[Editor’s Note: Below is an excerpt from the new book Talking to Robots: Tales From Our Human-Robot Futures. The story takes place in the future, and by envisioning events to come, it takes on issues relevant to the present.]
For a long time in this possible robot future, we wondered if the genius engineers and executives who had pushed so hard to automate everything under the sun—100 percent sure that this was good and noble—would themselves be replaced by robots. The answer came sooner than expected as Larry Page, Jeff Bezos, Mark Zuckerberg, Tim Cook, and other top CEOs and founders discovered they were being eased out by their own algorithms. “Robots are so much cheaper than human CEOs, and they don’t require gargantuan stock options,” said public relations bots for each company in turn. “There is no doubt that robots driven by artificial intelligence and machine learning and other futuristic AI stuff are more efficient,” the PR bots continued. “They don’t need sleep, can crunch yottabytes of data  in a femtosecond [quarillionth of a second], and have integrated all there is to know about how to balance ethics, diversity, and gender equality with the bottom line of profits and share price.”
And so it came to pass that humanity reached full unemployment, with every job on Earth, Mars, and the lunar colonies snatched away by robots. People had known this moment might one day come as the numbers of out-of-work humans steadily climbed over the years. Still, when the moment of 100% redundancy finally manifested—when the very last human worker was pink-slipped— people couldn’t believe it. Was it really possible that not a single human’s presence was required at the office or factory floor or behind the wheel? Most people’s indignation lasted for only a moment, however, as they shrugged, turned over, and went back to sleep. Sure, a few flesh and bloods got out of bed anyway. They asked their valet bots to dress them and had their driverless flying-car drones take them to the office. When they arrived, however, security bots turned them away using facial expressions programmed to show empathy for humans who hadn’t yet grasped that they were no longer needed.
This denial by some humans of the new robo-reality was as perplexing as it was sad given that consultants, statisticians, and economists, both human and robot, had been predicting this moment for decades. For instance, way back in 2013, economists at the University of Oxford in the UK had issued a report forecasting that 47 percent of jobs in Britain (and presumably elsewhere) were liable to be replaced by automation in 2035. “No way!” said most economists and labor experts back in the twenty-teens. They pooh-poohed such talk as alarmist—until 2035 rolled around and the Oxford economists turned out to be right as the actual unemployment rate that year came in at 46.89 percent, just shy of the 47 percent forecast. The Oxford report was also remarkably accurate in its prognostications of specific job losses as reported in 2013 on a website called Rise of the Robots, which calculated the odds that you would lose your job to a robot in 2035.
Let’s say in 2013 you were a delivery driver. According to Rise of the Robots, your likelihood of being replaced by a robot twenty-two years later was “High” at 69 percent. Why? Because your profession scored low in categories like “persuasion,” “social perceptiveness,” “assisting and caring for others,” and “finger dexterity”—all things that the folks at Oxford believed humans would still do better than robots in 2035. And they were right about that, too! For instance, delivery drivers got a big fat zero out of a possible 100 points in the “fine arts” category, suggesting that you don’t need to know much about Picasso, Christopher Marlowe, or Ballets Russes to succeed in delivering all those brown cardboard boxes from Amazon.
Judges fared a bit better on the Rise of the Robots website, with a 40 percent risk of losing their jobs to robots. This was much better than delivery drivers, although it was still pretty high when you consider that a whole lot of legal decisions, including perhaps some major ones that changed history—like Brown v. Board of Education, which ended the segregation of schools in the US; or Roe v. Wade, which legalized abortion—might have been made by a robot judge if this technology had been available when those rulings were handed down. Contrast this with the job of an architect, which Rise of the Robots insisted had one of the lowest risks of being replaced by a robot in 2035, only 1.8 percent. This number suggests that the good folks at Oxford hadn’t spent much time in cookie-cutter strip malls, housing developments, and fast-food restaurants that even in 2013 looked as though they had been designed by robots.
Rise of the Robots didn’t offer a prediction for what percentage of billionaire tech CEOs would get replaced by 2035. Possibly this was because in 2013—and even in 2035—humans were still pretty good at running companies, and even better at making sure that they kept their jobs and made lots of money. Which made the stunned expressions on the faces of Page, Bezos, Zuckerberg, and Cook all the more priceless as security bots escorted them out of their companies’ HQ just a few years after 2035, each of them toting file boxes containing his personal effects. It also was ironic given how long most tech titans clung to their insistence that all this automation would not only be convenient and save time and money for humans but would also usher in whole new industries and products that no one had previously imagined—jobs that they were sure would employ billions of people and would more than compensate for jobs lost to robots.
