AI 中文总结
本研究利用10国在线职位空缺数据，结合NLP、大语言模型等方法，发现AI相关技能在核心STEM职业间趋同，但核心与其他职业间分化，强化了职业分层。
AI 中文摘要
人工智能（AI）正迅速成为当代劳动力市场的核心特征，但其扩散究竟是在不同职业间催生了共同的技能组合，还是加深了职业分化，目前尚不清楚。我们利用涵盖全球北方和南方10个国家的大规模在线职位空缺数据，研究与AI相关的技能需求如何重塑劳动力市场结构。结合自然语言处理、大语言模型和多层二分网络分析，我们绘制了新兴AI经济中职业、所需技能与职业阶段之间的关系图。研究发现，AI需求绝大多数集中在一个狭窄的技术核心领域，所有国家中约四分之三至五分之四的AI相关职位空缺位于STEM职业中。AI密集型工作始终强调Python、SQL、机器学习和数据分析，这在受AI影响程度较高的职业间形成了趋同。但这种趋同并未延伸至更广泛的劳动力市场，相反，AI相关技能大多仍局限于技术领域，且在劳动力市场入门阶段的需求最为强烈。这些研究结果表明，受AI影响的核心职业内部存在趋同，但该核心与其他职业结构之间存在分化。AI似乎并未让机会普及，反而强化了职业分层，抬高了入门门槛，并将AI应用的收益集中在具有先前技术优势的劳动者和职业群体中。
英文摘要
Artificial intelligence (AI) is rapidly becoming a defining feature of contemporary labor markets, yet it remains unclear whether its diffusion is producing a common set of competencies across occupations or deepening occupational divisions. We investigate how AI related skill demand is reshaping labor market structure using large scale online vacancy data from ten countries spanning the Global North and Global South. Combining natural language processing, a large language model, and multilevel bipartite network analysis, we map relationships between occupations, required skills, and career stages in the emerging AI economy. We find that AI demand is overwhelmingly concentrated within a narrow technical core, with approximately three quarters to four fifths of AI related vacancies located in STEM occupations across all countries. AI intensive jobs consistently emphasize Python, SQL, machine learning, and data analysis, generating convergence among highly exposed occupations. However, this convergence does not extend across the wider labor market. Instead, AI competencies remain largely confined to technical domains and are most strongly demanded at labor market entry. These findings reveal convergence within an AI exposed core but divergence between that core and the rest of the occupational structure. Rather than democratizing opportunities, AI appears to reinforce occupational stratification, raising barriers to entry and concentrating the benefits of AI adoption among workers and occupations with prior technological advantages.
