"What exactly is a "robot"? We might envision a Star Wars-like robot in the vein of R2-D2 or C-3PO, but a robot is any device or algorithm that does what humans once did, from mechanical combines and thermostats to dishwashers and airfare search sites." - Scott A. Wolla  (2018) Will Robots Take Our Jobs?

Eleven Reasons Why Robots will Not Make Jobs Obsolete:

  1. The law of demand: "As the price of something decreases, the quantity demanded of that good increase" (Scott A. Wolla, 2018).  Because AI technology reduces a firm's cost of producing goods/services,  firms are able to sell their product at a lower price, increasing demand for the product and releasing funds to ramp up production,  improve product quality,  expand marketing efforts and/or branch out into new products. This increases consumer demand for the firm's goods/services, increasing its need for labor, especially for workers who can monitor, manage, or otherwise exploit the AI technology. 
  2. Because AI technology reduces a firm's cost of producing goods/services, the firm will convert the savings into lower prices, higher wages and/or additional profits for shareholders, which has ripple effects throughout the economy by increasing overall supply and demand for goods/services, which in turn increases employment. As the McKinsey Global Institute noted, “new technologies have spurred the creation of many more jobs than they destroyed, and some of the new jobs are in occupations that cannot be envisioned at the outset; one study found that 0.56 percent of new jobs in the United States each year are in new occupations. Most jobs created by technology are outside the technology-producing sector itself.” For example, Graetz and Michaels (2015) examined the effect of industrial robots in 17 countries and found no negative impact on the total hours worked or demand for labour.
  3. Incorporating new technologies takes time, money, and a reliable revenue stream. Technical capability is not the same thing as economic feasibility. That's probably the number one reason why AI folks keep saying robots will take over and economists say nuh-uh.  Very often technological substitution does not take place as expected (Arntz and Zierahn, 2016).  For example, smart production machinery (aka "robots") can take several years to pay off - rather daunting for small to mid-sized employers who are just trying to survive the current competition and next business cycle. So making businesses put off making the investment in smart machines.
  4. Speaking of business cycles, one advantage of human workers is you can lay them off when times get tough. In a word: flexibility. One disadvantage of robots is you still have to make payments on that business loan you took out to pay for them, regardless of business conditions.
  5. Workers can take on new tasks when automatized tasks are removed from their job description.  For example, receptionists may take on additional administrative support or customer service tasks. This is what happened in the 1980s and 1990s as computers and computerized equipment became indispensable to a broad range of occupations. As noted by Alexandra Spitz-Oener (2006): "...most of the adjustment to the past computerization occurred through changing task structures within occupations, rather than changing employment shares between occupations."
  6. Very often the new tasks assigned to workers whose jobs have been partly automatized are newly created tasks, not just a reassignment of what used to be done by other employees. That's because technology expands the realm of the doable. In other words, when more is possible, more will be done to exploit the possible. For example, a receptionist may be assigned the task of updating the employer's blog or creating lists of marketing targets from the company's customer database.
  7. Technological change will generate additional jobs through demand for new products, such as ever more efficient and human-like personal assistants or home health robots. New products will create more jobs in engineering, sales, marketing, management, administration, customer service, maintenance/repair, etc. 
  8. Automation usually targets certain tasks rather than whole occupations, but some non-automatable tasks may be essential to the occupation and the employer (Arntz and Zierahn, 2016). If the nature of the occupation is such that its non-automatable tasks cannot become part of another job, the occupation will remain, relatively unscathed by automation.  For example, the job of truck driver is not just about driving from point A to Point B, something a fully autonomous vehicle could do (at least in theory).  Drivers also have to get in and out of trucks in order perform roadside checks and repairs, adjust loads that have become unbalanced, and unload cargo for direct deliveries. If truck drivers have extra free time when their vehicle is self-driving, they may be able to take on additional tasks in the truck cab, such as update records, contact current customers to discuss future needs, make sales calls, and schedule upcoming deliveries.     
  9. New technologies require qualified personnel to program, monitor, manage, and maintain them. Introduction of new technologies requires the supply of complementary skills in the labour market. That takes time and money.  In the meantime, smaller employers may prefer incremental technical improvements that are cheaper to implement and easier for existing employees to master. Gradual automatization of the workplace will allow more time for employers and workers to transition to new occupations, job tasks, and products.
  10. There will still be a market for tasks and services provided by humans. For example, no matter how competent your friendly robot nurse, nothing beats the human touch. Robots may eventually provide some types of medical care and treatment, but they are more likely to complement than replace healthcare workers. Just think of them as new members of the healthcare team. 
  11. A surprisingly large number of job tasks are hard to automatize. According to McKinsey Global Group (2017), these include:
  • Managing and developing people.
  • Applying expertise to decision making, planning, and creative tasks.
  • Interfacing with stakeholders, such as creditors, employees, government agencies, owners, suppliers, and the community from which the business draws its resources.
  • Performing physical activities and operating machinery in unpredictable environments.

How many occupations have a significant number of job tasks that are hard to automatize?  Difficult to say, but one way to explore this question is make use of the Department of Labor's Occupational Information Network (O*NET), which includes an extensive database on close to 1000 occupations. O*NET will generate lists of occupations in response to various queries, such as which occupations require high levels of complex problem solving.  I tried out several queries related to job qualities and requirements that would make an occupation hard to automatize (e.g., troubleshooting, judgment and decision-making) and each time got a list of over 500 occupations.  Don't take my word for it - explore O*NET for yourself. Here's the link for Advanced Search: https://www.onetonline.org/search/.    

Of course, some occupations will disappear or at least go into a steep decline as more and more job tasks become automated. But I tend to agree with economists Sar Levitan and Clifford Johnson, whose 1982 paper The future of work: Does it belong to us or to the robots ends with the following prediction:  "the period of adjustment which lies ahead may not be painless, but it seems that work is here to stay."

--

References:

A Future That Works: Automation, Employment, and Productivity; McKinsey Global Group January 2017

Arntz, M., T. Gregory and U. Zierahn (2016), The Risk of Automation for Jobs in OECD Countries: A Comparative Analysis. OECD Social, Employment and Migration Working Papers, No. 189, OECD Publishing, Paris. 

Graetz, Georg and Michaels, Guy (2015) Robots at work  CEP discussion paper, CEPDP1335. Centre for Economic Performance (CEP), London, UK.

Levitan, S.A. and  Johnson, C.M. The future of work: Does it belong to us or to the robots. Monthly Labor Review 1982, 105, 10–14.

Spitz-Oener, A. (2006). Technical Change, Job Tasks, and Rising Educational Demands: Looking outside the Wage Structure.  Journal of Labor Economics, 24(2), 235-270. doi:10.1086/499972

Wolla, S. A. (2018) Will Robots Take Our Jobs? Federal Reserve Bank of St. Louis/Economic Research