The AI second brain
When everyone has the same AI, the differential is your context.
For anyone tired of re-introducing themselves to a chat window every morning.
Try it right here ↓Free, like everything in the Hub. No signup to read it.
Chat is what happens in one conversation. A second brain is the system that remembers how you work: plain files and folders on your own computer that any AI can be pointed at — your context, your projects, your knowledge, compounding over time and owned by you rather than by whichever tool you happen to be renting this year.
The argument starts with a question Kyle asks every room: as these models get more powerful and everyone gets access to them, what is left that is yours? His answer has not changed in a year: the context you can provide. The game becomes building the system that lets AI reach your context in the most effective way possible.
The build can feel backwards — everything has spent a decade moving into the cloud, and this asks you to make files and folders on your own computer. That is precisely the point. Plain files are the one format every AI can read, so the system is tool-agnostic: Claude today, something else next year, and your context walks across untouched.
The working heart of it is an inbox folder you dump everything into — a report you found, a meeting transcript, a page you saved. Kyle's image: it is like sending it to your personal assistant to sort. Except the assistant is real now — a scheduled task sweeps the inbox every afternoon and files what it finds under rules you approved, flagging anything it is unsure about.
We can all get access to the same tools, we can all get access to the same AIs — so what's the differential? It's your context, your data, your frameworks. As the AI advances, you just repackage that in different ways. It's your IP. It's your value add.Kyle, March 2026 — he has put the same question to audiences, in almost the same words, for over a year.
- 000-START-HERE.mdOne file that gives the AI the map: what lives in each folder, what belongs where, and the filing rules. Written for the AI to read first, every time.
- 1-inboxThe dump zone. Anything you come across goes here without ceremony — sorting is the system's job, not yours.
- portfolioWho you are: an about-me, how you communicate, your goals, your current projects. This is the personal context portfolio, and three files is enough to start.
- projects · knowledge · archiveWork that is live, reference that endures, and everything whose time has passed. Numbered folders keep the hierarchy honest.
- The afternoon sweepA scheduled task: every afternoon, inspect the inbox, use START-HERE to decide destinations, move files under the approved rules, flag uncertainty.
- The weekly auditOnce a week, have it audit the whole brain: anything out of place, anything stale? Wrong context is worse than no context — this habit is what keeps it right.
The portfolio is the first folder worth building — these are the files Kyle has a room create, and the prompt that drafts them from what your AI already knows.
- 1 000-START-HERE.md — the map of the system, written for the AI: what is in each folder, and the filing rules
- 2 about-me.md — who you are, your role, your organisation
- 3 communication.md — your voice, your style, and how you want things explained to you
- 4 goals-and-priorities.md — what you are actually trying to achieve
- 5 current-projects.md — what is live right now, kept current
- 6 preferences, frameworks, knowledge — grow it from here, one file at a time
The bootstrap prompt — first draft from what it already knows
Using the memory and chat history available to you, draft my personal context files: who I am, my role, how I communicate, my goals, and my current projects. Write them one file at a time, in plain markdown. Do not invent anything — where you are not sure, ask me instead.
The personal context portfolio inside this came to Kyle via the AI Daily Brief podcast — identity, role, projects, relationships, tools, one page each — and he names the source every time he teaches it.
Next: The self-learning system.
Stop keeping up with AI. Build the system that keeps up for you.
Read The self-learning system →
AI Impact Hub