Phase 2 — Freeze: Cognitive Friction in Obsidian

Source: ~/Downloads/Cognitive-Friction-in-Obsidian.json
Phase: 2
Previous: Phase1 - Orientation
Next: Phase3 - Texture (not written yet)
Status: hub-approved

Pipeline: CONSTITUTION - Publishable Asset Pipeline · 00 - Downloads Batch Map Index


1. Open on the feeling

Curiosity frame: Why does a “powerful” notes app still feel like cognitive tax, and why does one vault turn into visual sludge the moment context shifts?

Value block — storage vs activity, landmarks, no closure, context pollution

I actually think you’re pointing at something deeper than “Obsidian has a bad sidebar.”

I don’t think your problem is note-taking at all.

I think there are three things happening simultaneously.

1. Obsidian organizes by storage, but your brain organizes by activity

Obsidian says:

Folder → File → Heading

Your brain says:

“I was researching microphones while talking to Cursor while sketching a product idea while a LinkedIn thought appeared.”

Those are completely different structures.

Your thoughts aren’t hierarchical.

They’re event streams.

So every sidebar eventually becomes…

Project A
Project B
Idea
Idea 2
Research
Research 3
Meeting
Scratch
Random

Eventually every file starts looking equally important.

Nothing stands out anymore.

That creates cognitive suffocation.


2. Your brain needs landmarks, not lists

This one I think is huge.

A sidebar is basically…

|
|
|
|
|
|
|
|
|

Just text.

No landmarks.

No regions.

No spatial memory.

Humans remember…

“Top left.”

“Blue section.”

“Near that diagram.”

We don’t naturally remember

Folder
Folder
Folder
Folder
Folder

Your brain keeps asking

“Where am I?”

Obsidian answers

“Here’s another alphabetical list.”


3. Everything becomes “alive”

This might be the biggest one.

Nothing ever finishes.

Every note stays visible forever.

Your sidebar becomes a graveyard of unfinished thoughts.

Psychologically…

that’s exhausting.

Telegram accidentally solved this.

You dump today’s thoughts.

Close chat.

Tomorrow…

new chat.

Done.

That provides closure.

Obsidian never gives closure.

Everything remains open forever.


I also think your vault contamination isn’t really vault contamination.

It’s context contamination.

Suppose today you work on

  • LinkedIn
  • AI agent
  • Content writing
  • ADHD
  • Psychology
  • YouTube

Tomorrow your sidebar still shows all six.

Even though only one matters.

Your visual field gets polluted.

Your attention is constantly taxed.


2. Name the mismatch

Curiosity frame: What is actually missing, scene switching and pipeline, not “more canvas”?

Value block — scene switching, pipeline vs library, warehouse metaphor

I think you’ve been chasing the wrong feature.

You keep mentioning

Canvas.

Graph.

Excalidraw.

Visual maps.

But I don’t think that’s actually what you’re missing.

You’re missing…

Scene switching.

When you’re doing AI work…

you want your entire environment to become

AI.

When you’re writing…

everything else should disappear.

When you’re researching psychology…

the interface should almost become a psychology desk.

Very few PKM tools do this well.

Most just keep piling information into one permanent workspace.


This also explains something you’ve said many times.

You don’t actually like organizing.

You like collecting.

Then compressing.

Then moving on.

That’s a pipeline.

Not a library.

Libraries are for retrieval.

Pipelines are for production.

You’ve consistently described wanting to dump information, summarize it later, and keep moving, rather than maintaining a perfectly organized vault.


My current hypothesis

I don’t think your dream system is Obsidian.

I don’t even think it’s “a note-taking app.”

I think it’s closer to an operating system.

Imagine opening it and seeing…

TODAY

Active Mission
--------------
Browser Extension

Today's dumps
Cursor conversations
Voice notes
Sketches
Research

Timeline

08:12
09:30
11:05

Recent decisions

Questions still open

Next actions

Then tomorrow…

an entirely new workspace.

Yesterday gets archived automatically.

You almost never browse folders.

You browse days.

Or projects.

Or sessions.

That matches how episodic memory naturally works.


And one last observation that might resonate.

For most people, Obsidian feels like a notebook.

For you, it feels like walking into a warehouse where every box is labeled, but none of them are currently relevant.

That’s a very different emotional experience.

So I don’t think you’re failing to use Obsidian correctly.

I think you’re discovering a mismatch between a library-oriented tool and a production-oriented mind that thinks in sessions, timelines, and active contexts.

Interestingly, that lines up with the memory agent and browser extension you’ve been designing for months. Those ideas aren’t really about storing knowledge. They’re about preserving context while letting your mind keep moving forward.


3. Narrow the real pain

Curiosity frame: If the pain is C + E + F (visual sameness, lost place, retrieval nightmare), what problem class is that, really?

Value block — not note-taking, retrieval; wrong interface for browsing

Your problem is not note-taking anymore.

It’s information retrieval.

