Think Forward : Think Forward In Practice

The Human-First AI Workbook: 21 Pages to Get Your School Ready for AI

Andrew Marcinek — Thu, 13 Aug 2026 19:22:29 GMT

Every school leader I’ve talked to this summer has asked me some version of the same question. What do we do about AI before September? And underneath that one there’s usually a second question they don’t ask directly, which is whether they’re about to get this wrong the same way we all got social media wrong. That fear is well earned. I’ve spent the summer answering it one pillar at a time, free, in this newsletter, and today I’m handing you the whole thing as something you can actually sit down and work through with your team. It’s called The Human-First AI Framework Workbook. Twenty-one pages, five pillars, and for each one the argument, the research behind it, the districts that got it right and wrong, and a worksheet your leadership team fills in together.

I built it as a workbook and not another essay because essays don’t produce decisions. This one is meant to leave you with three things in your students. The first is students who engage with AI thoughtfully and on purpose, instead of reflexively or not at all. The second is students who can read an AI policy, see who it protects and who it forgets, and argue for something better, because the rules governing this technology are being written right now by people who won’t live with them as long as your students will. The third is students who know how to push back on AI and understand that it is not an absolute source of truth. If your team works all the way through the book, you’ll walk out with a published statement of belief, a real audit of what your students are already doing, a lesson plan that shows AI both working and failing, a judgment protocol students use in class, and students seated on the committee that writes your policy. That’s your fall, in order.

Fluency-First AI Implementation

Andrew Marcinek — Fri, 17 Jul 2026 13:19:24 GMT

Tuesday’s post ended with a claim: I didn’t become a fluent AI user by being kept away from the tool. I became one through three years of intentional, daily, critical use — and our students deserve the same path, with adults who planned it with intention.

This deep dive is the plan. It runs on the Human-First AI Framework’s two layers: the Implementation Arc your leadership team walks, and the Fluency Progression your faculty and students build. The premise underneath both is the one thing most AI rollouts get backwards — the goal of implementation is not access. The goal is fluency. Access is a login. Fluency is why, how, and at what cost — and it has to be built in your adults before it can be taught to your students.

Your printable companion for this one is the Fluency-First Implementation Guide — the full sequence in TFS format, built to sit on the table at your next leadership meeting.

The Concept Before the Screen

Andrew Marcinek — Tue, 30 Jun 2026 16:06:28 GMT

This week’s anchor post made the argument: Los Angeles capped the clock and called it a screen-time policy, and a clock cannot tell the difference between a child wasting twenty minutes and a child learning something in twenty minutes. The cap measures dosage. It cannot measure understanding, and it teaches no child to use a tool well — or to be well while using it.

This is the other half. If the goal is real digital and AI fluency, here is what building it looks like, and here is the evidence that it works. Three lessons, one for each band, lower elementary through high school. Not a screen in any of them.

Run all three and a monitoring dashboard would log zero minutes of screen time — while your students walk away understanding more about how these tools actually work, and more about how to govern their own use of them, than they’d get from hours of compliant clicking. The screen was never the variable. The teaching is. Everything below is built to prove it and to hand you something you can run on Monday and defend to a board on Tuesday.