1 paper
Philipp Hugenroth, Valdemar Danry, Pattie Maes
As Large Language Models (LLMs) increasingly automate writing tasks, there is a growing risk of cognitive deskilling where users offload critical thinking to the system. To address…