The robots are coming for our jobs, artificial intelligence is ascendant, and invisible programs are taking over our lives. “Automation” is the word that comes up in each of those contexts and plenty more. It’s certainly one of the looming concepts of Our Times—a business imperative, an economic driver, a utopian ideal. We’re automating work, systems, services. Surveillance, commerce, manufacturing, policing. Everything, almost, or trying to.
But it’s a deceptively nebulous concept, one that strikes a chord in our psyche beyond definitions like ‘the technique of making an apparatus, a process, or a system operate automatically.” Understanding how our thinking about automation—and where the drive to automate stems from—should help us better grasp how it’s playing out today. It’s only relatively recently, after all, that we’d pin the push to automate on something like “business owners who want to capture more profits and cut labor costs.”
So, I set out to locate the origins of automation. I did so in a series of conversations and correspondence with scientists and scholars whose work focuses on subjects ranging from Greek mythology’s robots to the biological roots of abstract thought. The instinct to automate, classicists and zoologists seemed to agree, may be among the most ancient and universal human traits. In fact, one argued that it’s nothing short of the act that separates man from beast.
‘Automation,’ specifically, entered use in the mid-20th century. The ubiquity of the word has helped to obscure the fact that it has a weird, distinctly midcentury corporate vibe. Auto-mation; it’s almost Jetsons-esque. It’s a dash or two away from, say, smell-o-vision—the word itself is a retrofuture. It’s fitting that the first usage of ‘automation’ is generally attributed to a vice president at Ford, the company infamous for midwifing mass manufacture into industry. Delmar Harder created the first Automation Department at the company in the late 1940s and coined the term in the process. His mechanized factory, in which car parts were automatically transferred from station to station, was hailed as an engineering marvel; it also helped kickstart a widespread panic over impending job loss that eventually required the attention of not one but two congressional hearings.
But ‘automation’ wasn’t such a leap from ‘automatic,’ or ‘automaton,’ which both stem from the ancient Greek, αὐτός, auto—self.
“The first written use of the word ‘automaton’ in Western literature appeared in Homer’s Iliad, recounting the marvelous self-moving and intelligent machines fabricated by Hephaestus, the blacksmith god of invention and technology,” Stanford classicist Adrienne Mayor told me in an email. Mayor was writing me from the road, traveling in support of her new book, Gods & Robots, which examines the earliest cultural and technological manifestations of automatic machinery and artificial intelligence. “Writing in about 700 BC, Homer described gates of heaven that opened and closed automatically to admit the gods’ chariots; a fleet of driverless three-wheeled carts that delivered nectar and ambrosia to the god’s banquets; a bank of automated bellows that adjusted their blasts as required; and the crew of golden female androids endowed with artificial intelligence to anticipate Hephaestus’s every need,” Mayor told me.
Dr. Kanta Dihal and her colleagues at the Centre for the Future of Intelligence at the University of Cambridge have been doing some of the most in-depth work towards analyzing the historic narratives around artificial intelligence and automated systems. “What we found were the earliest stories were hopeful ones,” Dihal told me. “Both Greek myths about machines and the automata that they actually made, they’re all presented in quite a positive light. The earliest story, about Hephaestus’s golden handmaidens—they’re pretty much a mix of care bots and personal assistants. They’re meant to make the hard labors of a god much easier—which means they must be quite powerful.”
The earliest impulses to automate, then, were highly aspirational—automated servant bots and robotic caregivers were devices fit for the gods—and cast as boons to their masters. Automation was desired, and not just for care, but for security, and protection. But, Mayor notes, these benefits were reserved solely for the gods—when automated machines came down to earth, the populace was more wary.
“The story of the bronze robot Talos forged by Hephaestus and ‘programmed’ to defend the island of Crete appeared in writing around the same time, in a poem by Hesiod,” Mayor said.
“This was a warrior, essentially the first killer robot,” Dihal said, “it’s not a very positive story depending on which side of the robot you were on, but he was going to protect you from pirates.” Talos wasn’t controversial for his automated qualities, it seems—the boulder-hurling robot sentry was simply a boon to his master, too, and a danger to passersby not sensible enough to have robot armies of their own.
Remember, all of the early robot myths and automata—the drones, servant bots, and killer robots—described above are ancient. Automation is older than Jesus.
The poets who recorded the first iterations of robotica, Mayor said, “were drawing on even older oral traditions, which means that more than 2,700 years ago, people were able to imagine automatons and self-moving devices long before technological inventions made them feasible. So, the classical Greek myths show that animated statues and automatons were thinkable at a surprisingly early date, well before scientific innovations in mechanics existed.”
But were the Greeks unique in dreaming up the concept of automation itself? Probably not, Mayor said. “The impulse to ‘automate,’ to improve nature and amplify human powers, is extremely ancient and widespread.”
“I propose this: we”—humans—“are the beast that automates.” That’s a claim made by Dr. Antone Martinho-Truswell, a zoologist at the University of Sydney. “The bow and arrow is probably the first example of automation,” he wrote. “When humans strung the first bow, towards the end of the Stone Age, the technology put the task of hurling a spear on to a very simple device. Once the arrow was nocked and the string pulled, the bow was autonomous, and would fire this little spear further, straighter and more consistently than human muscles ever could.”
