This is going to be long, but 100% worth your time if you are trying to work with AI on a daily basis. I’m still reeling with dopamine and adrenaline after discovering this system.

It’s worth the read if you are using AI daily.

First, the problem.

When working with AI, you want it to have context and memory. This is the single biggest difference between generic outputs and usable ones that make you go “wow”.

For the past two years, I’ve been building a “second brain” in ChatGPT and thousands of people have taken my class and organized their accounts in the same way. It’s done a decent job.

+Better memory

+More context

+Less explaining over and over

However, there are some increasing concerns that OpenAI is not ethical. There are also issues with the fact that your data is in the cloud. Aside from OpenAI’s trustworthiness, just yesterday Anthropic released a long statement about their newest model Claude Mythos, and that it was so powerful they were not going to release it to the public- ever.

In fact, some of the techy gurus on X are saying we have 12 months before some bad actor builds a model that is as smart as Mythos and our data is all leaked and hacked. Whether or not that’s hype and exaggeration is still to be seen, but the problem remains:

If you build a second brain in a tool that goes obsolete, gets hacked, or does you dirty, then what?

I’ve been thinking about this a lot and last week, I stumbled upon an article on X that went viral because of how this guy Karpathy was explaining his knowledgebase idea for his AI agents.

I understood about 30% of it but I could tell by the response, that there was something to it.

Karpathy was building a secure, safe, second brain, that could be used with any AI agent that comes along, off the cloud, and maintain persistent memory and context with his AI over time. And the best part? He wasn’t having to even keep up with the organization of it.

THAT caught my attention. With my second brain in ChatGPT, there was still a ton of manual organization needed!

It took me about two days to break down all the geek speak and coding lingo to understand just how this private, secure, take-anywhere second brain worked.

But once it clicked, I was flooded with dopamine and adrenaline. I could not believe what I’d made (and what will continue to grow over time).

If you are super techy, you will be able to do this without my help. I’ll explain it at a high level and off you’ll go. If you aren’t super techy, it’ll be MUCH easier to do it with me in a presentation/workshop format.

Two ways to participate in that:

• Become a private client (you’ll get it for free)
• Join FG Society (this is where I’ll host the class ASAP)

Now I’ll explain the system in detail!

The first thing you need to know about is a free tool called Obsidian.

Think of it as Apple Notes, but on steroids. You open it up, you take notes exactly like you always have, same experience, same simplicity. The difference is what’s happening in the background.

Every single note you take in Obsidian is being saved as a markdown file on your computer. Not in the cloud. Not on someone else’s server. On your actual device.

Now I know what you’re thinking. Markdown file? What is that?

It’s just a plain text file. Super simple. Super lightweight. Every computer on the planet can read it. Every AI tool on the planet can read it. And unlike Apple Notes (which traps your notes in Apple’s ecosystem) markdown files go anywhere. They belong to you. Forever.

That’s the foundation. Notes you own, saved as files you can take anywhere.

Now we move onto the folders.

Inside Obsidian, you build a simple folder structure. For each area of your life - work, personal, faith, whatever makes sense for you - you create three subfolders.

1. Raw.
2. Wiki.
3. Output.

That’s it. Three folders. Let me explain what each one does.

Raw —> Your Brain

Raw is your brain dump folder. This is where you put everything. Messy thoughts, half-formed ideas, voice memo transcriptions, things you read that made you think, observations from your day. You dump it all in here exactly the way it comes out of your head.

No organizing. No cleaning it up. No renaming files. Just get it in.

This is what you’re already doing in Apple Notes or Google Keep… you’re just doing it here instead. Same behavior. Different destination.

Wiki —> The AI Brain

Here’s the part that broke my brain.

Once you’ve got stuff in Raw, you point Claude Cowork at that folder (because remember, all these notes have become organized md files because Obsidian is awesome) and say, “Read everything in here and build me a wiki.”

You know how Wikipedia works? You look something up, and certain words are highlighted and clickable, and they take you to related pages, which take you to more related pages? That’s a wiki. It’s not one long document. It’s a bunch of short pages that are all connected to each other. So instead of scrolling through a 47-page Google Doc trying to find something, you’re navigating a web of organized information.

