Teachers Can’t Teach What They Don’t Use
AI fluency starts with real life, not another training module
We all agree AI literacy matters. Our students are walking into an AI-saturated world, and they’re going to need to know how to use these tools well. So naturally, we’ve decided teachers should be the ones to teach them.
There’s just one problem: you can’t teach what you don’t actually use.
If we want teachers to raise a generation of confident, capable AI users, we need teachers who are already confident, capable AI users. Not teachers who attended one PD session on ChatGPT. Not teachers who can define “prompt engineering” but have never actually needed a prompt for anything. Real, lived, everyday fluency.
The Classroom-First Approach Isn’t Wrong. It’s Just Not Enough.
The obvious on-ramp for teachers is classroom use: lesson plans, letters of recommendation, scope and sequences, differentiation. And I want to be clear: those are genuinely great uses of AI. I’m not minimizing them. They save real time and they’re a fine place to start.
But they’re low-stakes, low-emotion tasks. They’re work. And work is exactly the kind of thing we’re all a little tired of by 3 p.m. Using AI to bang out a differentiated worksheet doesn’t make a teacher fall in love with AI. It just makes one more task go faster.
If we want teachers to get genuinely excited, the kind of excited that leads to real fluency, AI has to prove itself somewhere that isn’t work first. It has to help with their life.
Once it does that, something shifts. They stop thinking of it as a tool for a task and start thinking of it as a tool for thinking. That’s when it naturally starts showing up in their classroom, in their planning, and in how they talk to students about it, because now they’re speaking from experience, not from a training module.
AI has to prove itself somewhere that isn’t work first. It has to help with their life.
What This Actually Looks Like
A few examples from my own house:
My wife wanted to redesign our living room. She took a photo of the room as it was and asked AI for ideas: a new look, a shopping list, even a generated image of what it could become. She went back and forth a few times, refining the vision until it clicked. It gave her a rug, curtains, plants, pillows, a full color scheme, and a shopping list she could actually take to Target.
We didn’t buy every single item it suggested, but we bought into the concept, and the room got a genuine upgrade. One of my son’s friends walked in afterward and said, “Wow, you guys live rich.” We don’t have rich taste. We had a free interior designer who happened to run on a language model.
She’s done the same thing at the gym. Walking into a room full of machines can be overwhelming if you don’t know what half of them do. So she took pictures of the equipment, asked for a workout plan based on what was actually available, and answered a few follow-up questions about her goals. She walked away with a personalized plan, basically a personal trainer without the personal training price tag.
It doesn’t stop there. We’ve used it for weekly shopping lists and meal plans. I’ve pointed my phone camera out an airplane window and asked what I was looking at down below. My kids have had it generate custom, adjustable learning pages for themselves.
And when we couldn’t find a scorekeeping app for a card game called Hand and Foot, a game with quirky house rules passed down from family friends, not something you’ll find in any app store, I just had AI build one. It works beautifully, it’s got some fun extra features, and honestly, it looks polished enough that I could probably list it in an app store if I ever wanted to.
None of that is work. None of it required a PD day. It’s just people solving real, everyday problems and, in the process, quietly becoming fluent.
AI Is Not Just a Search Box
Part of what makes these examples click is that none of them are the thing most people still picture when they hear “AI”: type a question into a box, get an answer back.
- Snap a photo of a room, a gym, or a view out a window, then ask it to see what’s there and respond.
- Have it generate an actual image so you can visualize an idea instead of just reading a description.
- Have it build a small custom tool, like a scoring app, tracker, or generator, for a problem so specific that no app on the market solves it.
- Let it ask you clarifying questions, back and forth, until it understands what you actually want.
That’s a very different mental model than “search engine with better grammar.”
Life First, Then Work
The order matters here. AI for AI’s sake doesn’t stick. Another mandated tool doesn’t stick. But AI that redesigns your living room, plans your workout, or finally solves the one weird problem only your family has? That sticks. That’s the version of AI people actually go looking for again tomorrow.
Let teachers find that version first, in their own lives. The classroom uses aren’t going anywhere. But they’ll come easier, and go deeper, once a teacher already knows, firsthand, what this technology can actually do.