Zettelkasten over AI curation
On social, writers are pushing back on ‘AI curates my brain’ setups and praising Zettelkasten‑style atomic notes and explicit links as a more durable way to build knowledge you actually understand. (x.com) The argument is simple: small, linked notes preserve the reader’s intuition and searchability in ways black‑box AI summaries often don’t. (x.com)
Writers on social are arguing over a simple question: do you want a machine to “curate your brain,” or do you want notes you can actually trace back to your own thinking? Two recent posts pushed the debate into the open by praising Zettelkasten-style note systems over black-box artificial intelligence summaries. (x.com 1) (x.com 2) The split is not really about software. It is about whether knowledge stays in small, visible pieces you can inspect, or gets compressed into a polished answer whose path you cannot easily follow. (zettelkasten.de) (roamresearch.com) Zettelkasten is a German word that means “slip box.” The method is associated with German sociologist Niklas Luhmann, who published more than 70 books and nearly 400 articles while working from a giant paper note archive. (en.wikipedia.org) (openlibrary.org) The core unit in that system is the atomic note. An atomic note holds one idea, written clearly enough that it can stand on its own when you find it again six months later. (zettelkasten.de) (notedexapp.com) The second unit is the link. Instead of dropping a note into one folder and forgetting it, you connect it to another note with an explicit relationship, the way one Wikipedia page points you to the next page you need. (roamresearch.com) (obsidian.md) That explicitness is what many writers say gets lost in “artificial intelligence curates my brain” setups. A summary can feel useful in the moment, but it often strips away the trail of why one idea led to another and which source created the jump. (x.com) (zettelkasten.de) Searchability is part of the argument too. A vault of linked Markdown notes can usually be searched by phrase, title, backlink, and file, while an opaque summary depends on the model having made the right compression choices up front. (obsidian.md) (roamresearch.com) That difference shows up when you revisit an old topic. In a linked-note system, you can reopen the exact note, see the neighboring notes, and inspect the wording you wrote at the time; in a summary-first system, you often get a fresh answer that sounds confident without showing the intermediate steps. (roamresearch.com) (zettelkasten.de) The current backlash is also a reaction to the “second brain” boom of the last few years. Many note tools now promise automatic summaries, automatic links, and automatic insight generation, which shifts the user from building connections to approving suggestions. (atlasworkspace.ai) (zettelgarden.com) Even people who like artificial intelligence inside note apps often describe a narrower role for it. The more durable pattern is using the model to draft, summarize, or suggest, while keeping the final note small, editable, and manually linked by the reader. (zettelkasten.de) (github.com) That is why the praise for Zettelkasten keeps resurfacing. The method forces a person to decide what a note means, what other note it belongs with, and what wording is precise enough to survive outside the original article, book, or chat window. (zettelkasten.de) (soenkeahrens.de) The posts circulating now are really making a conservative claim. If you want knowledge that compounds over years instead of answers that vanish after one prompt, one-idea notes and explicit links are still beating the fantasy that a model can quietly organize your mind for you. (x.com 1) (x.com 2)
