I've been really productive with Cursor for both work and personal projects. However, I've noticed that while I learn new things with AI, they don't seem to stick. I think this is because I'm not engaging deeply enough with the solutions myself. I want to avoid just getting answers from AI — I need a system where AI acts more like a mentor. It should provide feedback like, 'this works, but here's why it might not be the best approach' or simply affirming my work. I'm hoping to create a setup where I can focus on real learning and avoid getting lost in unnecessary tweaks. Has anyone developed rules for this? Or do you have good resources from cursor.directory that promote learning over mere answers?
1 Answer
That's a neat idea! I’ve been using Cursor for a while as well, but I come from a coding background of 30 years. One trick is to ask for help in smaller pieces instead of big answers. For instance, use cmd+k (if you’re on Mac) to get targeted help. You might want to try writing code first, then highlight it and ask Cursor if it's the best way. The key is to avoid letting the tool do the work for you, as it can slow your learning.
Thanks for the cmd+k tip—I’ll try that! I’m also working on SwiftUI and have found that AI isn’t always accurate when I’m unsure what I need, leading to bad code unless I’m super precise. I want AI to teach me, but it often gives full solutions instead. It's tricky to get the balance right!