Building the Science Labs My Classroom Didn’t Have

I try to make my class as interactive as possible. That can be hard in my current setting: eight students, one teacher, and one aide in a self-contained autism classroom. We have a small classroom with no science setup, and classroom budgets are real. There isn’t enough money to buy a hands-on version of everything I want to do.

AI gave me an opportunity to build interactive experiences within those limits.

When I first started using AI, I was using it like everyone else: “Create a lesson plan on…” As I spent more time with it, I realized I could use HTML and other coding resources to make things interactive without actually knowing how to code.

I started doing this last year. This year, I’ve been refining the process into a workflow that gives me something useful to start with quickly. Then I mold it into what I actually want to use in the classroom.

Predict it, then try it

The first lab I used this year was my Species Interactions Lab. It helped students visualize the examples we were discussing and change variables within the model.

First, we made predictions about what would happen. Then we changed something in the virtual lab and watched to see whether what we predicted actually happened. Students could confirm or deny what they were thinking based on what they saw.

That’s the kind of interaction I’m looking for: something students can make a prediction about, try, and discuss.

It also doesn’t always work exactly how I pictured it. In the Species Interactions Lab, some of the interactions weren’t as clear visually as I wanted them to be. I’ve had to refine those visuals so they more closely match what really happens and make it easier for students to understand what they’re seeing.

That happens all the time. Going through a lab with students shows me things I didn’t notice when I tested it myself. Fortunately, I can go right back to the AI and make changes for the next class—or even for the next time the same students see it.

My curriculum is already part of the process

At this point, I don’t have to write a big prompt from scratch every time.

I have my whole yearly curriculum stored in a plan book I created with AI. I ask it for the next week’s lessons and have it create an interactive module based on the topic. It already has the context that I use New York State science standards and our district curriculum, so those are part of the starting point.

Then I refine it.

I’m looking for something functional, mildly entertaining, and connected to the curriculum and science standards. Sometimes the AI has an idea that could work, but I have to mold it into the way I see my students actually using it.

Having something generated quickly is useful. Knowing your students and deciding what will work for them is still a big part of the process.

These are models, and models have trade-offs

Before I use a lab with students, I run through it myself and test it. I know enough about the curriculum to recognize when something is glaringly wrong.

We also talk about the fact that all of these labs are models. Models aren’t perfect. They come with trade-offs based on what we’re trying to do.

A close-enough model can be useful in class as long as students understand that it doesn’t represent real life with 100 percent accuracy. I still need to check that it represents the science well enough for what we’re learning and that the simplifications aren’t getting in the way.

That’s another reason I keep refining them.

Having the option matters

I still create hands-on activities when I can. But I try to make an interactive experience for every lesson, even if we don’t end up using it.

That gives me the option to pivot.

If a student-run model would be the best choice that day, I have one available. If students need another way to visualize a concept, we can pull it up. I have another option instead of just lecturing to students who don’t want to be lectured to.

In a small classroom with a limited budget, having those options matters.

No experience required

You don’t need to know how to code to get started. Access to an AI tool that can generate code is enough to begin experimenting.

Your first version will probably need changes. Mine do. Try it yourself, tell the AI what isn’t working, and keep shaping it into something useful for your students.

That’s what I want other teachers to take away from this: no experience required. Anyone can jump in and start making interactive experiences for their students—or interactive projects for themselves.

You can browse mine at mrsajdak.com/labs.

A starter prompt to try

Help me build an interactive classroom activity about [topic] for [grade or course]. Students should learn [learning goal], aligned with [paste the relevant standards or curriculum requirements].

Create a browser-based activity in a single HTML file with the CSS and JavaScript included. Keep the controls clear and the text manageable. Give students something meaningful to change, a chance to predict what will happen, and observable results they can use to explain their thinking.

My classroom needs are [describe your students, devices, available time, and supports]. Prioritize a functional, scientifically sound model. Explain any simplifications and limitations so I can review them before teaching.

Include instructions for opening the activity and a short checklist of things I should test. Give me a useful first version, and then we’ll refine it.