Learning Technology Deeply - The Path to Becoming an Expert

Nowadays, learning has never been easier. With books, search engines, and LLMs, almost any piece of knowledge is just a few clicks away. However, access to information does not lead to true understanding. Many people spend years reading, watching lectures, and collecting resources, yet still find it difficult to solve unfamiliar problems or create something new.

This raises an important question: what does it actually mean to learn? If learning were simply the accumulation of facts, then people with the most information should always become the best thinkers. However, reality suggests otherwise. Experts are not distinguished merely by how much information they remember, but by how they organize knowledge, understand relationships, and apply principles to new situations.

Therefore, learning should be viewed not as a process of collecting more informations, but as a transformation of the way we think. A learner gradually moves from knowing what something is, to knowing how it works, and finally to understanding why it works. Through this process, external information becomes internal knowledge, practical ability, and eventually the capacity to create new ideas.

From Information to Mental Models

The first stage of learning is acquiring basic knowledge. This stage is often the most visible because beginners usually start by learning concrete facts: syntax, tools, frameworks, algorithms, or system components. A programmer may learn that a variable stores data, that a database uses indexes, or that an OS manage memory and processes.

At this stage, knowledge often exists as isolated pieces of information. A beginner may remember that a hash table provides fast lookup, but still not understand why it is efficient, what assumptions make this efficiency possible, or why another data structure may be a better choice in a different situation.

This distinction reflects a fundamental idea in cognitive science: human learning is not simply the storage of information. Instead, the brain gradually organizes information into structured mental models. These models allow us to explain how things work, predict outcomes, and connect new knowledge with previous experience.

This transformation happens in many areas of technology. Learning a ML framework is not the same as understanding ML. Memorizing API calls may help someone build a simple application, but understanding optimization, data representation, and model limitations allows engineers to design better solutions and adapt to new problems.

Therefore, the first stage of learning is necessary but incomplete. Informations provide the buidling blocks, but only when these pieces are connected into a structure can they become genuine knowledge.

From Mental Models to Practical Ability

Building a mental model is an important step, but it is still not the final goal of learning. A person may understand the concepts behind a system and still be unable to use the knowledge effectively in practice. The second transformation happens when knowledge becomes ability: when a learner moves from knowing what something is to knowing how to apply it.

This distinction has been discussed in philosophy as the difference between knowing-that and knowing-how, by Gilbert Ryle in The Concept of Mind. Knowing-that refers to possessing factual knowledge, while knowing-how refers to the ability to perform an action. These two forms of knowledge are closely related, but they are not interchangeable. A person can read hundreds of articals about swimming techniques and understand the physics of buoyancy, yet still be unable to swim without actual practice.

The same principle applies to engineering. A programmer may know that a mutex is used to prevent race condition, that memory management affects performance, or that database uses indexes to accelerate queries. However, real engineering ability requires much more than remembering these concepts. It requires experience with applying them in uncertain situations: identifying the cause of a bug, choosing an appropriate design, and understanding the trade-offs behind different solutions.

For example, a beginner learning C++ may first memorize that smart pointers help manage memory automatically. At this stage, smart pointers are simply another language feature. After writing larger programs, however, we encounters problems involving object lifetime, ownship, and resource management. We begin to understand why concepts such as RAII and ownship models exists. Through practice and failure, the information is transformed into pratical intuition.

This process explains why hands-on experience is essential in technical learning. Practice is not merely a way to apply existing knowledge; it is a mechanism for improving our mental models. When engineering debug complex systems, their understanding becomes more refined because they constantly compare their expectations with reality. Every failure reveals limitations in their current model and forces them to develop a more accurate one.

Therefore, the second stage of learning is the transition from passive understanding to active capability. Knowledge becomes valuable when it can guide decisions and actions in the real world. However, even practical ability has its limits.

From Practical Ability to Principles

Practical ability represents a significant improvement over simply possessing information. However, even experienced practitioners can still depend on patterns they have encountered before. A deeper level of learning begins when people move beyond applying existing techniques and start understanding the principles behind them.

This stage is about asking a more fundamental question: why does something work?

For example, a beginner may remember that arrays usually provide faster access than linked lists. An experienced programmer understands that this is not just because arrays use continuous memory, but because modern computer systems rely heavily on memory locality and cache behavior. The principle is not “always use arrays”; the principle is understanding how data organization interacts with the underlying hardware.

This ability to reason from principles is what allows experts to solve unfamiliar problems. When facing a new situation, they do not search for an identical example from memory. Instead, they analyze the fundamental constraints, identify relevant concepts, and construct a solution based on deeper understanding.

This idea can also be connected to the development of expertise. Experts often appear to have intuition, but this intuition is not mysterious. It is the result of years of experience transforming countless specific examples into abstract patterns. A senior engineer may quickly recognize that a system design will become difficult to maintain, not because they memorized a particular rule, but because they have developed a deeper understanding of coupling, abstraction, and complexity.

The journey from what to how and finally to why is therefore a journey from information to independence. It represents the gradual transformation of a beginner who follows instructions into an expert who can think, design, and create.