Use Cases for Prospective Sensemaking of Human-AI-Collaboration

Overview

• This study explores the potential of human-AI collaboration (HAIC) through interviews with 14 executives.
• The researchers identified 63 HAIC use cases and classified them using a matrix combining value chain and group work activities.
• Most use cases were related to firm infrastructure and technology development, with few pertaining to services, procurement, or logistics.
• HAIC is predominantly seen as supporting the choice and execution of group tasks, with more emphasis on choosing in supporting activities of the value chain.

Plain English Explanation

The researchers wanted to understand how businesses can work together with AI systems. They interviewed 14 business leaders and identified 63 different ways that humans and AI could collaborate. They grouped these use cases into a matrix based on the type of business activity (like operations or marketing) and the type of collaborative task (like choosing what to do or actually doing it).

Most of the use cases they found were related to the behind-the-scenes parts of a business, like managing technology and infrastructure. Very few were about customer-facing activities like services or logistics. Overall, the business leaders saw AI as being most helpful for choosing what tasks to do and then executing those tasks, rather than for negotiating or figuring out the best approach.

The researchers believe this information could help businesses plan how to effectively use AI to support their operations and strategic decision-making.

Technical Explanation

The researchers conducted semi-structured interviews with 14 executives to explore the potential of human-AI collaboration (HAIC). They identified 63 distinct HAIC use cases and classified them using a novel matrix that combined value chain activities and group work activities.

The majority of use cases were related to firm infrastructure and technology development, with very few pertaining to services, procurement, or logistics. HAIC was predominantly seen as supporting the choice and execution of group tasks, with more emphasis on choosing in supporting activities of the value chain. In contrast, primary activities like operations and marketing focused more on executing group tasks, with fewer use cases involving negotiating tasks.

The researchers propose that this taxonomy of HAIC use cases could serve as a tool for prospective sensemaking and to foster strategic managerial decisions around the deployment of AI in the workplace.

Critical Analysis

The study provides a valuable starting point for understanding how businesses currently envision the role of HAIC, but it has some limitations. The sample size of 14 executives, while appropriate for a qualitative study, may not be representative of all industries and organizational contexts. Additionally, the researchers acknowledge that the use cases identified may be biased towards more mature or widespread applications of HAIC.

Further research is needed to explore how these use cases may evolve over time, as well as to investigate potential barriers or challenges to successful HAIC implementation. It would also be interesting to examine how the perceived value and applications of HAIC may differ across various stakeholder groups, such as frontline workers, middle managers, and C-suite executives.

Conclusion

This study offers a nuanced and practical perspective on the current state of HAIC in the business world. By identifying a range of use cases and classifying them based on value chain and group work activities, the researchers provide a framework for organizations to consider how they might leverage AI to support their strategic and operational objectives. As AI technologies continue to evolve, this research can help guide managers in making informed decisions about where and how to integrate human-AI collaboration into their businesses.