Cognitive processes are defined by their accessibility to introspection—they are the categories of thought we can become aware of and reflect upon. Generative AI, however, lacks an awareness of its own thought processes and the introspective capacity to refine its mental content iteratively. While its outputs typically appear highly sophisticated, the absence of true cognition reveals itself in weaknesses in judgment, planning, and logical reasoning—subtle gaps that may be hard for novice learners to identify. Using examples from an introductory computational biology course, I demonstrate how these subtle deficiencies can affect collaborative exploration when students are tasked with extracting insights from biological patterns and constructing quantitative models to evaluate their significance. These observations point toward a potential solution: restructuring dialogues to cultivate introspective thinking within human-AI collaboration. By reintroducing cognitive processes into AI-mediated learning, we can create interactions that are not only more meaningful but also more educationally effective. Bio: Boris Steipe, MD, PhD, is a Professor Emeritus, Temerty Faculty of Medicine, University of Toronto, Canada. He was an Associate Professor of the Departments of Biochemistry and Molecular Genetics, University of Toronto, Canada (2001-2022), prior to which he received a PhD and Habilitation, and held academic positions in Germany at the Ludwig Maximilians University of Munich and the Max-Planck Institute for Biochemistry. His specializations include Structural and Computational Biology, Confucian Ethics, Japanese Aesthetics, and Artificial Intelligence in Education. He dedicates his work to bridge the domains of the sciences, the humanities, and the arts.