It’s a provocative argument, and, as Martinho-Truswell was quick to note, the latest in a long line of this-is-what-makes-us-different-from-the-animals distinctions. But it’s a useful one, I think, and does go some distance toward explaining humanity’s unique dominance over the global food chain as a product of a) our unique intelligence and b) the universal drive of all organisms to find ways to maximize their gains with minimal energy investment.
“If you think about what any reproducing organism that is under natural selection will do,” Martinho-Turswell told me, “it is going to try to maximize the effect it can get from the minimum investment. That is how you win the evolutionary game. Once we had mental capacities significant enough to produce systems that could do things for us, it is absolutely unsurprising that we totally bought into that paradigm and wanted to automate everything that we could.”
Plenty of animals use tools—even those that zoologists would term unintelligent, like sea urchins, which use tools despite lacking a brain. But even the smartest animals that do so—chimpanzees using a specific part of a stick to feast on termites, say, or Caledonian crows that craft sticks into hooked tools to probe for larvae—do not turn their innovations into automatic systems. As a result, gorillas still spend about half of their time finding enough food to stay alive, while we humans—living in “our layers upon layers upon layers of automated systems”—spend around 10 percent of our time doing so.
Simply put, automation is an evolutionary advantage. “If you can take the resources that you have and come up with some sort of silver bullet and that turns them into radically better efficiency for what you’re getting back, that is going to be evolutionary dynamite,” Martinho-Truswell said. “You’re going to do fantastically well, as we have. Our nearest relatives are all endangered because of us.” And our ability to automate.
He went so far as to posit that automation may be a universal biological impulse, if a species were intelligent enough to commence it.
“We started to pick up on the basic physics of the world around us, and that we could say, set up one object to strike another object and it would do the same thing every time,” Martinho-Truswell said. “You can then combine that kind of consequential thinking and our intelligence with the nearly universal drive to get the business of living as conservatively as possible. And you very quickly arrive at—let’s build things that will work for us.”
“I don’t actually think that would be unique to humans except as much as humans are uniquely intelligent on the planet right now,” he said. If there were another radically intelligent species isolated on an island, and they were pushed in similar ways, they too would automate. “It’s too obvious a solution to the problem of getting the job done without wasting energy to be unique in the grand scheme of things.”
“I just think it’s unique right now to us.”
Which may be why automation shows up in cultures around the globe.
“I have found evidence for similar ‘science fictions’ in ancient India and China, but so much has been lost or destroyed,” Mayor said. “We have the benefit of a body of classical Greco-Roman texts and artworks that managed to survive over millennia, but mythic thought experiments about techno-wonders may have arisen in many pre-modern cultures.” For instance, Ancient Hindu and Sanskrit texts describe a flying palace, known as “Vimana,” that was controlled by thought. And in the 7th century AD, the Chinese Buddhist monk Daoxuan “described a fabulous monastery defended by automatons in the form of men and animals.”
It’s further evidence of a universal impulse to automate. It’s no wonder that the Greeks would exult our labor-saving automatic ingenuity then—it’s what may have given us the edge over the beasts. Automation, humanity’s sacred ability.
It wasn’t until around the Middle Ages, Dr. Dihal said, that we humanfolk began to widely fear automated machines as forces that could run dangerously amok, and not until the Industrial Revolution as agents capable of eliminating jobs our very humanity. This, she notes, was when automation began to threaten the livelihoods of the educated and the upper classes. (Mayor, meanwhile, says that the fear that automation would be used as a tool by tyrants appears early, too, and that humans have been ambivalent about its impacts for nearly as long as we’ve been filled with wonder by them.)
Perhaps we can interpret some of those early myths as celebrations of our capacity to automate, and an implicit assumption that we’d take it further still. And that it would liberate us from drudgery.
“In one remarkable passage, Aristotle drew on the myth of Hephaestus’s fantastic creations to consider what the social and economic implications might be, if only ancient Athens possessed similar automatic objects, such as looms that could weave on their own and lyres that could play themselves,” Dihal told me. “He suggested that automation would abolish slavery.” (Though Aristotle, of course, owned slaves himself.)
Martinho-Truswell mused that we may even be automating past when it’s beneficial for our species’ survival. “For the last 1,000,000 years, the smartest thing to do if you came across a pound of butter was to eat the whole thing, evolutionarily speaking,” he said. “Today, that translates into obesity. Similarly, the impulse to automate may have advanced beyond evolutionary usefulness, and may, in fact, be destructive in its current environment.”
Today, that “positive light” that Dihal describes has largely vanished, as many view automation with, at best, suspicion. Our faith in automating our tools has darkened after past decades have seen the practice wielded by a select few by the employing class—today’s luxuriating gods, perhaps—to cut jobs, increase efficiency, to consolidate capital, and ultimately, exert control over who works and who doesn’t. Hence the fear when it was finally dubbed “automation” by Delmar Harder at Ford in the 40s—when robotization threatened to replace union jobs and push thousands out of the middle class.
There’s certainly no longer a wide sense that automation will yield universal, wondrous benefits. Humans used to want the robots to take our jobs, to see the emergence of AI. It was, it is not hyperbole to say, one of our foundational myths.
“The hope to be relieved from labor,” Dihal said, “to have a life of ease and leisure—it’s the oldest hope associated with artificial intelligence.”
This story is part of Automaton, an ongoing investigation into the impacts of AI and automation on the human landscape. For tips, feedback, or other ideas about living with the robots, I can be reached at firstname.lastname@example.org.