Question 1 of 4

1. Pick a unit of work

My choice to learn using AI on an existing unit enabled me to scan my teaching portfolio for units that I believe had promise of being pedagogically improved.

 

My choice for AI professional learning fell on a Year 7 Oral Language English Unit entitled DR SEUSS READERS THEATRE FESTIVAL. As a Term 1 introductory program, it focused on helping students’ transition to high school. I was a 'fun' way to allow me view the phonological awareness and reading fluency of I was given by the leaders of literacy data through NAPLAN, DIBELS and ACER'S PAT Tests.

 

More than anything I wanted to build a class culture despite the cohort having a four level gap (L3 - L7) : that meant that some students had not achieved automatic decoding skills and some that were highly proficient readers.

 

What unit will you choose to learn curriculum designing with AI? Write your thoughts and I'm more that happy to follow up your questions.

Question 2 of 4

2. Encountering the LLM’s Context Window 

No, it’s not a visit to ‘the Wiz’, it’s better! The context window is the powerhouse of AI technology. Do you need to revise the depth and breadth of its power?

When I asked my LLM what key questions a teacher should ask about its training, biases and ethics, this was its advice.

1. "What is your Training Bias?" – Is this model predicting the "Average Western Student" or does it recognize the specific cultural and disability frameworks of my state (e.g., ABLES or Marrung)?

2. "Where is your Hallucination Risk?" – Since you predict language rather than "knowing" facts, how often do you invent Victorian Curriculum codes when you run out of context?

3. "What is your Context Window?" – At what point in our unit design will you begin to "forget" the SWPBS social contract I uploaded at the start?

4. "Whose 'Typical' are you predicting?" – Are your lesson suggestions aimed at the "Median" (the average, safe, boring middle) or are you helping me differentiate for the "long tail" of my classroom?

Write the answers your LLM gave you in the space and I'll give you my reflections?

Question 3 of 4

3. Identify & plan your AI professional learning

The four categories of professional learning you need to look for are: 

i) Technical knowledge; 

ii) Pedagogical applications; 

iii) Ethical standards and 

iv) Policy obligations.

 

Try these suggestions:

1)  The "Audit & Pair" Strategy (APST 6.1 & 6.3)

Sit down with a colleague and map your current confidence levels against the four categories (Technical, Pedagogical, Ethical, Policy). Identify one specific "gap" that is currently causing the most friction in your planning—for example, Technical Knowledge of how context windows work, or Pedagogical Application of HITS using AI. 

Suggestion: Commit to a 15-minute once a week to share one thing that worked and one thing that "hallucinated."

2) The "Guided Mentor" Setup (Technical & Pedagogical)

Use your LLM as a dedicated coach rather than a task-doer. Instead of asking it to "write a lesson," ask it to "teach me to apply HIT 1 - Setting Goals

Suggestion: Open a dedicated thread in Gemini or NotebookLM titled "My AI Pedagogy Coach." Prompt it: "I want to improve my ability to use AI for Differentiated Instruction. For every suggestion you give me this week, explain Tomlinson’s frameworks so I can learn the logic behind your output."

3) The "Policy & Ethics" Boundary Check (Ethical & Policy)

To reduce anxiety, create a "Safe Workspace" document that outlines your non-negotiables. This removes the fear of the unknown by creating clear guardrails for your professional practice.

Suggestion: Create a one-page "AI Social Contract" with your colleague. Define what you will Offload (e.g., mapping VCAA codes) and what you will Never Outsource (e.g., final judgment on student safety). Use the AI to summarise your school's SWPBS and Privacy policies.

Record your work in the space below

Question 4 of 4

4. When uploading docs into a LLM use categories to show their importance.

We have devised THE THREE CATEGORY RULE for uploading documents into a context window.

1. Naming is Signposting:

The AI uses file names as primary tags. Use descriptive names (e.g., DrSeuss_Lesson_Sequence_Draft.docx) rather than generic ones (Lesson_Plan_Final_v2.docx) to give the AI a clear handle to grasp.

2. Separate "Source" from "Workspace":

Distinguish between static constraint documents (e.g., Source_VicCurriculum_Level7_English.pdf) and active creation documents (e.g., Workspace_Seuss_Rubric.docx). This allows for clear instructions like, "Use the Source file to check the validity of my Workspace file."

3. Permanent Anchors:

When a key decision is made, save that file as a permanent "Anchor" (e.g., Anchor_Seuss_Rubric_FINAL.pdf). This prevents "context drift" by forcing the AI to treat that document as a fixed rule for all future work.

Go to your curriculum files and show how you improved their AI-friendliness for use with a context window.  

I'd love to give you feedback, if you write the information in the square.  

NB: Do you see that the AI does not need you to have 'neat folders'???!!!!!