
What this covers
This is a seven-minute solo video from JA Westenberg that treats AI-assisted note-taking as a case of overproduction. The speaker starts from the 2007 jam study, in which shoppers offered 24 varieties bought at roughly a tenth the rate of those offered six. The video presents this as a precursor of what happens when AI agents file research briefs, critiques and link maps faster than anyone can read them. It then traces the problem through Christian Tietze's Collector's Fallacy from 2014 and Alvin Toffler's 1970 account of information overload. Toffler is credited with spotting the cognitive bottleneck but is said to have imagined a rising river where we now have a firehose, one people have pointed at themselves.
The middle of the video turns to the psychological cost. A 2007 University of London study is cited, which found that unread email carried a penalty comparable to losing 10 IQ points. The speaker extends this to vaults fed around the clock by agents, where merely knowing the notes exist produces a low-grade anxiety. A run of anecdotes follows: a writer whose weekly connection maps look like constellations but never feed any writing, a designer with 340 unread critiques after two weeks, and a consulting firm whose automated client briefings go unread. The speaker also compares the situation to a CEO with 50 tireless direct reports, and claims that most people's hit rate on saved notes is only 1 to 5 percent.
The prescription is the most contentious part. Better folders, tags and AI retrieval are dismissed as treating the symptom. Borrowing the logic of Taiichi Ohno's Toyota system, in which inventory counts as waste, the speaker proposes asking "Will I read this within 48 hours?" before generating anything. Other advice includes saving only with a reason attached, composting anything untouched for 90 days, and using AI in-session without archiving the results.
The speaker argues that generative AI has inverted the information problem from scarcity to overload, and that the bottleneck has shifted from production to consumption—people now generate vastly more information than they can meaningfully process, creating cognitive burden and system dysfunction rather than productivity gains.
- AI tools enable production of novellas' worth of organized content in minutes, far exceeding human reading capacity
- Collecting information feels like learning but isn't; automated collection without consumption creates cognitive burden masked as productivity
- The solution is deliberate constraint (just-in-time knowledge, not just-in-case), not better organization—the disease is overproduction, not poor filing
This asset isn't compiled yet
You're seeing its claims, ranked. Compile it to build the argument threads, weight them, and check each claim against your library — the full view.
Every note in a vault creates a small claim on attention not because it's being read, but because the user knows it exists; this ambient awareness of unprocessed information creates a persistent low-grade anxiety that has been documented by psychologists for decades.
“The problem is cognitive. Every note in your vault is a small claim on your attention. And it's not because you're reading it, it's because you know it's there. You know there's a research brief on AI regulation you haven't reviewed. You know there's a competitor analysis from 3 weeks ago that might be outdated. You know there are 14 meeting summaries you told yourself you'd synthesize into your quarterly review. This ambient awareness of unprocessed information creates a low-grade anxiety that psychologists have studied for decades.”
Christian Tietze coined the term 'collector's fallacy' in 2014, arguing that collecting information feels like learning but isn't; saving an article, clipping a passage, or filing a PDF creates the illusion of comprehension without actual understanding.
“In 2014, a blogger named Christian Tietze wrote about something he called the collector's fallacy. His argument was simple. Collecting information feels like learning, but it isn't. Saving an article to Instapaper feels like reading it, but it isn't.”
A 2007 study by researchers at the University of London found that the cognitive penalty of having unread emails sitting in an inbox was comparable to losing 10 IQ points; the emails didn't have to be urgent, they just had to exist unprocessed in peripheral awareness.
“In 2007, a team at the University of London found that the cognitive penalty of having unread emails sitting in your inbox was comparable to losing 10 IQ points. The emails didn't have to be urgent, they just had to exist unprocessed in your peripheral awareness.”
In a 2007 University of Minnesota study on choice paralysis, shoppers given 24 jam varieties bought jam at roughly 1/10 the rate of shoppers given six varieties, demonstrating that more choices lead to fewer decisions and more abundance leads to less action.
“In 2007, a team of researchers at the University of Minnesota ran a study on choice paralysis. They gave one group of shoppers six varieties of jam to sample, and another group 24. The group with 24 options bought jam at roughly 1/10 the rate of the group with six.”
Alvin Toffler coined the concept of 'information overload' in Future Shock in 1970, expressing concern that humans have a cognitive bottleneck limiting how much information they can process before resorting to pattern matching instead of comprehension.
“There's a concept in organizational theory called information overload, coined by Alvin Toffler in Future Shock in 1970. Toffler was worried about television, paperback books, the sheer volume of print media flooding post-war households. His concern was that humans have a cognitive bottleneck.”
Toyota's production system, developed by Taiichi Ohno in the 1950s, was built on the core insight that inventory is waste because stored parts cost money to store, track, and insure, and become obsolete while sitting in warehouses.
“Toyota's production system, developed by Taiichi Ohno in the 1950s, was built around a core insight: inventory is waste. Having parts sitting in a warehouse is a liability. It costs money to store, money to track, money to insure, and it becomes obsolete while it sits there.”
