That’s how I feel whenever I pop my head into Twitter/X. While it’s important to have some general awareness on the big changes, trying to follow every AI update keeps you in an unwinnable rat race.
Here’s the mindset shift that will truly change how you think about AI…
Don’t just learn about AI.
Teach AI to learn about YOU.
That’s the shift that most people are missing.
Here’s why.
How have you been using AI? Do you open a chat window and type a question?
That’s fine for "Who made Sting and how did Bilbo end up with it?" But it falls apart for the personal questions: “What should I focus on this week?” or “How does this idea connect to what I was working on last month?” or “What am I missing?”
Generic AI gives generic answers because you haven’t taught it to know you.
When you teach AI about you—your priorities, your efforts, your old notes, your new ideas—it truly does change the entire dynamic. You don't have to re-explain yourself every time. You’re no longer starting from scratch every conversation.
The result is that every conversation you have with AI improves significantly.
I’ll give you a quick example. I have a file called me.md. It started small. Now it’s around 2,000 words. It tells AI who I am, how I think, and what I care about. I’m no longer talking to a complete stranger. Now I’m talking to a colleague.
And me.md is just the starting point.
Over the past year, I’ve built an AI OS—an operating system for working with AI. It has three layers, and I think of them as three concentric circles:
AIOS is a form of Stewart Brand's concept of Pace Layers.
The core: Your Ideaverse. Your ecosystem of linked notes. It’s the timeless, most valuable part of your personal knowledge. It will always exist, with or without AI.
The middle layer: Maps and Manuals. A small set of files—me.md, a vault map, a skill map, a few others—that teach AI how to navigate your thinking and work for you. These files are the heart of an AIOS. They are the translation layer between your notes and any AI tool.
The outer layer: Tools. Claude, ChatGPT, whatever comes next. These tools are the disposable part. You can swap them out at any time.
Your thinking isn’t swappable, but the tools always are.
When you build an AIOS, you get to decide how much you share.
Maybe you’re fine with sharing everything. Maybe you want granular control over what AI sees. And maybe you want AI to read files, but don’t want it to write in your files. There are ways to control what AI can access. In this way, you build an AI thinking partner on your terms.
On Thursday, I’m going to show you what all of this looks like in practice.
In 60 minutes, I’ll walk you through the principles, the frameworks, and a live demo of an AI thinking partner that knows me—my notes, my ideas, my voice. You’ll see me.md in action, the maps and manuals layer, and how IDI keeps me in the driver's seat.
If you’ve been wondering what it looks like to use AI on your terms, this is the session.
Talk soon,
Nick
P.S... In the next email, I'll show you the different parts of an AI thinking partner: the Producer AI, the Creative AI, the Inner Guide AI, and the Synthesizer AI.
P.P.S... Want to revisit past emails in this AI Primer series? Here is the current list: