Introduction

In Part 1, we explored how curiosity is a muscle that requires exercise, and how every major change in how we process information from writing to AI has faced the same criticism: that it’ll make us more stupid. If history is anything to go by we’re probably wrong about AI too, but only if we use it strategically and intentionally.

This post is about how to do that. Specifically, how to combine AI with evidence-based learning strategies to amplify your curiosity rather than atrophy it.

The Challenge: Cybersecurity Training Isn’t Built for Everyone

Here’s my personal frustration: I’m primarily an audio and hands-on learner. I learn best when I can listen to concepts explained conversationally, then immediately apply them in a practical exercise. This is how the military trains, its called on the job training, someone explains to me how to do something and I do it, they watch me do it and correct me as I go.

But Cybersecurity, for the most part, is not built for this.

We utilise courses, certifications or tabletops exercises, anything but on the job training. My issues is that most courses/certs are:

  • Heavily text-based (long documentation, PDFs, written labs)
  • Video lectures that can’t adapt to my pace or answer my direct questions
  • Linear progressions that assume everyone learns the same way
  • Mostly limited in interactivity

I’ve spent years fighting my own learning style, trying to force myself to learn from formats that don’t work for me. And I know I’m not alone. Everyone learns differently. Some people love reading technical documentation. Others need visual diagrams. Some need to break things repeatedly until they understand how they work.

The frustration is real. I have spent years knowing I need to learn something, wanting to learn it, but struggling to learn it because the format doesn’t match my brain.

This is where AI has genuinely changed things for me.

The Learning Cycle Framework

A colleague recently shared this video with me: How to Study for Exams - An Evidence-Based Masterclass

He had also made a flow chart to summarize the very long video, which was too good for me not to share. It breaks learning into three phases that you cycle through continuously:

┌─────────────────────────────────────────────────────────┐
│                  THE LEARNING CYCLE                     │
│  ┌──────────────┐    ┌──────────────┐    ┌──────────┐   │
│  │ Phase 1:     │    │ Phase 2:     │    │ Phase 3: │   │
│  │ Understanding│───▶│ Remembering  │───▶│ Focusing │   │
│  │              │    │              │    │          │   │
│  └──────┬───────┘    └──────────────┘    └────┬─────┘   │
│         │                                     │         │
│         └──────────────────◀──────────────────┘         │
└─────────────────────────────────────────────────────────┘

Phase 1: Understanding
├── Step 1: Scoping the Subject
│   ├── Examine the subject overall
│   ├── Subcategorise the different topics
│   └── Build a macro view (tree view)
│       ├── Trunk = Subject
│       ├── Branch = Sub-topic
│       └── Leaves = Individual Points
├── Step 2: Can you reason from first principles?
├── Step 3: Can you explain this concept to a 5 year old?
│   ├── Keep the language simple
│   ├── Minimize technical terms
│   └── Drill down to the concept in few simple words
└── Step 4: Ask yourself, does this make sense?
    └── How did we get from Point A to Point B? Ask WHY

Phase 2: Remembering
├── Active Recall
│   ├── Step away from the information source
│   └── Ask: What have I just read? Can I phrase in my own words?
├── Note Taking System
│   ├── Cornell Note Taking Method
│   ├── Handwriting vs Typing
│   │   ├── Handwriting: Slower, forces focus on key concepts
│   │   └── Typing: Faster, but danger of transcribing vs recalling
│   └── Take notes beyond immediate study material
├── Spaced Repetition
│   ├── Increase repetition = longer retention
│   ├── The more effortful learning feels, the more you retain
│   └── The retrospective revision timetable
│       ├── Tackle most difficult first
│       ├── Organise by topic
│       └── Colour code by difficulty
├── Interleaved Practise
│   ├── Split time within one study session
│   ├── Mix up practise topics
│   └── Avoid focusing on one skill in isolation
├── Things to Avoid
│   ├── Re-reading notes passively
│   └── Highlighting in isolation
├── Flashcards
│   ├── Minimise amount made
│   ├── Pass through content with Cornell method first
│   └── Flashcard only points you struggle with
└── Memory Techniques
    ├── Mnemonics (elaborative encoding & imagery)
    └── Memory Palace (associate facts with familiar places)

Phase 3: Focusing
├── System-based vs Motivation-based
│   └── Motivation is required for things we don't want to do
│       ├── Punishment
│       └── Reward
├── Pomodoro Technique
│   ├── Work for 25 minutes
│   ├── Break for 5 minutes
│   └── After 3-4 sessions, take 30-60 min break
├── Work/Life Balance
└── Studying Together
    ├── Give everyone time to think
    └── Discuss answers together

This is a cycle. You don’t just understand something once, then memorise it, then you’re done. You continuously move through these phases as you deepen your knowledge.

