- “most people’s maps, including my own, looks like a junk drawer with a search bar” — the state of most people’s digital thinking made viscerally concrete
- “your cognitive architecture is now your competitive advantage — the better that you think, the more value you can bring because of the way that you use ai” — clarity of thinking is the new leverage
- “your file structure is a map of your cognitive architecture” — how you organise externally reveals how you think internally
the creator realised his disorganised file folders were a direct reflection of his disorganised thinking — and that fixing the outside required fixing the inside first. his solution was a three-folder structure (atlas/thinking, projects/building, end products) that maps to how he naturally works, how claude is structured, and even how his school communities are organised. the deeper insight is that ai isn’t just a productivity tool — it’s a mirror that reveals your cognitive patterns, and the bottleneck in the ai age isn’t the tool but the clarity of your thinking.
- your file structure is a mirror of your mind — disorganised folders are disorganised thinking made visible, and you can’t fix one without fixing the other
- three folders is enough — atlas (thinking), projects (building), end products (done) covers everything and forces clarity about what mode you’re actually in
- the bottleneck is thinking, not tools — in the ai age, the core human skill is directing, not executing, which means cognitive clarity is the real competitive advantage
the creator wants to help people realise that the problem with their ai outputs isn’t the ai — it’s the disorganisation of their own thinking — and that a simple three-folder structure can dramatically improve both their ai results and their mental clarity.
- file structure as cognitive architecture — your folders externalise how you think
- three-folder framework: atlas (thinking), projects (building), end products (done)
- head, hands, heart — thinking, building, serving
- the mirror principle — as above so below; fix the inside, the outside follows
- the sophistication trap — smart people overcomplicate what is actually simple
- ai as a mirror, not just a tool — ai reveals how you think, not just amplifies output
- cognitive architecture as competitive advantage in the ai age
- directing vs executing — the core human skill shifts to clarity and direction
- reduce all file folders to three: thinking, building, end products
- before filing anything, ask: is this a thinking doc, a building doc, or a finished product?
- use claude chat for thinking and flushing out ideas
- use claude co-work for active building and project work
- use claude code for finished, deployable end products
- stop building and pause when things feel messy — it’s a signal to reorganise first
- ask claude to audit and reorganise your existing file structure using the three-folder framework
- use ai not just to get outputs but to observe what your outputs reveal about how you think
This video is uncomfortably accurate about where I am outside of obsidian. The vault itself is relatively organised, but my broader file system — Finder, Downloads, Claude projects, content drafts — is exactly the junk drawer Dan describes: “final version 3 (no, this is the real final one).” The three-folder framework (Atlas/Projects/End Products) maps almost perfectly to how the vault is already structured conceptually — private thinking vs active content vs published — but I’ve never named it that cleanly, so things leak between layers. The deeper point — AI reveals how you think, not just amplifies what you do — is the one worth sitting with longest. The scattered claude code prompts I sometimes write are a mirror of scattered thinking, not a tool problem.
- file structure as cognitive architecture — the vault is decent, but Finder/Desktop/Downloads is the actual junk drawer; the outside hasn’t caught up with the inside
- three-folder framework — Atlas (thinking/raw ideas), Projects (actively building), End Products (done/published) maps to private notes vs content drafts vs ryeones.com output
- AI as a mirror — when Claude gives me scattered outputs, the prompt was probably scattered; the tool is reflecting my thinking back at me
- sophistication trap — smart people overcomplicate; I’ve built some vault queries and automations that are complex for the sake of it rather than because they work better
- directing vs executing — the real skill shift; claude code workflows are pushing toward this but I’m still spending too much time executing rather than directing
The second brain discourse always pulls toward more complexity — more plugins, more templates, more structure. This video cuts the other way entirely: three folders, that’s it. What’s interesting is Dan landed here after going deep into agents, which mirrors where I keep getting tempted to go. The “stop building when things feel messy and reorganise first” rule is genuinely useful — I tend to push through the mess and compound it. The fix-the-inside-the-outside-follows principle is the same logic behind the daily note structure and 12 week year: not the system itself, but the clarity of thinking it enforces.
Short, punchy, and more philosophical than the title suggests. The Claude chat/co-work/code parallel is a genuinely clever structural payoff. Would have benefited from a concrete before/after of what the reorganisation actually looked like in practice. Better as a mindset reset than a how-to guide. ★★★☆☆
- does the Atlas/Projects/End Products structure already exist implicitly in the obsidian vault, or is the vault itself just one of the three folders — and the problem is everything outside it?
- where does soffcopy sit right now — is it still an Atlas idea, or has it crossed into Projects? what would the move actually look like?
- if AI reveals how I think, what are the claude code workflows I’ve built actually showing me about my own cognitive patterns?
- which parts of the vault setup are genuinely load-bearing and which are complexity for its own sake — the sophistication trap question?
- “directing vs executing” — in seeksophie content work, am I the director or still mostly the executor? what would actually changing that look like?
- obsidian — the tool this framework most directly applies to; worth auditing folder structure against Atlas/Projects/End Products
- 12 week year — same underlying logic: simplify to what moves the needle, cut everything else
- mindset and identity sculpting — “fix the inside, the outside follows” is the same thread
- soffcopy — where in the three folders does this actually live right now?
- steal like an artist — the “as above, so below” mirror principle echoes the copywork philosophy
- before filing any document, ask: is this thinking, building, or a finished product? — put it in the right folder, not the nearest open one
- when AI outputs feel scattered or wrong, stop and reorganise the prompt/context before retrying — it’s a thinking problem, not a tool problem
- is today’s work thinking, building, or finishing? am I in the right mode?
- audit the obsidian vault against the three-folder pattern — is Atlas/Projects/End Products already implicit, or does structure need clarifying?
- apply the three-folder structure to Finder/Desktop/Downloads — not the vault, the actual chaos outside it
- identify one thing currently in “Atlas” (idea stage) that’s ready to move to Projects — is soffcopy it?
- ask Claude to review my recent chat/co-work/code usage patterns and flag where I’m using the wrong mode for the task