You have read the books. Done the journaling. Built the routines, designed the morning protocol, tracked the habits, watched the YouTube videos on deep work and time blocking and systems thinking. You have deleted Notion and rebuilt it from scratch twice. You feel like you understand productivity better than most people. And yet you look up and two years have passed and the actual things you wanted to do are still not done.
This is the self-improvement treadmill. Moving constantly, going nowhere.
What is actually happening
The pull toward self-improvement content is not random. It gives you the feeling of progress without the exposure of doing the real thing. When you read about discipline or design a new system, the project is still perfect. The moment you actually execute, it is not. The system is a protected space where nothing can fail yet.
Psychologist Peter Gollwitzer has documented this in research on what he calls "symbolic self-completion." When people have an identity goal -- "be a productive person," "be a person who ships things" -- they compensate for not doing the goal by doing activities that symbolize the goal instead. Reading productivity books, building systems, watching talks about execution all signal to yourself and others that you are someone who is working toward being that person. Except the signaling absorbs the motivation that would otherwise go into the actual work (Gollwitzer & Wicklund, 1982).
The second mechanism is what researcher Ayelet Fishbach calls "goal licensing." Once you feel you have taken meaningful steps toward a goal -- even preparatory ones -- the brain grants you permission to ease off. Two hours spent reorganizing your task system registers as productive work, so the actual project gets a pass for the day. The more elaborate the preparation ritual, the more licensing it grants (Fishbach & Dhar, 2005).
This is why deleting Notion does not fix it. The tool is not the mechanism. The mechanism is that preparation and action both feel like forward motion, and preparation is always lower risk.