Mensch-KI-Teaming: Mensch und Künstliche Intelligenz in der Arbeitswelt von morgen

Abstract

Der Einsatz von Künstlicher Intelligenz (KI) in Produktion und Fertigung wird die Arbeitswelt und die Fabrikfertigung verändern. Dieser Beitrag widmet sich der Frage, wie die Arbeit mit KI im Interesse des Menschen gestaltet werden kann. Dafür werden verschiedene Perspektiven auf menschzentrierte KI vorgestellt. Nachfolgend wird das Modell SMART-er Arbeitsgestaltung eingeführt [37, 43]. Dieses dient als Heuristik, vor dessen Hintergrund Kriterien für die Entwicklung menschzentrierter hybrider Intelligenzen diskutiert werden. Abschließend werden fünf Gestaltungsschritte abgeleitet, die Unternehmen bei der menschzentrierten Gestaltung von Mensch-KI-Teams helfen können.

The Job Perception Inventory: considering human factors and needs in the design of human–AI work

“…Such an adaptive task allocation brings enormous benefits for the human and the quality of the work (Tausch et al, 2020 ), and also increases the complexity in the workplace. Given the increasing complexity and possibilities of usage, it is necessary to address the identity-forming and motivation-promoting factors of a workplace while implementing intelligent technologies to meet the needs and capabilities of the employees, and thus to be able to benefit from the synergies of cooperation (Kluge et al, 2021 ). For this purpose, the present work contributes by deducing an inventory from existing findings of psychological work research, which should accompany human-centered implementation in the workplace.…”

“…The way an AI is evaluated depends on its functions, interaction capabilities, accuracy, and task appropriateness (Parker and Grote, 2020 ). Especially in a human–AI teaming workplace, the possibility for interaction and collaboration is crucial, as the workplace design is aimed at collaboration (Kluge et al, 2021 ). Furthermore, successful collaboration and task distribution are important for the perception of the workplace after the AI introduction (Sadeghian and Hassenzahl, 2022 ).…”

“…In the past, principles of design, implementation, and application of technical systems often focused on the technology itself and the associated strengths without considering the consequences or strengths of the human part (Kluge et al, 2021 ). That needs to change.…”

“…A useful approach, which attempts to unite the strengths of an AI system with those of the human, is human–AI teaming (Huchler, 2022 ; Mirbabaie et al, 2022 ). Human–AI teaming describes the cooperation between the human and technical parts, which is characterized by mutual support, collaboration, transparent thought and action processes, and a common situational understanding (Kluge et al, 2021 ). This collaboration makes it possible to compensate for errors and to achieve better results (Jarrahi, 2018 ; Dubey et al, 2020 ; Kluge et al, 2021 ) because the capabilities and strengths of an AI system and a human complement each other ideally (Huchler, 2022 ): On the one hand, the strengths of AI can be identified in rational, analytical decision-making and the achievement of quantitative goals (Parry et al, 2016 ).…”

“…To overcome these barriers and enable a teaming approach, we developed an instrument to implement AI with a focus on humans with their needs, motives, and capabilities. Basing an AI implementation on this unlocks synergies in the workplace (Wilkens et al, 2020 ; Kluge et al, 2021 ) and promotes wellbeing, motivation, job identity, and in turn performance and organizational outcomes (Parker and Grote, 2020 ; Huchler, 2022 ).…”