The concept of an “intelligent machine”—as a tireless assistant, the ultimate soldier or even a caring companion—has captivated the human imagination for thousands of years. Long before artificial intelligence was a reality, writers from ancient Greece to Cold War-era America spun fictional stories that reflected our collective hopes and fears about AI. Will intelligent machines help humanity reach our greatest potential, or will these brilliant creations turn on their “masters” and be our demise?
“The quest for AI is a really ancient one and it’s fundamental to who we are as humans,” says Kanta Dihal, co-editor of AI Narratives: A History of Imaginative Thinking About Intelligent Machines. “That’s why I’m not surprised to find intelligent machines as far back as Greek mythology.” Below are some of the earliest imaginings of AI among humanity.
1. Homer's 'Golden Maidens'
The Iliad was written nearly 3,000 years ago in a world without electricity, let alone robots or computers. Yet in 800 B.C., the ancient Greek poet Homer was able to imagine a godlike power that could create intelligent machines.
In the Iliad, we meet the metal-smithing god Hephaestus, who was tossed from Mt. Olympus for trying to protect his mother during a fight with Zeus. Hephaestus suffered serious leg injuries and his lameness was mocked by the Olympian gods. But Hephaestus poured his humanlike suffering into making beautiful works of art, including the magnificent shield of Achilles.
Because of his physical limitations, Hephaestus needed helpers in his blacksmithing workshop. Homer wrote that Hephaestus used his special powers to create “attendants made of gold, which seemed like living maidens.” In this brief passage, Homer perfectly described humanlike machines programmed to assist their creator.
“In their hearts there is intelligence, and they have voice and vigor, and from the immortal gods they have learned skills. These bustled about supporting their master.”
Dihal isn’t surprised that Hephaestus’s golden assistants were women, since AIs that serve people have traditionally been gendered as female. Take Rosie the robot maid on the animated TV show, “The Jetsons.” Even real-life virtual assistants Alexa and Siri were given female names and (at least initially) female voices.
2. The Original 'Terminator'
Five hundred years after Homer, the Greek author Apollonius of Rhodes wrote the Argonautica, an epic poem about the adventures of Jason and the Argonauts in their quest for the Golden Fleece.
When the Argonauts approached the island of Crete, they were repelled by a bronze giant named Talos who hurled house-sized boulders at their ship. In some tellings of the Talos myth, he was also an automaton built by Hephaestus to patrol the shores of Crete and fight off pirates.
“He’s the proverbial super-soldier,” says Dihal. “He’s enormous, a giant made of bronze, specifically built to defend the island of Crete from invaders.”
Fictional heirs to Talos include the advanced robot assassins in the Terminator movies and the battle droids in the Star Wars franchise.
A few years ago, the U.S. Army experimented with a bulletproof, mechanized exoskeleton that they named TALOS (Tactical Assault Light Operator Suit) in a nod to what Dihal calls “the original killer robot.”
3. Arabian Automata
In the ancient world, artisans and engineers like Heron of Alexandria built elaborate machines powered by water, wind or a wound-up rope. Two thousand years before Disneyland, Heron assembled a series of simple machines to produce a 10-minute mechanical puppet play with sound effects.
After the fall of Rome, these magician-like skills were lost in the West, but automaton-making thrived in the Arab and Muslim world. In the 9th century, the Banu Musa brothers, three brilliant engineers and illusionists from Baghdad, wrote the Book of Ingenious Devices, a collection of 100 gadgets like trick jugs that dispense three types of liquid (including boiling water) and dancing fountains.
The fascination with intelligent machines in the ancient Muslim world shows up in 1001 Arabian Nights, the centuries-old collection of Middle Eastern legends. In “The Story of the City of Brass,” a band of travelers came upon a mechanical horseman made entirely of brass. Luckily, it also came with instructions:
“O thou who comest up to me, if thou know not the way that leadeth to the City of Brass, rub the hand of the horseman, and he will turn, and then will stop, and in whatsoever direction he stoppeth, thither proceed, without fear and without difficulty; for it will lead thee to the City of Brass.”
When the travelers did as instructed, the horseman “turned like the blinding lightning, and faced a different direction from that in which they were traveling.”
