Column: Robots Will Take Your Job – Threat or Blessing?

Fear and discontent has been turned against the machines since the dawn of the Industrial Age. The Luddites rebelled in nearly 19th century England, named after the weaver Ned Ludd who smashed the stocking frames that took his job. Luddite has since come to mean an opponent of new technology.

Economists regard automation and innovation as increasing productivity. It lowers prices, which in turn increases demand, and makes the economy grow. The last two hundred years of rising productivity has not lead to widespread unemployment. The idea that machines would take over all the jobs has been rejected as the Luddite fallacy.

Recently, the Luddite issue has come back in fashion. Nobel laureate Paul Krugman expresses his sympathy for the Luddites, while senior maverick Kevin Kelly says that robots will – and must – take our jobs.

Robots are acquiring smarts, while becoming easier to operate and program. Sensors are becoming cheaper and more powerful. Robots are increasingly becoming software, artificial intelligence probabilistic algorithms deriving meaning from large amounts of data, that could be run in a metal frame as well as a computer mainframe. A program does not need to know exactly how a task should be done, just mimic the behavior under similar circumstances. The shift in productivity is not linear any longer, but exponential, as robots are affecting manufacturing, services and knowledge workers simultaneously.

Automation seems to be hollowing out the skills of the middle class. In the Euro-zone almost 4.3 million low-pay jobs have been gained since recovery started in 2009, but the loss of mid-pay jobs continued with a total of 7.6 million jobs disappearing, according to AP. Meanwhile, computers increase the productivity of skilled workers, such as managers, engineers and consultants.

Occupations requiring a high degree of creative intelligence and social interaction are less susceptible to automation at present. But the bureaucracies created in the 20th century are basically huge systems of computation, so administration and routine decision making are being hit. Big scale editing is increasingly performed by bots, software applications that run simple and repetitive pattern recognition tasks over the internet. Wikipedia’s content is maintained by hundreds of correcting bots. In fact, bots were once permitted to create and write whole articles. Quill, the software of the Narrative Science company is able to collect data, like sport statistics, financial reports and housing prices and compile them into news stories, operational reports, customer statements and investment research. The software does not produce Pulitzer prize journalism, but neither do most journalists.

Robots do the jobs we won’t, because they are dull, dangerous or require skills humans cannot match. Kevin Kelly rightly points out that there probably will be many new jobs rising from automatisation. The problem is not that robots are taking the jobs, but that so few new jobs are being created. But it should be noted that the jobs added are not the same kind of jobs that disappeared, and require new sets of skills.

We cannot race against the machines, as Kelly notes. Giving up on human labor, as Krugman does, won’t put workers in a better position either. It is up to humans to decide what robots should do. Will robots be designed to drive down costs? Will robots take the form of cheap, easily programmed helpers to enhance the ability of individuals? The latter path would be more consistent in spreading the benefits of the knowledge economy and increased productivity to a larger part of society. Then, a renewed take on manufacturing and services would provide us with both efficiency and purpose.

Waldemar Ingdahl
CEO, think tank Eudoxa