Cyborgs, Centaurs and Self Automators: Human-Genai Fused, Directed and Abdicated Knowledge Co-Creation Processes and Their Implications for Skilling
Posted: 12 Sep 2024
Date Written: August 08, 2024
Abstract
Research on human-AI collaboration has focused primarily on task- specific interactions between knowledge professionals and AI at the point of decision making. Yet generative AI is a general purpose technology that broadens the possibilities of human-AI interactions, and changes the nature of interaction to be conversational and iterative. We set out to explore human-GenAI knowledge collaboration in a problem solving workflow. We share findings from a field study of 244 global management consultants from the leading firm, Boston Consulting Group, as they engaged with GenAI (Chat GPT-4) for business problem solving. We found that knowledge professionals may engage in one of three distinct types of human-AI collaboration that we conceptualize as fused co-creation (“cyborgs”), directed co-creation (“centaurs”), and abdicated co-creation (“self-automators”). We describe these three different types of knowledge co-creation, and demonstrate that each of these types has implications for professionals skill transformation around task-related capabilities and AI-related capabilities. Professionals who engaged in directed co-creation (centaurs) upskilled themselves in using AI for task-related capabilities. In contrast, professionals who engaged in fused co-creation (cyborgs) newskilled themselves by acquiring and developing new GenAI-related capabilities. Professionals who engaged in abdicated co-creation failed to gain new skills during their human-AI collaboration. Understanding fused, directed, and abdicated co-creation and their important skilling outcomes is critical to understanding the future of problem-solving work and knowledge work more generally.
Suggested Citation: Suggested Citation