Prompt Engineering
Hvatahönnun
Anyone can chat with AI. They can direct it.
What they'll learn
- Prompts that work Use structured prompting to get reliable, repeatable results from AI.
- Verify, don't trust Check AI output for mistakes instead of taking it at face value.
- Use AI honestly Apply AI to schoolwork and projects in a way that's honest and allowed.
- Reusable workflows Build prompts and simple workflows they can reuse again and again.
A personal 'prompt playbook' and an AI-assisted project with its process documented.
The 13-week journey
-
Sprint 1 — How models read prompts
Tokens, context and why phrasing changes everything; the anatomy of a strong prompt (role, task, constraints, examples). Ends with a weak-to-strong prompt rewrite set.
-
Sprint 2 — Reliable results
Few-shot examples, step-by-step prompting, structured output; verifying claims against sources; catching and fixing hallucinations. Ends with a tested prompt for a real task.
-
Sprint 3 — Honest power-use
AI as tutor vs ghostwriter; disclosure habits; building a reusable prompt playbook and a simple multi-step workflow for a project they care about.
-
Week 13 — Showcase
Live before/after demos: the same task with a naive prompt and an engineered one, with results compared.
What we cover
Every topic, unit by unit — so you know exactly what your child builds and learns.
How models read prompts
- Tokens, context and why phrasing matters
- The anatomy of a strong prompt: role, task, constraints, examples
- Common prompt mistakes
- Weak-to-strong prompt rewrites
Reliable results
- Few-shot examples and step-by-step prompting
- Asking for structured output
- Verifying claims against sources
- Catching and fixing hallucinations
Honest power-use
- AI as tutor versus ghostwriter
- Disclosure habits for school and projects
- Building a reusable prompt playbook
- A simple multi-step AI workflow
Showcase & the playbook
- Documenting before/after results
- Presenting the playbook
- Live naive-vs-engineered prompt demos
A documented before/after showing weak vs engineered prompts, plus a real project built with disclosed AI help.
The course that turns ‘kids using ChatGPT’ from a worry into a skill. Teens learn to direct AI to expert results — structured prompting, verification, honest disclosure — and leave with a personal playbook and the judgement to use AI well.
The hooks
Kid hook: “I can get AI to do exactly what I want — and I know when it’s lying.” Parent hook: “They learn the honest, skilful way to use AI for school — disclosure and verification built in.”
Who it’s for
Any 13–15 who already uses AI (most do); pairs well with AI Makers. Thrives: writers, researchers, makers, and anyone who wants leverage without losing authorship.
Outcomes — by the end, students can
Write structured prompts that get reliable results; verify AI output against sources; recognise and fix hallucinations; use AI honestly with disclosure; build reusable prompts and a simple workflow.
Tools & compliance
School-managed AI access with caps, 13+ accounts with signed parental consent, source-checking tools; camera-free; honesty and disclosure norms taught and enforced.
Where this course fits
Pairs with AI Makers; sets up AI App Builders and the 16+ agent track.
Parent questions
Isn't this just teaching kids to use ChatGPT for homework?
It's the opposite of cheating — they learn the line between AI tutoring them and AI doing the work, and to disclose it. Schools value this skill.
Is prompt engineering a real skill?
Yes — getting reliable, verifiable results from AI is one of the most useful skills they can carry into school and work.
Do they need coding?
No coding required, though it pairs naturally with AI Makers for those who want to build.
The first lesson is a free trial.
Book a no-commitment trial — pay nothing if it's not a fit.
Book a free trial