HOW TO...
Generate new curriculum ideas & plans
Curriculum development can be a lengthy process. These examples show how AI can help teachers quickly model their ideas.
Thinking allowed:
- Can AI help model different curriculum options to support teachers’ debate and decisions about the curriculum they want to teach?
- Can AI generate a unit of work tailored to the specific needs and circumstances of a particular teacher, department, or students?
- Is AI training data and algorithms sufficient for the AI programme to recognise and understand UK educational concepts and practice?
- Is AI training data and algorithms sufficient to enable the AI programme to respond in a way that is sensitive to the different age and ability profiles of students
AI Prompts and Responses
The pace of change in the field of AI is speedy and relentless. So it is worth mentioning that this was written in July 2025 and most of the research into prompts and AI responses took place in the last year.
This research sought to explore the extent to which art and design teachers might use well-designed AI prompts to generate support materials and pragmatic programmes of study from abstract ideas and aspirations. In the course of this research, 100’s of individual lesson plans and about 80 separate units of work have been generated.
Purpose
The purpose of this website and of these prompt and response trials is to open the door to any art and design teacher who wishes to explore the use of AI in their professional work.
It offers not solutions but examples which will help teachers make their own enquiries. The website is founded on the principle that teachers are at their best when teaching their own curriculum, and these examples illustrate how teachers can take any curriculum idea and quickly and easily model it against their own circumstances in order to support their discussions and decisions.
Process
Initially the process involved the search for the one best prompt to generate the one best curriculum plan. However, in March there was a damascene conversion when it was realised that the search for the one ‘perfect’ AI response was foolhardy, as well as unrealistic. Bea Wohls explained at a recent conference at the University of the Arts London, that
“all datasets are created by humans and are never perfect”.
So the approach, methodology and rules changed, and the view was taken that in generating curriculum plans, the purpose was to inform teachers’ own choices and decisions rather than providing answers.
Principles
We have progressively built into these prompts the following principles:
- AI should not be used to generate THE curriculum, the one solution. However, teachers could utilise AI to model a range of examples, which would contribute to their debates and decisions about the curriculum they decide to offer.
- These prompts provide opportunities for teachers to personalise prompts to explicitly respond to their own particular circumstances, experience, aspirations and those of their students
- These prompt examples usually offer to create at least two, or more, examples of any unit of work. In this way, these examples underscore the notion that teachers should always make their own choices and not simply rely on a single (fallible) AI response.
- That these prompts should include the instruction within the central task that the purpose of the enquiry was to inform the teachers decision making, rather than to provide a single solution.
- That these prompts should always include the requirement that the response should include a statement [That AI is not sentient…etc]. This will serve as a constant reminder that AI is fallible and we humans must always critically evaluate the outcomes of these algorithms before we use them.
Prompts
The prompts and responses on these pages have been evolving in parallel over the last few months. Some have evolved from the writers interests and curiosity, some have evolved from a shared strategic enquiry from a group of colleagues. Most have arisen from discussions and events generated by NSEAD.
The most important outcome of this research will be the prompts. The responses illustrate the nature of the AI outcomes, but they will have been generated for fictitious schools, to demonstrate how teachers should input their own context.
The prompts used have been developed using two strategies. Firstly they followed typical guidance about prompt engineering using the headings: Role; Context; Reason; Format; Style. These are explained on the page headed ‘Prompts‘. These prompts were tested and modified in different contexts.
A second strategy was developed when exploring less well defined concepts, for instance, the distinctiveness of thinking and behaving as designers rather than artists. This involved a two prompt strategy. Firstly the desired outcome or curriculum learning, was described and placed into the prompt with the instruction that the AI programme generate the most appropriate prompt to lead to a unit of work that would deliver that outcome. The second stage was simply to use the resulting prompt in a couple of AI programmes.
The final prompts used in these pages draw from experience of both strategies and increasingly take pains to provide clear information and instructions related to the particular issues being explored.
The final prompt template has also been created to have the above principles hardwired into it.
Examples
The prompts and AI responses are given below in full. The prompts will typically ask for several different examples, and some of these examples may be for an entire programme of study. For some of these examples, this will include many lessons and pages. However, it will not be necessary to read everything. These are not intended to be adopted, but skim reading will be enough to put the prompt into perspective and adapt it.
