I've been learning to code for about two months and recently started experimenting with AI as a learning aid. I'm unsure how much I should rely on it. When I ask for broad explanations, it sometimes skips important details, makes mistakes, or jumps into advanced topics that aren't relevant to what I'm currently studying.
For example, I'm struggling with asynchronous programming. Instead of focusing on the event loop, promises, and the basic mechanics, the explanation quickly moved into software architecture. I'm not sure whether that information is useful background or simply too much too soon.
Would it be better to use AI for specific questions, debugging explanations, researching documentation, testing ideas, or explaining individual pieces of code? How can I use it without letting it replace the process of learning and solving problems myself?
4 Answers
Use AI as an interactive rubber duck or an error explainer, not as your entire curriculum. Ask focused questions such as what a particular function does, why an error occurs, or how one small example works. For asynchronous programming, stick to the event loop, promises, and a few small experiments before worrying about architecture. Official documentation and writing isolated examples will help you build a much stronger mental model.
Don’t let AI write your programs for you. Try solving problems first, then use it to clarify a confusing error, compare your reasoning with another explanation, or point out a mistake in code you already understand. You can also use regular searches and official documentation alongside it, since AI answers can sound confident while still being incomplete or wrong.
A practical middle ground is to use AI like a smart search tool: ask about a specific language feature, request a simpler example, or have it explain a line of code. Avoid vague prompts like “teach me programming” or relying on generated solutions you haven’t worked through yourself. Keep the scope narrow and make sure you can explain the result afterward.
There’s a good argument for minimizing AI at the beginning. Struggling through bugs and figuring things out independently is where a lot of programming knowledge comes from. If you do use it, treat the response as a suggestion that you need to verify, not as an authority. Later, it can be useful for reviewing code, spotting redundant sections, or discussing possible performance improvements.
So it’s better to ask about something specific or use it to test my understanding, rather than asking for a complete explanation of a large topic?