Is Distance from Innovation a Barrier to the Adoption of Artificial Intelligence?

15.04.2024 14:15 – 15:30

INSTITUTE OF ECONOMICS AND ECONOMETRICS SEMINAR

(jointly with: James Bessen, Iain Cockburn)

Abstract:

We investigate whether online vacancies for jobs requiring Artificial Intel-ligence (AI) skills grow more slowly in U.S. locations farther from AI “inno-vation hotspots.” To do so, we create a dataset of AI publications (research papers and patents) and define hotspots based on locations’ cumulative num-ber of AI publications by 2006. The source for job vacancies is online job advertisements scraped by Burning Glass Technologies from 2007–2019. With a hotspot defined as a commuting zone with at least 1000 AI publications, a 10% greater distance from a hotspot (about a standard deviation) reduces a commuting zone’s growth in AI jobs’ share of job advertisements by 3–5% of median growth. Distance from a hotspot plays no role if a commuting zone is itself a hotspot, but distance is a greater barrier the greater a hotspot’s share of publications that are patents rather than research papers. Analysis by occupation, industry and AI type suggests that the type of job posting for which distance is a barrier is jobs adapting AI for use in a new setting. We do not find convincing evidence for an effect of distance on the adoption of AI, perhaps because there is as yet little adoption.

Lieu

Uni Mail
Room M 3250
Boulevard du Pont-d'Arve 40
1205 Geneva

entrée libre