Feels Like Magic, But It’s Math

Note 001 — why this exists, and what you’ll get out of it.

Meta
AI
Demystified
Every week someone senior leans in and quietly asks me what AI is actually doing. Cribb Notes is my answer — Cliff’s Notes for the stuff everyone pretends to understand.
Author

Mark Cribb

Published

May 23, 2026

Once a week, somebody senior and genuinely smart leans across a table, drops their voice so the room won’t hear, and asks me the same question:

“Okay… but what is AI actually doing?”

They’re never embarrassed without reason. They run departments. They sign the contracts. And somewhere along the way the technology got wrapped in so much hype, jargon, and vendor theater that asking the basic question started to feel risky.

Here’s the secret I tell every one of them: the question isn’t dumb, and the answer isn’t magic. It’s math — often surprisingly simple math, dressed up in intimidating clothes.

That conversation, on repeat, is why Cribb Notes exists.

What this is

Cribb Notes is Cliff’s Notes, by Cribb — short, plain-language notes on the things everyone in tech pretends to already understand. AI, mostly. But also the broader enterprise: the systems we actually run companies on, the IT decisions that quietly cost millions, the way the movie business and higher ed are getting rewired in real time. And every so often, the human side — what coaching martial arts for thirty years, or chasing a drum fill I still can’t quite land, taught me about how people (and machines) actually learn.

One idea per note. About a three-minute read. No homework, no prerequisites, no “well, actually.”

Why me

Fair question. The internet is not short on AI takes. Most of the people writing them do exactly one of these three things. The reason I think this is worth your inbox is that I’m stubborn enough to do all three:

  • I build it. I ship real software — an enterprise platform and a stack of working systems, a lot of it on weekends, with AI riding shotgun.
  • I govern it. I wrote my company’s enterprise AI governance framework and chair the hard conversations about what we’re actually allowed to point this stuff at.
  • I teach it. I teach cloud and data to thousands of university students a year, and I’m grinding through a doctorate on the side because apparently I hate sleep.

The interesting view is at the intersection — where the hype meets the budget meets the actual code. That’s the seat these notes are written from.

The one promise

I will not waste your time, and I will not make you feel stupid.

Every note has a single job: hand you one clear idea you can use in a meeting on Monday — whether you’re an executive deciding where to place a bet, a student trying to break in, or just a curious human who’s tired of nodding along.

You won’t leave here an engineer. You’ll leave here un-foolable — able to tell the real thing from the sales pitch, and to ask the next sharp question instead of the safe one.

“Feels like magic, but it’s math”

That line is on the front door for a reason. Almost everything that looks like sorcery in this field — a model that writes, a system that predicts, a chatbot that seems to understand — comes apart into steps you can follow, once somebody bothers to show you the gears. And here’s the part nobody mentions: seeing the gears doesn’t ruin the wonder. It upgrades it. Fear turns into judgment. Judgment is the thing this whole moment has been missing.

So that’s the deal. Every week: one idea, gears exposed.

Pull up a chair. And if there’s something out there that’s been quietly confusing you — the thing you keep nodding along to in meetings — tell me. Odds are you’re not the only one, and odds are it makes a better note than whatever I had planned.

Feels like magic. It’s math. Let’s go.

— Mark