Table of contents

Intro

There are many reasons to go to university after high school: getting a better job (less true than what society wants us to believe), becoming an expert in a field we’re interested in, or just curiosity.

My reason was to not waste time, and to improve my critical thinking about the resources I find on the internet. There are a lot of people out there trying to make money out of ignorance, which I don’t approve of, but I understand.

Learning how to learn

I think the most important skill I got from my 3 years of bachelor + 2 years of master in computer science engineering is the ability to learn how to learn, and to tell the difference between high-quality material and bad stuff.

I started in the last years of high school, because I was sick of losing so much time on stuff I would forget the month after (I don’t know about you, but that was my constant feeling before university). So I read different books, tried different study methods and bought different courses, to understand how the human brain actually learns new things.

This field is full of people who just want to waste your time and money too, but there is one single book that made me really understand the key, which is Make it Stick: The Science of Successful Learning.

This book does something very different from the others: it doesn’t give you a specific study method. Instead, it makes you understand how the human brain actually works and retains information long-term.

Basically the book says the key to learning something lies in the effort of trying to retrieve it once you kinda forgot it, which at first I didn’t really understand. I wanted a specific method to follow for everything, the secret recipe to apply in every context so I could be a super-human.

This is a lie, a lie that malicious people use against you to manipulate you into buying what they say.

Here comes an important lesson in life: nothing comes for free, and shortcuts usually make the road to success longer.

So after reading this book several times I realized that all of us are different. Different fields require different study methods, and different people require different study methods too, so if we consider n the number of fields and m the number of different kinds of people, the number of study methods would have a complexity of O(m*n), which is too high.

A good strategy would be to give some principles, like that book does, and let you figure out what works for you in which context. And here we are: that’s exactly why I chose to go for university and used 5 years of my life to actually understand, in the field I love, how to learn different kinds of things in depth.

Obviously I decided to go for this career also to understand which part of computer science I prefer to go deep into: learning a bit of everything (BFS) to then go deeper in the exact part I like the most (DFS).

Effort

The most important key-word is effort, because a lot of the time we are tempted to do the things we feel safe with, like: reading and repeating, highlighting everything with multiple colors, reading everything from start to end multiple times (holy time losing).

All of these are not always bad, but they’re not really effective, because you stay in your comfort zone while doing them. That’s where the biggest enemy comes in, the illusion of learning: after reading a chapter 5 times in a row you feel like you know it, but once you get tested on it you magically don’t know anything, and you just lost a lot of time.

Professors are usually very good with this, giving mandatory exercises to students and making them test themselves. But once you’re in university it’s difficult to make them mandatory, and more than that, right now with LLMs this got far worse.

So it’s really important to test yourself and go out of your comfort zone, embracing effort with anything that can help this process of retrieving stuff:

• flashcards
• exercises
• reading a chapter and then describing it in your own way, without the book in front of you, and then finding the gaps you didn’t fill (I find this very effective, I always use it when I’m reading technical books of any kind)
• whatever kind of testing

AI

In my opinion, using AI tools to learn could be really beneficial, because you have the possibility to personalize your learning based on how you understand things. For example, when I want to learn a new complex topic I know nothing about, I like to start by asking a chatbot “how would you explain this thing to a 5 year old kid”, which usually helps me visualize it before.

On the other hand, these kind of tools can also be a shortcut, one I’ve fallen for myself, that makes the fall even harder when the illusion of learning ends.

Something I found out while using AI agents to make my flashcards is that for the questions it’s really important to give the right context, based on how you want to learn something. When I want to learn how to talk about a topic, I ask for questions that make me talk a bit more and give 2-3 examples. When instead I want specific definitions and direct facts, like “what does the CIA triad stand for?”, I make it understand what I want by giving it context and a few examples of the questions I have in mind; in prompt engineering this is called few-shot prompting. (Don’t ask for too many at once by the way, the quality drops fast).

But there is a line here I try not to cross. Deciding what I should be asked is part of the effort: it forces me to figure out what actually matters in a chapter, and that alone already teaches me something. So the agent writes the questions, but I’m the one deciding what they have to test, and I throw away the ones that don’t test anything real. The moment I let it decide what’s important for me, I’m back in the illusion of learning, just with nicer flashcards.

Conclusion

If it’s simple to learn, you’re probably doing something wrong. Embrace the effort, avoid the illusion of learning, try different things to understand what suits you the best, and most importantly avoid people that make use of your weaknesses to sell you magic stuff. Nothing is easy, and if it’s easy usually it’s not worth it.