The Zero Marginal Cost Society: The Internet of Things, the Collaborative Commons, and the Eclipse of Capitalism (19 page)

That goal is now within sight with the introduction of driverless vehicles. The prospect of eliminating human driving with the substitution of driverless vehicles operating on smart roads, once relegated to the realm of science fiction, will soon be operational. There are currently over 2.7 million truck drivers in the United States alone.
23
By 2040, driverless vehicles, operating at near zero marginal labor costs, could eliminate much of the nation’s truckers. (Driverless vehicles will be discussed in greater detail in chapter 13.)

Automation, robotics, and artificial intelligence are eliminating human labor as quickly in the white-collar and service industries as in the manufacturing and logistics sectors. Secretaries, file clerks, telephone operators, travel agents, bank tellers, cashiers, and countless other white-collar service jobs have all but disappeared in the past 25 years as automation has driven the marginal cost of labor to near zero.

The Hackett Group, a consultant on back-office employment, estimates that 2 million of the jobs lost in the United States and the European Union since the onset of the Great Recession have been in human resources, finance, information technology and procurement, and that more than half of those job losses were the result of technology displacement due to automation.
24

Automation is also making deep inroads into the retail sector, which employs one out of every ten Americans. While back-office work, warehousing, and shipping were surefire candidates for automation, observers of the retail industry long assumed that salespeople, at least, would be spared the ax because of the very social nature of their relationships with customers. Wishful thinking.

Vending machines and kiosks now sell everything from bathing suits to iPods and even gold coins. In 2010, $740 billion in retail sales were transacted by self-service machines. Industry watchers expect that figure to jump to $1.1 trillion by 2014.
25

Walmart already has self-checkout terminals and plans to install 10,000 additional terminals in more than 1,200 of its stores by the end of 2013. The giant retailer is also expanding its Scan and Go self-checkout system in 40 of its stores in the Denver, Colorado, area. Shoppers scan the bar code of the products they take off the shelves on their iPhone app before putting them in their shopping carts. When they are finished shopping, they press the “done” button and the app provides them with
a custom QR (quick response) code. The self-checkout terminal scans the QR code on the smart phone, adds up the price of the items, and asks the customer to pick a payment option.
26

Despite efforts by brick-and-mortar retailers to automate more and more of their operations to reduce their labor costs, they continue to lose ground to online retailers whose marginal labor costs are heading to near zero. On the surface, brick-and-mortar sales appear healthy, if not robust. They made up 92 percent of retail sales in 2011, versus only 8 percent online.
27
It’s only when we probe a little deeper and look at growth rates that we begin to see the handwriting on the wall. According to the National Retail Federation, brick-and-mortar sales are only growing at 2.8 percent per year while online retailers are growing at 15 percent per year, raising the question of how long physical retailing, with its significant fixed costs and sizable payrolls, can compete with much lower marginal labor costs online.
28
The casualty list is already mounting. Borders books and Circuit City, once giant brick-and-mortar retailers, have already fallen victim to the low marginal labor costs of online retailing. With online retail stores expected to double by 2020, many more brick-and-mortar retailers, already stretched by falling profit margins, are likely to succumb to virtual retailing.
29

Brick-and-mortar retailers are caught in a bind, unwittingly becoming free showroom spaces for customers to browse and handle products and clothes that they then purchase online. Price-check iPhone apps allow customers to scan a product in the store and then compare it on the spot to prices online, where they are sure to get the item at a cheaper price at Amazon or other virtual retailers—often with free shipping thrown in.

Some brick-and-mortar retailers are fighting back against “fit-lifters” who try on clothes and shoes in their stores to get the right size and then order them online. Gary Weiner, the owner of Saxon Shoes in Virginia and a board member of the National Shoe Retailers Association, shares the concern of a growing number of retailers who resent what they call showrooming—the practice of online customers using stores to try on things they then buy online. Weiner says that it’s common for young people to come in the store and say, “My mother sent me in to get my size fitted so she can buy them online.”
30
A few stores are beginning to charge a browser fee to discourage fit-lifting. Other retailers worry, probably rightly so, that if they charge for browsing and fitting it will only result in consumers abandoning their stores altogether.
31

A number of retailers are attempting an accommodation with online shopping by encouraging their customers to buy from their online shop and pick up their merchandise in the brick-and-mortar store—so the stores become, in effect, mini-distribution centers. Still, such efforts are likely to be only a stopgap measure because of the steep overhead costs in maintaining brick-and-mortar operations.

Many of the big-box retailers, including Best Buy, Target, and Walmart, will likely attempt to get ahead of the curve by pushing more
of their business online. Others—especially traditional department stores like Macy’s, Nordstrom, and Neiman Marcus—are going to pare down or simply die off as more and more retail goes virtual. Online clothiers already offer virtual fitting. Online customers can create a virtual model of themselves, providing information on their size, gender, age, chest, waist, and hip size. Using a mouse, the customer can even check the fitting from different angles.

A growing number of retail industry analysts are forecasting the imminent death of large segments of the brick-and-mortar retail trade. Jason Perlow, technology editor at ZDNet, says that convenience stores like 7-Eleven, drug stores like Walgreens, and supermarket chains like Kroger will continue to keep their doors open, along with high-end specialty and luxury stores like Crabtree & Evelyn, and a few big box stores like Walmart. Much of the brick-and-mortar retail business, however, is going to shrink, especially as a younger generation weaned on purchasing online comes of age.