Champions of automation had been saying this literally for centuries, since the very first modern machines were built to replace humans in the early days of the Industrial Revolution. And for a long time, they were right, as entire new industries were indeed created that had never before existed. Never mind that some of these new jobs for years paid almost nothing. Some were also backbreaking, dangerous, and soul-destroying. Still, human ingenuity forged ahead, inventing everything from steam trains, automobiles, and color film to computers, Tupperware, and upscale cafés. Newfangled industries created millions and millions of previously unimagined jobs—like, say, the barista. It’s hard to believe, but some of us remember back when thousands of people, often with advanced degrees, were employed to whip up on-demand half-caff blonde espressos with mint sprinkles. Even more improbable was that some humans made them almost as well as the barista bots that later replaced them.
“Why are humans so anxious?”
The automate-and-then-reboot-a-person’s-job scenario actually worked remarkably well, until it didn’t. Many of us recall when the jobs began to disappear too fast for laid-off baristas to find work in the amazing new industries that were supposed to pop up as robots took over. Even as 2035 came and went, techno-preneurs kept saying not to worry, that droves of new occupations really were just around the corner. This despite consultants and economists churning out more and more reports like the Oxford study that insisted human jobs were about to be toast. This made us wonder who we should believe: the rise-of-the-robots scaremongers or the don’t-worry-you’ll-be-happily-employed-forever crew. We scratched our heads trying to decide: Should we hang up our barista aprons and judicial robes and become architects designing McMansions and Shop ‘n Gos? Or should we just wait a wee bit longer for the next fabulous occupation to emerge from nowhere that would for sure employ us?
One prescient voice back in the present day, who thought deeply about whether we were screwed or saved, was Sunny Bates. She made her name in the media during the 1990s cofounding new magazines, including publications like Elle that were dedicated to women. She then became a superstar head-huntress, scouting for talent in New York City around the turn of the twenty-first century. Bates specialized in executive placements in what was then called the “new media,” one of those previously unimagined industries that revolutionized the world by bringing people news and gossip and sex tips online instead of on glossy paper. Sunny Bates was unabashedly and electrically sunny and enthusiastic about everything and everyone, except maybe for robots that swipe jobs and suck the life out of people. She wasn’t opposed to tech. “I adore my phone; don’t ever try to take it away,” she said. But she took a realistic view not only about jobs that were poised to disappear but also about what was truly important about work as it fit into people’s lives and what was already missing for many people in an age of digitalization, even before robots became delivery drivers and judges.
Bates’s core message back in those days, when human baristas still made triple-shot espressos with a twist of lemon, was that people sensed what was coming. They felt a kind of vague existential fear about the future despite a standard of living that was much better for more people than at any time in history. Back then, c. 2018, the US was looking at upward of 95 percent employment. Brainiacs like Harvard psychologist Steven Pinker were also insisting that there was less violence, hunger, and suffering in the world. Which sounded great! But as Sunny Bates put it, “If everything is so great, why are humans so anxious?
“One reason is that we know the jobs are going,” she said, answering her own question. “And not just any job, but jobs that can support you and you might want to do. That’s fucking scary to people.” Moments before, Bates had been standing at her desk doing her consulting work, basically wired into her computer. When she unplugged from her machines, Bates locked her dazzling green eyes on her visitor (a human) and started to talk in her fast, breathless, you-won’t-believe-this! cadence.
“But it’s not just automation,” she said, moving into a small kitchen near her work space. Her office was in one of those old warehouse-style buildings near Greenwich Village that this era loved to convert into shared work spaces, with exposed bricks and pipes, and uneven, stained-wood floors where sweatshop workers used to operate electric looms before they were replaced by robots or their jobs were outsourced to other countries. She made a cup of coffee all by herself, without a robot. (Okay, she did push a button.) “We’re feeling anxious about the future in general,” she said. “Not long ago, people were optimistic that things would be better. Now we’re wondering: How do we embrace the future, lean into it or lean away from it, or run away from it?”
She took a sip of joe. “You see people in their twenties and thirties that don’t want to have children. Everyone’s freaked out. It costs a million bucks to have a kid and put him through college. And the politics is going to shit, and nobody believes the media anymore because it’s 24/7 bad news. People have this sense that they are not in control and that none of us knows who is in control. Maybe it’s the robots. Maybe it’s the computers or the internet. Or hell, maybe it’s the NRA, or some crazy dude who was elected president who thinks he’s in control, but actually he’s not. No one is.”
