• Subscribe

Artificial intelligence: Is your job at risk?

Speed read
  • Waves of computerization will continue to impact the labor market
  • Occupations requiring greater creativity and human socialization are most resistent to computerization 
  • Track your odds of replacement with interactive website

In early 19th century England, a group of workers destroyed weaving machinery to protest the use of new technology to drive down the cost of skilled labor. A similar disruption is happening today. Robots are displacing factory workers, and administrative work is rapidly being automated. Technology is changing work – and we must change with it.<strong>Luddites</strong> revolted against the advent of machines during the nineteenth century. What is the place for the human in our brave new world?

The future of employment

Carl Benedikt Frey and Michael A. Osborne of Oxford University wondered which jobs are most susceptible to computerization. Drawing upon advances in machine learning (ML) and mobile robotics, they developed a way to predict which of 702 occupations are most at-risk.

In recently published research, they predict that 47 percent of total US employment is at a high risk for computerization, and they believe there may be two waves of computerization separated by a technological plateau.

The first wave is likely to affect workers in transportation and logistics as well as those in office and administrative support. The computerization of manufacturing occupations will continue as robots get better at performing non-routine tasks. Occupations in services, sales, and construction are also predicted to fall victim to automation.

47 percent of total US employment is at a high risk for computerization.

Frey and Osborne note that the growing market for personal and household service robots is an indication that these types of service jobs are at risk. They also see sales occupations like cashiers, rental clerks, and telemarketers moving to computerization as they do not require a high degree of social intelligence. The practice of prefabricating materials in factories is also likely to reduce the need for workers on construction sites.

According to their study, this wave of automation will be followed by a slowing of job computerization, due to engineering bottlenecks. These bottlenecks involve getting computers to perform tasks requiring creativity and knowledge of human heuristics. The second wave of computerization is expected to come when these bottlenecks are overcome and computers gain mastery of more creative, human social behaviors.

Beyond routine tasks

Tasks can be categorized as routine or non-routine and then further determined to be manual or cognitive in nature. A routine task is one that follows explicit rules that can be accomplished by machines. Non-routine tasks are too complex or not sufficiently understood to be specified in computer code. These task categories can be either manual (physical work) or cognitive (knowledge work) in nature. 

Until recently, occupations most at-risk for computerization were those requiring only manual and cognitive routine tasks. Industrial robots have already taken on the routine tasks of most factory workers, and today’s more advanced robots have improved sensors and manipulators that allow them to perform non-routine manual tasks.

For example, General Electric has developed robots that can climb and maintain wind turbines, and in hospitals, surgical robots with greater flexibility will soon perform a greater number of operations.

Today, the trend is spreading to jobs commonly defined as non-routine. Activities like driving a car in city traffic or deciphering handwriting used to be considered non-routine, but advances in ML have made them automatable.

Computers are also getting better at performing non-routine cognitive tasks thanks to their capacity to quickly calculate large data sets. Additionally, algorithms are not burdened with the need to eat or sleep, so they can work longer hours and with greater efficiency.

<strong>Friend or foe?</strong> Is automation the greatest threat or greatest boon to our labor force? Courtesy Alaa Fadag; Indiana University.

Credit card fraud detection is an example of a completely automated non-routine cognitive task which takes advantage of a computer’s capacity to detect trends in big data and make impartial decisions.

Law firms use sophisticated algorithms to perform the work of paralegals by scanning thousands of legal briefs for pre-trial research. In health care, knowledge gleaned from medical evidence reports, patient records, and medical journals is being used to automate diagnostic tasks.

Increasingly, computers are learning to recognize interconnected hierarchies of features as human brains do (e.g., edges, ➡️ textures ➡️shapes ➡️ objects ➡️ semantic concepts➡️ meaning ➡️ knowledge).

This so-called deep learning model, constructed after our brain’s neural networking by which we create knowledge, promises to accelerate this already speeding trend toward automation.

Preparing for the changes

Frey and Osborne mention that regulatory concerns, political activism, and resistance to technological change may slow the process of computerization. They predict, however, that low-skill and low-wage occupations are at the highest risk.

What can you do to stay relevant in this new work environment?  The new world of work requires flexibility and the willingness to change, as well as creativity and social skills. Consider going into the sciences, healthcare, or education.

In the meantime, to see the likelihood of your job being done by a machine, check out this online tool developed using Frey and Osborne's research.

Join the conversation

Do you have story ideas or something to contribute? Let us know!

Copyright © 2017 Science Node ™  |  Privacy Notice  |  Sitemap

Disclaimer: While Science Node ™ does its best to provide complete and up-to-date information, it does not warrant that the information is error-free and disclaims all liability with respect to results from the use of the information.

Republish

We encourage you to republish this article online and in print, it’s free under our creative commons attribution license, but please follow some simple guidelines:
  1. You have to credit our authors.
  2. You have to credit ScienceNode.org — where possible include our logo with a link back to the original article.
  3. You can simply run the first few lines of the article and then add: “Read the full article on ScienceNode.org” containing a link back to the original article.
  4. The easiest way to get the article on your site is to embed the code below.