The first cohort of the Second Brain System has officially started.

If you missed my emails or posts over the last few months, here's the context: this is a peer-building cohort where solo consultants and strategists are building their AI team. We're starting with designing a second brain and building a meeting assistant, then expanding from there. It feels weird to call it a "group program" or "online community." It's more like peer-building, where we're doing the work together, live.

Yesterday's lesson was a milestone. Not just because we're building something new, but because of how we're building it.

At the top of the session, I did a quick survey. I wanted to know what tools people were actually using so I could design the system to work with their reality, not some ideal scenario.

And what I found was really interesting.

Everyone (literally 100% of the cohort) is using Google Drive for file management. Some people also use Dropbox or OneDrive, but Google Drive is the anchor.

Virtual meeting platforms were evenly split between Zoom and Google Meet, and people use both depending on who they're meeting with.

But AI note-takers? That's where we had the most variation. Which makes sense because note-takers are the tool people are most likely to switch.

File storage and meeting platforms are sticky. Once you commit to Google Drive or pick Zoom as your main platform, you're pretty locked in.

But note-takers are nimble. They're also where you can end up in vendor lock-in hell if you're not careful about how you capture those notes and transcripts.

So what we did was set up Google Drive as the central brain for all meeting information. Transcripts, notes, all of it lives there.

The note-taker feeds it, the meeting assistant we're building will read from it, but the foundation stays stable no matter what tools come and go. Your data, your control, your system.

Here's the thing about getting organized: if anyone on your team (a new hire, a vendor, anyone new to the business or not) can look at information and know what it is and have an expectation for what it means, that means AI is going to be able to do that too. That's the whole point of getting organized. When the structure is clear to humans, it's clear to AI. Here's what that looks like in practice.

Now, this cohort of business owners, nobody has time to perfectly name their files and put them in the right place. If we rely on humans to do it, it won't be done consistently. At some point it will just stop getting done.

The nicely-named transcripts you see here are what happens when the meeting assistant has the knowledge on how to name our files the right way./

Here's what surprised everyone on the call: Google Drive files actually have descriptions. And AI natively reads your descriptions.

It's more likely that all of your LLM tools will be able to read the file descriptions on Google Drive files, even if they're not native Google files. If it's a PDF or another file type in your Google Drive, you can put a description on it, then any AI or integration you have can read that description and make sense of what is there.

While this feature has existed for a while, in reality, who's going to be writing these descriptions? Who has time to write file descriptions for all these files?

This is what the assistant helps us do. These descriptions aren't summaries, though they might read like one. They're SEO-style, AI-optimized descriptions that each business owner can customize to fit the words they would use.

It's optimized for searchability. When a human goes to Google Drive to find something, the words in the description matter. If I want to come back to certain types of meetings later, it matters what words are in the description so I can find things. Putting the full phrase of the day ("Friday") in the description instead of the file name means that if I remember what day of the week it was and who I met with, I can find that information.

That's the foundation everyone designed for their meeting assistant yesterday. And yes, everyone did it live on the call.

I'm pretty sure this was the first online program I've been part of where we did a systems build live together. And I'm really happy I'm leading it.

Every single person finished the session with their Second Brain folder structure in place and ready for the automation we're building on Friday. Not "here's homework to do later." We did it right there, with support when needed and questions answered in real-time.

One cohort member who's taken several AI and automation courses shared feedback later: this feels better paced and more substantive. Other courses were overly simplistic with heavy reliance on rigid templates rather than teaching customization or systems thinking. This one balances conceptual theory—patterns, data organization, how information actually works—with hands-on building.

It's about becoming better systems builders, not just template consumers.

The conversation in the chat was really validating as well:

"I just let out a deep MMMMMMM! I get it!" "OMG, no more homework! Clutch." "This is going to be so helpful!"

The feedback I'm getting is that this feels different from other programs. And that's intentional.

I spent a lot of time thinking about how to structure this to get the most out of 90 minutes. It's not about consuming more content. It's about actually building something that works by the time you leave the call.

Oh, and about that note-taker decision. I started recording my thoughts about which tools to recommend. I was going to turn it into a comparison doc for the community.

But then I realized, you know, who wants to sit down and read through all that nuance and all those caveats?

So I ended up building an interactive tool instead. You answer a few questions about your tech stack and meeting needs, and it scores the different note-takers based on how well they'll integrate with what you're building.

One of those "this could have been a post" moments that turned into "this should actually be a tool."

Tuesday was peer-building light. Friday, we're going all the way in. We're building the automation workflow live—everyone in their own tech stack, their own accounts, their own systems. I'll walk through the setup for 20-30 minutes, then cohort members will share their screens and build theirs live with the group.

This is where my hypothesis on peer-building gets its first real test.

Peer-building. No one gets left behind.

I'm calling this peer-building for now. If you've ever demoed anything live, like a presentation, then you know the feeling. So fingers crossed!