Struggling with AI prompts? I got you!

Struggling with Gen AI prompts?
Sometimes being “detailed & specific” isn’t enough!

“I'm a 10th grade history teacher and our class is reviewing the French Revolution. Create an escape room activity that supports a review of key events of the French Revolution.”

This is what I consider an overload for GenAI. I’m sure there is a more technical term, but you get the picture. What is produced tends to give you “inspiration” for creating an escape room but still a lot of work for you, the teacher. It really isn’t that helpful. It's just too complex of an ask.

Break it down.

My friend Justin Germishuys, explained that when you use something called iterative prompting, you can break down a BIG ASK of the GenAI into smaller, connected steps. Justin does a much better job explaining this. 😅

So instead ask a broader question.

 “What are the steps to creating an escape room activity for the classroom?”

GenAI (I’m using ChatGPT so answers may vary) will produce a step by step guide on how to create an escape room activity.

This is where it gets good. It’s like a “choose your own adventure”.

Now you can be more specific.

“Help me with step 1: write objectives for an escape room for 10th grade history for a review of key events of the French Revolution”

Boom. Objective written. Human in the Loop check, does it look good?

“Help me with step 2, creating a story theme on the French Revolution.”

Boom. Story written. Check it to make sure you like it and copy and paste that sucker into a doc.

“Help me with step 3, create an activity for my escape room.”

Boom. One activity and answer key (thank goodness) produced. Double check if you think the kiddos can solve or if it fits your curriculum.

Repeat. Repeat. Repeat.

Broad —> Narrow.

You got this!

Respond to this email and let me know how you changed your prompt game by thinking BIG! Rooting for you and your students. -M.

Beep. Boop.

Ok, 👋 baiii teacher bestie!

Need more guidance? I can help you, friend! Here’s how:

This includes a discovery call of your exact problem and work together to customize a solution based on your needs. Included in the package is a blueprint for implementation.