The 12 Prompts of Xmas
A Festive Guide on How to Use AI in Instructional Design in 2025
It’s that time of year again, and what better way to wrap up the year that AI shook-up instructional design with a festive guide to how use AI in instructional design?
The 12 Prompts of Xmas
🎵 On the first day of Xmas, my LLM gave to me...
An Understanding of My Course Topic
What it does: Creates initial suggestions for the content required to achieve learning objectives.
Why it works: With the right prompting & input, LLMs that are fine-tuned on peer reviewed research (e.g. Consensus GPT) are capable of sourcing and summarising reliable, robust information on most topics.
Example Prompt:
You are a subject matter expert in [topic]. I am an instructional designer. You will:
1. First, search the research and give me a summary of the most influential and well-cited peer-reviewed research on [topic].
2. Then, search the research and give me a summary of the most influential and well-cited peer-reviewed research on [topic] specifically for [learner type] who need to be able to [goal / outcome]. Note: DO NOT try this in generic AI tools like ChatGPT 40 - they will just make stuff up. Instead, use research tools like Consensus GPT.
🎵 On the second day of Xmas, my LLM gave to me...
Differentiated Learning Content
What it does: Adapts content for different audiences and complexity levels.
Why it works: LLMs are trained to understand concepts at multiple levels of complexity. Their attention mechanisms can maintain focus on key concepts while adjusting language complexity - like a skilled interpreter who can explain quantum physics to both scientists and children. Also, by providing the baseline content, you maintain control of the source of AI’s “understanding” and “thinking”.
Example Prompt:
Convert the following technical manual into three versions:
1. An executive summary with a strategic focus.
2. A manager's guide with an implementation focus.
3. End-user instructions with step-by-step guidance.
Throughout, you must maintain consistency in critical information, only adjusting: - Technical terminology - Detail level - Example complexity - Supporting context
[insert original document]🎵 On the third day of Xmas, my LLM gave to me...
Multiple Assessment Types
What it does: Creates a range of question banks, knowledge checks etc.
Why it works: LLMs’ are good at pattern recognition and repetition, which means they’re pretty good at understanding structures like questions and generating variations on this theme. By defining constraints like the specific assessment types and outputs required, we can help to optimise the quality of AI’s responses.
Example Prompt:
Create 15 suggestions for how to asses [learning objective]. You must generate:
- 5 "near miss" multiple choice questions
- 5 scenario-based questions & activities
- 5 application questions & activities
For each question you suggest, you must:
- Provide sample answers & common error/misconceptions
- State the related Bloom's level
- Explain the conditions in which it will be most effective, e.g. type of learner, mode of delivery
🎵 On the fourth day of Xmas, my LLM gave to me...
Rapid Project Documentation
What it does: Creates project documentation, e.g. templates for reporting.
Why it works: With the right direction and inputs, LLMs do a pretty good job of creating common structured documents.
Example Prompt:
Using the course outline below, create the following documentation set for my e-learning project:
1. Timeline template: state how long each state will likely take, based on the outline. For each approximation, provide a rationale.
2. Status report: create a status report document with suggested milestones and checkpoints to enable me to track and monitor the project.
3. Review checklist: create a checklist to use for each milestone, using the RACI model.
[Insert course outline] 🎵 On the fifth day of Xmas, my LLM gave to me...
Multiple Variations on a Theme
What it does: Creates multiple versions of learning examples, scenarios and practice questions based on an exemplar.
Why it works: LLMs excel here because they're trained on millions of educational examples. Their transformer architecture can maintain consistent structure while varying content - like keeping the "skeleton" of a good learning example while changing the details. Think of it as having a master chef who can make infinite variations of the same recipe while keeping the essential cooking technique intact.
Example Prompt:
Based on the exemplar below, generate 5 customer service scenarios about [specific skill].
