Grounded in the learning sciences, Generative Learning (GL), is a learning method or process that encourages students to actively generate information and make connections between new and existing knowledge. This active participation in the learning process promotes a deeper understanding of the instructional material, fosters long-term retention of knowledge, and cultivates critical thinking skills and problem-solving abilities. The magic lies in the “generation” process, where learners actively make sense of the material rather than passively receiving information. In this paper, we argue that Generative Artificial Intelligence (GenAI) can be used as a Mindtool (knowledge representation tool) to facilitate GL by enhancing and augmenting learning rather than replacing the learning process. More specifically, we describe how GenAI can be used as a learning strategy or study buddy to support knowledge organization and comprehension monitoring in varying degrees of complexity; as a collaborative thinking tool to foster teamwork and facilitate project-based activities by encouraging the sharing, discussion, and integration of spatial representations of content in order to construct a more cohesive and comprehensive knowledge structure; as a possibility engine that helps students explore different ways of expressing ideas by generating alternative responses; as a Socratic opponent that challenges students to develop and refine their arguments; as a personal tutor that provides personalized feedback; as an exploratory research engine that allows students to explore and interpret data; as a motivator that proposes games and challenges to engage learners; and as a dynamic assessor that can evaluate students' knowledge in real time, allowing for tailored generative learning activities (GLA) based on the students' current understanding. The paper ends with the conceptualization and application of a pedagogical model or framework that can be used to design and support GLA using GenAI technologies.
Dabbagh et al. (Mon,) studied this question.