Download
Abstract
Artificial intelligence is on the mandate of almost every organization today, transforming productivity and redefining the boundaries of work (Singla et al., 2025). Almost 60% of jobs in advanced economies are expected to be impacted by AI (IMF, 2024) - yet the sentiment towards how workers and their tasks will be affected remains ambiguous. This paper aims to assess how AI shapes the skill premium through analysis of changes in wage inequality as measured by the Theil Index. Our regression results reveal that the scale and complexity of AI models (measured by AI model parameter counts) exacerbate wage gaps, especially among low-skilled workers, but AI research and development (using AI patent filings for proxy) show no significant short-term effect on inequality. This suggests that technological advances alone do not immediately translate into labor market disparities, but the complexity of models and the problems they are able to solve may disproportionately disadvantage lower-skilled workers. Following an analysis of conservative, baseline, and optimistic projections of AI parameter growth over 5 years, we anticipate that wage inequality is likely to continue harming low-skilled workers more than their high-skilled counterparts even in the future. Employment growth also has significant implications on disparities between occupations, likely due to skill-biased demand.
Acknowledgments
- We would like to thank Dr. Carl Heese (HKU) for his guidance in this project.
- Finally, I’m extremely grateful for my groupmates, Malisha and Swarna.