For each:
1. Vary customer personas and emotions
2. Include different product types
3. Add environmental factors (time pressure, language barriers)
4. Specify learning outcomes
5. Include success criteria.
[paste exemplar to frame AI's "thinking", tone etc e.g.
Title: The Rushed Business Traveler's Lost Laptop
Scenario: A visibly stressed business traveler approaches the hotel front desk at 10:45 PM. She has just discovered her laptop is missing after a series of back-to-back meetings. She has a crucial presentation at 8 AM tomorrow and needs access to her files immediately.
Customer Persona:
- Sarah Chen, senior executive
- Emotional state: Anxious and frustrated
- Status: Platinum rewards member
- Time constraint: Critical (presentation tomorrow)
Product/Service:
- Premium business suite accommodation
- Guest services
- Lost and found procedures
Environmental Factors:
- Late night (limited staff available)
- Language element (guest has slight accent, indicating English is second language)
- Time pressure (urgent need for resolution)
- Business center is typically closed at this hour
Learning Outcome:
Demonstrate ability to handle time-sensitive issues while maintaining professional composure and following security protocols.
Success Criteria:
1. Remains calm and professional despite customer's stress
2. Follows security protocols while expediting process
3. Offers immediate alternative solutions (temporary laptop, business office access)
4. Communicates clear timeline and next steps
5. Documents incident properly
6. Demonstrates empathy while maintaining boundaries
Key Learning Points:
- Balancing urgency with security procedures
- Managing expectations under pressure
- Creative problem-solving within policy constraints
- Clear communication with stressed customers]🎵 On the sixth day of Xmas, my LLM gave to me...
Ideas for Activity Design
What it does: Generates new ideas and encourages creativity by suggesting a range of ideas for learning activities and exercises.
Why it works: LLMs aren’t great at instructional design but with the right expert input and curation, they can be a great tool for inspiring new ways of thinking and designing.
Tip: by working objective by objective, you keep focus an control over AI’s “thinking” and outputs.
Example Prompt:
You are an expert instructional designer. You will look at the learning objective below, suggest activities as follows:
1. Individual activity to achieve the objective
2. Pair exercises to achieve the objective
3. Group activities to achieve the objective
4. Reflection tasks to achieve the objective
5. A wildcard task I am unlikely to have considered but which will achieve the objective
For each suggestion you must:
1. State the pros and cons for [insert info on target learner]
2. Approximate the duration
3. List the materials required
4. Include instructions
5. Suggest variations for different types of learners
[insert learning objective] 🎵 On the seventh day of Xmas, my LLM gave to me...
Assessment Criteria & Feedback
What it does: Generates rubrics for assessing learner performance.
Why it works: With a clear input and expert checking, LLMs can rapidly turn assessments intro scoring criteria.
Tip: by working activity by activity, you keep focus an control over AI’s “thinking” and outputs.
Example Prompt:
You are an expert instructional designer. I will give you a learning activity and a description of my target learners. For each activity, you will suggest an assessment criteria for scoring the learner's performance in the activity. For each activity you must provide:
1. A definition and example of exceptional performance.
2. A definition and example of excellent performance.
3. A definition and example of good performance.
4. A definition and example of poor performance.
For each definition you must provide example copy of the feedback you would provide the learner. The feedback must:
1. State what they did well.
2. State how they can improve and/or expand their learning.
3. Provide actionable next steps.
[insert learner description, e.g. age, start point, goal]
[insert learning activity] 🎵 On the eighth day of Xmas, my LLM gave to me...
Resources for My Course
What it does: Creates supplementary, supporting learning resources like guides and job aids for designed courses.
Why it works: When we define what LLMs should synthesise and how, they do a pretty effective job of organising and presenting information it in a variety of common formats.
Example Prompt:
You are an expert instructional designer. I will give you a learning design including a description of the target learner. You will use only this information to create:
1. A quick start guide
2. An FAQ document
3. A glossary of key terms
Throughout, you must focus on practical application and follow quick reference principles.
