AI’s Local Patterns of Innovation

Matheus Eduardo Leusin, Björn Jindra, Daniel Hain

Publikation: KonferencebidragPaperForskningpeer review


The concept of relatedness has been crucial to understanding the geography of innovation. It explains local specialisations patterns such as the probability that a location develops a given technology. Yet we still lack an understanding of how technological relatedness changes over time, and how these changes affect local technological development. At the example of Artificial Intelligence (AI) we investigate (i) how specialisation patterns linked to AI change over time, (ii) how these changes affect the local exploration of AI at country level, and (iii) how new capabilities related to AI are created by countries. Thereby, we focus on the US, Japan, South Korea, and China as the four countries leading AI development during the observation period (1974 - 2018). Using patent data, we apply a technological space perspective coupled with specialization indices to identify the dynamics occurring at local and technological levels. We find that the technological evolution of AI has little association with how it was locally developed. Instead, the local development of AI relates to countries’ existing knowledge bases, even in cases when it was weakly related AI
Antal sider37
StatusUdgivet - 2021
BegivenhedDRUID21 Conference - Copenhagen Business School, Frederiksberg, Danmark
Varighed: 18 okt. 202120 okt. 2021
Konferencens nummer: 42


KonferenceDRUID21 Conference
LokationCopenhagen Business School


  • Artificial intelligence
  • Technological space
  • Evolutionary economic geography
  • Technological relatedness
  • Knowledge complexity