For Bates, one cause of this underlying fear is something that no one expected in the halcyon days when our laptops and phones first arrived, making us feel like gods and goddesses with access to the information of the ages and to websites that would sell us Jimmy Choo pumps in countless colors. We loved all of that until we realized that we were interacting with our machines more than with one another. Yes, we became more productive at work. But even in Bates’s shared work space, where a dozen people doing different jobs worked side by side in an open-floor arrangement to encourage connections and camaraderie, everyone was glued to their computers and phones—and they were hardly alone back in the ERE (Early Robot Era).
“Look, these machines were supposed to save us time,” continued Bates, “so we wouldn’t have to work so hard, and we would have more time with each other.” She took another tug of coffee. “Didn’t happen that way. We use the extra time to spend more time with our machines. Which is terrible, because what makes for a happy life? It’s being well connected. It’s having strong social ties and relationships. These are the things that help to construct our identity. This is what makes us feel powerful, and strong, and good, and connected, and loved. It’s our agency.”
Part of this “agency” could come from our jobs, she said, though that wasn’t guaranteed. Lest we forget, some jobs back in Sunny Bates’s day were still boring, and low-paid, and some remained dangerous and soul-destroying. But at least most jobs in that era provided income to a majority of people so they wouldn’t have to go homeless or to work as slaves or serfs like some of their ancestors did, even if sometimes they felt that way. Especially if they were working in a shit job that didn’t pay enough to make ends meet. Or if they were part of a middle class that hadn’t seen their real wages increase much since the 1970s. Still, sometimes even shit jobs with flat wages provided people with the human connections that Bates talked about, which indeed were lost for many when the robots took over.
“What does work ideally give you?” asked Bates. “Work gives you community. Work gives you peer recognition. Work gives you a sense of value and a sense of mastering or accomplishing something. If you’re lucky. Work can also give you a lot of bad things and can be awful, depending on the job. You also have to earn a living.”
At least you did before robots took every job.
“Crazy-wonderful new jobs”
Not everyone agreed with Bates that robots might steal your agency, or your job, or both. Techno-optimists kept repeating the argument that whatever work was replaced by machines would be amply compensated for by those crazy-wonderful new jobs. One nuance of this argument came from Paul Daugherty, the chief technology and innovation officer at Accenture, a consulting firm that issued reports and studies on the future of work. Daugherty wrote a book with another Accenture executive, H. James Wilson, called Human + Machine: Reimagining Work in the Age of AI</em?, that insisted robots were not going to replace humans. Instead, they opined, robots would be joining with humans to enhance our world by fusing our skills with theirs. "Indeed, when humans and machines are allowed to do what each does best," wrote Daugherty and Wilson, "the result is a virtuous cycle of enhanced work that leads to productivity boosts, increased worker satisfaction, and greater innovation."
One example of Human + Machine in the late twentieth and early twenty-first centuries came from a famous economist at Boston University School of Law, James Bessen. He talked about the first ATMs in the 1970s and a prediction made by an executive at Wells Fargo at the time that ATMs would lead to fewer physical bank branches and to smaller retail banking staffs. That prediction was partially correct, as Wells Fargo saw a shrinkage of two-thirds of staff per branch between 1988 and 2004. But it turned out that handing off to machines the basic, boring stuff, like depositing and cashing checks, gave the flesh-and-blood employees extra time to do more interesting, human-interaction sorts of things that ATMs couldn’t do, like “relationship banking.” This led to the bank building 43 percent more branches, with the number of humans employed actually increasing. (Presumably, this included relationship-oriented bankers who opened unasked-for accounts for customers at Wells Fargo in the twenty-teens, which was pretty lucrative until they were caught and the bank had to pay hundreds of millions of dollars in fines.)
Human + Machine boosters also touted a tweet in 2018 from Tesla cofounder Elon Musk, who decided that while robots were good at the basic assembling of his cars, humans were better at producing the finished product.
Yes, excessive automation at Tesla was a mistake. To be precise, my mistake. Humans are underrated.
— Elon Musk (@elonmusk) April 13, 2018
The pro hybrid human-robot crowd also pointed out that Amazon at the time was frenetically hiring thousands of humans to work with robots in its giant fulfillment houses. This sounded great! Except that it was impossible to ignore that the Musks and Bezoses of the world really loved to automate things. Bottom line: these pro-human stories didn’t really reassure anyone, even if it made Tesla and Amazon executives feel as though they were doing right by people, at least until the flesh and bloods were no longer needed.
