Reversing Knowledge Gains, Notes on Notes, Actionable Problems ![rw-book-cover](https://curtismchale.ca/favicon.ico) 🔗 **Source:** [Reversing Knowledge Gains, Notes on Notes, Actionable Problems](https://curtismchale.ca/2026/09/19/reversing-knowledge-gains-notes-on-notes-actionable-problems) > [!note]+ Document Notes > tags: [death-of-expertise, note-taking, productivity-critique, ai-and-learning, personal-knowledge-management] > author: Curtis McHale > source: curtismchale.ca > reference: https://curtismchale.ca/2026/09/19/reversing-knowledge-gains-notes-on-notes-actionable-problems > date created: 2026-09-19 18:07 > > > [!info] Generated analysis — not the author's words. Sections below are machine-written > > from the captured source. Only text inside `> [!quote]` blocks is verbatim from the author. > > Do not quote or attribute any other line to the named author. > > # 🏷️ Title: Reversing Knowledge Gains, notes on notes, actionable problems > <!-- deep-read:sections start --> ## 🧭 Table of Contents - [📝 Summary](#-summary) - [📖 Main Content](#-main-content) - [🗒️ Notes](#️-notes) - [📚 References](#-references) - [🗣️ Paraphrase](#️-paraphrase) - [📰 Newsletter](#-newsletter) - [✅ Action Items](#-action-items) - [🧪 Hypotheses](#-hypotheses) - [📊 Methodology](#-methodology) - [❓ Questions](#-questions) - [🔮 Future Research](#-future-research) - [💡 Implications](#-implications) - [🧠 Learning Styles Assessment](#-learning-styles-assessment) - [📊 Accessibility Assessment](#-accessibility-assessment) - [🌍 DEIB Assessment](#-deib-assessment) - [✅ Logical Validity](#-logical-validity) - [🛡️ DARVO Analysis](#-darvo-analysis) - [🎯 Conclusion](#-conclusion) ### 🔗 References - Not applicable to this source — the capture contains no outbound links. ## 📝 Summary - A three-item linked-list post connecting Tom Nichols's The Death of Expertise, a 2022 self-reflection on note-taking, and a line from Designing Your Life. - The first item extends Nichols's argument — that skimming headlines online produces false confidence and unwillingness to dig deeper — to AI as de facto search, arguing AI output is aggregated skimming reclassified as knowledge gain. - The second is uncomfortably self-directed: site analytics indicate readers come for notes about how to take notes rather than for thinking about anything else, which McHale connects to his earlier arguments that tool obsession is a fetish and productivity rewards are hollow. - The third quotes Designing Your Life page 9 — 'If it's actionable, it's not a problem' — and draws the sharper version: if you have merely chosen to do nothing about something, your inaction is the problem. - Stakeholders: anyone running a personal knowledge management practice, and specifically anyone whose PKM audience wants systems rather than substance. - Risk: all three items are assertions from reading rather than arguments with evidence, and the AI claim in particular is offered as a wondering rather than a finding. - Bottom line: a short, self-critical post whose real subject is that PKM audiences reward tooling over thinking — including the author's own. ## 📖 Main Content - The Nichols extension is the most interesting and the least supported. The original argument concerns skimming producing false expertise; applying it to AI requires the claim that reading a synthesis is epistemically equivalent to skimming sources, which is arguable in both directions and is not argued here. - The counter-case the post does not engage: a well-grounded synthesis with citations may produce better calibration than skimming headlines, because it surfaces disagreement the skimmer never sees. Whether AI output does that is an empirical question, not a definitional one. - The analytics observation is the post's genuine contribution because it is self-implicating. An author who writes about note-taking notices that the note-taking content is what sells, and names that as a problem with the audience and with the incentive. - Linking that to his own prior arguments — 'obsession over tools is a fetish', 'terrible rewards for productivity' — makes it a consistent position developed over years rather than a passing complaint. - The Designing Your Life quote is the sharpest item and the most immediately usable: reclassifying an unaddressed situation from 'problem' to 'inaction' relocates agency, and the follow-up question ('Why aren't you taking action?') is the operative part. - Format matters here. This is a linked-list post with book citations including page numbers (The Death of Expertise p.20, Designing Your Life p.9), which is better sourcing discipline than most