TL;DR:
- Learning gets easier when you find the right perspective, make things concrete, and work through them step by step.
- Understanding comes from asking why each step works, not only from applying formulas or memorizing facts.
- AI can shorten feedback loops and give different perspectives on concepts, which helps you understand them.
- If you want to learn something, you have to do it yourself.
Start with the right perspective
When I try to understand something, I first search for the perspective that helps me understand the topic. That perspective can vary from person to person, because everyone brings different assumptions and prior knowledge to a new subject. It usually helps to make things concrete. How does a concept look when applied? What is a simple example?
Learn by solving
Getting to a solution usually involves several steps, and each step often contains smaller components. I like to think of it as a tree:
Problem
├── Step 1
│ ├── Component A
│ └── Component B
├── Step 2
│ ├── Component C
│ └── Component D
└── Step 3
└── Component EBy trying to reach the solution yourself, you find out which components are actually needed and whether you can apply them. Working through a topic step by step lets you spend enough time on each part. The more topics you understand this way, the easier it gets to learn new ones, because at a certain level of granularity many concepts repeat. Mathematical basics, for example, show up across engineering and physics. Matrix operations matter in quantum computing as well as in closed-loop control. Over time, you build a broadly useful set of rules that helps you solve problems.
Ask why it works
At university, applying concepts is often enough to pass exams at the beginning. Later on, it helps much more to understand why things work the way they do. For each step in a solution, ask why it is the right step, which assumptions it depends on, and which concept it is based on. If you learn for an exam, this sometimes overlaps with memorizing things. That is especially true for natural phenomena that we ultimately have to accept. In those cases, it can make more sense to focus on what we can actually control.
Use AI as feedback
At all of these levels, the latest AI models can help. They give each student access to a wide range of different perspectives on the things they need to learn. If you explain a concept in your own words and get feedback from AI, the learning loop becomes much shorter because you can immediately test your understanding and ask follow-up questions.
But if you want to learn something, you still have to do it yourself. If you need mathematical basics to apply them in an exam, letting AI solve equations for you does not help. But if you are trying to understand how semiconductors work and do not yet understand pn junctions, AI can help by explaining a prerequisite concept, such as electric fields. You may already have learned that in another class. If not, learning it now will help in many other areas, because the basics often stay the same.