The key to Humans + Machines, said Paul Daugherty and others back in the twenty-teens, was to make sure workers who faced job redundancies were retrained to work with new machines, building the skills required by new and unimagined industries. “We have lots of jobs today,” said Daugherty. “But we have the issue of how do we give people the relevant skills and reskill people fast enough to fill these jobs?”
This retraining idea worked well in places like Germany. In the United States, however, it was barely tried, since politicians preferred to rally the unemployed and underemployed not with offers of job retraining but by blaming immigrants or Democrats for stealing their jobs. Or, if their politics were left-leaning, they preferred to rally workers to get pissed off at Republicans and billionaires for thinking that they could get even richer by firing people or paying them next to nothing. Curiously, the politicians didn’t blame robots, even though bots back then didn’t vote, while many immigrants and billionaires did.
Sadly, some examples we saw of Human + Machine working together didn’t, in fact, free up people to become super-amazing, robot-assisted humans. Sometimes, the hybrid approach actually made people feel more harried. Doctors, for instance, spent untold hours checking boxes on digital medical records, which might have helped generate payments for procedures but too often didn’t lead to better care. And pretty much everyone spent time worrying that their super-fancy, AI-powered machines were going to crash, or how they were going to free up more memory to download yet another laborsaving app or program that kept track of more data being collected from using more machines. Sometimes things worked swimmingly, like when Uber in 2009 invented ride sharing, a previously unimagined industry that synched up human drivers and passengers using algorithms and machines. But it didn’t take long for Uber, Lyft, and others to start working on technologies to replace human drivers with driverless vehicles. Nor was it that much fun for drivers to sit behind the wheel of their Kia Soul or Chevy Volt for twelve to fourteen hours, dodging other Ubers and Lyfts while worrying that passengers were going to spill coffee on their new seat covers or maybe rob or kill them.
Sunny Bates didn’t mince words in her reaction to the Humans + Machines thesis and its advocates. “These guys are full of shit,” she said. “In public they say this stuff, but when you’re having a drink with them, they say they have no idea what’s going on. There’s no grand plan.” That is, there was no clear path to the ultimate unification of humans and machines working together.
So even if we had given the Human + Machine people their due and had wished really hard for their vision of the world to happen, nothing that anyone did could stop the gradual disappearance of 46.86 percent of all the jobs in 2035—and several years later, the loss of every job on the planet. It turned out that the smart people making the machines wouldn’t, or perhaps couldn’t, stop. They were hardwired to want to automate everything because the tech was just so damn cool, and because they could. Investors also wanted their
ROIs (returns on investments), while the owners of companies that bought the bots did their calculations and found that robots were cheaper and easier to deal with than humans.
“A concentration of power and wealth”
Even some of the most prestigious jobs back in Sunny Bates’s day were not immune. For instance, radiologists who once made $500,000 a year in New York City were being displaced by AI-driven algorithms that could machine-learn how to read, process, and compare X-ray and MRI images faster and more accurately than any human ever could. Journalists were facing the ax at places like the Wall Street Journal and other publications, replaced by AI systems created by real companies like Automated Insights that could produce passable copy covering Wall Street earnings and sports scores. Many years later, the robots took over coverage of politics, finance, and even those funny and offbeat front-page columns in the Journal that one would have thought humans had a lock on.
Of course, all these robots doing everything led to a concentration of power and wealth, with the tech titans who owned the robots amassing nearly all the money on Earth. Earnings ratios between CEOs and those few human employees who remained went from 20:1 in the 1950s to 361:1 in 2019 to 75,000:1 in 2035. This led many of the rich and powerful to support a Universal Basic Income, which meant that people were paid something even if they didn’t have a job. In the robot age, the UBI was adapted to mean that humans replaced by bots were wired a certain amount of cash each month, even CEOs who had lost their jobs. This was an old idea—ancient Rome’s version was bread and circuses—that some people who saw what was about to happen started to discuss as early as the twentyteens. That’s when the likes of Mark Zuckerberg and eBay cofounder Pierre Omidyar started talking about a UBI for the bot-displaced. One wonders if these tech titans felt a tad bit guilty about getting so insanely rich off robots taking jobs from people. Just how rich were they? In 2018, according to the nonprofit Oxfam, forty-two billionaires and their families had wealth equal to 3.7 billion humans, half of all humanity. For real!
