what Tiago Forte got right (and what AI breaks)

by Ray Svitla

Tiago Forte probably did more for the concept of personal knowledge management than anyone since David Allen. and I think the PKM community has been unfair to him — picking apart PARA’s edge cases while ignoring the genuinely important things he contributed.

so before I explain what AI breaks about his framework, let me explain what he got right. because it’s more than most critics acknowledge.

what he got right

1. he named the practice.

before Forte, personal knowledge management was something nerds did in private. he gave it a name — Building a Second Brain — that resonated with millions of people who had never heard of Zettelkasten and never would. naming a practice is how you build a movement. and the movement, whatever its flaws, brought millions of people to the realization that managing your own knowledge is a skill worth developing.

that’s not nothing. that’s a genuine cultural contribution.

2. CODE is a better model than most realize.

Capture, Organize, Distill, Express. this four-step process gets criticized as too simple, but the simplicity is the point. most people never get past step one — they capture endlessly and do nothing with it. CODE says: organizing isn’t enough. distilling isn’t enough. you have to express — create something, produce output, act on what you know.

the Express step is the most underrated idea in the entire BASB framework. it’s the action requirement. knowledge that doesn’t produce expression is knowledge that’s rotting in storage.

this maps directly to the narcissism trap problem: if your PKM system doesn’t push toward output, it becomes a procrastination machine.

3. actionability as organizing principle.

PARA’s real insight isn’t the four categories. it’s the principle that information should be organized by how close it is to action, not by what it’s about. a note about “management” doesn’t go in a “management” folder. it goes in the project where management is relevant right now.

this was genuinely original. every other system organized by topic, by type, by source. Forte organized by proximity to doing something. that’s a philosophical contribution, not just a filing system.

4. progressive summarization.

the idea that you should distill notes in layers — first highlight, then bold the highlights, then extract the essence — is a beautiful cognitive practice. it forces re-engagement with material. each pass filters signal from noise. by the third pass, what remains is genuinely useful.

this survives AI. not because AI can’t summarize — it can, better than you — but because the act of summarizing yourself is how you understand. the process is the product.

what AI changes

now the harder part.

capture is automatic. Forte’s first step — Capture — was born in an era where information was abundant but your ability to save it was limited. you had to choose what to capture. AI makes this choice unnecessary. you can capture everything — every conversation, every article, every meeting — and let AI sort it later.

this doesn’t eliminate the need for capture. it eliminates the need for selective capture, which was most of the cognitive work in step one.

organize is semantic. PARA’s organizational structure — Projects, Areas, Resources, Archives — exists because humans can’t search intuitively across thousands of unstructured notes. AI can. semantic search makes folder structure optional. the elaborate filing system that PARA prescribes is overhead in a world where “find my notes about X from last quarter” just works.

distill is what LLMs do. progressive summarization — reading, highlighting, bolding, extracting — is literally what LLMs are trained to do. ask any model to distill a 5,000-word article to its core insight and it’ll do it in seconds, probably better than your third-pass highlights.

but — and this is important — the human value of distillation isn’t the output. it’s the process. you understand material by summarizing it yourself. outsourcing summarization to AI is faster, but it’s the cognitive crutch version: the output is identical, the learning is absent.

express survives. this is the critical one. AI can capture, organize, and distill. it cannot express for you — not authentically. expression is the human step. the creation of something new from what you know. the synthesis that only happens when a specific person, with specific values and specific experiences, produces something the world hasn’t seen.

AI can help with expression. it can draft, edit, suggest. but the core act — deciding what to express and why it matters — remains human. and this is where Forte was more prescient than he maybe knew: the entire value of a knowledge management system collapses into step four.

the missing step

CODE has four steps. it needs a zeroth step: Identity.

who is doing the capturing? who decides what matters? who is the person whose expression the system is trying to enable?

Forte assumed this was obvious — it’s you. but “you” isn’t a fixed entity. “you” in a creative phase captures differently than “you” in an execution phase. “you” at 25 expresses differently than “you” at 40. and no version of PARA accounts for this.

this is the gap self.md fills. the identity layer sits before CODE. it answers: given who I am right now — my tensions, my current state, my context — what should I capture, how should I organize, what should I distill, and what should I express?

CODE without identity is a universal machine — it works the same for everyone. CODE with identity is a personal machine — it works differently for you than for me, and differently for today-you than for yesterday-you.

the generous verdict

Forte built something that mattered. he brought millions of people to care about their knowledge. he created a framework that was good enough to use and simple enough to adopt. and his emphasis on expression — on output, on action, on not letting knowledge rot — anticipated the narcissism trap years before personal AI made it acute.

what AI breaks is the mechanical parts: filing, organizing, summarizing. what AI preserves is the philosophical parts: actionability, progressive engagement, the imperative to express.

and what AI adds — the thing Forte couldn’t have anticipated because the technology didn’t exist — is the routing layer. the identity-aware intelligence that sits between who you are and how your knowledge serves you.

Forte built the second brain . the next step isn’t a better second brain. it’s a first self that knows how to use it.

→ why your second brain is dead — the storage paradigm’s evolution → every PKM framework is wrong — PARA in context with Zettelkasten, GTD, and more → the routing layer — what sits on top of CODE → identity as protocol — the zeroth step

Ray Svitla stay evolving