Monday, October 25, 2021
logo economy journal
< view full issue: Robotica
Will Oremus

Does technological innovation hurt society?

Senior technology writer, Slate

Innovation is supposed to be the remedy for economic quagmire. What if, on the contrary, it was in fact the cause? More concretely, is it possible that the increasing automation that affects all sectors, from factories to retail to journalism, was actually destroying more jobs than it creates?

010597608 prevstill

This is a question that economists and workers have been asking at least since the Industrial Revolution. And in the past, the response was generally a frank and direct "no".

It is true that automation makes certain low-skilled jobs obsolete but also induces new categories of jobs that are the opposite, such as engineers or equipment managers, through the banking sector or blogging. Its main effect is to increase productivity, which is supposed to raise the level of income and stimulate the demand for new products and services.

Has the nature of technical progress changed?

In any case, the jobless economic recovery that we are experiencing at the moment, as well as a longer-term trend towards income and wealth inequality, lead some intellectuals to consider whether the latest wave of automation is not different from the ones before. Andrew McAfee and Erik Brynjolfsson, MIT researchers, are not alone in seeing a "great decoupling" between productivity and wages, especially since 2000, as technology distances itself from training human skills. In other words, wage earners are losing the race between education and technology. This could be exacerbating an older trend, in which capital has been gaining an advantage over work since the 1970s.

Replacing manual labour with machines on farms and factories was different, some worry. These stupid and extremely specialized machines needed to be supervised by humans in all their stages of operation. But the 21st century has seen the arrival of much more intelligent devices, capable of performing tasks that, so we previously thought, could never be automated.

Wage earners are losing the race between education and technology. This could be exacerbating an older trend, in which capital has been gaining an advantage over work

Computers today can respond to your orders, organize your diary, sell you shoes, recommend a movie and target you with publicity. Those in the future will diagnose your illnesses, write your newspaper articles and even drive your cars. When even the highly qualified intellectual professions run the risk of being replaced by machines, which human jobs can still survive? Those of politicians perhaps and of course those of business and management. In other words, the rich are going to be even richer, and the rest will be left behind.

All this has brought the concept of "technological unemployment" into university discourse, eighty years after John Maynard Keynes invented the term. On Aug. 6, Pew Research and Elon University released a report titled "Artificial Intelligence, Robotics and Future Professions”. This report compiles and synthesizes opinion poll results for 1,900 economists, operational experts, analysts, and other theorists who are faced with a big question: "Will automated and interconnected artificial intelligence applications and robotic devices make more jobs disappear of those they create"? The results of the survey are fascinating. Nearly half the people surveyed (48%) believe that smart computers will destroy more jobs than they can create. The other half predicts exactly the opposite.

The middle class in danger

This lack of consensus among experts on such a crucial and apparently simple issue is a concern. Mainly because history and dominant economic models make one aspect of the issue very clear: the one that recons that society will adapt, that new jobs will appear and that technology will make the economy stronger. Even Keynes asserted in 1930 that technological unemployment would be no more than a "temporary phase of adaptation." This view of things has been accepted for so long that its detractors are labelled pejoratively "Luddites."

Luddites were English weavers who destroyed and burned mechanical looms and spinning wheels on the principle that technology represented a fundamental threat to the welfare of mankind. Who would have thought then that in the 21st century, half of the great American experts would they share the luddites’ view of the primary effects of automation on the labour market?


"Automation is like Voldemort: the terrible force no one can name," says one of the people polled, quoted in the "Pew" report. Well-paid jobs will become increasingly rare", says Mark Nall, programme manager at NASA. "I'm not sure if jobs will completely disappear altogether," says Justin Reich of Harvard's Berkman Centre for Internet and Society, "but the jobs that remain will be far worse off and more precarious than they are today."

Is this grim prospect justified? I raised the issue with Andrew McAfee, who has contributed to launch the current debate on the effects of automation with the appearance in 2011 of the e-book Race Against the Machine and its sequel (The Second Age) in 2014, co-written with his colleague, MIT researcher, Erik Brynjolfsson.

McAfee, a specialist in the effects of information technologies in business, shows an unqualified optimism about the impact of technology on economic growth. “I'm super optimistic about the size of the pie. However, its effects on employment and on the middle classes in particular create some concern”, he says.

“It is quite clear at this level that the middle class is being squeezed. The balance between capital and labour is in the process of being modified, and the best studies analyze this by clearly identifying information technologies as the cause”, McAfee notes.

The balance between capital and labour is in the process of being modified, and the best studies analyze this by clearly identifying information technologies as the cause

McAfee believes that the education system must evolve to prepare young people for a world in which most jobs will be automated. What is not so clear is that this could happen very quickly.

“The closer you look, the more it seems that certain of these transitions have lasted for about two or three years, at the end of which the whole world ends up assuming them. They have been long and difficult in certain respects and require more consistent policy responses” says Columbia University economist Joe Stiglitz, who believes that if the Great Depression was so long it could have been due to the mechanization of agriculture.

It is quite possible that the current fears are only the result of an underestimation of the ability of capitalist economies to adapt and prosper. We cannot deny that they somehow did so in the past. And remember that half the people surveyed by Pew still think that everything will be fine. However, we have no guarantee that the future will look like the past. The fact is that a large number of intelligent people who study these problems believe that this time things could be different, which is very disturbing for those at the wheel of public life. And one last thing, even if Keynes was right (that the sufferings provoked by automation are temporary and the economy ends up adapting) it is worth remembering in the spirit another famous quote from the same great economist. "In the long run, we will all be dead."

Will Oremus is a writer and journalist. He writes Slate's technology column and contributes to the Future Tense blog, which covers technology, social media and automation and its relationship to society and politics.

Original translation to spanish by Peru Erroteta



Ronda Universitat 12, 7ª Planta -08007 Barcelona
Tlf (34) 93 301 05 12
Inscrita en el Registro Mercantil de Barcelona al tomo 39.480,
folio 12, hoja B347324, Inscripcion 1




Aviso legal - Cookies