As the adoption of artificial intelligence (AI) in the workplace is expected to surge, understanding the dynamics of employee-AI interaction becomes crucial. In a pioneering study, Chris McComb and Jon Cagan conducted a head-to-head matchup between human teams and hybrid teams comprising humans and AI agents to design a fleet of delivery drones under dynamic constraints.
McComb, an associate professor of mechanical engineering, elaborated, "We compared teams of five humans against hybrid teams of three humans and two AI teammates to analyze failure modes in human-AI teams." The experiment, conducted on the online platform HyForm, aimed to assess team communication and fleet success.
Surprisingly, the human-AI hybrid teams performed on par with human teams, showcasing adaptability to unforeseen events. Published in the Journal of Mechanical Design, the study highlighted the crucial role of hybrid teams in handling unexpected challenges.
Initially, human teams exhibited significantly higher communication levels than hybrid teams. However, communication within hybrid teams surged when unforeseen constraints were introduced. Notably, much of the increased communication involved interactions between humans and AI teammates, indicating the emergence of human "AI handlers" to guide AI agents.
Despite their success, members of hybrid teams expressed concerns about their cohesion and effectiveness. Cagan, a professor of mechanical engineering, emphasized the significance of human adaptability in hybrid teams, stating, "The flexibility of human-AI teams can be driven by the human side of the partnership, negating the need for perfectly adaptive AI partners."
McComb underscored the essential role of humans in human-AI teaming, highlighting the importance of training future engineers to collaborate effectively with AI agents. The study sheds light on the potential of hybrid teams in navigating the evolving landscape of AI integration in the workplace.
