Built Once, Taught Twice: One Prompt Battle, a Corporate Hour of Code and a Youth STEM Camp
In this blog
An 18-slide AI fundamentals platform built for a WWT Hour of Code prompt-a-thon was rebranded, re-leveled to a 9th-grade reading level and run again for a STEM summer camp, without rebuilding a single slide. The competition at the end stayed exactly the same. What changed was who was in the chair.
Most curriculum gets built once and used once.
A team spends weeks designing a workshop for a specific audience, runs it, and when a different audience asks for something similar, the whole thing starts over from a blank page. New slides. New examples. New build. The work was good, and then it evaporated.
Two engagements this year went the other way.
Engagement One: The Hour of Code Prompt Battle
The first was a WWT prompt-a-thon, built in the spirit of Hour of Code. An 18-slide AI fundamentals progression hosted in a Lovable platform, facilitator-led, with participants working live in the tool the whole way, ending in a competition called the Prompt Battle.
Not a lecture with a quiz. A progression that builds toward a contest, where the thing being judged is the quality of what you asked for.
Engagement Two: The same platform, different room
Then a request came in from the Hope Ignites STEM summer camp.
The obvious move would have been to build a second platform. Youth audience, different brand, different reading level: start over.
Instead the existing platform was rebranded and re-leveled. New colors to match the camp. Every slide brought to a 9th-grade reading level, because comprehension was the constraint that mattered, not content coverage. The 18-slide sequence stayed. The facilitator path stayed. The Prompt Battle stayed.
Hours of work, not weeks.
The three pillars
Whichever room it runs in, the whole arc resolves into three ideas, stated plainly on the final slide:
Security: What you put into an AI system matters as much as what you get out of it.
Responsible AI: The tool produces output. The human owns it.
Effectiveness: Input quality drives output quality. Vague in, vague out.
That third pillar is the throughline of every AI education program in this portfolio, and it's the one that survives contact with any audience. A director understands it. A ninth grader understands it. The examples change; the principle does not.
The progression
Eighteen slides, moving in four stages.
What this thing actually is. Capabilities mapped as a central node with branches. A three-level pyramid with a "you are here" marker. A five-rung ladder with a hallucination trap built into it. GPT and LLM broken into stacked acronym rows with a next-word-prediction example. The machine learning loop, and the neural network as an honest black box.
Where humans and machines differ. A 2×2 grid separating human work from AI work. Two warning cards on caution and bias. A red fact-check banner that's impossible to scroll past.
How to actually prompt. A good/not-good checklist in two columns. Vague versus specific, shown before and after. A fill-in-the-blank user-story template. A rough → refined → final loop. And the one that lands hardest with a younger audience: a split panel showing a human reading a request the normal way next to an AI reading it completely literally, using a peanut butter sandwich as the example.
Who is responsible. Slide 17 puts the author first, three review steps after, and closes on a single line: You are accountable for the output.
Then slide 18, "You Are Ready to Prompt," collapses everything into the three pillars and hands off with a banner: You're ready. The Prompt Battle starts now.
Re-leveling is not dumbing down
The camp version didn't cut content. Every one of the 18 slides survived.
What changed was the reading level and the examples. Slide 17 still says you are accountable for the output; that sentence works at any age, and softening it for teenagers would have removed the most important idea in the deck. The peanut butter sandwich example, meanwhile, does more work in a room of ninth graders than it does in a room of directors.
That's the actual skill in re-leveling: knowing which parts are the content and which parts are the packaging.
The visual layer
For the camp version, each slide got an infographic embedded above the text, visual first, so the graphic is what a participant sees before they read anything.
Eighteen infographics is normally a design project. Here it was a prompt system: 18 self-contained NotebookLM prompts, one per slide, each with the style line baked in so any single prompt can be pasted alone. Above them, a Master Style Block: one editable block that retunes the entire set. Decide the background should be light instead of dark, change it in one place, regenerate all eighteen.
The constraints were deliberate: colors matched to the camp brand, but no logos and no brand words inside the graphics themselves, so they stay clean and reusable. Landscape orientation. Minimal text. Each graphic complements the slide copy rather than repeating it. Green and gold reserved for yes/no and caution states, so a participant learns to read the color before they read the words.
The output isn't 18 images. It's a template for generating a consistent visual layer on any slide-based platform, for any audience, in an afternoon.
The reuse argument
Here's what makes these two engagements worth writing about together, and it isn't the infographics.
One build served two completely different audiences. A corporate Hour of Code event and a youth STEM camp ran on the same 18-slide sequence, the same platform, the same competition format. The curriculum didn't get thrown away and rewritten; it got pointed somewhere new.
That changes the economics of every AI literacy program in the portfolio. A platform built once for one client becomes an asset that can be re-leveled for a school district, a nonprofit, a graduate program or an internal cohort. The slides are agnostic about audience. The only things that change are the reading level, the examples and the brand colors.
What's live
The camp platform: https://steam-camp-prompt-battle.lovable.app
An 18-slide AI fundamentals sequence, now proven across two audiences from a single build, with a facilitator path and a Prompt Battle at the end.
Same platform. Same competition. Different room.
That's what reusable curriculum looks like when it's built as a system instead of an event.