Practice With Less AI Makes Perfect: Partially Automated AI During Training Leads to Better Worker Motivation, Engagement, and Skill Acquisition

Mario Passalacqua (), Robert Pellerin, Esma Yahia, Florian Magnani (), Frédéric Rosin, Laurent Joblot () and Pierre-Majorique Léger
Additional contact information
Mario Passalacqua: MAGI - Département de Mathématiques et de Génie Industriel - EPM - École Polytechnique de Montréal
Robert Pellerin: MAGI - Département de Mathématiques et de Génie Industriel - EPM - École Polytechnique de Montréal
Esma Yahia: LISPEN - Laboratoire d’Ingénierie des Systèmes Physiques et Numériques - Arts et Métiers Sciences et Technologies
Florian Magnani: CERGAM - Centre d'Études et de Recherche en Gestion d'Aix-Marseille - AMU - Aix Marseille Université - UTLN - Université de Toulon, ECM - École Centrale de Marseille
Frédéric Rosin: LISPEN - Laboratoire d’Ingénierie des Systèmes Physiques et Numériques - Arts et Métiers Sciences et Technologies
Laurent Joblot: LISPEN - Laboratoire d’Ingénierie des Systèmes Physiques et Numériques - Arts et Métiers Sciences et Technologies
Pierre-Majorique Léger: HEC Montréal - HEC Montréal

Post-Print from HAL

Abstract: The increased prevalence of human-AI collaboration is reshaping the manufacturing sector, fundamentally changing the nature of human work and training needs. While high automation improves performance when functioning correctly, it can lead to problematic human performance (e.g., defect detection accuracy, response time) when operators are required to intervene and assume manual control of decision-making responsibilities. As AI capability reaches higher levels of automation and human-AI collaboration becomes ubiquitous, addressing these performance issues is crucial. Proper worker training, focusing on skill-based, cognitive, and affective outcomes, and nurturing motivation and engagement, can be a mitigation strategy. However, most training research in manufacturing has prioritized the effectiveness of a technology for training, rather than how training design influences motivation and engagement, key to training success and longevity. The current study explored how training workers using an AI system affected their motivation, engagement, and skill acquisition. Specifically, we manipulated the level of automation of decision selection of an AI used for the training of 102 participants for a quality control task. Findings indicated that fully automated decision selection negatively impacted perceived autonomy, self-determined motivation, behavioral task engagement, and skill acquisition during training. Conversely, partially automated AI-enhanced motivation and engagement, enabling participants to better adapt to AI failure by developing necessary skills. The results suggest that involving workers in decision-making during training, using AI as a decision aid rather than a decision selector, yields more positive outcomes. This approach ensures that the human aspect of manufacturing work is not overlooked, maintaining a balance between technological advancement and human skill development, motivation, and engagement. These findings can be applied to enhance real-world manufacturing practices by designing training programs that better develop operators' technical, methodological, and personal skills, though companies may face challenges in allocating substantial resources for training redevelopment and continuously adapting these programs to keep pace with evolving technology.

Keywords: Human-centered AI; training curriculum; motivation; self-determination theory; industry 5.0 (search for similar items in EconPapers)
Date: 2024-03-03
New Economics Papers: this item is included in nep-ain and nep-neu
Note: View the original document on HAL open archive server: https://hal.science/hal-04487695v1
References: View references in EconPapers View complete reference list from CitEc
Citations: View citations in EconPapers (2)

Published in International Journal of Human-Computer Interaction, 2024, pp.1-21. ⟨10.1080/10447318.2024.2319914⟩

Downloads: (external link)
https://hal.science/hal-04487695v1/document (application/pdf)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:hal:journl:hal-04487695

DOI: 10.1080/10447318.2024.2319914

Access Statistics for this paper

More papers in Post-Print from HAL
Bibliographic data for series maintained by CCSD ().