Goldman Sachs economists suggest that over 300 million jobs could be affected in the next decade, while McKinsey’s analysts call people to update their skills portfolios amid potential risk of labor displacement. The rapid introduction of AI in production technologies and services has been seen as the harbinger for job losses specially for low skill workers. But is this fear completely justified? Are there other factors that will interfere in favor of human dexterity in labor markets? What is the prospective scenario regarding the income distribution? This article aims at answering these questions by analyzing current developments and forecasts comprehensively, using numerous sources, but not failing to consider however, the historical relationships between innovation and introduction of new technologies with distributional displacements on labor markets.
Historically, technology has been either the worst enemy or the best friend of a worker, its ultimate effect conditioned on the nature of the job or task to be performed. In the most routine, manual, non-cognitive tasks technology managed to displace workers as early as society has lived. The earliest and most accounted example is in the industrial revolution; in the XIX century, with the advent of steam-driven machines and other innovations, many low-skilled workers were displaced, and this caused the famous luddite riots in England in the time. However, the introduction of these technologies created new jobs for another part of the population, those qualified to run the machines, to provide adjustments, and to perform periodical repairs. This combination of effects was taken almost as a rule of innovation. Therefore, from what history may teach us, the negative effect that technologies signify to labor markets through displacement effects has for the most part been at least partially counterbalanced by higher demand for other jobs, both through the creation of new ones and through complementarities of technology and labor. This evidently was and still is today a cause of a very skewed income distribution, primarily favoring high-skilled individuals. Another rule, rather trivial, is that with innovation technology has always “raised the bar” for humans, calling for a continuous upward revision of their skill portfolios.
Yesterday there was the risk of automation in manufacturing and clerical processes. Today instead the bar is higher than ever with artificial intelligence: now the risk of displacement is existent in occupations which we may consider high-skilled, such as accountants, data scientists and software developers. According to McKinsey global institute partner Anu Madgavgakar, humans can be outperformed, since AI language models like ChatGPT are already capable of writing codes more accurately and rapidly than humans. Madgavgakar however stresses that that it is vital to view AI’s incidence not merely as a replacement but a productivity enhancing effect. This is of course true as there are industries where workloads cannot be directly replaced but streamlined with AI. For example, financial analysts and market researchers can rely on the use of AI for finding initial trends, identifying over or underperforming portfolios, etc.
Nonetheless, there are many industries where AI has already been adopted in direct replacement of humans in some tasks. For instance, in the media industry: Tech news site and Buzzfeed have been using only AI softwares to write articles and develop travel guides, respectively. Moreover, in the legal industry, the role of legal assistants, according to the Goldman Sachs report, is in a long-term precariousness given its high component of automatable tasks, i.e. managing and synthetizing large amounts of information. Overall, the report predicts an effect on over 300 million jobs around the world, but flipping the coin, it will potentially account for an increase of 7% of GDP, and potentially a reinstatement effect of displaced workers, by either jobs created directly from the use of AI, or indirectly through a greater aggregate demand.
So how peremptory is the need to update our skill portfolios?
AI technologies are now able to produce work indistinguishable from the human mind, and this can be a great challenge which needs to be harnessed by society. Young people in the active workforce should strive to boost their productivity in function of AI technologies, that is, not only learning new skills, but getting the most out of them with the help of AI. For instance, this article was written by a human, but the bit of empirical evidence was suggested by an open AI software. Finally, today more than ever, access to education is very wide, and learning new skills, even if these are more complex, does not require an excessive complication related to money, time, or location.