How I Keep Learning Without Forgetting Everything

Whenever I learn something useful, I try to capture it.

Sometimes I write it as a blog post. Sometimes I publish it on GitHub Pages. Sometimes I keep it in my private knowledge base — what people often call a “second brain”.

For my personal knowledge base, I use Logseq. It helps me connect ideas, notes, experiments, commands, references, and reflections together.

Writing things down forces me to clarify my thinking. If I cannot explain something clearly, it usually means I have not fully understood it yet.

A blog is not only for others. It is also a conversation with my future self.

Some skills are easy to forget if we only use them once.

For skills that matter, I practise them repeatedly using spaced repetition. I revisit them after a few days, then after a few weeks, and later after a few months.

Once a skill is deeply practised and used in real situations, it becomes hard to forget.

This is especially true for practical skills such as Linux commands, Vim shortcuts, shell scripting, programming patterns, debugging techniques, and architecture decision-making.

The goal is not to memorise everything. The goal is to build strong mental muscle memory.

I do not usually choose a technology only because it is popular.

Before selecting an option, I prefer to do some due diligence. I read the documentation, compare alternatives, test small examples, and play around with the tool for a while.

This helps me understand not only the strengths of a technology, but also its limitations.

Many tools look great in a demo. But after using them in a real scenario, you start to see the hidden complexity: deployment issues, operational cost, learning curve, ecosystem maturity, community support, security concerns, and long-term maintainability.

That is why I believe technical decisions should be based on both research and hands-on experience.

One of my strongest learning principles is this:

Learn skills that can be stacked.

Some skills give you a return only in one narrow area. Other skills become foundations for many future skills.

For example, Linux, Vim, Python, shell scripting, programming fundamentals are highly stackable skills.

Linux helps you understand servers, containers, DevOps, cloud platforms, and production systems.

Vim improves editing speed and helps you work comfortably in remote environments.

Python can be used for scripting, automation, data science, AI, testing, and backend development.

Data engineering and big data knowledge make it easier to understand modern AI systems, because AI depends heavily on data pipelines, storage, processing, and governance.

When skills can be stacked, every new skill becomes easier to learn because it connects with something you already know.

Many people focus only on learning more.

I try to focus on building a learning system.

My system has four parts:

This system helps me avoid starting from zero every time.

Over the years, my notes, blogs, code snippets, experiments, and technical decisions have become a personal knowledge asset.

The real secret is not talent.

The secret is accumulation.

Write things down. Practise what matters. Test before choosing. Learn stackable skills.

Over time, these small habits compound.

And when knowledge compounds, learning new things becomes faster, deeper, and much harder to forget.