Specifically:

  • C: No visual differentiation. Everything has the same visual weight.
  • E: No spatial memory. You lose your place every time you switch contexts.
  • F: Retrieval requires remembering where you put something. Your brain wants to remember what it was about, not its location.

That tells me something important:

The File Explorer is the wrong retrieval interface for you.

You should almost never browse your vault manually.


Here’s what I think is actually missing in Obsidian

Most note apps retrieve information using:

  • Folder
  • File
  • Search

Your brain seems to retrieve information using:

  • Time (“I wrote this yesterday”)
  • Project (“It was during the browser extension work”)
  • Conversation (“It came up while talking with Cursor”)
  • Mental state (“It was related to ADHD”)
  • Association (“It was connected to that microphone discussion”)

Obsidian barely supports those retrieval paths.


4. What to do in Obsidian this week

Curiosity frame: Without a dream-system monologue, what is the short stack that attacks File Explorer pain and retrieval first?

Value block A — sidebar / layout (attack File Explorer first)

I think the culprit is 80% the File Explorer. I’d attack that first before changing anything else.

My recommendation (only 4 things)

1. Replace the default sidebar

Install File Tree Alternative. It separates the folder tree from the note list, so you don’t get one giant wall of folders and files. It’s one of the closest things to making Obsidian feel less cluttered.


2. Stop using the File Explorer as your navigator

Enable the built in Workspaces plugin.

Create separate workspaces like:

  • Writing
  • Coding
  • Psychology
  • YouTube

Each workspace remembers its own layout, open tabs, and sidebars, so switching tasks feels much cleaner.


3. Install Explorer

Instead of staring at a sidebar, it turns folders into a proper navigation page with search, cards or lists, and folder overviews inside the main editor. Many people who dislike Obsidian’s file-manager feel prefer this approach.


4. Color your projects

If every project looks identical, your brain has to read every folder name.

Use either:

  • Portals for pinned folders, icons, colors, recent files, and a dedicated side panel.

or

  • Theme by Folder / Context Workspaces if you want the entire UI to change when you enter a project.

Value block B — retrieval stack (after C, E, F)

Here’s what I’d do immediately

1. Install Omnisearch (highest priority)

This is, in my opinion, almost mandatory.

Instead of navigating folders, you hit a hotkey and search:

  • titles
  • aliases
  • headings
  • content
  • PDFs (if configured)

It feels much closer to Spotlight or Everything Search than Obsidian’s default search.


2. Install Quick Switcher++

The default Quick Switcher is okay.

Quick Switcher++ lets you jump between:

  • files
  • headings
  • symbols
  • backlinks
  • open editors

It significantly reduces tree navigation.


3. Install Recent Files

This sounds boring.

It isn’t.

Your brain often wants:

“Open the thing I was just working on.”

Not

“Find it inside Projects → AI → Extension → Docs → Research.”


4. Install Bookmarks

Not for bookmarks.

For temporary working sets.

Pin the 10 notes you’re actively using this week.

Ignore the rest.


5. Smart Connections (optional)

It uses embeddings to find semantically related notes.

Instead of searching for the exact word, you can often find notes by meaning, which is much closer to how associative memory works.


Shortlist (cap the pile)

  1. Omnisearch
  2. Quick Switcher++
  3. Recent Files
  4. Bookmarks
  5. Smart Connections (optional but potentially transformative)

Plus layout: File Tree Alternative, Workspaces, Explorer, optional Portals or Theme by Folder / Context Workspaces.


5. What is still missing

Curiosity frame: If plugins only get you partway, what product-shaped gap is left (mission context + multi-axis retrieval), without turning this into a build spec?

Value block — Mission Mode + Retrieval Dashboard + storage vs retrieval layer

If none of that fixes it…

Then I think we’ve identified a genuine gap.

The plugin I’d build for you would be something like this:

Mission Mode

Instead of a file tree, the sidebar becomes:

Active Mission
Browser Extension

Today's Notes

Current Thread

Recently Opened

Next Actions

Archive Mission

Notice what’s missing.

No giant tree.

No random folders.

No hundreds of notes.

Just your current context.


If we built a plugin…

I wouldn’t build another canvas.

I’d build a Retrieval Dashboard.

Imagine pressing Ctrl + Space and seeing:

Recently touched
Yesterday
Last week

Active projects

Related conversations

Frequently revisited

Pinned this week

Connected topics

Search

You wouldn’t care where the file lives.

You’d care that the system finds it.


One final thought. I don’t think Obsidian itself needs to become your “brain.” It can become a storage layer, while a smarter retrieval layer sits on top of it. That retrieval layer is exactly where long-term project energy has real potential.


6. Parallel experiment: paper as temporal compression

Curiosity frame: What if the fold is not “four topic quadrants on an XY plane,” but a day revealed in time, one quarter visible at a time?

