Does Career Friction Always Hinder New Hires’ Performance? New PhD Hires in the AI Industry
Academy of Management Proceedings · 2026-07 · 2026 (1)
DOI: 10.5465/amproc.2026.11352poster ↗Kurz gefasst
Fachartikel in der Zeitschrift Academy of Management Proceedings, 2026-07. Verfasst von Joanne Cao.
Abstract (Originalsprache)
PhD graduates are increasingly moving from academia into industry, yet we still have limited understanding of how they perform during the job transition, particularly given their anticipated career friction. Drawing on an information-processing perspective, I argue that early performance depends on how well task demands align with the type of information processing new hires are familiar with. Specifically, I propose that tasks requiring internal information processing are associated with greater performance disadvantage for new PhD hires than for other new hires, such as when tasks involve coordinating interdependent components and navigating firm-specific routines. In contrast, tasks that rely on external information processing are associated with smaller performance disadvantages for new PhD hires, such as when tasks involve acquiring and integrating knowledge beyond the firm. I test these arguments using data on newly hired deep learning engineers at a leading AI firm, where performance is observable through GitHub activity after entry. Measuring performance disadvantage as the required amount of iteration to complete tasks, I find support for my hypotheses.
Zitieren
APA
Cao, J. (2026). Does Career Friction Always Hinder New Hires’ Performance? New PhD Hires in the AI Industry. Academy of Management Proceedings, 2026(1). https://doi.org/10.5465/amproc.2026.11352poster
Vancouver
Cao J. Does Career Friction Always Hinder New Hires’ Performance? New PhD Hires in the AI Industry. Academy of Management Proceedings. 2026;2026(1). doi:10.5465/amproc.2026.11352poster
BibTeX
@article{Cao2026,
author = {Cao, Joanne},
title = {Does Career Friction Always Hinder New Hires’ Performance? New PhD Hires in the AI Industry},
journal = {Academy of Management Proceedings},
year = {2026},
volume = {2026},
number = {1},
doi = {10.5465/amproc.2026.11352poster}
}
EUVIMED