Previous studies have identified the importance of access to electricity and the socio-economic environment for poverty reduction, but the comprehensive interplay and synergistic relationships between them remain unclear. Using data collected from Myanmar, this study establishes and tests conceptual models to explore the paths from access to electricity and socio-economic environment to capability and income poverty reduction. The results of structural equation modeling show that capability poverty reduction acts as a complete mediator between access to electricity and income poverty reduction, and plays a partial mediating role between socio-economic environment and income poverty reduction. Analysis of fuzzy-set qualitative comparative analysis demonstrates that lack of electricity is a key barrier impeding poverty alleviation efforts. Moreover, five effective configurations of poverty reduction factors are identified for Myanmar by considering the heterogeneity of different regions. This indicates that individual factors could not ensure a highly effective poverty reduction and different factors need to be appropriately configured for achieving the synergistic effects. These findings contribute to practical insights into poverty reduction policy making and sustainable development for developing countries.

Artificial intelligence (AI) has seen tremendous improvements in capabilities recently, yet how AI impacts different labor markets remains unclear. Leveraging the launch of ChatGPT, this study aims to elucidate how AI influences freelancers across different online labor markets (OLMs). Employing the Difference-in-Differences method, we discovered two distinct scenarios following ChatGPT’s launch: 1) displacement effects in translation & localization OLM, reducing freelancers’ work volume and earnings; 2) productivity effects in web development OLM, increasing freelancers’ work volume and earnings. Theoretically, we developed a Cournot competition model to explain and identify that an inflection point exists for each occupation. Before this point, human workers benefit from AI enhancements; beyond this point, human workers would be replaced. Additional analyses across various occupations consistently demonstrate the two scenarios caused by AI. Further investigation into the progression from ChatGPT 3.5 to 4.0 reveals three evolving patterns of AI’s effects, thereby reinforcing our inflection point conjecture.