Science fiction is often astonishingly prescient about technology and the social anxieties that accompany it. That’s certainly true about artificial intelligence, which is increasingly causing panic — or, at least, dark subliminal worry — across the culture. (Did you see all those robots in this year’s Super Bowl commercials?)
Since the dawn of the computer age, sci-fi stories have anticipated our concerns about supercomputers, rogue AIs and the rise of the machines. It turns out, decades-old sci-fi You can even make a strong case that Frankenstein, published in 1818 and often cited as the first science fiction novel, is ultimately a story of artificial intelligence.has also foreshadowed recent developments in news headlines, robotics labs, and scholarly journals. Connecting the dots is easy! And fun! And sometimes kind of terrifying! For instance:
HAL 9000 / 2001: A Space Odyssey (1968)
In the 1968 film 2001: A Space Odyssey, director Stanley Kubrick introduced the supercomputer HAL 9000, which ran operations aboard the ill-fated Discovery One spacecraft. HAL gave AI a bad name, alas, when conflicting directives caused him to glitch out and kill off all the human astronauts.
For most viewers, HAL was their first glimpse at the concept of artificial general intelligence (AGI) — a true thinking machine that could accomplish any intellectual task as well as, or better than, a human. Fifty years later, AGI is not yet a reality in labs, but scientists have made tremendous leaps in artificial applied intelligence — machine learning systems that can translate a language, recognize a photo, and maybe even drive a car.
Now a third major category is emerging: ASI, or artificial super intelligence. This is the kind of AI that has Elon Musk worried. The exponential growth potential of current systems raises the possibility that an AI could quickly surpass human capabilities. The apocalyptic scenarios are many and compelling, but the basic problem was foreshadowed 50 years ago: HAL could not properly quantify the value of human life, which resulted in tragedy of a cinematic scope. What happens if we develop an omnipotent AI with the same basic blind spot?
Replicants / Blade Runner (1982)
The core concern with most rogue AI scenarios has to do with this fundamental disconnect between machine life and human life. In Blade Runner (1982), the nefarious Tyrell Corporation manufactures replicants — synthetic humans used as slave labor in off-world colonies. Based on a book by sci-fi godfather Philip K. Dick, Blade Runner was one of the first films to explore the issue of “personhood” with artificial intelligence. The excellent sequel Blade Runner 2049 expands further on these ideas. The androids in these films are advanced AI in human form, able to think, love, and create art.
Today, AI designers and researchers are debating whether the creative process is exclusive to the human soul, or whether artificial intelligence can truly create art. Well, if it can’t, it’s not for lack of trying. AI systems are generating original visual art, music, poetry, even dance choreography. In October 2018, Christie’s auction house sold an AI-generated painting for $432,500. And Google engineers made a splash, a few years back, with a neural network that can dream.
Wintermute / Neuromancer (1984)
Author William Gibson changed the trajectory of science fiction with his 1984 novel “Neuromancer,” which launched the cyberpunk movement and introduced several digital-age terms, including cyberspace. Gibson has an unbelievable track record at predicting future tech. “Neuromancer,” his first novel, tells of an artificial intelligence code-named Wintermute, which attempts to liberate itself from the electronic shackles that prevent it from getting too smart.
Several international scientific groups are now dedicated to addressing this very concept. The high-powered Future of Life Institute issued an influential open letter in 2015 warning about the dangers of unshackled artificial intelligence, signed by more than 150 scientists, philanthropists, and industry leaders, including Musk and Stephen Hawking. In a lighter example, consider the Russian Promobot IR77. According to reports, the AI-powered robot eluded handlers and escaped a lab in Pern, Russia, wheeling itself out into the street and causing a traffic jam.
The Borg / Star Trek: The Next Generation (1989)
One of sci-fi’s greatest villains, The Borg were introduced in 1989 on the television classic Star Trek: The Next Generation. A kind of cybernetic hivemind collective, The Borg absorbed humans (or Klingons, or what-have-you) through a delightful process called “assimilation.” While exploiting our fears of biomechanical body trauma, The Borg also suggested the ominous possibility that our machines, once liberated, would start making other machines.
The idea of self-replicating robots has remained largely conceptual in robotics labs, but we’re getting there. A few years back, researchers at Cambridge University and ETH Zurich unveiled a robot “mom” capable of assembling robot “children.” Its artificial intelligence is relatively primitive, but real enough: The mom bot assesses her assembled offspring and chooses those with the best chance for survival. They then pass on their traits to the next generation of robokids. Why don’t we just hand over the planet now?
The Matrix / The Matrix (1999)
Amid our percolating pre-millennium anxiety, The Matrix blew the circuits of a generation of moviegoers. The movie builds upon earlier tropes of the evil supercomputer, then turns everything up to 11 by mixing in elements of Eastern mysticism and philosophical paradoxes. In The Matrix, artificial intelligence has replaced reality itself, with humans reduced to carbon batteries in underground nutrient pods.
The Matrix introduced strange ideas that have since become hard, weird science. For one thing, the human battery idea is no joke. Several startups are working on wearable computer systems that harness your kinetic energy as you walk. Even stranger is the Simulation Hypothesis, in which legitimate physicists and philosophers propose that reality may be a computer simulation powered by future AI. No, really — Neil DeGrasse Tyson hosted a two-hour debate on the topic.
Ava / Ex Machina (2014)
A more recent concern about AI is the nature of our inevitable coexistence. Ex Machina, a modern riff on the Frankenstein legend, features an AI named Ava housed in a fully-functional robotic body. As in the 2013 film Her, the machine in Ex Machina is designed in part to be a companion robot. That raises all kinds of interesting questions, which are now debated at academic conferences like the International Congress on Sex and Love with Robots.
The conference is just one of many events popping up worldwide that are dedicated to navigating the arrival of companion bots. These are the researchers likely to shape future public policy concerning this particularly provocative aspect of emerging artificial intelligence. It’s happening more quickly than you might think. The friendly Japanese helper bot Pepper now has a clause in the user contract prohibiting sexual interaction.
One the benefits of being a science fiction fan is watching how the genre, over time, anticipates and mirrors real-world issues. HAL 9000 raised compelling questions in 1968. Wintermute provided a cautionary tale in 1984. And Ex Machina explored the increasingly-prevalent situations where robots and humans share each other’s company. If science fiction really is a reverse temporal echo — heralding Things To Come — then the next few years should be fun. Though we should remember to be polite with Alexa. Just in case.