WORKERS TRAINED AND ENGINEERED
[Content warning for this chapter: animal cruelty]
“We have used pigeons, not because the pigeon is an intelligent bird, but because it is a practical one and can be made into a machine, from all practical points of view.”
—B. F. Skinner
One of the most enjoyable experiences in Claire Spottiswoode’s life was the first time she walked through the woods, letting a little bird lead her to honey. Spottiswoode is a zoologist at the University of Cambridge and the University of Cape Town. She’s done extensive field research in the savannas of southern Africa, learning how the Yao villagers communicate with a bird called the honeyguide.
The honeyguide is one of the few avians that can digest wax. To access their food of choice, they have evolved to attract the attention of humans, then lead them to beehives. Once people have harvested the sweet, golden honey, the honeyguide gobbles up the exposed comb and grubs. The birds and humans form a perfect team: the honeyguides are far better at finding beehives, which are often located high in the trees, but they need human help to open them.
The honeyguide-human collaboration goes back to at least the 1500s, but some zoologists think that we’ve searched for beehives together for closer to 1.9 million years. Honeyguides aren’t the only animals that we’ve partnered with—we’ve harnessed animals’ unique skills to help us with tasks for millennia. Some, like the honeyguide, have evolved in a way that happens to be useful to people, and others have been intentionally domesticated and bred to live and work with us, their entire genetic lineages changed in the process.
The reason we’ve partnered with animals is not because they do what humans do. We’ve partnered with them because their skills are different from ours, and because we have much more to gain from combining their strengths with our own. In the same way, technology can and should be a supplement to our own abilities, a way to find the honey we could never reach alone. But this is not how we currently think about robots.
One muggy midsummer day, I stood outside the Baltimore/Washington airport and summoned a ride from the Lyft app on my phone. An older-generation red Prius pulled up nearly immediately and I slid into the back seat, relieved that, despite my delayed flight, I was going to make it to my destination with a few minutes to spare. We cruised down the highway toward Baltimore. My driver, Debbie, was listening to an R&B station, but turned down the volume once the music switched to advertisements so that we could chat. When I told her what I did, she asked the question I’ve discussed with nearly every driver I’ve taken a ride with in countless cities and countries over the past ten years: “How long will it take before I’m replaced by a robot?” We spent the next twenty minutes talking about robots and jobs. Debbie, who was close to retirement age, said that she had heard on the news that all human work would be replaced by robots. She was hopeful she could drive for a few more years and retire before it happened, but she was worried for her grandchildren. Suddenly, Debbie realized that the navigation system had failed and we had gone fifteen minutes out of our way. I wound up being late for my panel, but I was glad that Debbie and I had time to chat.
With a big surge of interest in artificial intelligence and robotics in the past few years, the press is eagerly speculating about our future with robots, with headlines like “Will Robots Steal Your Job?,” “The Robots Are Coming, Prepare for Trouble,” and “Welcome, Robot Overlords. Please Don’t Fire Us?” In 2013, a widely promoted University of Oxford study predicted that almost half of all employment in the United States was at high risk of being replaced by robots and AI within ten to twenty years, and others have predicted even greater vulnerability. Technology is advancing at a breathtaking pace, they say. And robots, the story goes, will soon be able to do everything that humans do, while never tiring, never complaining, and working twenty-four hours a day. A 2017 Pew Research study showed that 77 percent of Americans think that during their lifetime, robots and AI will be able to do many of the jobs currently done by humans. According to Pew surveyor Aaron Smith, most people “are not incredibly excited about machines taking over those responsibilities.”
Not only are we on the cusp of the robot job takeover, say the headlines; some believe the robots will take over more than our jobs. Artificial intelligence, they claim, is on the threshold of outsmarting us. Respected thinkers have raised concerns about artificial superintelligence, predicting that robots could outpace human intelligence and wreak havoc on the world. From Stephen Hawking to Elon Musk, these high-profile individuals have sounded the alarm on what they view as the greatest threat to humanity, fanning the flames of latent fears. It’s easy for people to get on board with the robot takeover narrative, at least in the West. After all, most of our mainstream science-fictional portrayal of robots has been around precisely this topic, from 2001: A Space Odyssey to Ex Machina.
