Instrumental and experiential attitudes toward (A.I.) augmented decision-making at work

In augmented decision-making, defined as a process wherein human judgment is complemented with decision-support systems powered by artificial intelligence (A.I.-DSS), employees are expected to monitor and sometimes override system outputs to enhance decision-making performance. Despite the growing use of these costly technologies in organizations, they often fail to add value as employees seem unwilling to delegate some of their tasks to A.I.-DSS or monitor its outputs. Past research has shown that employees differ in their attitudes toward (collaborating with) emerging technologies, and that these attitudes can facilitate or hinder effective technology use. Drawing on literature from technology acceptance (TAM) and user experience (UX), this study qualitatively explored whether employees hold both instrumental (i.e., related to consequences like performance) and experiential (i.e., related to experiences of the process) attitudes toward augmented decision-making, and whether these two types of attitudes differ in terms of their antecedents and outcomes. Seventeen semi-structured interviews with A.I.-DSS users from various organizations revealed that experiential attitudes were mentioned more frequently, but were significantly less positive than instrumental attitudes. In terms of antecedents, instrumental attitudes were primarily mentioned in relation to technology (A.I.-DSS) characteristics, whereas experiential attitudes were also related to task and individual characteristics. As for outcomes, instrumental attitudes were solely associated with employees’ intentions to use A.I.-DSS, while experiential attitudes were also mentioned in relation to employee absorption, motivation and stress. These findings highlight the importance of distinguishing between instrumental and experiential attitudes toward augmented decision-making at work. • Workers engage more in experiential than instrumental evaluations of augmented decision-making • Employees engage more often in experiential than instrumental evaluations of working with (AI) decision support systems • Employee experiential attitudes were significantly less positive than instrumental ones. • Instrumental attitudes mostly related to DSS characteristics, experiential ones also to task and individual characteristics • Experiential attitudes also had more varied outcomes (stress, motivation, absorption, usage) than instrumental (usage) ones • We conclude that experiential attitudes are important and should be considered in future research