Key points are not available for this paper at this time.
ABSTRACT Intelligent tutoring systems (ITS) are reshaping adult learning in STEM by providing adaptive, data‐driven instruction across classrooms, workplaces, and informal environments. In the context of ITS, this article compares generative AI, which creates personalized explanations and practice materials, with explainable AI, which focuses on transparency and learner trust. Generative models can scale content creation and simulate authentic problem‐solving, while explainable systems help learners understand and calibrate their own knowledge. Together, these approaches offer new opportunities for self‐directed and lifelong learning but raise challenges related to correctness, utility, and interpretability. Integrating both forms of AI into ITS may redefine how adults engage with STEM education and training.
Zarestky et al. (Tue,) studied this question.