short-form blogging. - Commercial context is disclosed openly at the top and bottom — membership and a free book club with roughly 3,000 readers — and does not shape the content. - Contested by construction: none of the three items is developed far enough to be assessed as an argument. That is the form's limitation, not a flaw in execution. ### 🗒️ Notes - Three unconnected items sharing an underlying theme: the gap between the appearance of intellectual work and the substance of it. - Directly relevant to Courtney's vault practice and the readwise-deep-read pipeline itself — a system that generates extensive notes is exactly the thing McHale is warning produces the feeling of knowledge without the thinking. - That self-application is the most useful way to read it: these notes are only worth their cost if they get used for thinking, not filed as evidence of having read. - The actionability reframe is the one item worth acting on directly; it converts a stalled item from a problem into a decision. - Page-numbered citations make both book claims checkable, which is unusual for the format. ## 📚 References - Not applicable to this source — no primary source is cited in the capture. ## 🗣️ Paraphrase - Nichols argues in The Death of Expertise that people who skim online headlines come to believe they are experts and become unwilling to listen or dig deeper. - McHale wonders what Nichols would say now that AI is the default search for many, since AI output is still essentially aggregated skimming being counted as knowledge gain. - Revisiting a 2022 post, he asks whether he is better known for notes about note-taking or for thinking about other topics. - His analytics answer the question: readers come for how to take notes, rarely for deeper thinking on other issues. - He connects this to earlier arguments that tool obsession is a fetish and that people want productivity's hollow rewards rather than engagement with ideas. - Quoting Designing Your Life: if it is actionable, it is not a problem. - His extension: if you have merely chosen to do nothing yet, your inaction is the problem — so the question is why you are not acting. ## 📰 Newsletter - Curtis McHale published three short thoughts, and the second one should sting anyone running an elaborate note-taking system. - His analytics say readers come to him for notes about how to take notes, and rarely for his thinking about anything else. He has been making this argument for years — tool obsession as fetish, productivity rewards as hollow — and now his own traffic proves the point. - The first item asks what Tom Nichols would make of AI as default search, given The Death of Expertise is about skimming producing false confidence. Worth noticing that reading a cited synthesis and skimming headlines are not obviously the same thing, which McHale does not argue either way. - The one to actually use is from Designing Your Life: if it's actionable, it's not a problem. If you have simply chosen to do nothing about it yet, the inaction is the problem — so ask why. ## ✅ Action Items #### DevRel - **Education:** The tools-versus-thinking tension applies directly to developer education, where tutorial traffic reliably exceeds conceptual content traffic. #### OSPO - Not applicable to this source — nothing in it bears on OSPO work. #### Community - **Contributors:** Audiences rewarding process content over substance is a pattern worth naming when planning community programming. #### WordPress - **Training:** Learn WordPress faces the same incentive — how-to content outperforms conceptual content — and should decide deliberately rather than by traffic. #### The WPOCC - **Programs:** Worth considering whether community programming rewards visible activity over durable thinking. #### FAIR - Not applicable to this source — nothing in it bears on FAIR work. ## 🧪 Hypotheses - Content about knowledge-management tooling receives more traffic than content applying that knowledge, across PKM-focused sites generally. - People who obtain information from AI synthesis show similar overconfidence effects to people who skim search results, measurable by calibration testing. - Reading a synthesis that surfaces disagreement produces better calibration than skimming sources that do not, which would falsify the post's implied equivalence. - Reframing a stalled item from 'problem' to 'inaction' increases the rate at which it is resolved or explicitly abandoned. ### 📊 Methodology - No study; this is a linked-list post citing two books with page numbers and one of the author's own prior posts. - The analytics claim is first-party data described