A new study by the International Labour Organization (ILO) has concluded that Generative Artificial Intelligence (AI) is more likely to increase jobs than destroy them by automating some tasks rather than taking them over entirely.
Generative AI is a type of artificial intelligence system capable of generating text, images or other media in response to commands, with potential applications in industries, art, writing, software development, product design, healthcare, finance, gaming, marketing and fashion.
The study, “Generative AI and Jobs”, argues that most jobs and industries are only partially exposed to automation and are more likely to be complemented than replaced by the latest wave of generative AI, such as chatGPT.
Therefore, he argues that “the biggest impact of this technology is likely to be not job destruction, but rather potential changes in the quality of jobs, in particular work intensity and autonomy”.
Administrative work is the category with the highest technology exposure, with almost a quarter of the tasks considered highly exposed, and more than half of the tasks with medium-level exposure.
In other occupational groups, such as managers, professionals and technicians, only a small share of tasks are considered highly exposed, while about a quarter had a medium level of exposure.
The study, released by the ILO from its headquarters in this Swiss city, documents notable differences in the effects on countries at different levels of development, linked to economic structures and existing technology gaps.
It shows that in high-income countries 5.5 % of total employment is potentially exposed to the automation effects of technology, while in lower-income countries the risk of automation affects only 0.4 % of employment.
The lowest income countries are 27 in sub-Saharan Africa and Asia, according to the World Bank’s income classification, while high-income countries are mostly in Europe, the Pacific such as Australia or Japan, North America or Uruguay in South America.
In between are the lower-middle-income countries, such as Bolivia, Egypt or Pakistan, where the impact of AI on employment would be 1.3 % according to the ILO, and the upper-middle-income countries, such as Argentina, Malaysia or Russia, where the impact would be 2.4 %.
On the other hand, the potential for employment gains from the application of AI averages 13% globally and is almost the same across countries, “suggesting that, with the right policies, this new wave of technological transformation could offer significant benefits to developing countries”.
The study finds it likely that the potential effects of Generative AI will differ significantly between men and women, as more than twice as many women’s employment could be affected by automation.
This is due to the over-representation of women in administrative work, especially in high- and middle-income countries.
Given that these jobs have traditionally been an important source of female employment as countries developed economically, one of the outcomes of Generative AI could be that certain administrative jobs may never emerge in low-income countries.
The study highlights that the socio-economic impact of Generative AI will largely depend on how its diffusion is managed, and argues for the need to design policies that support an orderly, just and consultative transition.
“Worker voice, empowerment and adequate social protection will be key to managing the transition,” otherwise “there is a risk that only a few countries and well-prepared market participants will benefit from the new technology,” it said.
The paper concludes, however, that “the outcomes of the technology transition are not predetermined. It is humans who are behind the decision to incorporate such technologies, and it is humans who must guide the transition process”.