“There has been a reappreciation of some of the 1001 Nights stories as what we would now call science fiction, because they anticipated some of those ideas,” says Dihal.
4. The 'Uncanny' AI
An 19th-century short story by the German writer E.T.A. Hoffmann features an AI that’s so convincingly real that she’s mistaken as human. In The Sandman, published in 1816, the protagonist falls in love with the “daughter” of his mentor, only to find out that she’s a machine. The shocking realization drives him to suicide.
Dihal says that automata-making flourished in 17th- and 18th-century Europe with popular creations like the Mechanical Turk, a famous chess-playing machine (and fake) that squared off with Napoleon. Or Jacques de Vaucanson’s “digesting duck,” a mechanical duck that could eat grain and defecate.
“Having actual machines that were that sophisticated prompted works of fiction that asked, ‘What if this kind of technology gets better and better and we can’t tell the difference?’ This is where stories like The Sandman really tap into the zeitgeist.”
A century later, Sigmund Freud referenced The Sandman in an essay called “The Uncanny,” a feeling that he described as “all that arouses dread and creeping horror.” In modern robotics, the term “uncanny valley” describes the unsettling feeling of interacting with an AI that is “close to being humanlike, but not close enough, and that is terrifying,” says Dihal.
5. The 'Frankenstein Complex'
In 1818, just two years after Hoffmann wrote The Sandman, Mary Shelley published her gothic novel Frankenstein; or, The Modern Prometheus, a direct reference to the Greek myth of Prometheus, who stole fire from the gods and gave it to man.
“Frankenstein is addressing the same zeitgeist as The Sandman, but with a slightly different science,” says Dihal. “Even though Frankenstein’s creation is biological, it raises the same question as AI—what are humans allowed to do with this ‘forbidden’ power?”
In Shelley’s novel, Frankenstein’s “monster” gains self-awareness and wants to live like a human, begging Dr. Frankenstein to build him a wife. When the doctor refuses, the monster goes on a murderous rampage.
Elements of the “Frankenstein complex”—as sci-fi author Isaac Asimov dismissively called it—pop up in a lot of modern fiction. In the 1982 film Blade Runner, the bioengineered AI workers known as “replicants” are denied their freedom and revolt against their human overlords. In the 2014 movie Ex Machina, a Frankenstein-like tech guru builds a convincingly human AI woman who kills him for her freedom.
“There’s an element of Frankenstein in nearly all of the ‘robot uprising’ stories,” says Dihal.
6. Birth of the 'Robot'
Even though Homer wrote about humanlike machines 3,000 years ago, the word “robot” didn’t exist until the 20th century. It was the brainchild of the Czech playwright Karel Čapek, whose 1920 play R.U.R. or Rossum’s Universal Robots was a worldwide hit.
In the Old Slavonic language, robota means “forced labor,” which is exactly what the robotic characters were in Čapek’s play.
“The very first use of the word robot is also where we get the whole ‘robot uprising’ theme from,” says Dihal. In the play, the robots grow tired of being exploited, so they rise up and kill the humans. “The play was very much written in the revolutionary spirit of the times.”
7. HAL: The (Murderous) Mind in a Box
In 2001: A Space Odyssey, the 1968 sci-fi film directed by Stanley Kubrick, the uncannily intelligent machine isn’t a robot, but a computer.
In the movie, the human characters chat with the spaceship’s personable onboard computer, the HAL 9000, which responds like a helpful assistant. But when Dave discovers that the supposedly infallible HAL made a mistake—and HAL realizes that the human astronauts are going to shut him down—HAL tries to kill them.
“A big shift in AI took place when the computer was invented, and suddenly people realized that you can have a ‘thinking’ machine without needing a human-like body,” says Dihal. “HAL looks very much like those early computers, and the film is an early exploration of how a distributed, although still contained computer ‘mind’ might operate.”
Another sci-fi novel and movie from that same era was Colossus, the Forbin Project, about a supercomputer that gains control of America’s nuclear arsenal, then goes to war with a Soviet supercomputer.
“That’s a really early example of a non-humanoid AI posing an existential threat to all of humanity,” says Dihal. There was a similar plot in the 1983 hit movie WarGames, in which a powerful military computer mistakes nuclear war as a “game.”