Will We Really Lose Half our Jobs to Automation?



The following is an excerpt from GigaOm publisher Byron Reese’s new book, The Fourth Age: Smart Robots, Conscious Computers, and the Future of Humanity. You can purchase the book here.

The Fourth Age explores the implications of automation and AI on humanity, and has been described by Ethernet inventor and 3Com founder Bob Metcalfe as framing “the deepest questions of our time in clear language that invites the reader to make their own choices. Using 100,000 years of human history as his guide, he explores the issues around artificial general intelligence, robots, consciousness, automation, the end of work, abundance, and immortality.”

One of those deep questions of our time:

When the topic of automation and AI comes up, one of the chief concerns is always technology’s potential impact on jobs. Many fear that with the introduction of wide-scale automation, there will be no more jobs left for humans. But is it really that dire? In this excerpt from The Fourth Age, Byron Reese explores the prospect of massive job loss due to automation.


The “jobs will be destroyed too quickly” argument is an old one as well. In 1930, the economist John Maynard Keynes voiced it by saying, “We are being a afflicted with a new disease . . . technological unemployment. This means unemployment due to our discovery of means of economising the use of labour outrunning the pace at which we can find new uses for labour.”

In 1978, New Scientist repeated the concern:

The relationship between technology and employment opportunities most commonly considered and discussed is, of course the tendency for technology to be labour-saving and thus eliminate employment opportunities—if not actual jobs.

In 1995, the refrain was still the name. David F. Noble wrote in Progress without People:

Computer-aided manufacturing, robotics, computer inventories, automated switchboards and tellers, telecommunication technologies—all have been used to displace and replace people, to enable employers to reduce labour costs, contract-out, relocate operations.

But is it true now? Will new technology destroy the current jobs too quickly?

A number of studies have tried to answer this question directly. One of the very finest and certainly the most quoted was published in 2013 by Carl Benedikt Frey and Michael A. Osborne, both of Oxford University. The report, titled The Future of Employment, is seventy-two pages long, but what has been referenced most frequently in the media is a single ten-word phrase: “about 47 percent of total US employment is at risk.” Hey, who needs more than that? It made for juicy and salacious headlines, to be sure. It seemed as if every news source screamed a variant of “Half of US Jobs Will Be Taken by Computers in Twenty Years.”

If we really are going to lose half our jobs in twenty years, well, then the New York Times should dust off the giant type it used back in 1969 when it printed “MEN WALK ON MOON” and report the story on the front page with equal emphasis. But that is not actually what Frey and Osborne wrote. Toward the end of the report, they provide a four-hundred-word description of some of the limitations of the study’s methodology. They state that “we make no attempt to estimate how many jobs will actually be automated. The actual extent and pace of computerisation will depend on several additional factors which were left unaccounted for.”

So what’s with the 47 percent figure? What they said is that some tasks within 47 percent of jobs will be automated. Well, there is nothing terribly shocking about that at all. Pretty much every job there is has had tasks within it automated. But the job remains. It is just different.

For instance, Frey and Osborne give the following jobs a 65 percent or better chance of being computerized: social science research assistants, atmospheric and space scientists, and pharmacy aides. So what does this mean? Social science professors will no longer have research assistants? Of course they will. They will just do different things, because much of what they do today will be automated. There won’t be any more space scientists? Pharmacists will no longer have anyone helping them?

Frey and Osborne say that the tasks of a barber have an 80 percent chance of being taken over by AI or robots. In their category of jobs with a 90 percent or higher chance of certain tasks being computerized are tour guides and carpenters’ helpers.

The disconnect is clear: some of what a carpenter’s helper does will get automated, but the carpenter helper job won’t vanish; it will morph, as almost everyone else’s job will, from architect to zoologist. Sure, your iPhone can be a tour guide, but that won’t make tour guides vanish.

Anyone who took the time to read past the introduction to The Future of Employment saw this. And to be clear, Frey and Osborne were very up-front. They stated, in scholar-speak, the following:

We do not capture any within-occupation variation resulting from the computerisation of tasks that simply free-up time for human labour to perform other tasks.

In response to the Frey and Osborne paper, the Organization for Economic Cooperation and Development (OECD), an intergovernmental economic organization made up of nations committed to free markets and democracy, released a report in 2016 that directly counters it. In this report, entitled The Risk of Automation for Jobs in OECD Countries, the authors apply a “whole job” methodology and come up with the percent of jobs potentially lost to computerization as 9 percent. That is pretty normal churn for the economy.

