Cognitive effort is a form of labor, and unsurprisingly, people tend to favor less demanding forms of cognition and other mental shortcuts [18, 51].
Unfortunately, this human tendency can lead to unintended or dangerous outcomes because humans are susceptible to a wide variety of cognitive biases such as confirmation bias
Confirmation bias [41]
refers to the interpreting of new evidence in ways that confirm one’s existing beliefs
For XAI, this manifests as practitioners only superficially examining explanations instead of digging deeply, leading to over-trust, misuse, and a lack of accurate understanding of the outputs [31]
Forcing users to cognitively engage through some small task before showing a system’s output yielded the highest performance in a comparative study [21]
Train conductors in Japan famously point and call out important information on their journeys—a cognitive forcing method which has reduced human errors by nearly 85% [45].
Realistically, how much will users actually cognitively engage with the magnitude of generated outputs to ensure that they are correct and aligned with their intentions?