Get AI summaries of any video or article — Sign up free
Zettels, AI, and a Book Even I Didn't See Coming thumbnail

Zettels, AI, and a Book Even I Didn't See Coming

5 min read

Based on Zsolt's Visual Personal Knowledge Management's video on YouTube. If you like this content, support the original creators by watching, liking and subscribing to their content.

TL;DR

Zettelkasten chains don’t require finding an “ideal” parent note; linking to the least versatile option is enough to create meaningful trains of thought.

Briefing

Zettelkasten’s “train of thought” doesn’t require finding the perfect parent note—success comes from linking to the least versatile (least “fit”) option, turning the order of discovery into a personal fingerprint. After struggling to build coherent chains because the process kept demanding an “ideal parent,” the creator learned that Zettelkasten is designed to work with imperfect matches. The sequence in which ideas appear, get observed, and get linked to the least versatile candidate is what makes the resulting chain uniquely theirs, rather than a universal structure they’re trying to force.

That shift reframes how they think about note-making and writing. Instead of treating note connections as a correctness problem, it becomes a creativity problem: the chain is “yours” because it reflects your path through ideas. The insight also echoes a well-known idea attributed to Steve Jobs—dots can’t be connected looking forward, only looking backward—supporting the claim that perfection in forward planning isn’t necessary. For future work, the creator says they’ll focus more on building true chains rather than “pools of thought,” using the least-versatile linking strategy to create a more structured storyline. If a second book happens, they expect to be better prepared because the methodology will already be more tightly organized.

The second major takeaway centers on AI, shifting attention away from privacy or “fraud” concerns toward a more fundamental risk: giving up agency. The creator acknowledges the obvious worry—sharing personal notes with AI can leak sensitive information—but highlights a larger danger raised in workshop discussions: dependence that gradually replaces independent thinking. The mechanism is familiar in everyday life. People offload memory by searching for book titles or facts on demand; they remember not the content itself, but how to retrieve it. That’s manageable. The danger escalates when AI becomes a source of decisions—brainstorming ideas, proposing initiatives, and shaping the direction of work.

A corporate example illustrates the pattern: when a boss speaks first, subordinates tend to align with that perspective, and “group think” replaces genuine contribution. With AI, prompting it first to brainstorm can similarly steer outcomes toward what the model produces, leaving less room for the user’s own perspective. The creator isn’t claiming AI is inherently catastrophic—writing their book with AI as a tool is part of their reality—but they found that without a clear story line and constraints, AI output meandered, repeated itself, and lacked an arc. The core message and narrative structure still have to come from the human mind; AI can support the writing process, but initiative must originate with the user.

Alongside these insights, the creator announces a book pre-order on Amazon (release set for June 15) and promotes cohort 13 of a visual thinking workshop, starting April 12, centered on Tiny Experiments by Ann Lur Lant—framing their approach as a series of small experiments that accumulate into a coherent methodology.

Cornell Notes

The creator’s Zettelkasten breakthrough is that building “train of thought” chains doesn’t depend on finding the ideal parent note. Instead, the system works by linking each new idea to the least versatile (least “fit”) candidate, and the order of discovery becomes the user’s fingerprint. That same mindset—imperfect inputs, personal structure—also informs their writing plans.

On AI, the key risk isn’t mainly privacy or misinformation; it’s losing agency. Offloading simple recall to tools is one thing, but prompting AI first for brainstorming and decisions can steer thinking toward model-driven group think. Their own experience writing with AI shows that a clear human story line is required; otherwise AI output meanders, repeats, and lacks an arc.

Why does “least versatile” linking matter in Zettelkasten, and how does it change the goal of note connections?

The creator previously tried to find the “best parent” note for each new idea, which produced anxiety and resulted in “pools of thought” rather than chains. The new lesson is that Zettelkasten is built to work without perfect parent selection. The practical rule is to link to the least versatile candidate note. The resulting chain becomes personal because it reflects the user’s discovery sequence—what ideas show up, what gets noticed, and what gets linked next. In that framing, note linking is not a correctness test; it’s a way to preserve the user’s unique path through ideas.

How does the Jobs-style “connect dots backward” idea support the Zettelkasten insight?

The creator connects the least-versatile strategy to the idea attributed to Steve Jobs: you can’t connect dots by planning ahead; you connect them by looking back. That supports the claim that perfection in forward planning isn’t required. Instead of forcing an ideal structure while new ideas are still forming, the chain can emerge after the fact—built from whatever link choices were available at the time, including the least versatile option.

What’s the difference between using tools to retrieve information and using AI to shape decisions?

The creator draws a line between searching for information and searching for decisions. When someone forgets a book title, they can search it and continue the conversation; the person still contributes by choosing what to discuss next. But with AI, the prompts can move from “find facts” to “brainstorm ideas,” “generate initiatives,” and “decide direction.” That shift matters because it can replace the user’s own initiative with the model’s output.

How does “boss speaks first” illustrate the agency problem with AI?

In a corporate setting, when the boss speaks first, subordinates often align with the boss’s perspective, even if they intend to contribute. The result is reduced freedom of thinking and a drift toward group think. The same dynamic can happen with AI: if a user prompts AI first to brainstorm, the user may end up driven by the AI’s suggested outcomes rather than their own perspective, effectively outsourcing the creative starting point.

What did the creator learn from writing their book with AI about maintaining a human narrative arc?

AI can support drafting, but the creator found that without a very clear story line—what they want to convey and where the narrative should go—AI output tended to meander, repeat, and fail to produce a coherent arc. Their contribution is the message and structure; AI assistance affects the mechanics of writing, not the originating intent.

Why is Tiny Experiments positioned as relevant to their methodology?

The creator frames their work—channel content, the book, and the visual thinking workshop—as a sequence of small experiments that accumulate into a methodology. Tiny Experiments by Ann Lur Lant is presented as a match for that approach, reinforcing the idea that small, iterative trials can produce real, usable results rather than requiring a perfect plan from the start.

Review Questions

  1. In Zettelkasten, what does “least versatile” linking replace, and what does it preserve about the user’s thinking process?
  2. What does “giving up agency” mean in the context of AI prompting, and how is it different from using tools to look up facts?
  3. How can a clear human story line prevent AI-assisted writing from becoming repetitive or directionless?

Key Points

  1. 1

    Zettelkasten chains don’t require finding an “ideal” parent note; linking to the least versatile option is enough to create meaningful trains of thought.

  2. 2

    The order in which ideas are discovered and linked becomes the user’s personal fingerprint, making the chain inherently theirs.

  3. 3

    Planning perfect connections in advance isn’t necessary; connections can be made by looking back at what emerged.

  4. 4

    The biggest AI risk highlighted is not only privacy or fraud, but gradual loss of agency when AI is used to generate decisions and direction.

  5. 5

    Prompting AI first for brainstorming can produce group think by steering outcomes toward the model’s suggestions rather than the user’s own perspective.

  6. 6

    AI can assist writing, but the human must supply the core message and narrative arc; without a clear story line, AI output can meander and repeat.

  7. 7

    Small experiments are treated as a practical method for building a coherent personal knowledge workflow over time.

Highlights

Zettelkasten’s “train of thought” can be built by linking each new idea to the least versatile parent—no perfect match required.
The sequence of idea discovery and linking choices is what makes the resulting chain uniquely personal.
The agency risk with AI grows when prompts shift from retrieving information to generating decisions and brainstorming direction.
Writing with AI still depends on a human-defined story line; otherwise the draft loses arc and coherence.
Tiny Experiments is positioned as a direct fit for a methodology built from iterative, small trials.

Topics

Mentioned