I've found AI to be a mixed bag while learning to code. Initially, it was really helpful, but I often spent ages fixing its errors. I've started using prompts like "explain this line" and running the code in sandbox editors, which lets me understand the output better. This approach really boosted my learning! I'm curious, what tools or techniques do you guys use to make AI more of a tutor rather than just a code generator?
4 Answers
I’ve got this rule where I only use AI for coding tasks I could do myself, but it would take ages. Like, I’ll use it to generate boilerplate code or figure out where an error is coming from. If you don’t really understand the code you’re working with, spotting mistakes is tough and that’s not a great way to learn. It's all about the trial and error process, really.
Good prompts are key! I tell AI, "You’re my tutor, and I’m a programming student. I want to learn how this works, so don’t do the work for me. Show me how to write (whatever)." It’s helped me a lot!
I hardly ever ask AI to actually generate code for me; instead, I ask it to explain concepts. For example, I might say, "What is a coroutine in C?" I see AI more like a quick-access library that can give me summaries. Google’s starting to do that too, which is neat!
I think of using AI to take some of the cognitive load off. It can suggest exercises or help explain why things aren’t working. But if you let AI write your code for you, you’re missing out on the effort that helps you learn! Remember, putting in the work is what makes the learning stick.
Absolutely! Embracing the struggle is part of the process. It’s tough now, but all that effort pays off later when you really get it.