Distributed Atrophy: How AI Shapes and Shrinks Our Cognitive Habits

The integration of generative AI into everyday cognition raises urgent questions about its long-term impact on human mental processes. This paper introduces two theoretical models—the Distributed Atrophy Model and the Cognitive Cost Comparator—to examine how habitual reliance on large language models (LLMs) may reshape cognitive effort allocation, memory encoding, and metacognitive self-trust. Drawing on research in cognitive offloading, transactive memory systems, and digital tool use, the paper outlines potential downstream effects of AI-mediated cognition, including representational decay and reduced cognitive autonomy. Rather than framing AI as inherently corrosive, it emphasizes the importance of intentional use and system design to safeguard human cognitive agency.