The Impact of AI on Work and Income

The Impact of AI on Work and Income

The spread of artificial intelligence is expected to have effects on the labor market and income distribution different from previous technological revolutions

Eduardo Felipe Matias

The concern about the impact of automation on the labor market due to the use of machines and, more recently, algorithms in tasks previously performed by humans has been a historical constant, which even led economist John Maynard Keynes to formulate, in 1930, the expression "technological unemployment." 

Physical automation after the Industrial Revolution and, in recent decades, the integration of computing into professional activities, by targeting routine tasks, predominantly affected laborers and workers with lower levels of education. The latest advances in AI—such as generative AI—on the other hand, extend to cognitive functions, capable of performing tasks that require subtle data analysis and creative problem-solving, which until now have been dominated by highly skilled and better-paid white-collar professionals. 

Therefore, the spread of AI has the potential to accentuate the effects of previous technological revolutions, contributing to an even broader and deeper transformation in the labor market and income distribution. 

This is the conclusion of a study released by the International Monetary Fund (IMF) earlier this year, which found that nearly 40% of global jobs are exposed to AI. In advanced economies, the exposure is higher: about 60% of jobs would be impacted, due to the predominance of the services sector and mature industries with occupations focused on complex cognitive tasks. However, although about half of these jobs may be negatively affected, the other half tends to benefit from the productivity gains brought by AI. In emerging economies and low-income countries, often still reliant on manual labor and traditional industries, the total exposure would be 40% and 26%, respectively. In Brazil, it would be 41%. 

The study distinguishes between "high exposure and high complementarity" occupations, such as managerial roles, which benefit from the automation of certain tasks, and "high exposure and low complementarity" ones, such as administrative and technical support, which are at greater risk of being adversely affected. In advanced economies, 27% of jobs fall into the first category and 33% into the second, compared to 16% and 24% in emerging economies and 8% and 18% in low-income countries. 

Higher education levels, typically found in developed countries, are associated with a larger share of jobs in the high-exposure and high-complementarity category. This better positions them to take advantage of growth opportunities resulting from the use of AI, which mitigates their greater risk of labor displacement. 

Emerging economies and low-income countries are likely to experience less immediate impact, but may not have access to AI productivity gains due to a lack of infrastructure and qualified human capital. In the services sector, new technologies could lead to the relocation of activities from less developed regions to more technologically advanced countries—call centers in emerging economies, for example, might be replaced by generative AI solutions. The result would be an exacerbation of economic inequality between countries. 

AI could also accentuate income disparity within each country. This depends primarily on how much AI substitutes or complements workers, and, in the latter case, how it does so. With high complementarity, higher-paid workers could expect disproportionately higher income growth, which would lead to increased inequality. Additionally, as AI adoption shifts tasks previously performed by labor to capital—machines—there will be a decline in labor’s share of total income. 

Many compare the current period to the transition from an agricultural to an industrial society, which saw the largest labor movement ever witnessed. Optimists believe that, just as the Industrial Revolution created many new jobs, the same will happen now with AI. However, that transition occurred over a century, spanning multiple generations. The AI revolution, which may have a similar scale, is expected to happen within the lifetime of some professionals. People whose jobs are automated will have to retrain in a few years, only to perhaps see their new professions also automated shortly after, facing yet another period of requalification. 

The training and relocation of these workers is a complex and time-consuming process. New tasks require new skills, and it cannot be assumed that the labor market adjustment will happen quickly. Responses to this, including discussions about overhauling education systems and implementing a universal basic income, are likely to intensify in the coming years.

Eduardo Felipe Matias is the author of the books “Humankind and its borders” and “Humankind against the ropes”, winners of Premio Jabuti, and coordinator of the book “Startups Legal Framework”. PhD in International Law from the University of Sao Paulo, he was a visiting scholar at the universities of Columbia, in NY, and Berkeley and Stanford, in California, and is a partner in the business law area of Elias, Matias Advogados.

Article originally published in Portuguese at the Broadcast of Estadão/Agência Estado.

#column #Broadcast #Estadao #AgenciaEstado

#AI #technologicalunemployment #generativeAI #jobs