The Offloading Ladder: How Experience Changes What Learners and Practitioners Should Hand to Artificial Intelligence
Year: 2026
Type: preprint
Abstract: Learning involves struggle, and large language models make it easy to offload the cognitive work that produces learning. Recent studies show that this offloading also lowers learners' persistence and interest, although offloading repetitive work can leave people more time for the high-order work that people and institutions value. To protect learning, I argue that the room to offload grows with experience, because a task's order is set not by the task alone but by the person's experience with it. Building on Bloom's taxonomy and research on higher-order thinking, I call a task low order for a person when they already have the knowledge or procedure it needs, and high order when they have to work out what to do. The task stays fixed, but through repeated practice, it moves from high to low order for the person, so its order differs from one person to the next and for one person over time. I propose the offloading ladder, a model of what a person can offload as their experience with a task grows. The ladder stops at the core, the work a person needs to keep doing to judge the tool's output. I close with implications: rules for AI use should depend on the person as well as the task, assignments should aim at the core of the work, experienced people should keep the core, and AI companies should build tools that adjust how much of the work they do to the user's experience.
Author: Mete Akcaoglu
Language: English
Subfield: Information Systems
Field: Computer Science
Domain: Physical Sciences
SDG: Quality education
Open Access status: green
OpenAlex ID: W7223837901