Part IV — Meta-skills
How to actually learn. The skills underneath all the other skills — practice strategies, AI as a learning accelerator, asking good questions, escaping the tutorial trap.
Part IV — Meta-skills
The skills underneath all the other skills.
You can finish Part I and have built five real projects, and still be stuck. Not because you missed a topic, but because the process you used to learn the topics was wrong — passive reading instead of active building, framework hopping instead of depth, asking ChatGPT for answers instead of explanations.
This part is the meta-curriculum. Read it whenever you're stuck — and read it once preemptively, before you're stuck, so the patterns are loaded when you need them.
What's in this part
- How to actually learn — Active recall, spaced repetition, build > read, the "feels easy" trap.
- AI as a learner — When to use ChatGPT/Claude/Cursor to accelerate learning vs. when they're stealing the reps you need.
- Asking good questions — Stack Overflow, Discord, your senior engineer. The reproducible-example skill.
- Escaping the tutorial trap — The most common stall pattern: finishing tutorials forever and never building. The exit ramp.
The single highest-leverage decision a beginner can make is reading Part IV early. Not after they're stuck — before. Most of "I tried to learn to code and gave up" stories are meta-skill problems, not technical-skill problems.