Daniel Wentsch
@klickreflex
Freelance designer and web dev from Freiburg, Germany.
Daniel Wentsch
zettel.org
LLMs become powerful when you give them enough context to work with. Not a prompt, not a conversation—actual accumulated data over time. That context comes from years of Day One entries. Daily capture happens through Rosebud—voice notes with immediate AI feedback. When I want deep analysis, I turn to Claude Code.
Article
You don’t need interesting material. You’re not performing for an audience. You can write about why you chose the lunch you chose, or what you noticed on the walk to the store, or a single, unmarked moment in a conversation that stuck with you. The point isn’t to produce great content. The point is to practice putting thoughts into words.
Joan Westenberg • Why You Should Write Every Day
My daily writing is usually mundane. I’ll work through why I’m annoyed about something in my work or home life, or try to figure out whether I actually need to move to a new note taking app or just want to want to move note taking apps for the sake of it. The topic matters less than the act of forcing vague feelings into specific words. Once
... See moreJoan Westenberg • Why You Should Write Every Day
I’ve been writing 1000 words every day for about 10 years now, and I’m still not always sure what I’m doing. Arguably, most of that output is terrible. Some days I manage maybe two hundred words before I run out of things to say and end up describing what I had for breakfast.
But those three years have changed how my brain works more than any other
... See moreJoan Westenberg • Why You Should Write Every Day
Above all, do not lose your desire to walk: every day I walk myself into a state of well-being and walk away from every illness; I have walked myself into my best thoughts, and I know of no thought so burdensome that one cannot walk away from it. Even if one were to walk for one’s health and it were constantly one station ahead—I would still say:... See more
Kierkegaard, 1843.
Third, and this is harder, you have to learn to metabolize feedback differently. When reality contradicts your model, it’s not a failure signal; it’s data. Neutral, immutable data. That's the whole point of making models in the first place, to test them against the world and update them. But you can only get that data by acting. The model serves
... See moreJoan Westenberg • The Map Is Not the Territory
First, you have to get comfortable with the idea that your model will always be incomplete and partially wrong. Always. You could spend your entire life refining your understanding of how to write a novel, and you'd still learn more from writing one bad novel than from all that study. The territory is infinitely richer than any map can capture.
... See more