Generative AI Workshop at Yan Oi Tong Tin Ka Ping Primary School — INSPIRE Gets Primary Students Actually Using GenAI

From 14–16 and 20 July 2026, the INSPIRE team ran a series of Generative AI workshops at Yan Oi Tong Tin Ka Ping Primary School — 8 sessions in total, 1.5 hours each, split between P1–P2 (4 sessions) and P3–P4 (4 sessions). The core goal was to give primary students the chance to genuinely use generative AI — not merely to have heard the term "AI," but to personally experience, understand, and practise GenAI and Large Language Models (LLMs).

From "Having Heard of AI" to "Actually Using AI"

For many primary students, AI is a familiar yet abstract term — they have heard of it, but have rarely used it in a generative AI or LLM way. Our workshop was designed to fill exactly this gap. Led by INSPIRE instructor Melvin, students were introduced to the concept and applications of AI, with a strong focus on prompting skills. We primarily used Google Gemini as our teaching medium, letting students practise on a real tool.

Using Everyday Analogies to Unpack How AI Works

Teaching young students to understand AI requires methods that match how they think. We turned abstract concepts into examples they can see and touch.

For example, to explain how AI "learns," we compared human learning to machine learning: a teacher giving both questions and answers is like "supervised learning," while a student observing and finding patterns on their own is like "unsupervised learning."

Yan Oi Tong Tin Ka Ping Primary GenAI workshop material — comparing human learning with machine learning: supervised vs unsupervised

Going further, we used distinguishing fruits as an example: red and round is an apple; yellow, long and curved is a banana — this is exactly how AI learns to recognize things through large numbers of "labelled" examples. One relatable analogy, and students instantly grasped "supervised learning."

Yan Oi Tong Tin Ka Ping Primary GenAI workshop material — explaining supervised learning by distinguishing apples and bananas

Understanding Generative AI: Same Prompt, Different Results

A key concept in the workshop was helping students realize that entering the same prompt can produce different results each time. Through fun Gemini-generated images and "AI brain-teasers," students saw AI's diversity and creativity with their own eyes — this is the very essence of "generative AI." The experience also nurtured their ability to compare and judge the strengths and weaknesses of different LLMs.

Yan Oi Tong Tin Ka Ping Primary GenAI workshop material — the same prompt generating different results, showing the diversity and creativity of generative AI

Prompting Skills: Who, What, When, Where — Clear and Specific

To get AI to accurately understand our intentions, writing good prompts is key. We taught students a simple, memorable framework — When, Where, Who, What: a good prompt should clearly state the time, place, people, and event.

Yan Oi Tong Tin Ka Ping Primary GenAI workshop material — prompting tips: when, where, who, what

Beyond the basics, we emphasized that prompts must be clear, detailed, and specific, adding environmental details and output-format requirements. For instance, the same sentence "a male maths teacher teaching maths" — simply by adding different styles (cute / sci-fi / fairytale / cartoon / watercolour / realistic) — produced completely different images. Through this, students learned that the more precise the prompt, the closer the AI's output comes to what they imagined.

Yan Oi Tong Tin Ka Ping Primary GenAI workshop material — adding style and output format to generate six different styles of a maths teacher

A Four-Day Journey: From Understanding to Creating

The whole workshop was designed to progress step by step, taking students from understanding to creating:

Day 1｜Understanding Generative AI — building a basic understanding of GenAI and learning to distinguish the benefits of different LLMs.

Day 2｜Q&A Experience — students completed pre-class homework: preparing 10 questions to "test" the AI, in order to understand and try applying prompting skills.

Yan Oi Tong Tin Ka Ping Primary GenAI workshop — instructor Melvin inviting eager students to answer during the interactive Q&A

Day 3｜AI Image Creation — students used Gemini (Nano Banana 2) to generate images, learning to use clear, precise prompts so the AI accurately rendered the picture in their minds.

Day 4｜AI Coding Experiment (Building a Mini-Game) — students used Gemini to generate a mini-game, then personally debugged it, doing their best to make it smoother, more fun, and more user-friendly — experiencing the full create–test–improve cycle.

Yan Oi Tong Tin Ka Ping Primary GenAI workshop — instructor Melvin demonstrating the AI-generated mini-game Train Dispatcher, students building on iPads

Yan Oi Tong Tin Ka Ping Primary GenAI workshop — a student playing the AI math mini-game she generated with Gemini (11 × 6 = ?)

Yan Oi Tong Tin Ka Ping Primary GenAI workshop — a Gemini-generated math RPG game with a character, coins and HP system

Teamwork and Collective Brainstorming

In some segments, students worked in groups to encourage team spirit, while our teaching assistants helped pool and organize their ideas — so that every idea had a chance to be heard and realized. We placed strong emphasis on brainstorming, teaching students how to harness AI to turn their own creativity into finished work. We also designed accompanying worksheets to support both preparation and in-class learning, making the experience more structured.

Yan Oi Tong Tin Ka Ping Primary GenAI workshop — students working hands-on with iPads and dedicated worksheets while a teaching assistant helps

How Does This Benefit Primary Students?

Early exposure to generative AI brings multiple benefits to a young learner's growth:

Building digital literacy — in an age where AI is woven into daily life, learning early to use AI tools correctly, safely, and effectively lays a strong foundation for future study and life.

Sharpening critical thinking — through the "same prompt, different results" experience and comparing the strengths of different LLMs, students learn not to blindly trust AI output, but to actively judge and verify.

Sparking creativity and expression — from idea to AI-generated images and games, students learn to express their thoughts clearly (writing good prompts) and turn their creativity into finished work with their own hands.

Building problem-solving confidence — in the debugging process of building a mini-game, students experience the try–fail–fix cycle, nurturing a growth mindset that isn't afraid of failure and takes initiative to solve problems.

Conclusion

This Generative AI workshop at Yan Oi Tong Tin Ka Ping Primary School took primary students one real step forward — from "having heard of AI" to "actually using AI." Through relatable analogies, tailor-made teaching design, ample hands-on opportunities, and idea-to-reality missions, students not only mastered prompting skills and the basics of GenAI, but also built the confidence to experiment and create. The one question we most wanted to leave them with was: "Has this sparked your curiosity about the world of AI?" That curiosity and confidence is exactly the educational philosophy INSPIRE has always pursued — helping every student believe that their creativity can be brought to life by their own hands.