Speaker
Description
Artificial intelligence (AI) is increasingly seen as a key factor in boosting future productivity. However, the important question now shifts from AI's general potential to its varying effects in different countries. Recent analyses by Moody's predict that AI could increase labor productivity by about 1.5% each year across a group of 106 nations, leading to a total growth of around 15% over ten years. However, these gains are not certain. They will not be shared equally, nor will they be socially neutral. The current research raises a crucial question: under what conditions does exposure to AI lead to real productivity gains at the national level? The argument focuses on the interaction of three main factors: the types of jobs in the workforce, demographic aspects, and the effectiveness of institutions. Countries with a large number of jobs that work well with AI are likely in a better position to take advantage of productivity benefits. Meanwhile, aging populations might see different results based on their labor market conditions and their ability to handle changes. The study will use a cross-country panel dataset that includes about 80-120 economies, depending on data availability. It will combine AI-exposure metrics from IMF datasets with economic, demographic, and institutional indicators from the World Bank, OECD, and United Nations. The analysis will use regression methods to find the relationship between AI exposure and productivity growth. It will include interaction terms to see how factors like aging, social safety, education systems, and labor market conditions affect this relationship. The study aims to show that AI does not change economies in the same way for everyone; instead, it creates advantages based on specific conditions. Some countries may turn AI exposure into inclusive productivity growth, while others may face rising inequality, labor market disruptions, and increased social division.