Perlow says that while brick-and-mortar retail will not disappear, in “ten years hence [the] retail footprint will be a shadow of its former self and heavy competition from online will allow only the strongest brick-and-mortar businesses to survive.”
32

As in other industries where automation is quickly reducing human labor, virtual retailing is following suit. At best we can say that the future does not look good for the 4.3 million workers in brick-and-mortar retail as we move closer to near zero marginal labor costs and a near workerless world.
33

Even Knowledge Workers are Expendable

By 2005, the anecdotal evidence of automation replacing workers in the manufacturing and service industries was no longer an object of curiosity. Automation had become pervasive. Everywhere we turned, it seemed, workers had disappeared and we found ourselves surrounded by intelligent-machine surrogates talking to us, listening to us, directing us, advising us, doing business with us, entertaining us, and watching over us. Early on, the experience of a workerless presence was often amusing, sometimes irksome, and even eerie. Now it’s commonplace. Still, it was not until around 2010 that an avalanche of new books came out with alarming titles like
The Race Against the Machine
,
Light at the End of the Tunnel
, and
Automate This
, warning of automation’s impact on jobs. Their authors took to the talk shows, and their message of a coming workerless world began to gain attention in social media outlets, even attracting some comments from policy makers, think tank researchers, economists, and President Barack Obama.

We are just beginning to hear the rumble of what’s likely to become a global policy debate on automation and the future of jobs. In part, that
discussion is starting to happen because of the jobless recovery that followed the Great Recession. The disconnect between a rising GDP and diminishing jobs is becoming so pronounced that it’s difficult to continue to ignore it, although I’m still somewhat amazed at how few economists, even at this stage, are willing to step forward and finally acknowledge that the underlying assumption of classical economic theory—that productivity creates more jobs than it replaces—is no longer credible.

The other reason I suspect that the great automation debate may be about to take flight is that the new innovations in the use of Big Data, the increasing sophistication of algorithms, and advances in AI are, for the first time, crawling up the skill ladder and affecting professional work itself, long considered immune from the forces of automation and the advances of technology displacement. Computers are being programmed to recognize patterns, advance hypotheses, self-program responses, implement solutions, and even decipher communication and translate complex metaphors from one language to another in real time with accuracy approaching that of the best translators in the world.

Advances in AI are now being employed across a range of professional disciplines to increase efficiency and productivity and reduce human labor. EDiscovery is a software program that can sift through millions of legal documents, looking for patterns of behavior, lines of thought, concepts, and the like, at speeds that would trump the best Harvard-educated lawyers and with crisp analysis that even the most well-trained legal scholars might miss. The savings in labor cost is equally impressive.

The
New York Times
journalist John Markoff cites the example of a blockbuster lawsuit in 1978 involving five television studios, the U.S. Justice Department, and CBS. The studios’ lawyers and paralegals had the unenviable task of reading through more than 6 million documents over months at a cost of $2.2 million in labor time. In January 2011 BlackStone Discovery, a Palo Alto, California, enterprise, analyzed 1.5 million legal documents using eDiscovery software for less than $100,000. Bill Herr, an attorney at a U.S. chemical company who used to pack an entire army of lawyers in an auditorium to read documents for weeks at a time, says that “from a legal staffing viewpoint, it means that a lot of people who used to be allocated to conduct document review are no longer able to be billed out.”
34
Mike Lynch, founder of Autonomy, another eDiscovery firm, calculated that with the new search software, one lawyer can do the work of 500 lawyers, and with greater accuracy. Using eDiscovery software, Herr found only 60 percent accuracy when attorneys were doing the research, leading him to gripe: “Think about how much money had been spent to be slightly better than a coin toss.”
35

Very few professional skills are being spared the long arm of IT and Big Data–crunching by algorithms. Knowledge workers of every stripe and variety—radiologists, accountants, middle managers, graphic designers, and even marketers—are already feeling the heat as pattern-recognition
software begins to penetrate every professional field. Mike McCready is the head of a startup company called Music Xray, a firm that uses Big Data and algorithms to identify potential musical hits. The company, which has secured recording contracts for more than 5,000 artists in less than three years, uses sophisticated software to compare the structure of a song to songs previously recorded to assess its potential to break out and hit the charts. His company already has an impressive track record of spotting songs by unknown artists and accurately predicting their success. A similar software program developed by Epagogix analyzes movie scripts to project box office hits for the film industry.
36
Its success in identifying winners has made algorithm assessment standard fare in the industry. In the future, these kinds of forecasting tools will eliminate the need to hire pricey marketing agents to conduct expensive focus-group encounters and other marketing-research initiatives, the accuracy of which might pale against the crowdsourcing accuracy of Big Data filtered by algorithms.

Big Data and algorithms are even being used to create copy for sports stories that are chatty, chock full of information, and engaging. The Big Ten Network uses algorithms to create original pieces posted just seconds after games, eliminating human copywriters.
37

Artificial intelligence took a big leap into the future in 2011 when an IBM computer, Watson—named after IBM’s past chairman—took on Ken Jennings, who held the record of 74 wins on the popular TV show
Jeopardy
, and defeated him. The showdown, which netted a $1 million prize for IBM, blew away TV viewers as they watched their
Jeopardy
hero crumble in the presence of the “all-knowing” Watson. Watson is a cognitive system that is able to integrate “natural language processing, machine learning, and hypothesis generation and evaluation,” says its proud IBM parent, allowing it to think and respond to questions and problems.
38

Watson is already being put to work. IBM Healthcare Analytics will use Watson to assist physicians in making quick and accurate diagnoses by analyzing Big Data stored in the electronic health records of millions of patients, as well as in medical journals.
39

IBM’s plans for Watson go far beyond serving the specialized needs of the research industry and the back-office tasks of managing Big Data. Watson is being offered up in the marketplace as a personal assistant that companies and even consumers can converse with by typed text or in real-time spoken words. IBM says that this is the first time artificial intelligence is graduating from a simple question-and-answer mode to a conversational mode, allowing for more personal interaction and customized answers to individual queries.
40