The instinctive response to information overload is to organize better (more folders, better tags, smarter search, AI-powered retrieval), but this treats the symptom rather than the disease—the disease is overproduction, and the solution requires less production, not better organization.
“The instinct is to organize better. More folders, better tags, smarter search, AI-powered retrieval so you can query your own vault like a database. And this helps a little, but it's treating the symptom. The disease is overproduction. You don't need a better filing system for 4,000 notes, you need to not have 4,000 notes.”
The fundamental bottleneck in AI-augmented knowledge systems is not generation (which is easy and cheap) but consumption (which remains a human constraint), and this bottleneck is not a technical problem but a human problem.
“The generation is easy. The consumption is the bottleneck. And the bottleneck isn't a technical problem, it's a human problem.”
A critical threshold question before using AI to generate anything should be: 'Will I read this within 48 hours?' If the answer is no, the content should not be generated; if needed later, it can be generated then, since Claude's ability to produce research briefs on Tuesday is identical to its ability on Friday.
“It's worth asking before telling AI to generate anything, 'Will I read this within 48 hours?' If the answer is no, don't generate it. If you need it later, you can generate it later. The information isn't going anywhere. Claude's ability to produce a research brief on Tuesday is identical to its ability to produce one on Friday.”
As generative AI tools get faster and cheaper, generation keeps getting faster and agents keep getting more capable of producing content in multiple formats with deeper connections to existing work, but the constraint was never production—it was always going to be attention, and attention doesn't scale.
“And the tools keep getting faster. The generation keeps getting cheaper. The agents keep getting more capable of producing more things in more formats, with more connections to more of your existing work. And the constraint was never going to be generation. It was always going to be attention. And attention doesn't scale.”
The organizational dysfunction that occurs when companies hire too fast (e.g., 10 to 50 people in a quarter) creates a failure mode where too many documents are produced and not enough are read, causing the organization to drown in its own productivity—a pattern that generative AI has now replicated at the individual level.
“Think about what happens when a company hires too fast. You go from 10 people to 50 in a quarter. Everyone's producing work, memos and specs and strategy docs and Slack threads and Notion pages and Confluence wikis. The output looks great. The org chart fills in. The shared drive balloons. But nobody can find anything.”
In the individual-as-overproductive-organization metaphor, a person is like a CEO of a 50-person org where all 50 people report directly, are tireless, don't require breaks, produce clean copy, and never question whether anything they produce is actually needed—documents keep arriving, folders keep filling, backlinks multiply, but the CEO keeps deferring review and never catches up.
“You're either the CEO of a 50-person org where all 50 people report to you, and they're tireless, and they don't need lunch breaks, they produce clean copy, and they never once stopped to ask, 'Does anyone actually need this?' The documents keep arriving, the folders keep filling, the backlinks keep multiplying, and you keep telling yourself you'll review it all later. But you won't.”
People using integrated AI and note-taking systems report widespread inability to find information in their vaults, with about 80% of notes being effectively useless despite impressive organizational structures, revealing a disconnect between organizational sophistication and actual utility.
“Scroll through any Obsidian forum, any Notion subreddit, any productivity community, and the same confession surfaces over and over. 'I have thousands of notes that I can't find anything.' The vaults are, by any reasonable measure, impressive, but they're also, by any honest measure, useless about 80% of the time.”
Toffler was correct about cognitive bottlenecks but underestimated the scale; while he imagined overload as a rising river, current generative AI creates a firehose of information, and crucially, people have deliberately pointed that firehose at themselves.
“Toffler was right, but he was thinking too small. He imagined overload as a river rising. What we have now is a firehose pointed directly at your face, and you are the one who turned it on.”
For most people auditing their own knowledge vaults or note systems, the hit rate of notes actually used to inform decisions or change thinking is roughly 1 to 5%, with the remaining 95-99% constituting 'dead weight with good metadata'—storage is free but the cognitive burden is not.
“Try running an audit on your own vault or note system. How many have you opened in the past 90 days? And of those, how many did you actually read, not just glance at? And of those, how many changed how you thought about something or directly informed a decision? I think for most people, the hit rate lands somewhere around 1 to 5%. And the rest is dead weight with good metadata.”
When scaled from an email inbox to an entire knowledge management system fed by tireless AI agents 24 hours a day, the ambient cognitive burden becomes severe because the vault never stops filling, creating persistent and escalating anxiety.
“Now scale that from an email inbox to an entire knowledge management system that's being fed by tireless AI agents 24 hours a day. The vault isn't a second inbox, it never stops filling.”
A product designer set up Claude to automatically generate design critiques for every screenshot saved to a folder, but after 2 weeks accumulated 340 unread critiques, creating cognitive burden and anxiety rather than productivity gains.