Below are examples of prompts and AI responses illustrating how AI programmes can generate a variety of different curriculum options and other materials to support art and design teachers.
This is a simple enquiry and was done quite early on. It seeks address the misconceptions that often arise about the requirement to teach a demonstration lesson during an interview for a teaching post. Facebook, in particular, contains repeated queries and suggestions about this, which often completely miss the point.
JCQ has stated that any work generated by AI cannot be seen as evidence of a students unaided achievement and will, therefore, not be awarded any marks. However, it has declared that AI can be used for research for NEA submissions. This is a suggested prompt that might be used by a student in KS 4 or 5 to support their research. The prompt takes care to include within it the evidence that the outcomes are intended for research and offers guidance to students about how they can guard against accusations of improper use. This is a delicate matter and the consequences of misunderstanding are serious for the teacher and students. This is dealt with on the page Exams: Rules and Options.
This is an example of the one key prompt that teachers can use to explore curriculum options. It uses a single six-lesson Yr 9 unit of work from the new OAK curriculum about portraiture, and shows how this can be modified and adapted by three different (hypothetical) schools to reflect their particular circumstances.
The first school is represented as being in an area with a high proportion of families from a traditional Islamic heritage. This prompt seeks find three different units or work to replace the OAK curriculum.
The second school is represented as being located in an area of London with a large number of students involved in ‘Carnival’ as an expression of their identity. This prompt asks for three options which use mask and costume as a focus for their version of this portraiture unit.
The third school is represented as being located in the midlands and follows a traditional fine art curriculum. The prompt asks for three suggested units related to developing drawing skills and which have a drawing as an outcome.
These prompts have generated a total of nine units of work (54 lessons) tailored to the individual context of three different schools, all from a single OAK curriculum unit, which was based on the context of no school in particular.
This is an example of a scheme of work which was developed from a casual email. The email in question described an NSEAD event that involved teachers using drawing to explore the Church Street market near Edgeware Road in London. The email describing the discussion that ensued, exploring issues of race, culture, the underempowered, and the disenfranchised. The email text was simply copied into the prompt and ChatGPT was instructed to generate one term scheme of work for Year 9 – ‘Reclaiming the Narrative’. The email text included:
“We used Church Street Market off Edgeware Road as a focus to review our own (each teacher’s) curriculum to explore how to make it not like existing curricula and not generic. Exploring race, diverse cultures, the needs of the community, the soft skills necessary for those coming from such a culturally diverse community with the lowest life expectancy in Britain (14 years below the average), exploring how to empower the unempowered and disenfranchised, prevent youngsters being drawn into gangs, drugs, crime, radical politics etc, respecting diverse cultures and exploring creativity to see the art and design in everything – while avoiding the imposed curriculum of normality and middle class values.
It was a good old-fashioned curriculum CPD-type event. Teachers were energised and laughing. Many said it had solved various problems they had been feeling with some of what they were doing.”
This is a simple prompt which explores the notion of a unit of work to introduce students to the potential of AI and to learn about issues such as hallucinations and bias. The prompt asked for two examples of a 6 lesson unit of work in Year 7 and two in Year 9. All four units were balanced taking account of potential benefits and disadvantages of AI. The aims were for students to learn;
- Understanding AI Concepts: How AI works, its strengths (ideation, variation) and weaknesses (bias, hallucination).
- Prompt Engineering Basics: Learning to communicate effectively with AI.
- Critical Evaluation: Analyzing AI outputs for bias, accuracy, and aesthetic quality.
- Ethical Considerations: Safe use, data privacy, and avoiding anthropomorphism.
- Human Creativity First: Emphasizing AI as a tool to augment, not replace, human artistic skill and intention.
This is a simple example of an AI generated cover lesson plan for a photography lesson. The teacher is assumed to be a non-specialist.
This is the second of two prompts following this theme. The first (below) was rather casual and lightweight so this second prompt took the theme more seriously. It began by compiling a full list of as many significant female artists as possible between the years of 1850 and 2025. This provided a more robust data set from which to explore common themes and a continuous narrative of art which did not refer to the male orthodoxy. The resulting scheme of work is interesting for both male and female students.
This AI research took place in three prompt stages. The AI programme used throughout was Gemini.
- Use AI to identify a comprehensive list of notable female artists in this time frame and begin to identify what might be deemed a ‘female perspective’.