New technologies often inspire concern, but perhaps not quite in the same way as robots. According to tech philosophy and ethics scholars Peter Asaro and Wendell Wallach, our robot narratives throughout history are about good robots turning evil, either turning against their genius creators, like Frankenstein’s monster, or turning against human civilization at large. Is this because robots inherently pose this threat? It’s worth noting that this fear seems culturally specific. Karel Capek’s famous 1920s play about the uprising of robot factory workers was performed in both Western countries and Japan. But while the West embraced its negative messages in our robot narratives, Japan gravitated toward friendlier robot portrayals in popular culture, like the famous cartoon Astro Boy. In the 1960s, Japan began to view robots as a potential driver of productivity and growth, and when robotics played a big role in Japan’s economic revival, it inspired a positive image of robots as nonthreatening and helpful to humans.
Many of my colleagues in robotics are weary of the Western trope that the robots will take all the jobs and become our overlords. The news media often reports on their work in ways that are clickbaity and alarmist, complete with an obligatory picture of the Terminator. I’ve heard curse words directed at the public intellectuals who extol the dangers of robot takeovers, and complaints that the big-name alarmists are mostly physicists, philosophers, and CEOs who don’t have in-depth knowledge of artificial intelligence or robotics. But the Cassandras tend to shoot back that the people who actually work in the field aren’t the best judges of broader trends. One night at a conference, I watched Sam Harris, a writer and philosopher with a degree in neuroscience, get on a small stage in front of about a hundred roboticists from some of the top research centers in the world and argue that artificial superintelligence was a significant and likely danger to humanity, and that the technologists who disagreed weren’t able to see the forest from their position among the trees. The ensuing uproar was monumental.
I was still thinking about his words the next day, as I rested my head against the back of my seat in a fancy black car driving smoothly down the empty early morning highway toward the airport. “What do you think about self-driving cars?” I asked the young, clean-shaven driver in a black suit and tie. He kept his eyes on the road. He told me that he had gone through a year of training to be a professional black-car driver, a lot of which was about more than just driving. He said he was trained to handle unanticipated situations, like protecting his passengers from attacks or violence, and that, if we got into an accident, his first aid skills could save my life. He asked me, solemnly: “Can a robot car do CPR?”
DIRTY, DULL, DANGEROUS
It’s not that robots aren’t capable or smart—like animals, their physical and sensory abilities are often better than ours. But before I get into the animal world, I want to put the current state of robotics into perspective. Because it’s important to understand that robot abilities differ from human abilities in significant ways.
The first practical robot we put to work was a robotic arm called Unimate. Devised by inventor George Devol in the 1950s, Unimate was set up at General Motors in New Jersey to maneuver the blistering hot die-cast car parts that were dangerous for workers to handle. This factory arm was the ancestor of industrial robotics, the technology still used in manufacturing today, and it defined how we would come to view the function of robots in industrial settings.
Robots have classically been delegated jobs that qualify as one of the three Ds: tasks that are dirty, dull, or dangerous for humans. Industrial robots like Unimate ushered in a shift toward automating certain tasks that were high risk or required repetitive grunt work. The machines were accurate, and incredibly strong. Robots could do heavy lifting and take over difficult work in areas with toxic fumes or other health hazards. But they were also fairly crude and limited to very specific tasks, and they themselves were dangerous machinery to be around, necessitating cages and other safety measures to keep humans away.
After the success of welding car parts, the market for industrial robots exploded, as did innovation around what else we could use robots to do. Companies started exploring using industrial robots for tasks like packaging, palletizing, basic transport, and loading. Farming industries also got in on the action. On today’s farms, agriculture robots spray crops, plant seeds, pull weeds, and even deal with the delicate job of picking fruit. After robots permeated our industrial world, it wasn’t long before they made their way into other workplaces.
Figures from George Devol’s patent on the first robot arm, filed in 1954
Copyright © 2021 by Kate Darling