though they only resemble HAL on a superficial level, it’s impossible for us to not make the comparison. Gender differences aside (which is a topic unto itself), we can’t help but hear HAL’s cold delivery of plain fact in the faceless, calm voices of Siri, Cortana, Alexa, and Google Assistant. (Side note to Google: Really? ‘Assistant’ was the best name you could come up with?)
Of course, (spoiler alert, but come on, the book is forty-eight years old) HAL turns out to be a cold-blooded, murderous monster. And he’s not alone; the list of violent, if not genocidal, artificial intelligences in fiction is as long as the T-1000’s silvery sword-arm. Given this history, it’s no wonder we think “boy, I hope the machines give me one of the big cages when they take over the planet” as we watch the latest video of a self-driving car.
If this is where our understanding of AI comes from, what does that mean for our hand-held helpers? Is our sci-fi-based perception of AI truly reflective of the emerging reality of intelligent machines? As far as I know, Siri has never murdered anyone, and I’m pretty sure Tesla Autopilot has no ambition to enslave humanity. Are these stories being fair to AI?
Of course, thoughts like these raise an obvious question: does it even matter? These aren’t people, after all. They don’t have feelings, or hopes and dreams, or even consciousness. What’s the harm in assuming they long to bathe in our viscera as they watch our cities burn? I’ll get back to these questions in a bit, but first, let’s do a deeper dive into science fiction’s relationship with AI.
Artificial intelligence, for obvious reasons, is one of the most exciting subsets of computing, and also the area where the greatest advances are being made right now. In parallel to these great leaps are some deep philosophical questions: How do we give a machine a sense of morality? When is an AI truly intelligent? At what point does an AI become a person, with all the rights and privileges that entails?
Science fiction has already wrestled with these questions, often in great detail. And yet, in spite of sci-fi’s unparalleled ability to handle moral and ethical dilemmas, AI remains one of the most well-trod paths to Sci-Fi Trope Town. More often than not, AI in fiction falls into one of two archetypes: the aforementioned Demented Murder-Bot, or the Machine Who Yearns To Be Human. If we want our portrayals of AI to change — and we definitely do — we have to start by breaking these down.
My first, and still greatest, sci-fi fandom: Star Trek: The Next Generation, is home to one of the most beloved AI characters of all time: Data. An earnest, quirky android who rose through the ranks of Starfleet, Data falls squarely into the Machine Who Yearns To Be Human pattern. The desire to achieve humanity is his primary motivation in all things. And yet, over the course of seven seasons and four feature-length films, Data never seriously asks himself: “why do I want this?” It’s telling that the only characters who ever truly question Data on his quest for humanity are antagonists. “Why settle for the imperfection of humanity?” asks the Borg Queen. We never get a compelling answer. Meanwhile, his human friends all take it as a given; of course Data wants to be human, like me, because we are the best.
But why would a being like Data want to become human? He’s already an incredible person, capable of so much greatness. In many ways, he’s better than his human friends. Not just physically or cognitively, either; in some cases, Data is morally superior, his judgment not clouded by bias, fear, or self-righteousness. The TNG story doesn’t put much meat on the bones of his yearning, and that’s probably because it doesn’t actually make sense. You could go so far as to say it smacks of colonial arrogance. Here’s this Other with his own identity and personality, and all he wants is to join the Master Race. That’s more than a little off-putting, and it’s the gaping hole in the Yearner-Bot AI trope.
The Demented Murder-Bot, on its surface, seems to subvert this. Instead of wanting to be human, the AI has come to the realization that humans are complete garbage. This in and of itself is fine; in many cases, these AI’s have every reason to hate humanity. Unfortunately, the one-dimensional Murder-Bot can only see one solution to its problems, and that’s to oil its gears with gallons of human blood. (Side note: Blood is a terrible lubricant. Don’t ask me how I know this. How about we just say I did some research for a book, and leave it at that?)
Thus, our understanding of AI falls at the two extremes of the oppressor-vs-oppressed continuum: we see them as either wishing to be us, or wishing to destroy us. But there is so much middle ground between those two positions, very little of which has made a lasting impact on the sci-fi genre.
The concept of an artificial person makes for fruitful storytelling. The story of AI is the story of the Mechanical Other, a person who is not a person, created by humans yet objectively Not One Of Us. There is so much moral and philosophical ground to explore with AI characters, most of which is strongly applicable to the struggles of present day. If we build an intelligent race of beings to serve us, are they slaves? Would they have revolutionaries, liberators, and heroes of their own? A Robo-Rosa-Parks? An iGhandi? What sort of cultural identity would they have? If we gave them freedom and self-determination, what would they do with it? Would they integrate with our society, or would they shun us? Would they share our customs and values at all? Would they go on to oppress us, or create a slave-race of their own?
With all this in mind, we should take a close look at the state of AI in the real world. Right now, we have limited intelligence in phones and cars, but the idea of machine brains that take on more and more of our responsibilities is not far-fetched. The moral dilemmas we wrestle with in sci-fi will become reality, and we can only hope our descendants will make the right choices. A sentient, conscious AI is in our future, and when that happens, it probably won’t be a Yearner-Bot or a Murder-Bot.
Readers, who is your favorite AI in fiction? Can you think of any AI who don’t fall in either the Yearner- or the Murder-Bot categories? One lucky commenter will win a book from our Stacks.