Embodying algorithms, enactive artificial intelligence and the extended cognition: You can see as much as you know about algorithm

• Published in: Journal of information science Vol. 49; no. 1; pp. 18 - 31
• Main Author: Shin, Donghee
• Format: Journal Article
• Language: English
• Published: London, England SAGE Publications 01-02-2023; Bowker-Saur Ltd
• Subjects: Algorithms; Artificial intelligence; Cognition; Cognition & reasoning; Embodied cognition; User experience; human–artificial intelligence interaction; Algorithm experience; explainability; heuristic and systematic process; enactive algorithm; embodied cognition
• ISSN: 0165-5515; 1741-6485
• DOI: 10.1177/0165551520985495

The recent proliferation of artificial intelligence (AI) gives rise to questions on how users interact with AI services and how algorithms embody the values of users. Despite the surging popularity of AI, how users evaluate algorithms, how people perceive algorithmic decisions, and how they relate to algorithmic functions remain largely unexplored. Invoking the idea of embodied cognition, we characterize core constructs of algorithms that drive the value of embodiment and conceptualizes these factors in reference to trust by examining how they influence the user experience of personalized recommendation algorithms. The findings elucidate the embodied cognitive processes involved in reasoning algorithmic characteristics – fairness, accountability, transparency, and explainability – with regard to their fundamental linkages with trust and ensuing behaviors. Users use a dual-process model, whereby a sense of trust built on a combination of normative values and performance-related qualities of algorithms. Embodied algorithmic characteristics are significantly linked to trust and performance expectancy. Heuristic and systematic processes through embodied cognition provide a concise guide to its conceptualization of AI experiences and interaction. The identified user cognitive processes provide information on a user’s cognitive functioning and patterns of behavior as well as a basis for subsequent metacognitive processes.