Modelling Hybrid Human-Artificial Intelligence Cooperation: A Call Center Customer Service Case Study

As autonomous systems become an essential part of augmented decision-making in the workforce, we have the opportunity to change the relationship between human and machine into a more collaborative one. The future of industry, commercial and public services point in a direction where humans and artificial intelligence (AI) increasingly work together. AI systems are increasingly extending and enriching decision support by complementing and augmenting human capabilities. To further elevate this partnership, we need to form organic human-AI teams that communicate with, adapt to, and learn from each other. We create a new human-in-the-loop hybrid spectrum, that expands existing definitions of human and machine teaming. For a given situation and a team of humans and AI systems, we are interested in testing variations on human-AI cooperation outcomes. We examine a call center use case to determine how variations in human and machine teaming affects average handle time and response quality outputs that affect customer service. We have evaluated three scenarios: 1) human-only, 2) AI-only, and 3) human + AI collaboration. Under the parameter space we studied, we found that human + AI collaboration is optimal.