“A product designer sets up Claude to automatically generate a design critique for every screenshot she saves to a specific folder. Smart, right? Automated feedback. Except after 2 weeks, she has 340 design critiques she hasn't read. And the sight of the folder gives her a stomach ache.”
A writer set up automation to generate weekly connection maps between all notes added to his vault, creating beautiful visualizations that light up like constellations on Obsidian's graph view, but he never uses the notes to actually write anything despite showing the maps to people at parties.
“A writer sets up an automation to generate weekly connection maps between all the notes added to his vault that week. The maps are beautiful. Obsidian's graph view lights up like a constellation. He shows it to people at parties. He never once uses the notes to actually write anything.”
Every document an AI agent produces carries a hidden cost: someone has to read it, or decide not to read it, or feel guilty about not reading it; this cost is invisible at generation but appears later as ambient behind-ness, a sense of owning a library faster than it can be kept up with.
“Every document an AI agent produces carries a hidden cost. Someone has to read it, or decide not to read it, or feel guilty about not reading it. That cost is invisible in the moment of generation. It shows up later as a vague sense of being behind, of having a system that's smarter than you are, of owning a library you can't keep up with.”
By 2025, a critical mass of people achieved a configuration where Claude or ChatGPT were integrated into note-taking systems, calendars, and reading lists, enabling automation of tasks like summarization, filing, and note linking that created the illusion of having a research team and librarian working in parallel.
“Somewhere in 2025, a critical mass of people hit a configuration sweet spot. They had Claude or ChatGPT hooked into their note-taking systems, their calendars, their reading lists. They could say, 'Summarize this paper and file it under research and AI governance,' and it would happen.”
A consulting firm automated client briefing document generation by pulling recent news, financial filings, social media activity, and industry trends into three-page summaries before every call, but consultants never read the briefings because they're already on back-to-back calls with six other unread briefings waiting.
“A consulting firm sets up AI-generated client briefing docs. Before every call, an agent pulls the client's recent news and financial filings and social media activity and industry trends into a three-page summary. The summaries are good, but also never read, because the consultants are already on back-to-back calls and have six other briefings waiting.”
Future you has never been grateful for speculative research briefs generated in advance; the people who have gotten a handle on information overload converge on using AI in-session for immediate tasks, asking questions, using answers, and letting conversations disappear rather than archiving.
“But future you has never once been grateful for speculative research brief. The people who have gotten a handle on this tend to converge on the same set of habits. They use AI in session for the thing they're working on right now. They ask questions, get answers, use the answers, and let the conversation disappear.”
People who manage information overload deliberately save things only with a reason attached, explicitly connecting saved content to current work (e.g., 'This is the research for the blog post I'm writing today', 'This is the financial model I need for Thursday's board meeting').
“If they need to save something, they save it deliberately, with a reason attached. 'This is the research for the blog post I'm writing today. This is the financial model I need for Thursday's board meeting.'”
People who manage information overload successfully delete aggressively, treating their knowledge base like a garden rather than an attic; things that haven't been touched in 90 days get composted, the anxiety of deleting lasts about 20 minutes, and the relief lasts weeks.
“They delete aggressively. They treat their knowledge base like a garden, not an attic. Things that haven't been touched in 90 days get composted. The anxiety of deleting lasts about 20 minutes. The relief lasts weeks.”
The just-in-time principle—produce what you need, when you need it, in the quantity you need it—has proven successful in manufacturing but most people using generative AI are doing the opposite: running a just-in-case knowledge factory that generates content speculatively in case it's relevant, with no guarantee of future utility.
“The just-in-time principle says, 'Produce what you need, when you need it, in the quantity you need it.' Too many people using generative AI are doing the opposite. They're running a just-in-case knowledge factory. Generate the brief just in case you need it. Summarize the article just in case it's relevant.”
The critical difference between pre-AI and AI-era information overload is that the collection phase has been almost entirely automated; users no longer have to do the collecting themselves—they can instruct AI to research topics and save findings, removing the final friction point between intention and accumulated unprocessed information.
“The gap between collecting and knowing has always existed. What's changed is that the collection side has now been almost entirely automated. You don't even have to do the collecting anymore. You can tell an AI to research X and save your findings, and the collecting happens without you lifting a finger.”
When vaults get smaller and deletion becomes routine, hit rates go up, stomach aches go away, and the system becomes useful rather than pathological.
“The vaults get smaller. The hit rates go up. The stomach aches go away.”
Generative AI tools in 2026 can produce text, diagrams, structured data, slide decks, wiki pages, linked knowledge graphs, and formatted reports faster than humans can read them, with a single Claude session capable of outputting 10,000 words of organized, tagged, cross-referenced content in under 5 minutes—roughly equivalent to a short novella every 5 minutes on demand.
“The generative AI tools available in 2026 can produce text and diagrams and structured data and slide decks and wiki pages and linked knowledge graphs and formatted reports faster than any human can read them. A single Claude session can output 10,000 words of organized, tagged, cross-referenced content in under 5 minutes.”