
Foundations of Generative AI in Education
- Session
- Day 1 · 9.00 – 10.30 am
- Focus
- Concepts & landscape
- Outcome
- Your own AI value map
What generative AI actually is
A generative AI model has read an enormous amount of text, images and audio, and learned the patterns in it. When you type a prompt, it predicts the most plausible continuation, piece by piece. That single sentence explains almost every strength and every weakness you will meet.
Because it predicts patterns
It is excellent at
- Drafting, rewriting and restructuring text.
- Explaining one idea at five different levels.
- Turning messy notes into an organised document.
- Producing many variations quickly.
Because it predicts patterns
It is unreliable at
- Precise facts, citations, page numbers and statistics.
- Anything about your specific school, students or policies it was never given.
- Counting, marking to a strict scheme, or arithmetic in long text.
- Knowing when it is wrong — it never sounds unsure.
Myths versus realities
| The myth | The reality | |
|---|---|---|
| Knowledge | It knows everything and searches the internet. | It generates plausible language. Unless a tool explicitly searches or reads your file, it is working from patterns, not sources. |
| Accuracy | If it sounds confident, it is correct. | Fluency and accuracy are unrelated. Invented references and dates are the most common classroom risk. |
| Teaching | It will replace teachers. | It removes preparation drudgery. Pedagogy, relationships and judgement stay with you. |
| Cheating | Detectors will catch AI-written student work. | Detectors are unreliable and can accuse honest students. Redesign the task instead of policing the text. |
| Privacy | Anything I paste is private. | On free consumer tiers your input may be used to improve models. Never paste identifiable student data. |
The tool landscape for education
There are thousands of AI tools, but for teaching you only need four categories. Pick one reliable tool per category and learn it well.
Category 1
General assistants
Category 2
Video creators
Category 3
Slides & visuals
Category 4
Documents & data
Where AI adds real value
- 1
Preparation and planning
Lesson outlines, differentiated versions of the same material, worked examples, analogies for a difficult concept, starter and closure activities. - 2
Content creation
Slides from your notes, a narrated video from a script, worksheets, reading passages at a chosen level, diagrams and infographics. - 3
Assessment support
Question banks across cognitive levels, rubrics, distractors for MCQs, and draft feedback comments you then verify and personalise. - 4
Administration and reporting
Summarising long policy documents, drafting parent communication, turning result tables into a readable narrative.
Where AI should not be in charge
Hands-on task
Map AI onto your own teaching week
Goal: Leave this module with three specific tasks you will hand to AI first.
Finished this part?
Mark it complete to update your progress meter and move on.