INT6066 Chap.8 Generative AI and Human AI Co Design of Lessons
Generative AI and Human AI Co Design of Lessons
A reading week that studies how you behave
The reading week in this course is built around generative artificial intelligence and its implications for education, with resources split into three deliberate categories: short videos to build a basic understanding of the concepts, applications and underlying technology; introductory readings giving both the technical and the educational view; and academic papers on the promise and challenges of the technology in education and on its role in collaborative problem solving.
Then the course does something more demanding than ask you to read.
It puts you inside a design system running on agents together with retrieval-augmented generation, and studies three things: whether the help makes your design better or more original, how you deal with what it offers by rewriting, picking, judging, taking or turning down, and whether working this way shifts what you understand about teaching and about knowledge that spans subjects.
The focus is put plainly: not what the machine turns out, but the thinking, judging, adapting and working-together you do while it does.
The task, and the eight components
The instruction has three parts. Settle on a topic or a real problem that genuinely needs two or more of science, technology, engineering, the arts and mathematics.
Co-design a lesson that clearly states its target learners, learning objectives, the disciplines involved, the learning activities, the technology or interactive tools used, how learners interact with that technology, how the design supports learning, and the assessment methods. Then reflect. The eight components are not a formatting requirement; they are the eight places where an under-specified design becomes visible.
A list of authentic contexts is offered to start from, running from air quality monitoring and campus energy use through waste sorting and irrigation to earthquake-resistant structures and wearable health monitoring.
Accept, modify, reject or develop further
You are told explicitly that you are encouraged to revise generated content and should not simply accept what the system proposes.
Five alignment questions are supplied to test each substantial suggestion: whether it aligns with your learning objectives, fits your chosen pedagogical approach, meaningfully integrates the disciplines, suits your target learners, and genuinely supports learning. Four decisions are then available, and the two that matter are the middle ones.
Accepting tells you nothing about your own judgement and rejecting throws away the sound part of a suggestion, whereas modifying forces you to say which test failed and what the minimum change is.
That sentence is exactly what the reflection needs and exactly what a marker can assess.
The risks named, and the record the reflection depends on
The named risks are over-reliance, reduced independent thinking, incorrect or inappropriate generated content, homogenisation of lesson designs, and misalignment between suggestions and the pedagogy or objectives you chose.
Homogenisation is the one that catches careful people, because a plan can be internally consistent and indistinguishable from what everyone else produced.
The structured reflection then asks about process, about knowledge and judgement, about role and agency across designer, editor or evaluator and co-creator, about critical reflection on the limits, and about whether the experience changed what instructional design means to you and which parts still have to be done by the teacher.
None of that can be reconstructed honestly a week later, which is why recording each decision with one reason is part of the task rather than tidiness.
What this chapter covers
- 01
Three categories of reading-week resource and what each is for
- 02
What the study is actually interested in observing
- 03
The eight components a co-designed lesson has to carry
- 04
The five alignment tests, applied one suggestion at a time
- 05
Accept, modify, reject or develop further, and why modify is usually right
- 06
Five named risks, including homogenisation of designs
- 07
Designer, editor or evaluator, and co-creator as positions on one axis
- 08
What the course asks you to declare, and where
Run the five alignment tests, then decide
- 5Run all five tests and mark each pass or fail.
- 2Decide accept, modify or reject and say why.
- 3Write the specific modification for each failed test.
Key terms
- Human Machine Co-Design
- A way of working in which a designer and a generative system develop a lesson together, with the designer testing, revising and deciding on each suggestion.
- Design Homogenisation
- The risk that generated assistance produces internally consistent plans that are indistinguishable from one another across a whole class.
- Alignment Test
- One of the five checks applied to a suggestion, covering objectives, pedagogical approach, disciplinary integration, learner fit and support for learning.
- Interdisciplinary Integration
- The requirement that two or more disciplines each carry part of the learning, rather than one supplying content while another supplies decoration.
Generative AI and Human AI Co Design of Lessons FAQ
Am I allowed to use generative tools in this course?
You are encouraged to explore them sensibly and asked to cite or acknowledge their use explicitly. For the written individual work you must declare any use, submit a word processor file with tracked changes, and the report notes rule out generated text in the submission itself alongside copying from online sources.
Designing with a tool and submitting text written by one are treated differently, so confirm the current wording on Moodle before you submit.
What should I actually record while co-designing?
For each substantial suggestion, write the suggestion, the test it failed or passed, the decision you took and one reason. An entry like suggestion four, rejected, not suitable is useless later; an entry naming the test, the evidence and the replacement answers the process, judgement and role questions in the reflection at once.
Does using a generative tool weaken my design?
Only if you stop deciding. The named risks are over-reliance, reduced independent thinking, incorrect content, homogenised designs and drift away from your chosen pedagogy, and every one of them is a failure to test a suggestion rather than a property of the tool. A design that passed every test on its first pass is worth a second look.
Assessment move
Keep one page with four columns while you design: the suggestion, the test it failed, the decision, and one reason. Fifteen entries is enough for the whole reflection, and writing them as you go costs about a minute each while reconstructing them later is impossible to do honestly.
Working through Generative AI and Human AI Co Design of Lessons in INT6066? Sia is AskSia’s AI Education tutor — ask any INT6066 Generative AI and Human AI Co Design of Lessons question and get a clear, step-by-step explanation grounded in how INT6066 is taught and assessed. Read this chapter free, then take your hardest questions to Sia.
