Impact of Artificial Intelligence on Economics based on Labor Market View with use of ChatGPT

Abstract

Large language models (LLMs), like ChatGPT, are an example of artificial intelligence (AI) that is upending labour markets by automating jobs that were previously thought to be within the realm of human intellect. The present research uses a fixed-effects panel regression and machine learning (ML) prediction models to empirically investigate how the adoption of AI affects employment and salaries. Using panel data from 2018-2024 in 23 economies, we analyze employment and wage growth dynamics and sectoral shifts associated with AI adoption. Overall, we find heterogeneous results: AI reduces employment in routine and low-skill roles but generates employment in AI-complementary roles generating overall net productivity (or overall economic gains) for higher-income economies. In predicting outcomes, ML models like Navie Bayes and XGBoost clearly outperformed linear regression in predicting outcomes and identifying nonlinear threshold effects and emphasizing the importance of education and economic development in our conclusions. ChatGPT was utilized in numerous aspects of this study, including to synthesize literature, generate classified interpretations of case study outputs, and provide summary feedback for iteration on explanations of the model outputs. The research experience will be published, along with our results which would underscore the notion of flexibility in adapting and updating rescaling policies, timing for staged adoption of AI, and utilizing $A I$ in an augmentative fashion as opposed to full mode substitution.

Ähnliche Arbeiten