The pandemic appears to be quietly accelerating the automation of jobs. Nearly two-thirds of senior executives surveyed last year said they were increasing investment in automation and artificial intelligence amid the crisis. And some of the greatest growth has been in jobs—such as delivery app and warehouse work— directed by computers or so optimized that the human component is effectively robotic. 

Concerned about what this means for society, or how you and your offspring might navigate this? New York Times tech columnist Kevin Roose aims to answer those questions in his book Futureproof, due out next week. 

(You can read all of our book briefings here.)

Roose assesses the arguments about the impact of automation and AI on workers, concluding that almost no job is inherently protected from being automated away. And he challenges tech optimist dogma, such as the idea that humans and AI will collaborate rather than competing. (Research has repeatedly shown that AI is more effective alone than when paired with humans.) 

What concerns him even more, however, is the enthusiasm that executives seem to have for doing so in the pursuit of cost savings, even when it results in job losses and little improvement in the work being done. 

Roose recounts the story of one AI startup founder selling “Boomer Remover” automation that would allow companies to lay off “old, overpaid middle managers.” And he describes how Infosys president Mohit Joshi told him clients were discussing shedding 99% of their workforces through automation.

Chillingly, Roose notes that remote workers “are already halfway automated,” showing up as Slack avatars or little boxes on Zoom, their value measured by tasks completed, and their ability to contribute in subtler, more human ways to an organization greatly circumscribed.  

In the face of all this, he lays out a central line of inquiry for the book: “What can humans do much, much better than even our most advanced AI?” 

Among Roose’s answers:

  • Humans can handle surprises better. Occupational therapists, police detectives, and emergency-room nurses, for example, work amid changing and unpredictable conditions.
  • Humans are better at meeting social needs. The travel agents who have survived the automation of vacation booking are those focused on providing unique experiences for travelers, such as wilderness adventures. “Jobs that make people feel things are much safer than jobs that simply involve making or doing things,” he writes. (p. 71)
  • Humans are better at jobs that Roose describes as “scarce,” which are not needed on a regular basis or involve extraordinary talent. Top athletes, performers, and chefs are in this category. “If someone would pay to watch you do your job, it’s probably safe,” he notes. (p. 73)

Roose proposes nine rules for navigating automation. Among them:

  • Leave handprints. People still value human artisanship and service, which is why we’re willing to pay a premium for handmade ceramics or to be served a cappuccino in a cafe. 
  • Don’t be an endpoint. When you’re in a job overseen by a machine, such as gig work or a warehouse, you’re vulnerable to automation.   
  • Treat AI like a chimp army. AI makes mistakes, breaks things, is prone to replicating biases—so we should be careful about giving it more power than it can handle.
  • Build big nets and small webs. Social safety nets help people through the dislocations brought by automation, and community-level support can further help.
  • Learn machine-age humanities. Roose catalogs skills essential for the future, including the ability to remain focused amid distraction, to rest and stave off burnout, to be discerning about information, and to anticipate future consequences of technology.
  • Arm the rebels. Roose contrasts Henry David Thoreau, who retreated to the woods in the face of industrialization, with his 19th-century contemporary Sarah Bagley, a labor activist. He argues for activism in the face of automation. 

To be sure…

  • Many of Roose’s recommendations implicitly assume the reader has the power to switch jobs or control their work conditions, something that’s likely not true for many low-wage workers. 
  • Roose argues that it’s ultimately up to humans to decide how to use the technology. But it’s unclear that those with the power to do so will moderate its impact on other people. And it’s also not certain that a broad shift to the work he identifies as uniquely human is possible, or if there’s enough of it to go around.
  • Roose dismisses the idea that technology careers should be the focus of education, referring to “STEM supremacists.” But, at least for now, fields such as data science are among the fastest growing, and such areas are sure to evolve. 
  • The most useful part of the book is the five-page appendix, where Roose shares his own goals based on the nine rules in the book. They’re very thoughtful and specific.
  • Arguably the second most useful part is Roose’s description of how he reduced his reflexive focus on his phone—tangential to the question of automation—accomplished with the help of a coach and a rubber band wrapped around it. 

Memorable anecdotes and facts:

  • “Phubbing” is a neologism for “phone snubbing,” avoiding interacting with someone else and instead using your phone. (p. 103)
  • Many of us have an aversion to being alone with our thoughts. In one experiment, 71% of men and 26% of women gave themselves an electric shock rather than sit by themselves in an empty room for 10 or more minutes. (p. 108)
  • People told researchers that automated massage chairs were more pleasurable when they believed a human had pressed the button to activate them. Research subjects in another experiment believed candy to be better tasting when told a human had selected it for them. (p. 123) 
  • Best Buy was struggling to compete against ecommerce companies when a new CEO, Hubert Joly, took over in 2012 and prioritized expert human customer service, paving the way for the company’s recovery. (p. 129)
  • Under Japan’s shukko practice, workers destined to be laid off can be at least temporarily loaned to another company. (p. 161)

Choice quotes:

  • “AI and automation were working well for some people—namely, the executives and investors who built and profited from the technology—but that they weren’t making life better for everyone.” (p. xvii)
  • “Executives, not algorithms, decide whether to replace human workers.” (p. xxvi)
  • “What the optimistic miss is that we don’t live in the aggregate, or over the longer term,” Roose writes, countering the assumption that automation will be a net positive. “We experience major economic shifts as individuals with finite careers and lifespans, and for many people, technological change hasn’t always resulted in better material conditions during their lives.” (p. 18)
  • “When it comes to avoiding machine replacement, what we do is much less important than how we do it.” (p. 35)
  • “In many of today’s workplaces, AI has been promoted to middle management. In industries from service to banking and food service, software now does the supervisory work of training workers, monitoring quality, and reviewing performance.” (p. 45)
  • “We may want to stop worrying about killer droids and kamikaze drones, and start worrying about the mundane, mediocre apps and services that allow companies to process payroll 20% more efficiently, or determine benefits eligibility with fewer human caseworkers.” (p. 57)
  • “If you could write a user’s manual for your job, give it to someone else, and that person could learn to do your job as well as you in a month or less, you’re probably going to be replaced by a machine.” (p. 69)
  • “Humans still want role models, and we want to be inspired by human greatness. That is why we cheer for Olympic swimmers, even though speedboats go faster.” (p. 73)
  • “The biggest risk of remote work, when it comes to automation, is that it’s much harder to display your humanity in the absence of face-to-face interaction.” (p. 140)

Roose has a great eye for stories and details and knows the tech landscape well, making this an easy, interesting book to read. Futureproof engages with the central questions about the future of work, and what individuals and employers can do about it. With automation accelerating, such questions are even more urgent.

You can pre-order at Bookshop.org or Amazon. (We may make a commission on any purchase.) All page numbers referenced above are for the hardcover edition.

You can listen to a podcast where Roose discusses the book

The handbook for this new era of business doesn’t exist. We’re all drafting our own as we go along—and now we’d like to start doing so together. You can sign up here to receive this briefing by email. You can read all of our book briefings here.