If ChatGPT and generative AI lives up to even a tenth of the hype around them, massive job losses may seem inevitable. But new economic data shows that the latest big jump in AI did not coincide with a reduction in jobs in affected industries, despite widespread fears of rapid replacement at the time.
In a new research paper, economists looked at the labor market in a number of European countries between 2011 and 2019. That’s when the AI technique deep learning emerged as a powerful way to perform tasks such as transcribing speech, analyzing images and making algorithmic recommendations for social feeds and e-commerce sites. At the time, it was widely expected that deep learning would have a broad and rapid impact on employment.
To find out what really happened, researchers from the European Central Bank, the Spanish central bank and the universities of Oxford and Pittsburgh used two established methods to measure how vulnerable professions are to AI-powered automation. Both involved examining the tasks employees perform and how they relate to the capabilities of algorithms. The researchers compared that information with survey data on EU workers that show the number of people leaving or joining various occupations in industries ranging from agriculture to financial services.
The main result was that industries where AI could be most useful saw no job losses. In fact, for more highly skilled jobs that are vulnerable to AI, such as office work that involves data, there was an increase of about 5 percent in the number of employees. The researchers say this supports the idea that new technology could increase demand for more skilled workers while also replacing those who do routine work. Less skilled workers did not appear to be significantly affected by software or AI.
While the fear of new technologies taking jobs is common—and perfectly understandable—economic research paints a complex, mixed picture. In general, many economists believe that automation can increase demand for jobs in general, as some recent studies have shown. However, if you’re not one of those more skilled workers, the advent of new technology can be a problem.
None of these studies make it possible to predict what effect ChatGPT or other generative AI technologies will have. They may be too new to cause noticeable changes. And it’s possible that generative AI will have a very different effect on jobs than before. “Although in the period of our analysis the association is positive, these results cannot be extrapolated into the future,” the authors of the new study write.
Generative AI is undeniably already causing some disruption. You don’t have to look far to find stories from individuals and companies already being changed by the kind of AI behind ChatGPT.
For example, some copywriters have to find a new career after clients replace them with generative AI tools. A number of publications are also experimenting with AI-generated content. The German tabloid Image recently cited AI as one of the reasons for the planned job cuts.
But let’s not forget that tools like ChatGPT are still untrustworthy collaborators, as they fabricate facts, reinforce prejudices, and otherwise misbehave. And it’s worth remembering that the rise of deep learning in the 2010s has prompted some AI experts to predict the disappearance of certain jobs, including radiologists – a prognosis that barely come out.
