Paradoxes Associated with the Introduction of AI-Powered Electronic Systems - A Case Study from Competitive Artistic Gymnastics
Loading...
Journal Title
Journal ISSN
Volume Title
School of Business |
Doctoral thesis (article-based)
| Defence date: 2022-09-23
Unless otherwise stated, all rights belong to the author. You may download, display and print this publication for Your own personal use. Commercial use is prohibited.
Author
Date
2022
Major/Subject
Mcode
Degree programme
Language
en
Pages
82 + app. 74
Series
Aalto University publication series DOCTORAL THESES, 114/2022
Abstract
The use of AI technologies in systems supporting human experts' performance in demanding analysis and decision-making situations has expanded rapidly and grown ubiquitous in various fields. While AI-powered systems' introduction provides new opportunities for enhancing work processes, it is an uncertainty-rife, ambiguous, and difficult process that may have numerous implications and significantly influence many facets of a stakeholder's life, requiring radical changes in the organization and people involved. However, many organizations, in their pursuit of greater productivity, efficiency, accuracy, reliability, profitability, speed of work processes, and new advanced capabilities for human experts, undertake it without considering these implications and diverse risks for the organizational system and its stakeholders – with system bias, errors, inaccuracy, non-transparency, disturbing of trust and humans' safety and privacy, cultivation of discrimination, and disruption of human interaction being just a few of the numerous possible effects. Furthermore, AI systems' implementation could become a major source of "paradoxical tensions" in the organization. If unable to cope with these tensions evoked by the implementation, key stakeholders may experience negative functionality-related and emotional consequences of discrepancies between their expectations and the real-world experience of using the technology. Expressing a resulting research interest in the transition from human-based expert systems to AI-powered ones, a rigorous qualitative study of possible risks and implications associated with an AI system's introduction was conducted as a doctoral project. The study, from the perspective of a broad set of stakeholders, was conducted in a unique research context: artistic gymnastics, where Fujitsu has recently introduced an AI-powered system to support the judging, thus aiding in expert evaluation. As the research process progressed, analysis revealed several paradoxical tensions associated with this new electronic judging system's deployment for artistic gymnastics, tensions not previously considered in information-systems studies. Explored in depth in the dissertation, these are articulated as "accurate AI is too exact," "'objective' AI can be biased," "even black boxed AI represents explainability," "an AI-based judging system for artistic gymnastics cannot judge artistry," "a system intended for humans lacks human interaction," "consistency requires AI's adaptability," and "automation requires augmentation and vice versa." Building on the existing body of knowledge of AI, the project contributes to scholarly understanding of the paradoxes associated with IT artifacts and shows that proceeding with a degree of caution in the process of their implementation is still necessary.Description
Supervising professor
Penttinen, Esko, Prof., Aalto University, Department of Information and Service Management, FinlandThesis advisor
Saarinen, Timo, Prof., Aalto University, Department of Information and Service Management, FinlandKeywords
paradoxical tensions, artificial intelligence (AI), sports judging, bias, explainability, artistry, hybrid automation–augmentation approach
Other note
Parts
-
[Publication 1]: Mazurova, Elena; Penttinen, Esko. 2020. Probing athletes’ perceptions towards electronic judging systems – a case study in gymnastics. In: Proceedings of the 53rd Hawaii International Conference on System Sciences (HICSS), Maui, Hawaii. Pages 4438–4447.
Full text in Acris/Aaltodoc: http://urn.fi/URN:NBN:fi:aalto-202009045279DOI: 10.24251/HICSS.2020.543 View at publisher
- [Publication 2]: Mazurova, Elena; Penttinen, Esko; Salovaara, Antti. 2021. Stakeholder-dependent views on biases of human- and machine-based judging systems. In: Proceedings of the 54th Hawaii International Conference on System Sciences (HICSS), Kauai, Hawaii. Pages 6327–6336.Full text in Acris/Aaltodoc: http://urn.fi/URN:NBN:fi:aalto-202101251568. http://hdl.handle.net/10125/71383
-
[Publication 3]: Mazurova, Elena; Willem Standaert; Penttinen, Esko; Tan, Felix Ter-Chian. 2021. Paradoxical tensions related to AI-powered evaluation systems in competitive sports. Accepted for publication in the Journal of Information Systems Frontiers in 2021.
DOI: 10.1007/s10796-021-10215-8 View at publisher
- [Publication 4]: Mazurova, Elena. 2022. Hybrid use of automation and augmentation in introducing AI applications for artistic gymnastics. In: Proceedings of the 30th European Conference on Information Systems (ECIS), Timisoara, Romania. Pages 1-14.
