Stop Archiving, Start Taking Action

Short Update

Hey everyone, this is Allan. Today’s video will be a little shorter because I have to catch a bus in about twenty minutes, and I need to take a shower first.

Second Brain Issue

I realized that my second brain wasn’t actually helping me connect my ideas. I had linked Cloud Code with the folder where I save my files and use Obsidian to read them, but I was mainly just archiving more content without making better connections between the things I put there. That’s a problem because, for me, a second brain is useful only to the extent that I take action from the content I study.

AI Integration

The idea of using AI with a second brain is actually really good. I see people saying that when you connect Cloud Code with Obsidian, that’s not a true second brain. The strategy of using a second brain existed before AI; it’s a staple in the productivity niche. A second brain is about organizing your files in a way that helps you memorize, save information, and create action based on what you consume.

Technical Setup

Instead of relying only on the default search tools in Cloud Code, I did a bit of over‑engineering for my amount of data, and it turned out to be useful. I set up a local SQLite database and created vector embeddings in the Obsidian folder that my computer can access. Cloud Code, by default, searches for the same word in different files and creates a connection, but that’s limited. Using embeddings in a vectorized database lets me group different files based on ideas, and it’s a lot faster and more precise to search based on concepts rather than just matching words.

Benefits

The vector approach proved better than what I was using before because it can connect and find information that repeats ideas across files. Based on those connections, I transform the ideas into files I call “principles.” Those principles are what really help me move into action.

Takeaway

I’m not using AI to make decisions for me; I’m using it to help organize all the information. As I mentioned in a recent video, we have access to a huge amount of information, and the problem isn’t getting more content, it’s getting the best content for your context and turning it into knowledge and action. If you’re using Obsidian with AI, stop focusing on retaining more data and start extracting action from it, because that’s what remains the human role in the long term. We are the ones who take action.