Hello World

What AI Is, and What It Is Not

Overview

# Overview

Curriculum.Units

How a generative model produces language by predicting the next word — and where that reasoning breaks down. Students frame an evidence-based question and write a structured prompt, separating what AI can reliably automate from what it cannot genuinely think through.

# Objectives

Curriculum.Units
  • 01 By the end, you can Frame a narrow question that can be settled with evidence, not opinion.
  • 02 By the end, you can Write a complete prompt — role, task, constraints, output format — and predict what the model will produce.
  • 03 By the end, you can Explain that the model predicts the next word rather than looking up facts, and point to where it may invent detail.
  • 04 By the end, you can Combine both to build and present a structured slide deck to the room.

# Sessions

Curriculum.Sessions
  1. What You Already Think About AI
    AI concepts
    • Warm-up Vocabulary Naming Drill The teacher shows words one at a time and the class says each aloud before learning what they mean.
    • Discussion AI Question Bank Each student writes one AI question, then the class sorts them into ones evidence can answer and ones it cannot.
  2. AI You Can Play With
    AI concepts
    • Game The Hive Mind Teams guess the most common search autocompletes, then see that AI copies the crowd, biases included.
    • Game Next Word Vote-Off Volunteers pick a next word for a sentence and the class votes, showing how AI predicts text.
    • Game Spot The Hallucination A chatbot writes a short bio with one made-up fact and students hunt for the invented detail.
  3. How AI Shows Up in the World
    Applied craft
    • Worksheet AI Sightings WebQuest Students follow a short set of teacher-picked links and write down three real places AI is being used and what it does in each one.
    • Discussion Sector Expert Jigsaw Each small group researches AI in one field like health, music, transport or gaming, then teaches what they found to the rest of the class.
    • Discussion Evidence Gallery Walk Groups pin their real AI examples on the wall and everyone walks around adding a sticky note with one question or fact.
  4. Build Your Slide Deck
    Deliverable prep
    • Build Slide Deck Sprint Students write one detailed prompt, have the AI build a slide deck from it, then present the result.
    • Role-play Corrupt A Wish The class grants a wish in the most literal, unhelpful way to show AI does exactly what you say.

Applied · Ethics

# Knowledge & Skills

ACA-Applied
K Knowledge · what you know
When an AI model makes up a detail (this is called a hallucination), it happens through the same process as when it gives a correct answer. It is not a separate fault.
Words learners need to know for the exam: training, tokens, transformer, prediction, generative, hallucination, algorithm, data, output.
A question that can be answered with evidence is narrow and specific.
S Skills · what you can do
Present the deck to the room. point w1d1-s2-p2
Explain what the letters GPT stand for and what the AI model does when it answers. from objectives

AI

# AI Concepts

ACA-AI.Concepts
Aligned with AI4K12 · Five Big Ideas in AI View framework →
AI4K12 · Five Big Ideas in AI
3-A-I

Humans vs Machines

Both humans and machines learn from examples, but they learn in different ways and at different rates.

AI4K12 · Five Big Ideas in AI
3-C-II

Large Datasets

More training examples generally help machine learning models learn better patterns, though very large datasets have practical challenges.

AI4K12 · Five Big Ideas in AI
5-B-II

Trust and Responsibility

For people to responsibly use and rely on AI systems, they must understand how those systems work and their limitations.

AI4K12 · Five Big Ideas in AI

Strong on Familiar, Weak on New

An AI model is good at summarising text, changing its format, working with familiar ideas, and writing a first draft in a familiar style. It struggles with new or unusual topics.

AI4K12 · Five Big Ideas in AI

Predicting Tokens, Not Facts

An AI model predicts the next word. It works with small chunks of text called tokens. It finds patterns in the text it was trained on and predicts from those patterns. It does not look up a stored fact.

AI4K12 · Five Big Ideas in AI

Specify Structure, Not Quality

A prompt (an instruction given to the AI) states the structure, length, colours and output format. It does not ask for a quality.

AI4K12 · Five Big Ideas in AI

What GPT Stands For

Learners can say what the letters GPT stand for and what the AI model is doing when it answers.

AI4K12 · Five Big Ideas in AI

Prediction Causes Hallucination

An AI model predicts the next word. It does not look up a stored fact. This is why it can make up details (hallucinate).

# AI Competencies

ACA-AI.Competencies
Aligned with UNESCO · AI Competency Framework for Students View framework →
Human-Centred Mindset
4.1.3

AI Foundations

Building fundamental knowledge of what AI is, how it works, and the technical and human considerations that shape its development.

# AI Skills

ACA-Tools.Skills
Aligned with Derived from course activity
Derived from course activity

Prompt Engineering for AI Agents

Iterating a prompt/system instruction until an AI agent's responses stop sounding generic and actually reflect the specific product or context. Evidence: "Does this information sound generic? Or does it sound actually tailored to the website?" (app-development, w3d3).

# Competency Goals

ACA-AI.Goals
Aligned with UNESCO · AI Competency Framework for Students View framework →
UNESCO · AI Competency Framework for Students
CG4.1.3.2

Training on Data and Algorithms

Grasping how AI systems learn from data using computational procedures and adjusting based on feedback.

UNESCO · AI Competency Framework for Students
CG4.1.3.3

Interdisciplinary Foundations

Recognizing that AI draws from computer science, statistics, philosophy, psychology, and social sciences.

UNESCO · AI Competency Framework for Students
CG4.1.3.1

Definition and Scope of AI

Understanding what constitutes AI and its various forms, from narrow task-specific systems to broader intelligent approaches.

UNESCO · AI Competency Framework for Students

Build a Full-Prompt Deck

Build a slide deck from a full prompt that sets the structure, length, colours and output format, and keep one look all the way through.

Deliverable

# A structured, presented deck

Projects.Briefs

Produce a slide deck that follows a fixed brief (a set of instructions) and keeps one look across every slide. A structured slide deck, made from one full prompt and presented aloud.

Transfers to: Holding a full creative brief, aesthetic and format, across a whole finished piece of work.

See past cohorts ship this in the Showcase →

# Assessed toward these aims

Projects.Rubrics

This deliverable is where students demonstrate

  1. A Explore how a generative model produces output, and where it fails
    Look out for
    • Prompt completeness: The prompt states the structure, length, colours and output format. The deck matches what the prompt asked for.
    • One look held across the deck: One look, meaning the same colours and fonts, runs across every slide. It is not a different look on each slide.
    • Explaining the model: The learner says what the letters GPT stand for, and describes the AI model as predicting the next word rather than looking up a stored fact.
    • Explaining invented details: The learner can point to where an AI model might invent a detail, and can say that giving the AI real facts (grounding the prompt) removes it.
    • Question can be checked with evidence: The question is narrow enough to be answered with evidence. It is not too broad, and it is not a matter of opinion.