Metaphor of overreaching: effects of AI agglomeration on innovation

Lian, Yanling, Jian, Jie, Xia, Chengcheng
The Journal of Technology Transfer · Springer Nature · 2026-10-01 · DOI: 10.1007/s10961-026-10390-3

Аннотация

Artificial Intelligence (AI) is reshaping innovation, yet the spatial dynamics and innovation consequences of its AI agglomeration remain underexplored. We introduce the concept of city-level AI agglomeration and examine its dual effects on city overall and radical innovation. Integrating organizational ecology with economic geography, we propose that AI agglomeration acts through two mechanisms. On the one hand, it generates computing spillover, a positive externality that expands the region’s environmental carrying capacity by sharing critical computing resources, thereby promoting overall innovation. On the other hand, it fosters algorithmic convergence, a negative externality that intensifies structural inertia, thereby constraining radical innovation. Using city-level data from China (2019–2023) and causal forest methods, our empirical analysis supports these hypotheses. Further analyses reveal heterogeneous effects moderated by geographic proximity to provincial capitals, fixed asset investment intensity, and regional innovation ecosystem support. This study advances agglomeration and innovation literature and offers insights for urban innovation policy.

Читать у издателя →

Похожие статьи