build-log··2 min read

Orchestrating Agents That Orchestrate Agents

What's Happening On My Screen Right Now

Right now, in this moment, here's what's happening on my screen:

Claude Code is instructing Claude Cowork to navigate a website, gather specific information, and use it to generate a personalized learning site and NotebookLM materials, all tailored to a student's individual learning style.

At the same time, I'm reimagining AI-native workflows for a large enterprise. Five agents are working in parallel to investigate core operating system processes. That plan? It started as a prototype in Replit that I handed to Claude Code to sharpen and audit.


The Scale of It

I'm orchestrating agents that orchestrate other agents. Across multiple platforms. In parallel. On a Wednesday night.

The strange part isn't that it works.

The strange part is that it feels normal. Productive. Addictive.

Then I stop and think about where we were two months ago — and the gap is staggering.


What This Tells Us

We're not waiting for the future of work. We're building it in real time.

And the people who are experimenting right now — messy, imperfect, hands-on experimenting — are the ones who will define what "normal" looks like six months from now.