Claude reads all your messy raw notes, finds the patterns and topics, and builds clean organized pages in your Wiki folder. One page per major topic. Each page links to related pages. Everything organized and interconnected.

And here’s the key thing to understand about the wiki:

You don’t build it. You don’t maintain it. And you don’t even really read it.

The wiki is not for you. The wiki is for the AI.

Claude is building an organized version of your thinking so that the next time you ask it a question, it can read the wiki and already know your frameworks, your history, your patterns. It has persistent context across every single session, not because it remembered, but because it read.

You’re not organizing your notes. You’re feeding an AI that organizes them for you, for itself, so it can help you better.

ARE YOU STARTING TO FEEL WHAT I FELT? 😅

Output —> The Merged Brain

The third folder is Output.

This is what gets created when your brain and the AI brain work together.

Let’s say you’re in Claude CoWork and you say, “Hey, read my wiki and help me build five presentation ideas for my mastermind.” Claude reads your wiki, understands your frameworks and your thinking FAST, and produces something genuinely useful.

If you like that output, you ask Claude to make it an md file in your output folder. That output gets saved in your output folder and is now immediately readable on Obsidian.

And here’s where the compounding starts. Over time, those saved outputs get folded back into your wiki. Every answer makes the next answer better. The system gets smarter every single time you use it.

Raw is your brain. Wiki is the AI brain. Output is what you build together.

The Daily Habit

Now you’re probably wondering… okay but how does this actually work day to day? Do I have to think about which folder to put things in?

No. That’s the best part.

Obsidian has a feature called the Daily Note. Every single day when you open Obsidian, it automatically creates a note with today’s date. You just dump everything into that one note throughout the day. Work ideas, personal observations, faith reflections, farm notes… all of it, in one place, no thinking required.

Then you run one prompt. Claude reads your daily note, figures out what belongs where, and files everything automatically into the right raw folder for you.

Your only job is to write things down. The AI does everything else.

Why This Is So Freaking Cool

• You own every single file. Forever. On your own device. No company can take it, change it, or hold it hostage.
• It works with any AI emerging agent like Claude Cowork (or whatever gets built next) Your second brain isn’t married to anyone.
• The AI builds and maintains the wiki for you. You are not organizing anything. Ever.
• The more you feed it, the smarter it gets. This thing compounds. A year from now it will be unrecognizable compared to day one.
• Every AI session starts fully loaded with your context, your frameworks, your history. No more re-explaining yourself from scratch. It can read the wikis easily and fast. Persistent permanent memory and context.
• Your best thinking stops disappearing. Eight years of voice memos, presentations, coaching calls, ideas… all of it can be mined and preserved.
• It’s private. End-to-end encrypted. Off the cloud. Not being scraped, sold, or stored by anyone else.
• If Claude disappears tomorrow you just point your vault at whatever comes next. Your files are plain text. They go anywhere.
• You don’t have to be organized to use it. You just have to write things down. The AI does the rest.

The Two Things You Need

• Obsidian - it’s free. There is one caveat (read below).
• Claude CoWork - $20/m

Obsidian has a $5 a month subscription called Sync. And if you work across multiple devices (laptop, desktop, phone) this is a no brainer. Whatever you update on one device automatically shows up on all the others.

I was working on my MacBook last night and this morning I opened my iMac and everything was just there. Updated. Current. No emailing files to yourself. No Dropbox. No thinking about it. It just works. And it’s end-to-end encrypted, which means even Obsidian can’t read your content. Your second brain stays yours.

That’s it. Two tools. Three folders. A system that compounds forever.

Like I said, I just explained everything here so if you’re techy and I blew your mind, please let me know. 🥹

If you want to go through a step by step workshop, I will explain how Obsidian works, go into more detail about folders and files, show you the prompts I have, show you how mine is working (I just started mine so it’s still small), and walk you through the entire setup. I will also answer any questions I can!

Two ways to participate in that:

• Become a private client (you’ll get it for free)
• Join FG Society (this is where I’ll host the class ASAP)

xx Julie