[insert course outline and learner persona info]
🎵 On the ninth day of Xmas, my LLM gave to me...
Ideas for Instructional Strategies
What it does: Generates ideas to inform the selection of instructional strategies.
Why it works: With the right steering and validation, LLMs that are fine-tuned on peer reviewed research (e.g. Consensus GPT) understand enough about pedagogy to help generate ideas about which instructional strategy or strategies to use and why.
Example Prompt:
You are a subject matter expert in [topic]. I am an instructional designer. You will:
1. First, search the research and give me a summary of the most effective instructional strategies and methods to teach [topic] to [learner].
2. Then, using only the research you sourced, you will provide an evidence-based recommendation for which instructional strategy or strategies to apply in my design to teach the stated topic to the stated learner. Note: DO NOT try this in generic AI tools like ChatGPT 40 - they will just make stuff up. Instead, use research tools like Consensus GPT.
🎵 On the tenth day of Xmas, my LLM gave to me...
Engagement Strategy Ideas
What it does: Generates ideas for optimising learner engagement and interaction.
Why it works: LLMs can combine patterns from various engagement techniques and adapt them to different contexts and learner types.
Tip: by working objective by objective, you keep focus an control over AI’s “thinking” and outputs.
Example Prompt:
You are an expert instructional designer. Your task is to optimise my course design for learner engagement and interactivity.
I will give you a learning objective and a learner profile. For each learning objective, you will suggest:
1. Two discussion points which will help the defined learner to achieve the defined objective.
2. Two group activities which will help the defined learner to achieve the defined objective.
3. Two interactive exercises which will help the defined learner to achieve the defined objective.
4. Two reflection activities which will help the defined learner to achieve the defined objective.
For each suggestion you must include:
1. Notes on risks and benefits of the approach for engagement including variation options
2. Facilitation tips
3. Time requirements
[insert learning objective and learner profile]🎵 On the eleventh day of Xmas, my LLM gave to me...
A Way to Optimise Objectives
What it does: Helps refine and optimise learning objectives.
Why it works: With the right input and rubric, LLMs can understand the required structure, language & sequence of learning objectives, helping us to optimise learning objectives for impact.
Example Prompt:
You are an expert instructional designer which specialises in writing and sequencing learning objectives.
I will give you a set of learning objectives. Your task is to provide recommendations for how to optimise them for learner motivation and achievement.
To do this, you must always use the following rubric for writing and sequencing learning objectives:
1. Every learning objective has a verb from Bloom's taxonomy.
2. Learning objectives are sequenced from most simple to most complex, according to Bloom's taxonomy.
3. Every learning objective uses "you" language to directly address the learner.
[insert objectives]
🎵 On the twelfth day of Xmas, my LLM gave to me...
An Implementation Planning Doc
What it does: Creates implementation guides and rollout plans.
Why it works: With the right direction, LLMs can structure implementation steps and take into account various factors from their training on project management and instructional design.
Example Prompt:
You are an expert instructional designer which specialises in writing plans for the implementation and roll out of learning experiences.
I will give you a course outline. You will use the outline to create an implementation plan for the course The plan must include:
1. Rollout phases
2. Stakeholder communications
3. Risks & mitigations
4. Success metrics
5. Implementation overview: a summary of key implementation steps with approximate timeframes
[insert course outline] I'd love to hear how if and how these prompts work for you. Which ones might become your go-to helpers? Did you discover any interesting variations? Perhaps you found some completely new ways to use AI in your instructional design practice? Share your experiences and let's learn from each other as we navigate this exciting frontier.
Wishing you all a very Merry Christmas and a 2025 filled with innovative, impactful learning designs.
🎄✨🎁
Phil :-)
PS: If you want to get hands-on and learn how to get the most from AI in your day to day work, apply for a place on my AI & Learning Design Bootcamp.
PPS: You can download a full e-book version of this guide on my Gumroad site, here.