Reproduce “Large Language Models at Work in China's Labor Market” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Qin Chen, Jinfeng Ge, Huaqing Xie, Xingcheng Xu, Yanqing Yang
This paper empirically assesses the impact of Large Language Models on China's labor market, revealing a disproportionate influence on higher-skilled, non-routine cognitive occupations. It introduces a novel theoretical model that integrates information theory with a task-based framework to explain the distinct automation logic of LLMs, challenging the traditional routinization hypothesis and predicting a potential flattening of wage distribution.
Key insights
- LLMs disproportionately impact higher-skilled, non-routine cognitive tasks and occupations in China, deviating from the 'routinization hypothesis' that focused on routine, middle-skill jobs.
- The theoretical framework, dubbed 'AI learning theory,' provides a micro-foundation for understanding LLMs' distinct automation logic, showing their advantage in complex, data-intensive tasks due to neural scaling laws.
- LLMs could potentially lead to a flattening of the wage distribution and a decrease in the wage premium between skill levels by augmenting complex tasks, contrasting with previous information technologies that widened inequality.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.