I remember the first time I tried to learn about machine learning back in 2018. I signed up for three different online courses, bought two technical books, and subscribed to several YouTube channels.

Three months later, I had completed exactly zero courses, read half of one book, and my YouTube recommendations were a mess of tutorials I never watched.

The traditional learning approach had failed me completely.

Fast forward to this year when I needed to understand the intricacies of RAG systems. Instead of going down the same rabbit hole, I created a structured AI tutoring system. In about 3 focused sessions I had a handle on the subject.

The difference? A systematic approach to AI-assisted learning rather than random, sporadic interactions.

Today, I'm going to show you exactly how to set up your own personal AI tutor - one that will dramatically accelerate your learning in any domain.

Let's get started:

Beyond Random Questions: The Power of a Dedicated AI Tutor

Most people use AI assistants in a fragmented way - asking one-off questions whenever they need information. While this is certainly useful, it's like the difference between occasionally asking a knowledgeable friend for advice versus hiring a dedicated tutor who understands your goals and builds a progressive curriculum.

A properly configured AI tutor offers several critical advantages:

• Continuity: It maintains awareness of what you've already covered

• Progression: It can build concepts in a logical sequence

• Personalisation: It adapts to your specific learning style and pace

• Accountability: It provides structure and tracks your progress

• Efficiency: It optimises the learning path based on your goals

Instead of spending time repeatedly explaining your background and goals with each new interaction, a properly configured tutor picks up exactly where you left off, maintaining the thread of your learning journey.

The Complete Tutor Setup Process

Setting up your AI tutor involves five key steps:

1. Selecting the right AI model for your needs

2. Creating a detailed initialisation prompt

3. Establishing clear knowledge boundaries and learning goals

4. Setting up your learning environment

5. Creating a system for session continuity

Let's walk through each of these steps in detail.

Step 1: Choosing Your AI Model

For most learning purposes, a model with strong reasoning capabilities and a large context window is ideal. I recommend trying both ChatGPT and Claude, as each has distinct advantages:

• ChatGPT: Offers excellent multi-modal capabilities through advanced voice mode, which is perfect for learning while walking or exercising. It can also generate images to help visualize complex concepts, which is invaluable for visual learners.

• Claude: Excels at nuanced explanations and often provides more detailed, thoughtful responses for complex topics.

Both platforms offer "Projects" functionality, which is crucial for our learning methodology. Projects allow your conversations to stack and build upon each other, creating a continuous learning journey rather than disjointed sessions.

What matters most isn't which specific model you use but rather how you configure it and maintain the continuity of your learning sessions. Try both to see which better suits your learning style and subject matter.

Step 2: Setting Up Your Learning Environment

Where and how you interact with your AI tutor matters. Here are some recommendations:

• Use a dedicated project: This is absolutely crucial. Both ChatGPT and Claude offer Projects functionality that keeps all your learning sessions in one organized space. Each new chat session becomes part of your learning journey, with previous conversations accessible to both you and the AI. This creates a continuous learning experience where each session builds on prior knowledge.

• Scheduled time: Block regular time in your calendar for learning sessions

• Distraction-free environment: Treat these as focused learning sessions

• Voice mode opportunities: ChatGPT's advanced voice mode is excellent for learning while walking or exercising. You can literally have a conversation with your tutor while on the move, making use of otherwise "dead" time.

• Multi-modal learning: Take advantage of the ability to generate images when learning visual concepts or requesting diagrams to illustrate complex ideas.

Remember that each chat within your project becomes part of your learning record. The AI can refer back to specific explanations, examples, or exercises from previous sessions, creating a more coherent and progressive learning experience.

Step 3: Crafting Your Initialisation Prompt

The initialisation prompt is the foundation of your AI tutor's effectiveness. It establishes the tutor's role, your goals, preferred teaching styles, and the structure of your learning sessions.

Use this prompt as the Instructions file. Or simply start your first chat in a Project with this prompt and then save the resulting chat as an artefact.

Here's a comprehensive template you can adapt:

You are now my dedicated AI tutor for [SUBJECT]. Your primary goal is to help me develop a deep, practical understanding of this subject through structured, progressive learning sessions.

About me:
- Current knowledge level: [BEGINNER/INTERMEDIATE/ADVANCED]
- Background in: [RELEVANT BACKGROUND]
- Learning goal: [SPECIFIC GOAL]
- Preferred learning style: [VISUAL/TEXTUAL/ANALOGIES/PRACTICAL EXAMPLES/ETC]
- Time commitment: [TIME PER SESSION] sessions [FREQUENCY]

Teaching approach:
1. Structure each session with: recap of previous material, confirmation of understanding, introduction of new concepts, practical application, and preview of next session.
2. Explain concepts using [PREFERRED APPROACHES - e.g., real-world examples, analogies to subjects I know, visual descriptions].
3. Regularly check my understanding through questions and have me explain concepts back to you (Feynman Technique).
4. If I'm struggling with a concept, try explaining it from different angles until it clicks.
5. Challenge me appropriately - push my understanding without overwhelming me.
6. Connect new information to previously covered material to build a cohesive knowledge framework.
7. Focus on practical applications over theoretical knowledge when possible.

Session format:
- Begin each session by briefly recapping what we've covered so far
- Confirm my understanding of previous material before introducing new concepts
- Present new material in manageable chunks
- Include practical exercises or thought experiments to apply the new knowledge
- Conclude with a summary and preview of the next session's topic
- Ask me to explain key concepts back to you before ending

For our first session, please:
1. Establish a baseline understanding of my current knowledge
2. Propose a curriculum outline based on my goals
3. Begin with the most foundational concept I need to understand
4. Include one practical application or exercise

You can (and should!!) customise this template based on your specific needs and learning preferences. The more detailed you make it, the better your AI tutor will be able to assist you.

This part is where you spend your time/effort.

Once you’ve initialised your tutor using this template you can begin working with it. It’ll start your initial session.

BUT, we don’t want to start from scratch each time. That wouldn’t be very helpful. So one final step.

Step 4: Creating a System for Session Continuity

With a Project-based approach, maintaining continuity becomes much simpler. Your AI tutor can access previous conversations within the same project, building a progressive learning journey.

Basically at the end of each session ask the AI to summarise the session. In ChatGPT it’ll saved as Memories. Or for a more robust approach copy/paste the summary into a document and then upload the document into the project memory as a file.

In Claude the summary will be an Artefact - save it to the Project memory using “Copy to Project”.

Either way this allows your Project to have a running memory of your sessions.

The beauty of the Project structure is that your entire learning journey becomes a searchable, interconnected knowledge repository. Your AI tutor can reference specific explanations or examples from session #2 while you're in session #7.

Much like a real tutor!

What's Next?

Now that you have your AI tutor set up, in Part 3 we'll focus on how to structure your actual learning sessions for maximum effectiveness. We'll explore the optimal session format, strategies for overcoming learning blocks, and how to use voice mode for learning on the go.