When you’re first learning about, say, Kerberoasting, you need to understand what it is and how it fits into the broader landscape of Active Directory attacks (Phase 1). Then you need to practise detecting it, quiz yourself on the indicators, and drill the response process (Phase 2). And you need to manage your focus and motivation, taking breaks as needed (Phase 3).

But once you think you’ve got it, you cycle back: Can you teach it to someone else (Phase 1)? Can you still detect it a week later (Phase 2) without looking at your notes (Phase 3)?

This framework clicked for me because it acknowledges that different strategies work for different phases. You don’t need one “best” study method. You need to recognise which phase you’re in and apply the right techniques for that phase.

The School Problem: Blocking vs. Teaching

My father is a school teacher. We’ve had many discussions about how students are using AI and how schools are responding.

The typical response? Block it outright. Ban ChatGPT from school networks. Use AI detection tools to catch students who might be using it. Threaten academic integrity violations.

And I understand the impulse. Teachers are seeing students submit AI-generated essays they clearly didn’t write and that is cheating.

But here’s the thing: This is the same response schools had to calculators. To Wikipedia. To Google. The fear is always the same, students will use X tool to avoid learning, so we must ban that tool.

But banning never works. Students still have phones. They still have home internet. You’re not preventing access, you’re just stopping students from finding effective ways they can utilise this tool, and even worse, refusing to teach them how to safely utilise it.

What if instead we taught students when and how to use AI? What if we helped them recognise the difference between using AI as a crutch (copy-paste without understanding) vs. as an amplifier (using it to explore concepts more deeply)?

What if we integrated AI into the learning cycle intentionally?

The Critical Foundation: You Still Need to Read

Before we go further, I need to be extremely clear about something:

AI is not a replacement for engaging with the actual content.

You cannot skip reading the course material, throw it into an AI, and expect to learn effectively. You cannot rely on AI summaries as a substitute for understanding the primary sources.

Here’s why: If you don’t have a fundamental understanding of what a topic covers, you won’t know what to focus on. You won’t know what questions to ask. You won’t recognise when the AI is wrong or hallucinating.

It’s like trying to use Google without clicking on any of the results. The tool can only amplify what you bring to it.

My workflow always starts with going through the course material first. Even if it’s just skim reading or watching the videos at 2x speed whilst taking rough notes. I need that baseline understanding before AI becomes useful.

AI is a study tool for reinforcement, not a substitute for the classroom. If you try to skip the foundation and jump straight to AI-generated summaries and quizzes, you’ll end up exactly where the critics discussed in part one fear: appearing knowledgeable whilst understanding nothing.

How AI Fits Into This

Once you understand the learning cycle and have that foundational knowledge, AI becomes incredibly powerful. AI can help you apply these evidence-based techniques in ways that were previously impractical, expensive, or impossible.

Want to practise active recall? AI can generate custom quizzes focused on your specific weak areas.

Want to practise teaching (one of the best ways to solidify understanding)? Try explaining the topic to an AI role-playing as a clueless student who asks questions when your explanation isn’t clear.

Want interleaved practise? AI can mix up topics and question styles on the fly.

Want audio learning? AI can generate conversational explanations of complex topics.

The difference is deliberateness. You’re not asking AI for the answer, you’re using AI as a personal tutor, applying the learning principles which work best for you.

Next, we’ll walk through a specific workflow I’ve been using with NotebookLM that integrates all three phases of the learning cycle. It’s not the only way to do this, it’s one approach that works for me, but it’ll give you a starting point.

Final Thoughts

AI is a resource we haven’t fully figured out how to use yet. The concerns are valid - passive use of AI absolutely will make you less capable. But intentional use, combined with evidence-based learning strategies, can make you more capable than ever before.

The key is understanding how learning actually works. Understanding the cycle. Understanding that different phases need different techniques. Understanding that curiosity and active engagement are non-negotiable.

Tools don’t learn for you. They never have. Writing didn’t memorise things for Socrates’ students, it freed them from having to memorise everything so they could think about harder problems. In the same way AI won’t understand things for you, but it can help you develop your own understanding in more effective ways.

In Part 3, we’ll get practical. I’ll walk through my specific workflow using NotebookLM, including example prompts, techniques for each phase of the learning cycle, and how to adapt this to your own learning style.

Stay curious, keep learning, question everything.