qualitatively, with no figures given. - The AI extension of Nichols is explicitly framed as a wondering ('Now I wonder what Nichols would say'), not a finding, and should not be summarized as a conclusion. - To test the central AI claim: compare calibration on a topic after reading AI synthesis versus after skimming search results versus after reading primary sources. ### ❓ Questions - Is reading an AI synthesis epistemically closer to skimming headlines or to reading a review article? - Does surfacing disagreement in a synthesis improve calibration, and do current AI tools do it? - If an audience reliably rewards tooling content, is writing the other thing anyway a strategy or a hobby? - What distinguishes a note-taking system that produces thinking from one that produces only notes? ### 🔮 Future Research - Calibration studies comparing AI-synthesis reading against skimming and against primary-source reading are the obvious unbuilt research. - Whether extensive personal note systems improve or substitute for thinking is essentially unstudied despite a large practitioner literature. - The incentive structure of educational content — process outperforming substance — is measurable across platforms and rarely examined. ### 💡 Implications - Practice: apply the actionability reframe to any stalled item; it converts a vague problem into an explicit decision. - Practice: a note-taking system should be judged by what it produces downstream, not by its own completeness. - Policy: not applicable. - Education: educators whose how-to content outperforms their conceptual content face a real allocation choice rather than an accident. - Funding: audiences pay for systems more readily than for thinking, which shapes what independent writers can sustain. ### 🧠 Learning Styles Assessment - Reading level estimate: Flesch-Kincaid grade 8-9. Short linked-list blog post; three brief entries with citations to books. #### 👀 Visual learners - Present: No diagrams or images. - Improve: Not needed at this length. #### 🔊 Aural learners - Present: nothing in this source serves learners directly. - Improve: add material suited to learners if this source is reused for teaching. #### 🗣️ Verbal learners - Present: Concise, quotes sources directly with page numbers. - Improve: Each item is too short to develop; links carry the depth. #### 🧩 Kinesthetic learners - Present: The 'why aren't you taking action' question is directly usable on a real stalled item. - Improve: Nothing needed. #### 🧠 Logical learners - Present: The actionability reframe is a clean logical move. - Improve: The AI extension needs the equivalence argument it omits. #### 🧑‍🤝‍🧑 Social learners - Present: Links to a free book club with roughly 3,000 readers. - Improve: Nothing needed. #### 🧘‍♀️ Solitary learners - Present: Reflective and self-directed; suits private reading. - Improve: Nothing needed. #### 🌿 Naturalistic learners - Present: nothing in this source serves learners directly. - Improve: add material suited to learners if this source is reused for teaching. ### 📊 Accessibility Assessment #### Screen reader semantics - Present: Uses headings per item and proper footnote markup with return links. - Improve: Nothing evident needing improvement. #### Keyboard navigation - Present: not evaluable from the captured text alone. - Improve: assess keyboard navigation against WCAG 2.2 AA on the live page. #### Contrast and readability - Present: Not evaluable from the capture. - Improve: Standard check on the published page. #### Text resizing and reflow - Present: not evaluable from the captured text alone. - Improve: assess text resizing and reflow against WCAG 2.2 AA on the live page. #### Media and captions - Present: not evaluable from the captured text alone. - Improve: assess media and captions against WCAG 2.2 AA on the live page. #### Forms and interactions - Present: not evaluable from the captured text alone. - Improve: assess forms and interactions against WCAG 2.2 AA on the live page. #### Standards reference - Present: Footnote anchors with return links satisfy 2.4.1 Bypass Blocks patterns reasonably. - Improve: Not fully assessable from the text capture. ### 🌍 DEIB Assessment #### Cultural breadth - Present: Not addressed; the sources are two English-language US-published books. - Improve: The expertise argument has different force in contexts with different institutional trust baselines. #### Inclusive language - Present: Plain and direct. - Improve: 'Terrible rewards for productivity' assumes familiarity with the author's prior argument. #### Neurodiversity support - Present: Very short, three discrete items, easy