A Query Conundrum: The Mental Challenges of Using a Cognitive Assistant

Abstract

Cognitive assistants use vocal interfaces and artificial intelligence to assist humans with complex tasks. While much research has focused on the application of these devices, a few studies have addressed how these devices affect the way humans work. To fill this gap, this research studied the effects of a cognitive assistant on mental workload, frustration, and effort. Participants worked with a Wizard-of-Oz style assistant and completed the Wisconsin Card-Sorting Task and engaged in a peripheral-detection task in a two-sample study that compared participants (n = 21) who worked with the assistant to those who did not. Follow-up interviews were also completed. Results suggest that onboarding techniques, such as tutorials, are important for developing analogical trust before regular use. Additionally, results suggest that keeping the mental model of the CA clear, simple, and intuitive is important to reduce the mental effort that is required to account for the CA and interactions with it while working. Cognitive assistants offer a broad range of advantages but also have distinct challenges for users: primarily the lack of physical affordances that can be linked to functionality.

Focus and Modality: Defining a Roadmap to Future Ai-Human Teaming in Design

“…Working with an AI-as-Tool shifts the human contribution in the team towards higher-value work, such as sense-making and information management (Song et al, 2021;Maier et al, 2021). Such tools are disrupting industry today, and will result in new types of CAD capabilities for design, analysis and manufacturing.…”

“…In addition, AI assistance has been used at the concept generation (Camburn, Arlitt, et al, 2020), concept evaluation (Camburn, He, et al, 2020), prototyping (Dering et al, 2018), and manufacturing (Williams et al, 2019) stages,. Work has studied the impacts of AI assistance in aspects of engineering design, including decisionmaking, optimization, and computational tasks Rao et al, 1999), and its effects on mental workload, effort, and frustration (Maier et al, 2020(Maier et al, , 2021. Bang et al (2018) introduced DAPHNE as an intelligent cognitive assistant developed for providing support in system architecting, specifically for designing a constellation of satellites for Earth observation.…”

A Systematic Review of User Mental Models on Applications Sustainability

“…In addition, the study by Maier et al (2021) brings to light the fact that the impacts of cognitive assistants on humans have not been thoroughly researched, particularly in terms of the mental effort and the level of dissatisfaction they cause. This highlights how important it is to evaluate the impact that AI assistants have on human users in order to ensure that they successfully support and augment the cognitive processes of human users rather than overpowering or frustrating them.…”

“…application works and the actions a user might take to accomplish their goals in using the application (Maier, Abdullah, McComb, & Menold, 2021).…”