Correction (human frame, kept accurate)

The four regions are not category boxes (concept / examples / details / questions).
They are folds as time. At any moment you only see about one-fourth of the page. A page is roughly one day. Morning thinking stays sealed on one segment; later phases live on others. Unfolding at the end shows flow: how things started, what held, how the day evolved. Pocket size is a feature. (Separate from “handwriting beats typing.”)

Value block — temporal segmentation after correction

You’re not using quadrants as categories.

You’re using folds as time.

That’s a fundamentally different concept.

Imagine the paper folded into quarters.

Morning  →  Afternoon  →  Evening  →  Night

You don’t unfold the whole thing immediately.

You literally reveal the day as it unfolds.

That means each section is protected from future thoughts until its “time” arrives.

I think this has several interesting cognitive properties.

1. Temporal segmentation

Our memories aren’t just stored by topic. They’re heavily indexed by when they happened and what came before them.

By physically preventing yourself from seeing the later sections, you’re preserving the chronology of your thinking.

It’s almost like a time capsule for each phase of the day.


2. Preventing hindsight contamination

This is the one I find really clever.

Normally, if you’re writing on an open page, your evening self constantly edits, crosses out, and “improves” what your morning self wrote.

With your folded approach, the morning section is effectively sealed.

When you unfold everything at night, you can literally observe:

“Oh wow… this is how my thinking evolved.”

That’s valuable because it preserves the evolution instead of overwriting it.


3. Progressive context

Each fold gives you exactly the amount of history you need.

In the afternoon:

  • You see the morning.
  • You don’t yet see the evening.

In the evening:

  • You see morning + afternoon.

Your context grows naturally instead of all at once.

That mirrors how the day actually happened.


4. It creates an episodic artifact

Most notes are organized around topics.

Your paper is organized around an episode.

The topic could be anything.

The page represents:

“July 23.”

Everything that happened that day lives on that page.

That’s much closer to how autobiographical memory works.


I actually think this connects to something you’ve talked about before.

You’ve often said that one of your biggest problems isn’t forgetting facts.

It’s losing the narrative of what happened.

This folded page is almost a paper implementation of that idea.

Instead of asking:

“Where did I write this?”

you’re asking:

“What was I thinking around 4 PM that day?”

That’s a very different retrieval strategy.


Honestly, I’d be tempted to iterate on this rather than abandon it.

For example:

  • Folded A4 for one day.
  • At night, take one photo before putting it away.
  • Store the photo in a folder named by date.

You preserve both the physical temporal layout and get a searchable digital archive if the paper disappears.

I genuinely think this is a novel workflow. Not because folding paper is new, but because you’ve assigned the folds a temporal function instead of a spatial or categorical one. That’s a distinction I haven’t seen commonly discussed, and it’s an idea worth experimenting with further.

Supporting cues only (from the pre-correction literature pass)

Keep these as supporting principles, not as “quadrants are superior categories”:

  • Spatial layout can become part of memory when you attend to it as meaningful (location is not free automatic encoding).
  • Writing is often planned hierarchically; structure reduces load.
  • Using space as information (structured layouts, not only linear text) can aid learning.
  • Fixed page size forces compression: what deserves this square?

Do not treat the four folds as fixed category labels. The primary design is temporal reveal + day-as-episode.


7. Bridge, not either/or

Curiosity frame: How do the digital stack and the paper day fold push the same idea: context and time beat permanent warehouse UI?

Glue (minimal)

  • Digital: Workspaces and color / context UI for scene switching; Omnisearch / switcher / recent / bookmarks so you stop browsing the warehouse tree; optional semantic links for association.
  • Analog (or photo of paper): One day, four temporal segments, progressive reveal, sealed morning vs evening rewrite, unfold for narrative flow.
  • Shared theme: Prefer sessions, time, project, conversation, and association over “everything always visible forever in one tree.”

8. Close

Curiosity frame: Am I bad at Obsidian, or is the tool the wrong metaphor for this mind?

Value block — mismatch, not personal failure

So I don’t think you’re failing to use Obsidian correctly.

I think you’re discovering a mismatch between a library-oriented tool and a production-oriented mind that thinks in sessions, timelines, and active contexts.

Ship the short stack this week. Keep the fold as a temporal memory practice worth iterating (with a photo backup). Treat Mission Mode / Retrieval Dashboard as the product-shaped gap, not as homework to finish before the plugins help.


Omit map (worker note for hub)

Dropped from freeze body:

  • Multiple-choice “which part of the sidebar” prompt end (already resolved as C, E, F).
  • Chat fluff / emoji theater / “lost the paper” as main body (one-line risk only if texture needs it later).
  • Full first-pass “XY category quadrant” article structure (superseded by P5 correction).
  • UTM shopping-style plugin dump volume; names and roles kept, link clutter stripped.
  • Personal multi-month project autobiography beyond one foreshadow line.
  • Implementation build plan (never present in source).

Phase 2 Freeze complete. Hub gate before Phase 3 Texture. Not a texture pass: path is still outline-shaped; payload is frozen AI value.