to process. - Improve: The actionability reframe may land badly where inaction is a capacity issue rather than a choice; the framing assumes agency. #### Representation - Present: Not applicable at this length. - Improve: Nothing to improve. #### Psychological safety - Present: Self-critical rather than directed at readers; he implicates his own work first. - Improve: The 'your inaction is the problem' framing is blunt and could read as blaming in contexts of genuine constraint. ### ✅ Logical Validity #### Strong reasoning - The self-implicating analytics observation is the post's strongest move: rather than criticizing an abstract audience, he checks his own numbers and reports that they confirm the pattern he dislikes, which costs him something to say. - The actionability reframe is genuinely clarifying — relocating the problem from the situation to the inaction identifies where the decision actually sits, and the follow-up question makes it usable rather than merely clever. - Citing both books by page number is disciplined sourcing for a short-form post and makes both claims checkable. - Framing the AI extension as a wondering rather than a claim is appropriately modest given he supplies no evidence for it. #### Fallacies - **False equivalence** — located at: 'they're still essentially skimming a bunch of different texts aggregated into the AI return text and then calling it a knowledge gain'. Reading a synthesis and skimming headlines are treated as the same epistemic act, which assumes the conclusion. A synthesis that surfaces sources and disagreement may produce better calibration than skimming, which never shows the reader what was contested — and the post neither considers nor rules this out. Fix: State it as the open question it is, and specify what would distinguish a synthesis that builds understanding from one that substitutes for it. #### Tone of voice - Terse and unsentimental, with the register of notes made in public rather than an essay. - Self-critical rather than superior — the sharpest observation is aimed at his own audience and his own body of work. - Mildly weary about the productivity genre he is partly known for. ### 🛡️ DARVO Analysis - **Deny:** No instance present in this source. - **Attack:** No instance present in this source. - **Reverse victim/offender:** No instance present in this source. ## 🎯 Conclusion - Three short items whose common theme is the gap between apparent and actual intellectual work, with the author's own analytics offered as evidence against his own most popular content. - The AI extension of Nichols is a question worth holding, not a finding — the equivalence between synthesis and skimming is assumed rather than argued. - The actionability reframe is the piece worth using today: if something has gone unaddressed, the inaction is the problem, and the useful question is why. - Next step: apply it to one stalled item, and treat this deep-read pipeline as subject to the same test — notes are only worth their cost if they get used. ## 🔄 Refactor Appendix — Reversing Knowledge Gains, notes on notes, actionable problems ### 📚 Literature Atoms (H3) #### [H1] Analytics show readers want notes about notes > [!quote] > people come to me for notes on how to take notes. They rarely come to see my content for deeper thinking on other issues. **Context & use**: - Second item, reflecting on a 2022 post - First-party analytics offered as evidence against his own most popular work **Tags**: #xref #articleseed <!-- atomicity: 1 idea ✓ --> #### [H2] Inaction, not the situation, is the problem > [!quote] > If it's actionable, it's not a problem. **Context & use**: - Designing Your Life page 9, quoted directly - Extended by McHale: if you have chosen to do nothing yet, your inaction is the problem **Tags**: #xref #aphorisms <!-- atomicity: 1 idea ✓ --> #### [H3] AI search may reproduce the skimming problem > [!quote] > they're still essentially skimming a bunch of different texts aggregated into the AI return text **Context & use**: - First item, extending Tom Nichols's argument to AI as default search - Framed as a wondering; the equivalence is assumed rather than argued **Tags**: #xref <!-- atomicity: 1 idea ✓ --> ### 🧠 Permanent Candidates (H4) #### [P1] A knowledge system is judged by its output, not its completeness **Claim** (one sentence, Courtney's voice): Notes that are never used for thinking are a record of consumption rather than a knowledge practice, so the measure is what the system produces downstream. **Evidence pointer**: [[#H1]] **Scope & conditions**: Holds for personal knowledge work; does not apply to archival or reference systems whose purpose is retrieval rather than synthesis. **Links** (≥1 with annotation): - [[Readwise-deep-read]] — because applies-in this pipeline - [[Zettelkasten]] — because extends - [[Productivity-critique]] — because extends <!