Leapfrogging effect hypothesis: Generative AI as a permanent scaffold in higher education

This paper introduces the leapfrogging effect hypothesis, proposing that generative AI significantly expands students’ Zone of Proximal Development (ZPD) by permanently scaffolding procedural tasks and enabling earlier engagement with higher-order cognitive activities. Drawing upon foundational theories—including Vygotsky’s ZPD (1978), Cognitive Load Theory (Sweller et al., 2011) and Bloom’s revised taxonomy (Anderson & Krathwohl, 2001)—it argues that traditional educational practices, which often enforce procedural gatekeeping, require re-examination in the AI era. The paper integrates recent empirical and theoretical research with practical examples from writing, mathematics, and programming education, highlighting how thoughtfully integrated generative AI can democratize advanced cognitive opportunities, especially benefiting traditionally disadvantaged or struggling learners. It advocates incremental curriculum reform, emphasizing raised cognitive expectations, explicit instruction in AI literacy, and deliberate, discipline-specific integration of AI-supported learning experiences.