chatgpt-module-3-personalisation
ITIAN ChatGPT Academy
Technology Simplified — Solutions That Work
Personalise ChatGPT with Purpose
Shape how ChatGPT communicates, what preferences it can use and which context belongs in global settings, memory, individual conversations or project-specific instructions.
Welcome to Module 3
Personalisation works best when every setting has a clear job. This module helps you create a consistent experience without oversharing or expecting one control to do everything.
Choose a Communication Style
Understand personality and gradually available characteristics without confusing tone with capability or accuracy.
Write Useful Instructions
Create concise Custom Instructions for enduring preferences, working style and accessible output.
Control Remembered Context
Understand saved memories, chat-history reference, review controls and Temporary Chat.
Build a Personal AI Profile
Combine the right settings, test them with real prompts and keep an approved profile you can maintain.
Learning Outcomes
By completing Module 3, you should be able to:
Explain Each Personalisation Layer
Distinguish personality, characteristics, Custom Instructions, memory, project context and one-time prompts.
Choose Appropriate Settings
Select a base style and fine-tuning options suited to the way you want to work.
Create Strong Custom Instructions
Write clear, compact instructions covering language, teaching style, formatting and enduring preferences.
Manage Memory Deliberately
Review, add, correct, remove and disable remembered information when appropriate.
Protect Personal Information
Minimise sensitive details and understand how Temporary Chat and Data Controls differ from Memory.
Test and Maintain a Profile
Evaluate behaviour across several tasks, resolve conflicting instructions and schedule regular reviews.
Your Module Progress
Tick a lesson after completing its reading and practical activity. Progress is stored only in this browser when local storage is available.
Module 3 Completion
Seven lessons and the final knowledge check
0 of 8 completed — begin with Lesson 3.1.
Module 3 Lessons
Follow the sequence on your first visit. Each lesson builds one part of a maintainable personalisation system.
Personalisation Overview
Understand the different layers, what each one controls and which setting should hold which kind of information.
Personality and Characteristics
Choose a base style and fine-tune response qualities while recognising that capability and safety remain unchanged.
Custom Instructions
Write global guidance for language, tone, teaching style, formatting, recurring work and useful boundaries.
Memory Basics
Learn the difference between saved memories and chat-history reference, and when remembered context can help.
Managing Memory
Ask what is remembered, correct outdated details, remove unwanted memories and understand what turning Memory off does.
Personalisation and Privacy
Choose what not to personalise, use Temporary Chat appropriately and distinguish Memory from training and deletion controls.
Build Your Personal AI Profile
Combine the right settings, test realistic tasks, resolve conflicts and create a versioned profile you can maintain.
Module 3 Knowledge Check
Apply the personalisation layers to realistic style, instruction, memory and privacy scenarios.
Where Should This Preference Live?
Use the smallest personalisation layer that matches the scope.
| Layer | Best for | Example |
|---|---|---|
| Current prompt | One task or temporary exception | “For this answer, use a formal tone.” |
| Personality | General communication style | Friendly, Professional or Efficient. |
| Characteristics | Fine adjustments where available | More concise, warmer or fewer emojis. |
| Custom Instructions | Explicit preferences used broadly | Use NZ English and teach step by step. |
| Saved Memory | Useful details that may help future chats | A recurring learning preference you want remembered. |
| Project instructions | Rules and context for one project | Use the ITIAN template only inside the ITIAN project. |
Direct task instructions still matter
Personality, characteristics, Custom Instructions and memory work together, but a clear request in the current conversation can alter the response for that task. When instructions conflict, inspect the result and simplify the competing guidance.
Recommended: 16:9 • Captioned • Demonstration account only
Module Orientation Video
This video will compare the personalisation layers with the same sample prompt.
- Change personality and characteristics.
- Add one useful Custom Instruction.
- Review a saved memory safely.
- Show when a project instruction is better.
Module 3 Practical Challenge
Create and test your first versioned Personal AI Profile.
- Choose a base personality that suits most of your conversations.
- Write concise Custom Instructions for language, explanation style, organisation and important boundaries.
- Review saved memories and remove anything outdated, unnecessary or too sensitive.
- Decide which project-specific rules should stay out of global instructions.
- Test the profile with four tasks: an explanation, a technical checklist, a writing revision and a planning request.
- Record what worked, what conflicted and what you changed in version 1.1.
Evidence of completion: keep your approved instruction text, a memory-review checklist and a short test report without private values.
Official Sources and Further Reading
Reviewed against current first-party OpenAI guidance on 13 July 2026. Personalisation options can vary by plan, workspace, surface and gradual rollout.
OpenAI guidance used for Module 3
- Customizing Your ChatGPT Personality — base style, tone and interaction with other personalisation layers.
- Characteristics in ChatGPT — fine-tuning response style during gradual rollout.
- ChatGPT Custom Instructions — availability, settings and editing behaviour.
- Memory FAQ — saved memories, chat-history reference and management controls.
- Projects in ChatGPT — project instructions, files, chats and project memory choices.
- Data Controls FAQ — model-improvement preferences and account controls.
Ready to begin?
Start by mapping the personalisation layers before changing settings. That first lesson prevents global instructions, memories and project-specific rules from becoming a tangled collection.