# What is a real Second Brain ![[images/What-is-a-real-Second-Brain.webp]] **Creator:** Jake Van Clief (@JEVanClief) · **Format:** YouTube Short (vertical) · **Published:** 2026-07-01 · **Length:** 2:39 · [Watch on YouTube](https://www.youtube.com/shorts/aCtEdO4BGlY) > Caption source: YouTube's **auto-generated English captions** (en-orig), downloaded with **yt-dlp** on the first try. No TubeAlfred credits were used. The Short has no chapters, so the outline follows the speech. Most of it is filmed at a whiteboard, with cutaways to markdown files and a graph view on his monitor. Speech ends at about 2:36; the last few seconds are an end card. ## Outline - [0:00](https://www.youtube.com/watch?v=aCtEdO4BGlY&t=0s) Hook: "Most people are using their second brain wrong, and they probably don't even know it." He says he has built second brain systems for multi-million-dollar companies and has over 40,000 people learning his methods, and he sees the same mistake both in companies and among individuals. ![[images/What-is-a-real-Second-Brain-t0001.webp]] *Opening at his desk, with the on-screen caption "How to build AI Second Brains (like the BIG companies do)"* ![[images/What-is-a-real-Second-Brain-t0013.webp]] *Cutaway: structured markdown files on his monitor* - [0:16](https://www.youtube.com/watch?v=aCtEdO4BGlY&t=16s) **What a second brain is:** whenever you talk to your AI, it remembers past work, and you can throw research papers and notes into a memory base you can always reach. ![[images/What-is-a-real-Second-Brain-t0019.webp]] *The whiteboard: an AI model connected to a "Second Brain"* - [0:27](https://www.youtube.com/watch?v=aCtEdO4BGlY&t=27s) Most second brains are based on **graph theory**: a graph is made of vertices and edges (written G = (V, E)), also called **nodes** and **edges**. The dots are nodes and the lines are edges. But what do they mean, why do they connect that way, and what tells the computer to do that? ![[images/What-is-a-real-Second-Brain-t0031.webp]] *Whiteboard: nodes, edges, graph theory, G = (V, E)* ![[images/What-is-a-real-Second-Brain-t0043.webp]] *A knowledge graph on his monitor: dots are nodes, lines are edges* - [0:48](https://www.youtube.com/watch?v=aCtEdO4BGlY&t=48s) His rule of thumb: **nodes are nouns** (a person, a place, a team) and **edges are verbs**, the relationships or actions between nouns (running, getting, consuming). ![[images/What-is-a-real-Second-Brain-t0055.webp]] *Whiteboard: Nodes = Nouns, Edges = Verbs* - [1:01](https://www.youtube.com/watch?v=aCtEdO4BGlY&t=61s) Company example: each team is a node and each relationship between teams is an edge. The product team node *talks to* (edge) the data team node. ![[images/What-is-a-real-Second-Brain-t0067.webp]] *Drawing node-edge-node links between teams* - [1:17](https://www.youtube.com/watch?v=aCtEdO4BGlY&t=77s) Teams have **workflows**, for example collecting questions from a Slack channel. Sentence form: "the product team *runs* the Slack questions workflow, which *consumes* the ticket information and *produces* answers." ![[images/What-is-a-real-Second-Brain-t0085.webp]] *Whiteboard: "Product team runs Slack Questions, which consumes Tickets, produces Answers"* - [1:36](https://www.youtube.com/watch?v=aCtEdO4BGlY&t=96s) Each of those nouns and verbs can map to a set of files. **The common mistake:** most second brains map the whole sentence at once to a pile of random files. Good second brains break it down piece by piece. - [1:49](https://www.youtube.com/watch?v=aCtEdO4BGlY&t=109s) In practice, a product team folder holds markdown files that describe its processes, ideas, jobs and concepts, plus descriptions of where that data, those processes and ideas connect to everything else. ![[images/What-is-a-real-Second-Brain-t0127.webp]] *The resulting graph of nodes and edges on his monitor* - [2:01](https://www.youtube.com/watch?v=aCtEdO4BGlY&t=121s) Repeat this enough times and you get a large graph that describes how the company works, who its people are and what they're doing. It acts as a second brain, a memory the AI can navigate. - [2:17](https://www.youtube.com/watch?v=aCtEdO4BGlY&t=137s) Call to action: hours of free lectures on structuring folders, files and data, and "40,000 people" in his community. The link is in his bio (the Clief Notes Skool community). "Happy learning." ![[images/What-is-a-real-Second-Brain-t0145.webp]] *His Skool community's classroom ("Getting Started")* ## Key takeaways - A second brain is a **memory layer** an AI can navigate, not just a pile of notes. - Model your organization as a graph: **nodes = nouns** (people, teams, places, things) and **edges = verbs** (talks to, runs, consumes, produces). - **Decompose** sentences like "team runs workflow, which consumes X and produces Y" into separate files per noun and per relationship, rather than dumping the whole sentence into random files. - Per-team folders of markdown describe processes, ideas, jobs and concepts, plus explicit notes on how each connects elsewhere. Enough of these form a graph of the whole company. ## Things mentioned ### Concepts - **Second brain** (AI memory base / context layer) - **Graph theory**: graphs G = (V, E) of vertices (nodes) and edges; knowledge graphs - **Nodes as nouns, edges as verbs**; workflows (consumes / produces) - **Markdown files** organized in per-team folders ### Tools, products and examples - **Slack**: the example workflow collects questions from a Slack channel and answers tickets - **Skool**: his community, [Clief Notes](https://www.skool.com/cliefnotes/about) (linked in the description) - Related, not named in the Short: his **Interpretable Context Methodology (ICM)**, "folder structure as agent architecture" ([GitHub: RinDig/Interpreted-Context-Methdology](https://github.com/RinDig/Interpreted-Context-Methdology)); see also his video [ICM Explained (Folders Over Agents) in Five Minutes](https://www.youtube.com/watch?v=dZ8P7zevuGw) ## Biographies ### Jake Van Clief Creator of the YouTube channel **Jake Van Clief** (@JEVanClief, about 67.4K subscribers when this note was made). The channel tagline is "In a world where everyone is teaching you left and right, I teach up and down." He runs **Clief Notes**, a Skool community that teaches AI workflows, prompting, and organizing work into reusable markdown files for tools like Claude Code and Codex. It has Premium and VIP tiers, recorded "Tea Mastermind" sessions, and a Clief Notes AI assistant launched in August 2026. He created the **Interpretable Context Methodology (ICM)**, which replaces multi-agent orchestration with numbered folders and markdown context read by a single AI agent. In this Short he says he has built second brain systems for multi-million-dollar companies and has over 40,000 people learning his methods; those figures are his own claims. (Sources: the video; channel description; skool.com/cliefnotes; ICM README on GitHub.)