At the end of 2015, McKinsey & Company published a report entitled Four Fundamentals of Workplace Automation that came to similar conclusions as the OECD. But again, it had a number too provocative for the media to resist sensationalizing. The report said, “The bottom line is that 45 percent of work activities could be automated using already demonstrated technology,” which was predictably reported as variants of “45% of Jobs to Be Eliminated with Existing Technology.” Often overlooked was the fuller explanation of the report’s conclusion:

Our results to date suggest, first and foremost, that a focus on occupations is misleading. Very few occupations will be automated in their entirety in the near or medium term. Rather, certain activities are more likely to be automated, requiring entire business processes to be transformed, and jobs performed by people to be redefined, much like the bank teller’s job was redefined with the advent of ATMs.

The “47 percent [or 45 percent] of jobs will vanish” interpretation doesn’t even come close to passing the sniff test. Humans, even ones with little or no professional training, have incredible skills we hardly ever think about. Let’s look closely at two of the jobs at the very top of Frey and Osborne’s list: short-order cook and waiter. Both have 94 percent chance of being computerized.

Imagine you own a pizza restaurant that employs one cook and one waiter. A fast-talking door-to-door robot salesman manages to sell you two robots: one designed to make pizzas and one designed to take orders and deliver pizzas to tables. All you have to do is preload the food containers with the appropriate ingredients, and head off to Bermuda. The robot waiter, who understands twenty languages, takes orders with amazing accuracy, and flawlessly handles special requests like “I want half this, half that” and “light on the sauce.” The orders are sent to the pizza robot, who makes the pizza with speed and consistency.

Let’s check in on these two robots on their first day of work and see how things are going:

  • A patron spills his drink. The robots haven’t been taught to clean up spills, since this is a surprisingly complicated task. The programmers knew this could happen, but the permutations of what could be spilled and where were too hard to deal with. They promised to include it in a future release, and in the meantime, to program the robot to show the customers where the cleaning supplies are kept.
  • A little dog, one of those yip-yips, comes yipping in and the waiter robot trips and falls down. Having no mechanism to right itself, it invokes the “I have fallen and cannot get up” protocol, which repeats that phrase over and over with an escalating tone of desperation until someone helps it up. When asked about this problem, the programmers reply, snappishly, that “it’s on the list.”
  • Maggots get in the shredded cheese. Maggoty pizza is served to the patrons. All the robot is trained to do with customers unhappy with their orders is to remake their pizzas. More maggots. The robots don’t even know what maggots are.
  • A well-meaning pair of Boy Scouts pop in to ask if the pipe jutting out of the roof should be emitting smoke. They say they hadn’t noticed it before. Should it be? How would the robot know?
  • A not-well-meaning pair of boys come in and order a “pizza with no crust” to see if the robots would try to make it and ruin the oven. After that, they order a pizza with double crust and another one with twenty times the normal amount of sauce. Given that they are both wearing Richard Nixon masks, the usual protocol of taking photographs of troublesome patrons doesn’t work and results only in a franchise-wide ban of Richard Nixon at affiliated restaurants.
  • A patron begins choking on a pepperoni. Thinking he must be trying to order something, the robot keeps asking him to restate his request. The patron ends up dying right there at his table. After seeing no motion from him for half an hour, the robot repeatedly runs its “Sleeping Patron” protocol, which involves poking the customer and saying, “Excuse me sir, please wake up” repeatedly.
  • The fire marshal shows up, seeing the odd smoke from the pipe in the roof, which he hadn’t noticed before. Upon discovering maggot-infested pizza and a dead patron being repeatedly poked by a robot, he shuts the whole place down. Meanwhile, you haven’t even boarded your flight to Bermuda.

This scenario is, of course, just the beginning. The range of things the robot waiter and cook can’t do is enough to provide sitcom material for ten seasons, with a couple of Christmas specials thrown in. The point is that those who think so-called low-skilled humans are easy targets for robot replacement haven’t fully realized what a magnificently versatile thing any human being is and how our most advanced electronics are little more than glorified toaster ovens.

While it is clear that we will see ever-faster technological advances, it is unlikely that they will be different enough in nature to buck our two-hundred-year run of plenty of jobs and rising wages. In one sense, no technology really compares to mechanization, electricity, or steam engines in impact on labor. And those were a huge win for both workers and the overall economy, even though they were incredibly disruptive.


To read more of GigaOm publisher Byron Reese’s new book, The Fourth Age: Smart Robots, Conscious Computers, and the Future of Humanity, you can purchase it here.





Leave a Reply