Learning Beyond the Individual: Building Collective Capability in Human AI team
TL;DR
The paper proposes the CYCLE framework, which translates the framework into a practical operating model for L&D, including team simulations, decision-trace practices, peer challenge, AI debriefs, and measures of transfer at team level.
Abstract
Artificial intelligence is entering the workflow faster than most organizations are redesigning learning. This paper argues that the central L&D challenge is no longer individual AI literacy alone, but the collective capability of teams to reason, coordinate, challenge, remember, and improve with AI. An integrative review of peer-reviewed research and recent workforce studies is used to connect human-AI teaming, transactive memory, shared models, workplace learning, and capability development. The evidence is striking: 84% of executives expect regular human-AI collaboration within three years, yet only 26% of workers report being trained to collaborate effectively with AI; high-performing teams report higher AI use than other teams (78% versus 54%), but their advantage is also associated with trust, apprenticeship, agility, and human connection. The paper proposes the CYCLE framework: Clarify roles, Yield to evidence, Capture memory, Learn through correction, and Embed routines. It translates the framework into a practical operating model for L&D, including team simulations, decision-trace practices, peer challenge, AI debriefs, and measures of transfer at team level. The argument is deliberately human-centered: AI may accelerate access to knowledge, but collective capability develops only when people can question outputs, speak up, share judgment, and retain learning. The paper concludes with propositions and a field-research agenda for testing durable team capability over time.