The Dynamics of Science-Industry Knowledge Transfer During the Emergence of Deep Learning

We consider the dynamics of science-industry knowledge transfer during the 30-year emergence of deep learning. Deep learning constituted a paradigm shift in artificial intelligence and has been identified as the only true breakthrough in the 70-year history of the field. Using patent-to-paper citations as an indicator of knowledge transfer, we examine the factors that affect the likelihood that private technology developers absorb and build upon the research of pioneer scientists Geoffrey Hinton, Yoshua Bengio, and Yann LeCun. The data are 18,009 AI patent families. (2023-05-28)