Antecedents of Constructive Human-AI Collaboration: An Exploration of Human Actors’ Key Competencies

Abstract

Artificial intelligence (AI) has become an integral element of modern machines, devices and materials, and already transforms the way humans interact with technology in business and society. The traditionally more hierarchical interaction, where humans usually control machines, is constantly blurring as machines become more capable of bringing in their own (intelligent) initiatives to the interaction with humans. Thus, nowadays it is more appropriate to consider the interactive processes between humans and machines as a novel form of interdependent learning efforts between both sides, where processes such as critical discourses between humans and machines may take place (hybrid intelligence). However, these developments demand a shift in the understanding about the role of technology at work and about specific competencies required among human actors to collaborate constructively and sustainably with AI systems. This paper seeks to address this issue by identifying human actors' key competencies, which enable a more constructive collaboration between humans and intelligent technologies at work.

AI in academia: An overview of selected tools and their areas of application

“…These systems are extremely good at the specific tasks they are designed for, but cannot operate outside those tasks (Russell & Norvik, 2010). Furthermore, these systems do not possess the broader cognitive abilities or general intelligence of humans (Condello et al, 2021;Ducao et al, 2020;Goretzko & Israel, 2022;Laurent, 2018;Leszkiewicz et al, 2022;Liem et al, 2018;Park & Park, 2018;Schachner et al, 2020;Süße et al, 2021;Weidener & Fischer, 2023). Weak AI is characterized by its focused nature, excelling in well-defined tasks or areas of application, but lacking the ability to generalize knowledge or skills to other domains (Condello et al, 2021;Liem et al, 2018).…”

“…Weak AI is characterized by its focused nature, excelling in well-defined tasks or areas of application, but lacking the ability to generalize knowledge or skills to other domains (Condello et al, 2021;Liem et al, 2018). This is often contrasted with strong AI, which aims to replicate human-level intelligence and possesses general intelligence (Schachner et al, 2020;Süße et al, 2021). Weak AI systems are typically developed using machine learning algorithms and trained on large datasets to perform specific functions with a high degree of accuracy (Park & Park, 2018).…”