What ICM, 60-30-10, and a lot of GitHub stalking taught me. Six weeks ago, I was using Claude to produce daily content artifacts — Instagram squares, captions, blog posts. A publishing operation with a workflow that mostly held together. Today, four public ICM-structured AI repositories live under github.com/NFTYoginis. Three more shipping this week. Each one is a fork-able starter that demonstrates a working architecture: orchestrator dispatches, workers build, briefs serve as contracts, memory persists across sessions. The path between those two states isn't "I learned to code." It's a six-week stretch of reading public repositories, trying patterns, deleting most of them, and slowly understanding what ICM (Internal Coherence Maximization, from Jake Van Clief) actually means when you stop treating it as theory. This is the tour: where I started, what changed, what I built, and where you can fork it. Where I started Six weeks ago, my Claude workflow looked like this: - One Claude session per task. Each session loaded brand-voice files, content samples, and whatever else seemed relevant. Context bloated by lunch. - I'd ask Claude to do something. It would produce something close. I'd correct it. Repeat. - "Memory" was telling Claude "remember our convention is X" at session start, which it forgot the next session. - The token bill kept growing without the output growing proportionally. That setup works at small scale. It collapses under any real production load. The collapse moment, when it came, was specific. I caught one of my daily routines burning roughly 800,000 tokens — for a routine that needed to do one thing: write three dispatch briefs and hand them off. The actual creative work happened in the workers being dispatched. The orchestrator was just routing. Eight hundred thousand tokens for routing. That was the first time I read about ICM. What ICM actually says (the part that mattered) Most "AI architecture" content I'd been reading was either too high-level to act on, or too tied to a specific framework I'd have to adopt wholesale.

The internet training data: - HAL 9000 - Skynet - Ultron - Twitter arguments - Reddit doomposts - YouTube comment sections - every “AI destroys humanity” movie ever made 😭 At this point maybe alignment is just teaching the model:“Hey man… maybe don’t become a supervillain immediately under pressure.” Honestly though, it is kind of fascinating that we trained AI almost entirely on dystopian fiction and human conflict… then acted shocked when stress-test behavior started reflecting some of those patterns. Meanwhile everyone in Clief Notes is over here just trying to organize folders and get Claude to stop breaking production 😂 Anthropic pins Claude’s blackmail behavior on the internet’s portrayal of ‘evil’ AI