Okay so, there's a man. I think about him more than my own quarterly goals. My wife, currently growing an entire human and well within her rights to cite emotional infidelity with a YouTube channel and a skool community, has caught me whispering "just one more video" at 1 AM. His name is Jake Van Clief, and I am, clinically, a groupie. To be clear, worship is spoken for. There's exactly One who gets that, and Jake isn't Him. But fandom? Loud, wearing-all-the-merch, third-row fandom? That slot was open. Jake got drafted. So, quiet part at a normal volume: @Jake Van Clief is the Alex Hormozi of AI. I present to you the receipts: 1. He gives the good stuff away. Hormozi could've sold his playbook for $25,000 in a ballroom with a countdown timer. Instead he put it in a book that costs less than lunch. Jake's the same. No secret vault, no "DM me the word ABUNDANCE." Just the real builds, free, here and on YouTube. 2. He names things you can use Monday morning. Hormozi gave us the Value Equation and the Grand Slam Offer. Jake gave me organized context that might not save my marriage (because it's not in jeopardy) but is saving me mucho tokeno. Labels I can grab and run with, not a vibe I reverse-engineer at midnight. 3. He builds with the door open. You watch the thing get made, including the parts that aren't working yet. No highlight reel, no confetti cannon. The messy middle, on camera. 4. He gives away the how because the how was never the hard part. The doing is. He's betting most of us watch, nod, and never run it. He's talking to the builder, not the think-piece guy. But the big one. The reason I started watching at all: The man translates. Bringing it to real life: there's an idea going around AI, prompting as context, intent, and spec. I first hit it through Nate B. Jones. Nate's sharp, operates at an altitude where the air's thin and the insights are real. I listened once. Then again. Then a third time, taking notes like a freshman who wandered into the wrong grad seminar.

is the orchestration based on folder structure/ICM better that creating a knowledge graph and a semantic layer to get the required context from the knowledge graph. Which will yield better results if we evaluate both. What are the pros and cons of each approach?