AI-Powered Simulator Design Principles for Skill Transfer
Abstract
Interest in AI-powered simulations is growing faster than the instructional foundations for their design are developing. Research has focused mainly on technological feasibility and short-term outcomes within learning environments, while the transfer of skills to real work remains understudied. This article examines which AI simulator design principles support the development of professional skills and their transfer to workplace practice using the theory synthesis method (Jaakkola, 2020) across four areas: transfer of training, simulation-based learning, AI in training, and workplace learning. The analytical framework is the Baldwin and Ford model (1988). For each identified design characteristic, the mechanism through which it may influence transfer in traditional simulations and the limits of the existing evidence were established; changes introduced by AI were then analyzed, and a final design principle was formulated. Eleven design principles were identified and organized into a four-layer transfer-oriented architecture: target skill and performance criteria, pathway to autonomous practice, integration into the learning program, and workplace connectedness.