I keep everything.
Not in a hoarder way. In a systems way. Three decades in enterprise technology taught me that the most expensive thing in any organization is knowledge that walks out the door. Projects fail not because people don’t know things, but because the things they know aren’t connected to anything.
So earlier this year, I started building what some people call a “second brain.” I use Obsidian — a plain-text note-taking tool that lets you link ideas together the way your actual brain does. Not in folders. In connections. I’d been accumulating AI conversations for over two years across ChatGPT and Gemini — but they were scattered, disconnected, sitting in platform silos doing nothing. Obsidian is where I finally brought them together.
At last count, my vault has over 1,685 imported conversations with AI tools. Strategy documents, research notes, framework drafts, reading highlights, meeting reflections, and the raw material for a book I’ve been writing. Every idea is linked to related ideas. Every insight is connected to the source that sparked it.
It sounds impressive. It wasn’t.
How It Actually Started
The first version was a disaster.
I dumped everything into Obsidian with no structure. Hundreds of notes with no naming convention, no tags, no links. It was a digital junk drawer — the thing I’d spent my career warning clients about, except it was mine.
I couldn’t find anything. The search was useless because I hadn’t been consistent with how I named things. I had three different notes about the same concept with three different titles. I’d write something, forget I’d written it, and write it again.
This went on for months. I almost abandoned the whole thing and went back to scattered Google Docs and bookmarked articles like everyone else.
What Changed
Two things happened.
First, I stopped thinking about my second brain as a place to store information and started thinking about it as a place to think. That sounds like a small shift. It changed everything. Instead of filing notes away, I started asking: what does this connect to? What does this contradict? What question does this answer — or raise?
Second, I started feeding my notes into AI conversations. Not as a parlor trick. As a workflow. I’d take a cluster of linked notes on a topic — say, 15 notes about how enterprise AI governance actually works in practice — and give them to Claude as context. Then I’d have a conversation about what the notes were really saying. What patterns I was missing. What contradictions I hadn’t noticed.
The AI didn’t create the insights. The connected notes did. The AI helped me see what was already there.
The 1,685 Conversations
Here’s the part that surprises people: the most valuable thing in my second brain isn’t what I wrote. It’s the conversations.
Every time I work through a problem with AI — drafting a chapter, designing a framework, debugging a website, thinking through a strategy — that conversation becomes a note. Linked to the project. Linked to the concepts. Linked to what came before and what came after.
Over time, these conversations became a living record of how my thinking evolved. I can trace the exact moment I shifted from thinking about AI governance as compliance to thinking about it as competitive advantage. It’s in a conversation from September 2025, linked to three notes about failed governance implementations I’d seen in my career, connected to a Forbes article about vulnerability in leadership that changed how I think about tone and trust.
That chain of connections — the failed implementations, the article, the AI conversation, the insight — became a chapter in my book. Not because I planned it. Because the connections were already there, waiting for me to see them.
What This Has to Do With You
I’m not suggesting you need 1,685 conversations or an elaborate Obsidian setup. I’m suggesting something simpler:
Start connecting what you already know.
Most professionals have decades of experience locked in their heads. Lessons from projects that worked. Lessons from projects that didn’t. Intuitions about why things go wrong. Instincts about what matters. All of it disconnected. All of it underused.
A second brain — whether it’s Obsidian, Notion, Apple Notes, or a stack of index cards — is just a way of making those connections visible. And once they’re visible, AI can help you do something with them.
The people I worry about aren’t the ones who don’t understand AI. They’re the ones who have 30 years of experience and no system for connecting it. Because AI is about to make information abundant and cheap. The scarce thing — the thing that will matter — is the judgment that comes from connected experience.
Your experience is the asset. A second brain makes it usable. AI makes it powerful.
What I Got Wrong
I’ll be honest about the mistakes, because I’m still making them.
I over-engineered my system. I built elaborate tagging schemes that I abandoned within weeks. I created templates for every type of note and then ignored most of them. I spent more time organizing than thinking — which is the classic trap that every productivity system warns you about and every user falls into anyway.
The system that works for me now is simple: write the note, link it to one or two related notes, and move on. Let the connections accumulate. Don’t force structure. Let it emerge.
I also learned that a second brain without regular review is just a graveyard. The connections only matter if you revisit them. I spend 20 minutes most mornings looking at what I wrote last week, last month, three months ago. It’s the most productive 20 minutes of my day. Not because I find answers, but because I find questions I forgot I was asking.
The Bigger Point
We’re all about to have more AI capability than we know what to do with. Tools are getting cheaper, faster, more powerful every month. The bottleneck isn’t access to AI. It’s having something worth bringing to it.
Everyone’s worried about the bears — the big, dramatic AI failures that make the news. But the mosquitoes are the ones that actually get you. The decision you made based on an AI output you didn’t cross-reference. The pattern you missed because your knowledge was scattered across twelve tools and none of them talked to each other. The institutional wisdom that walked out the door because nobody had a system for keeping it.
A second brain is mosquito repellent. Not because it prevents AI from making mistakes — it doesn’t. But because connected knowledge gives you the context to catch the quiet errors before they compound.
Your experience. Your judgment. Your hard-won knowledge about how things actually work, not how the textbook says they should. That’s what AI needs from you. And a second brain is how you make it available — to yourself and to the AI tools you work with.
I’m 66. I’ve been building systems my whole career. The most important system I’ve ever built is the one that connects what I know to what I’m building next.
Start yours. It doesn’t have to be perfect. Mine certainly isn’t.
This is the second in an occasional series about building with AI — honestly, from someone figuring it out in real time. More at bearsandmosquitoes.ai.