Towards a Future Reallocation of Work between Humans and Machines Taxonomy of Tasks and Interaction Types in the Context of Machine Learning

Abstract

In today's race for competitive advantages, more and more companies implement innovations in artificial intelligence and machine learning (ML). Although these machines take over tasks that have been executed by humans, they will not make human workforce obsolete. To leverage the potentials of ML, collaboration between humans and machines is necessary. Before collaboration processes can be developed, a classification of tasks in the field of ML is needed. Therefore, we present a taxonomy for the classification of tasks due to their complexity and the type of interaction. To derive insights about typical tasks and task-complexity, we conducted a literature review as well as a focus group workshop. We identified three levels of task-complexity and three types of interactions. Connecting them reveals three generic types of tasks. We provide prescriptive knowledge inherent in the task/interaction-taxonomy.

To Use or Not to Use Artificial Intelligence? A Framework for the Ideation and Evaluation of Problems to Be Solved with Artificial Intelligence

“…To achieve research rigor, the coding process and initial framework were validated in subsequent discussions between five IS researchers and three student assistants [17]. Additional data sources regarding problem finding for digital technology contexts and the use of AI in organizations [i.e., 2,6,8,9,16,23,48,57,58] were also considered to compare results with existing knowledge. Based on this data and investigator triangulation [17], the final framework was formed.…”

Social Machines

“…So kann einerseits das System den Menschen trainieren (AI in the Loop) und andererseits der Mensch das System trainieren (Human in the Loop). Es können Probleme gelöst werden, deren Lösungen bisher typischerweise dem menschlichen Intellekt vorbehalten waren und beispielsweise strategische oder politische Entscheidungen betreffen [7,39].…”

Hybrid Teamwork: Consideration of Teamwork Concepts to Reach Naturalistic Interaction between Humans and Conversational Agents

“…insights and predictions), which rely on vast amounts of data and thereby qualify as criteria for human workers' decision making [1]. Moreover, teamwork between humans and machines will entail and allow the delegation and allocation of (sub-)tasks to one another [5,6].…”