-- atomicity: 1 idea ✓ --> ### ✏️ Fleeting Queue - [F1] Apply the actionability reframe to the oldest stalled item on my list #fleeting #inbox - [F2] Are these deep-read notes producing thinking, or accumulating? What would the test be? #fleeting #inbox - [F3] Is reading a cited synthesis closer to reading a review article than to skimming? Worth a real answer #fleeting #inbox ### → Things 3 (max 2 actions) - [ ] Pick one long-stalled item and answer 'why am I not taking action' #things <!-- deep-read:sections end --> 📄 **Full text:** [[Full Document Contents/Articles/Reversing Knowledge Gains, Notes on Notes, Actionable Problems.md|Reversing Knowledge Gains, Notes on Notes, Actionable Problems by Curtis McHale]] ## Highlights - Now I wonder what Nichols would say in the face of AI being the de facto search for many. While some may use it to increase their knowledge on a topic, they're still essentially skimming a bunch of different texts aggregated into the AI return text and then calling it a knowledge gain. ([View Highlight](https://read.readwise.io/read/01m2zdem3y41yxy4xxgjnym4y0)) ## New Highlights added 2026-09-21 at 5:45 PM - Now I wonder what Nichols would say in the face of AI being the de facto search for many. While some may use it to increase their knowledge on a topic, they're still essentially skimming a bunch of different texts aggregated into the AI return text and then calling it a knowledge gain. ([View Highlight](https://read.readwise.io/read/01m2zdem3y41yxy4xxgjnym4y0)) ## New Highlights added 2026-09-21 at 6:45 PM - Now I wonder what Nichols would say in the face of AI being the de facto search for many. While some may use it to increase their knowledge on a topic, they're still essentially skimming a bunch of different texts aggregated into the AI return text and then calling it a knowledge gain. ([View Highlight](https://read.readwise.io/read/01m2zdem3y41yxy4xxgjnym4y0)) ## New Highlights added 2026-09-22 at 8:52 PM - Now I wonder what Nichols would say in the face of AI being the de facto search for many. While some may use it to increase their knowledge on a topic, they're still essentially skimming a bunch of different texts aggregated into the AI return text and then calling it a knowledge gain. ([View Highlight](https://read.readwise.io/read/01m2zdem3y41yxy4xxgjnym4y0)) ## New Highlights added 2026-09-25 at 1:00 PM - Now I wonder what Nichols would say in the face of AI being the de facto search for many. While some may use it to increase their knowledge on a topic, they're still essentially skimming a bunch of different texts aggregated into the AI return text and then calling it a knowledge gain. ([View Highlight](https://read.readwise.io/read/01m2zdem3y41yxy4xxgjnym4y0)) ## New Highlights added 2026-09-25 at 1:46 PM - Now I wonder what Nichols would say in the face of AI being the de facto search for many. While some may use it to increase their knowledge on a topic, they're still essentially skimming a bunch of different texts aggregated into the AI return text and then calling it a knowledge gain. ([View Highlight](https://read.readwise.io/read/01m2zdem3y41yxy4xxgjnym4y0)) ## New Highlights added 2026-09-25 at 2:46 PM - Now I wonder what Nichols would say in the face of AI being the de facto search for many. While some may use it to increase their knowledge on a topic, they're still essentially skimming a bunch of different texts aggregated into the AI return text and then calling it a knowledge gain. ([View Highlight](https://read.readwise.io/read/01m2zdem3y41yxy4xxgjnym4y0)) ## New Highlights added 2026-09-25 at 4:30 PM - Now I wonder what Nichols would say in the face of AI being the de facto search for many. While some may use it to increase their knowledge on a topic, they're still essentially skimming a bunch of different texts aggregated into the AI return text and then calling it a knowledge gain. ([View Highlight](https://read.readwise.io/read/01m2zdem3y41yxy4xxgjnym4y0)) ## New Highlights added 2026-09-26 at 5:49 PM - Now I wonder what Nichols would say in the face of AI being the de facto search for many. While some may use it to increase their knowledge on a topic, they're still essentially skimming a bunch of different texts aggregated into the AI return text and then calling it a knowledge gain. ([View Highlight](https://read.readwise.io/read/01m2zdem3y41yxy4xxgjnym4y0))