Thinking with AI

Abstract

This article introduces the Thinking with AI framework, a four-phase model for developing artificial intelligence (AI) literacy in undergraduate education. The framework organizes student engagement with generative AI across a continuum: Passive Thinking, in which AI functions as a tool and the student as observer; Parallel Thinking, in which AI functions as a collaborator and the student as co-explorer; Deep Thinking, in which AI functions as a thinking partner and the student as synthesizer and evaluator; and Generative Thinking, in which AI functions as a design medium and the student as architect. The framework is grounded in current research on cognitive offloading, metacognitive laziness, AI literacy competency development, and human–AI collaboration and responds to documented institutional gaps in AI preparedness among students, faculty, and institutions. For each phase, the article describes the student role, the AI role, the cognitive load implicated, and a corresponding assignment design. The framework is positioned as a curricular tool that allows instructors to address AI literacy by treating cognitive engagement as the primary unit of pedagogical analysis.