Abstract
Turning to generative artificial intelligence to complete academic tasks is viewed by many as harmless cognitive offloading, a benign psychological practice. However, in this paper, I will examine the inaccuracies of this assumption. Cognitive offloading is a beneficial practice only when the delegator keeps three separate forms of control: executive control on how to implement a procedure, verification control on whether the output makes sense, and metacognitive control on what one knows and does not know. Generative systems generally deprive users of these control systems, meaning the practice in question should be more accurately considered as cognitive outsourcing. The most common reasons behind cognitive offloading are effort minimization and social loafing. That is, effort is an aversive experience, and essays or homework assignments are often anonymous products whose contributions cannot be distinguished from each other. In the paper, it is assumed that this comes from the experience of prediction error. The resulting lack of ability is doubly hidden: to the student, since his or her monitoring ability is being limited, and to the school, since it assumes that learning occurs because of the student’s good performance.
Keywords: cognitive offloading, cognitive outsourcing, metacognitive calibration, social loafing, desirable difficulties
Copyright © 2026 Scholar of Tomorrow. All SoT articles are distributed under the attribution non-commercial, with no derivative license. This means that anyone is free to share, copy, and distribute an unaltered article for non-commercial purposes, provided the original author and source are credited.