The article presents a model that helps understand how people switch between tasks when multitasking. The researchers used a cognitive architecture called ACT-R to create a model that predicts human behavior in multitasking situations. They focused on discretionary multitasking, where people choose when to switch tasks. The model accurately predicted task-switching behavior in a memory game and a subitizing task. It showed that humans interrupt themselves based on task demands and make decisions to switch tasks. The model's performance improved with task difficulty, showing how ACT-R can explain voluntary task switching.