After completing HTML and CSS, I moved on to JavaScript. The exercises are more difficult, and sometimes I can’t find the solution or idea by searching Google or the forum. Then I notice that I’m using AI more and more to get tips or have the syntax explained in more detail. I think the latter is perfectly fine because it helps me understand some things better. But when I get completely stuck, I use the generated code. I try to understand it, but I haven’t come up with the solution myself. That feels wrong because I didn’t come up with the solution. I pass the lab, but I don’t know if the learning curve is increasing.
How do you do it? How do you use AI for learning, or do you try to avoid it completely?
It’s not about AI, it’s about us, we naturally crave comfort, ease… No matter how much you’ll try to “use AI for learning“ you’ll kep falling into the same cycle:; get some info from it, feel satisfied, feel like you understood the concept, generate the code and swear the next time you get the exact challenge, you’ll pass on your own… next time, same cycle.
Avoid it if you’re learning, I’d rather peruse hundreds of documentations and articles, press the proffessionals in the forum so hard with my challenges , and write my own code.
AI gives the illusion of competence. No learning loop, just dopamine loop.
I think a good use of AI is when you have already solved a lab by yourself and want to know if you could have done it better / more efficiently.
I do it sometimes on the daily challenge and ask AI to refactor my (working) code. This way I have learned to use a few functions and modules I didn’t know about.
I’ve gone through similar learning phases with JavaScript. I see AI as a useful debugging tool rather than a full solution provider. I ask it to clarify concepts or find mistakes, but I always write the solutions myself. When using AI-generated code, I rewrite it to make sure I understand the logic. Using AI isn’t cheating if you’re genuinely learning; it’s how real developers work too. Thanks for mentioning this issue—many new learners struggle with it!
I mainly use AI to help me out, not to take shortcuts. When I get stuck, I ask AI for hints or explanations, but I still try to come up with the final answer on my own. It’s fine to use code generated by AI as long as I break it down and understand how it works—this is where I really learn. Over time, I’ve seen my problem-solving skills get better because I start to recognize patterns instead of just looking for the answers.
I would use a real human who can trust to be accurate. Human contact isn’t something to avoid and you do not have to be afraid or ashamed to ask questions to a human.
I absolutely would not let the LLM do any of the work for you. Learning how to do the ‘last 10%’ is hard and takes practice.
Save the LLMs for tasks you understand well and want to accelerate.
I get what you mean. It’s easy to fall into the habit of letting AI solve things for you, and then you feel like you didn’t really “earn” the answer.
What helped me was using AI only to explain concepts or point me in the right direction, not to write the full solution. If I’m stuck, I ask for a hint or an explanation of the logic, then try to write the code myself. That way I still struggle a bit, which is where the real learning happens.
And yeah, tools like MagicPost help me with writing, but for coding I try to keep AI in a “mentor” role, not a “do it for me” role.
You’re not alone in this. The key is finding the balance that still lets you learn.