- Use Ai to identify five key themes which, based on the artists identified, might represent themes which female artists have typically pursued.
- Use Ai to develop on of these themes into an explicit termly project and teaching plan.:
This is the first prompt in this two-part investigation. It was prompted by a visit to an exhibition by a little known artist, Maeve Gilmour, the wife of Mervyn Peake. The prompts and responses in effect illustrate a conversation with ChatGPT in which the AI programme invites further responses and enquiry. However, it remains casual and ultimately light weight. This led to the second prompt which initially gathered together a much greater set of references to women artists as the basis for the research. The prompts were also more directed towards the generation of a programme of study.
This arose from an NSEAD event for members at the Royal School of Drawing. This involved a life drawing class, an exhibition and a discussion. The day focused attention on drawing throughout. In discussion with Royal Drawing School tutors, the publication ‘Ways of Drawing’ was recognised as a important source of ideas about drawing. This grew from that. It takes a quote from the publication and builds a set of short term (4 lesson) drawing enrichment units that can be threaded through the curriculum in order to sustain and reinforce drawing as a key aspect of being in, and looking at, the world as an artist. The units draw on the different ways the the Royal Drawing School promotes the value of drawing as a way of engaging with the world.
The quote used was;
“Draw things that affect you. It is usually best to do what you like – to draw things that appeal and not to worry about being cool. It is also best not to pretend to be interested in things you are not, or to worry about the conventions of acceptability. Drawing is also about desire – a desire to understand, to record, to consume bodily; to love and hate, and to show others how much we love and hate things in our world: to see and be seen. This is no pathology or twisted craving, but a human need to meet one another. It is also about being able to be the ‘other’ and understand the ‘other’. “
This is a major part of this AI research. It was commissioned by the NSEAD SIG (Special Interest Group) which is seeking to refresh notions of ‘design’ in the art and design curriculum. This research took as a theme work by Professor Gary Granville describing ‘designerly thinking’.
Professor Granville and Dan China worked in partnership to see if AI could generate programmes of study which captured the distinctiveness of designerly patterns of behaviour and thinking, as opposed to the traditional approach to the subject from a fine art perspective. The research generated a great deal of words and explored the responses of different AI programmes. Initially AI programmes failed to recognise the muddle and grab of ‘wicked’ design processes, instead repeating the linear, predictable theory of ‘the design process’.
It became evident that AI was probably not going to make the imaginative leap required and so it was decided to manufacture it. This involved requiring two disruptive interventions to be inserted into the programme. These disruptive interventions compelled students to respond to changes in circumstances and discuss them, thereby further refining their solutions. This requires students to behave as designers rather than artists.
Other contributions came from Erin and Ella who shared their recent experiences of school and reviewed the materials generated by AI. They explained that there needed to be a pretext to introduce disruption and hence these units were modelled on the premise that student groups would form design agencies and the teacher would take on the role of the client. We are grateful to all our collaborators in this research.
The prompt included this text to illuminate the distinct aspects of designerly thinking:
‘Each programme should focus on developing designerly thinking through addressing “wicked problems” – complex challenges with no clear solutions and incomplete information.
The programme should develop students’ capacity for abductive reasoning: considering multiple pragmatic solutions when faced with incomplete data, and being comfortable with solutions that might be provisional or imperfect.
Each curriculum example should develop these specific capacities in students:
- Responding effectively to identified needs in defined social contexts
- Identifying challenges or opportunities in new/imagined settings
- Recognising and responding to ‘wicked problems’ in which unpredicable changes will occur at different stages in the process. This will require a cyclical process involving both lateral thinking, creativity and imagination complemented by process driven problem solving.
- Some collaborative work with peers
- Application of prior learning
- Critical thinking processes (brainstorming, reflection, critique)
- Evaluating solutions based on function, design, user engagement, and cultural factors
- Identifying follow-up challenges or extensions’
This unit takes on the ideas and structure of the previous example ‘behaving as designers’. However it locates the experience in the world of theatre design. This example enables students to retain the collaborative behaviours of designers but to do so in the context of a world of imagination and creativity.
This is an early version of the research into designerly thinking and behaviours. It illustrates the way that the AI programme tended to offer the traditional linear ‘design process’ as the only lens with which to view design. It was this simplistic linear design process convention that the other prompts sought to challenge and break.