Abstract
This paper presents a theoretical framework for understanding how generative AI may transform educational practice through what we term the leapfrogging effect. We propose that generative AI may function as a permanent scaffold, potentially enabling students to bypass traditional procedural prerequisites and engage earlier with higher-order cognitive tasks. This hypothesis emerges from a novel synthesis of Vygotsky's Zone of Proximal Development, Cognitive Load Theory, and Bloom's revised taxonomy. While each framework has been applied to educational technology separately, their integration provides unique explanatory power for understanding how AI scaffolding may expand cognitive access.
Drawing on theoretical analysis and illustrative examples from writing, mathematics, and programming education, we examine how traditional practices enforce procedural gatekeeping that delays advanced cognitive experiences, particularly for struggling or neurodiverse learners. The paper explicitly acknowledges that this hypothesis requires empirical validation and that current AI implementations exhibit significant variability across contexts. This contribution is theoretical rather than empirical: we articulate and justify a testable hypothesis to guide future research. We conclude by outlining directions for incremental curriculum reform and calling for sustained empirical investigation to determine under what conditions AI scaffolding enhances rather than undermines deep learning.
•AI may serve as a permanent scaffold bypassing procedural skills.•Vygotsky, Cognitive Load Theory, and Bloom are synthesized.•Procedural gatekeeping delays access to higher-order cognition.•AI scaffolding may reduce equity gaps in higher education.•The leapfrogging hypothesis requires empirical validation.