Post by Alexandre Rimthong

I don't think the biggest opportunity with AI is generating content. I think it’s improving judgement. Most people are focused on giving AI more context. I've been doing the opposite: I use AI to give _me_ better context. I've been building a lightweight "Chief of Staff" system around this idea. At the center of it is my Obsidian vault. It acts as a shared memory layer between me and my AI: I write in it, AI reads from it, and writes back into it. It's where I keep a running record of notes, decisions, and context in a format that's usable by both a human and a machine. From there, different agents plug into that layer: work signals (Slack, meetings, projects), health signals (sleep, training, recovery), even life admin (condo board noise, random overhead) and turn it into things like weekly recaps, meeting prep, coaching prompts, and decision briefs. Not dashboards. Not more inputs. Just what matters, and what should I do about it? The goal isn't to automate everything, it's to automate the boring parts of thinking so you can focus on judgement. This ties into something I've learned over time: You don't rise to the level of your goals. You fall to the level of your systems. (James Clear) Good systems become your floor. When things get messy, they make sure you still show up prepared, make good decisions, and don’t drop important threads. In practice, this means I walk into 1:1s or vendor calls with context already synthesized. Less digging. More deciding. Better coaching. I suspect this is where AI gets really interesting: not replacing thinking, but supporting it. I'm considering turning this into a deeper write-up (Substack?) or short video. Curious if that would be useful. And if you’re experimenting with similar setups, I'd love to compare notes.