How I run my work as a second brain: Obsidian + AI agents

At DigiMarCon in Cleveland I spent more time than I expected showing people what was on my screen: my Obsidian vault. People kept asking, so here is the short version of how I actually work. The longer origin story is the Cleveland lunch piece.

The “second brain” idea is not new — capture what you learn, connect it, reuse it. What is new is that AI agents now do the connecting and reusing for me.

The setup, simply

Capture everything in one place. Articles, meeting notes, half-formed ideas — they all land in one vault instead of scattering across ten apps. The friction of “where does this go?” is what kills most note systems.

Link, don’t file. Instead of burying notes in folders, I connect related ideas to each other. Over time the vault becomes a map of how I think, not a filing cabinet I never reopen.

Let agents do the synthesis. This is the part that changed the work. I point AI agents at my notes to draft summaries, surface connections I had forgotten, and turn raw capture into something usable. I stay the editor; the agent does the busywork.

The mistake to avoid

Most people treat AI like a vending machine — type a prompt, take the answer, move on. The real leverage is building a system the AI can work inside: your notes, your context, your judgment. The model is only as good as the brain you give it to read.

That is also the AI Builder standard: agents with a job, evidence, and an approval boundary. If you want the same coaching path, AI Builder Spotlight is the front door.