The rapid advancement of artificial intelligence technology, particularly large language models (LLMs) exemplified by ChatGPT, is significantly impacting the global labor market. By integrating competency theory with BERTopic topic modeling, this study analyzes 1,827 LLM-related job postings from recruitment platforms in China, revealing the core structural demands of the current talent market. The results identify six primary thematic categories for LLMs positions, dominated by AI product operation (37.71%) and large-model infrastructure development (18.98%), followed closely by data engineering and analysis (14.94%), application development and integration (13.57%), and multimodal AI development (13.03%). Although vertical-domain applications represent the smallest proportion (1.75%), they demonstrate a clear trend toward deep industry penetration. Furthermore, the identified positions reflect three notable characteristics: hierarchical technological structures aligned with evolving full-stack competencies, combining business acumen with innovative thinking, and multimodal expansion paired with vertical industry integration. Based on these insights, this paper proposes a project-based curriculum integrating technology, practical applications, and industry experience. Additionally, it emphasizes enhancing students’ competence and AI literacy. These recommendations aim to provide an empirical basis for cultivating versatile talents suited for the era of LLMs.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Data-Driven Analysis of Talent Demand for Large Language Models: Implications for Educational Reform from a Competency Perspective

  • Tang Zhuoyuan,
  • Wei Wei,
  • Liang Kai,
  • Yang Yao,
  • Lin Zhe,
  • Chi Kin Lam

摘要

The rapid advancement of artificial intelligence technology, particularly large language models (LLMs) exemplified by ChatGPT, is significantly impacting the global labor market. By integrating competency theory with BERTopic topic modeling, this study analyzes 1,827 LLM-related job postings from recruitment platforms in China, revealing the core structural demands of the current talent market. The results identify six primary thematic categories for LLMs positions, dominated by AI product operation (37.71%) and large-model infrastructure development (18.98%), followed closely by data engineering and analysis (14.94%), application development and integration (13.57%), and multimodal AI development (13.03%). Although vertical-domain applications represent the smallest proportion (1.75%), they demonstrate a clear trend toward deep industry penetration. Furthermore, the identified positions reflect three notable characteristics: hierarchical technological structures aligned with evolving full-stack competencies, combining business acumen with innovative thinking, and multimodal expansion paired with vertical industry integration. Based on these insights, this paper proposes a project-based curriculum integrating technology, practical applications, and industry experience. Additionally, it emphasizes enhancing students’ competence and AI literacy. These recommendations aim to provide an empirical basis for cultivating versatile talents suited for the era of LLMs.