The Effects of GenAI on Job Performance: Catch-up Effects or Masking?

Gen AI refers to algorithms that are capable of creating fresh content without being explicitly programmed (Brynjolfsson et al., 2023). Early empirical results of research on human-GenAI collaboration found support for the use of GenAI to augment human work, demonstrating a boost in performance for employees across many non-routine industries, such as programming (Peng et al. 2023), writing (Noy and Zhang 2023), and customer service (Brynjolfsson et al. 2023). GenAI has been found to mostly benefit lower-performing employees, while at the same time does not affect or even hinder the performance of those higher performing workers (e.g. Brynjolfsson et al. 2023, Noy and Zhang 2023). This is contradictory to the effects of previously-developed technology on human performance, which has been demonstrated to mostly benefit higher performers. However, it remains unclear why GenAI impacts performance differently from past technology. Moreover, it is undetermined if this “catch-up” effect will sustain in the long term, since evidence from longitudinal studies is scarce. Drawing from the Theory of Technology Dominance (Arnold and Sutton 1998), this research hypothesizes that, in contrast to the "catch-up” effect, using GenAI creates a masking effect that benefits lower-performing employees only in the short-term. Specifically, we propose that as GenAI is trained using the organizations’ knowledge base, it inherits the knowledge of the higher-performing employees. When the lower-performing employees follow suggestions from GenAI, they “mask” their performance behind this knowledge base of the higher-performing, thus creating an illusion of performance boost. We also argue that, as users apply results from GenAI more automatically, they spend less time on critical thinking. This hinders the learning process and will decrease the long-term performance. Theoretically, this study contributes to the body of literature on the impact of technology adoption on job performance by advancing the theory of technology dominance, highlighting the role of learning and transfer of knowledge on long-term change in job performance. Practically, this research provides important insights for application and implementation of GenAI in the workplace.