Capturing the Complexity of Cognitive Computing Systems: Co-Adaptation Theory for Individuals

Abstract

Cognitive computing systems (CCS) are the new generation of automated IT systems that mimic human cognitive capabilities. CCS reshape the interaction between humans and machines and challenge our traditional assumptions of technology use and adoption. This work introduces co-adaptation and defines it as the series of activities that a user and a system engage in simultaneously to make the system fit the user. Co-adaptation involves two types of adaptation: human adaptation and machine adaptation. Human adaptation refers to the user either changing their behavior to adjust to the technology or changing the technology to adjust to their use. Machine adaptation refers to the system adapting itself to fit users' needs. We use polynomial regression and response surface analysis to examine the impact of co-adaptation on individual performance. We add to previous work by offering a solid theoretical argument with supporting evidence that congruence between human adaptation and machine adaptation plays a critical role in determining the impact of technology use on individuals and their performance.