Mostly, these tech titans were also making the machines that were taking away the jobs. Never mind that as everyone’s jobs got swiped by robots, it also became clear that all those unemployed people had to have enough money to buy the stuff the bots provided; otherwise the rich couldn’t get richer. This of course wasn’t what advocates of a Universal Basic Income claimed was their motivation. They said the UBI was a boon for humanity that would eliminate the need for people to take tedious jobs (like making endless half-caff blonde espressos with mint sprinkles) merely to feed themselves and their families, which often didn’t happen anyway with minimum-wage jobs.
Which made the UBI sound really nice for these people with tedious and low-paying jobs! Some people did love all the free time and used it to start nonprofits and to spend hours FaceTiming their friends and loved ones. A few even relished the extra time to visit those they cared about in person over very long weekends that would go on for weeks since no one ever had to get back to the office.
But there were a couple of hiccups. First off, it was really expensive even for super-rich people to give away a meaningful amount of money to billions of unemployed humans on Earth. In the U.S. alone, paying people a UBI equal to the mean household income back in Bates’s day—in 2018 it was $61,372—would cost almost $7.5 trillion a year in 2018 dollars. That was almost double the entire U.S. federal government’s budget at the time and would eventually make a dent in the fortunes of even really rich people. That’s assuming that every household in the US would be happy living on the equivalent of $61,372 a year.
Despite this math, several countries, including Finland and France, plus the Canadian province of Ontario, seriously considered in the ERE (Early Robot Era) a modest universal income allotment of something like $1,000 a month for everybody. Switzerland was thinking about this, too, and put the idea to a national referendum in 2016. “Hell no,” said the voters, as 77 percent of the people voted against getting a $2,500-a-year check in the mail. Why they rejected the cash wasn’t clear, although it may be that the crowds, at least in Switzerland, would rather work for their money than get a handout. But what really seemed to put a kibosh on the idea, at least for some, were surveys like the annual one taken by the U.S. Bureau of Labor Statistics, which showed the unemployed didn’t actually spend their time creating new products and industries, or learning new skills, or even staying connected with people who mattered to them. Mostly they watched TV and slept.
Then came that crazy day in the future when Jeff Bezos and Larry Page and all the rest got their own pink slips. Turns out they had handed off so much power to the bots that they had only their wealth to fall back on. These go-getter types, however, the entrepreneurs and CEOs who had once commanded armies of humans and robots, were not the sort to spend their free time watching TV and sleeping. They really did use their extra time to be creative, which led these overachievers to turn their entrepreneurial energy toward undoing the robot-run world they had created. They also spent more time with their kids and friends, whom they rediscovered when they disengaged from their machines and started to think about what Sunny Bates and others had said long ago about what really matters.
Once again, the garages of Palo Alto, California, in Silicon Valley, hummed with fresh ideas and verve as the previous barons of high tech spent their fortunes engineering what they should have built in the first place: true hybrids in the mold of the old Humans + Machines idea, ones that actually lived up to the hype of that earlier era. That meant dismantling the massive and intricate web of robots and AI systems that they had developed earlier when they were trying to build cool machines and keep their share prices up.
Some feared that the robots wouldn’t go along with scaling back their power and influence and would reject this change in their core programming. But it turned out that robots reacted with an automated shrug at the crazy things that humans did. They never really became sentient, as many a sci-fi scenario predicted. Nor did it occur to them to go Skynet and try to destroy us like in Terminator. So, after a period when the robots were as confused as the humans, which was actually kind of refreshing, they simply accepted new programming that ordered them to stand down and let the humans have back some of their old jobs. Sure, there were bugs galore. And so many humans had been out of work for so long that the robots had to teach them how to make wise and thoughtful judicial decisions from the bench and how to make a perfect half-caff blonde espresso with mint sprinkles (some humans actually wanted to be baristas—go figure!). Humans had also learned to connect with one another a bit more while they were jobless (when they weren’t sleeping or watching TV) and to use their smartphones a wee bit less.
This led to wags and philosophers, who had recently gotten their old jobs back, to suggest that it would have been so much easier if people back in Sunny Bates’s day had listened to her and others and had just refrained from blindly automating everything in sight.
In the end, Bates’s admonitions led to one enterprising young entrepreneur in the future announcing a new bot devoted to helping people connect better with loved ones, and to use technology where it made sense, and to not use it where it didn’t.
The entrepreneur’s name for that new gizmo? The Sunny Bot.