Hi everyone,
I’m currently developing apps with Cursor, and while it’s been great for quickly getting functional code, there’s something that’s been bothering me.
Sometimes, the AI introduces tools, libraries, or coding patterns I’ve never seen before. My required features might be running without bugs, but I often don’t fully understand what some of the code is doing.
On one hand, trying to understand every new piece of code or tool can be very time-consuming and could slow down my project. On the other hand, I don’t feel comfortable leaving things in my codebase that I barely understand — I like to know exactly what’s going on.
I’m torn between:
-
Shipping quickly and accepting some “black boxes” for now
-
Slowing down to fully learn and understand everything before moving forward
How do you balance speed and learning when using Cursor (or AI coding tools in general)?
Do you just trust the generated code until later? Or do you make sure to understand everything before you move on?
Would love to hear your workflows, mindsets, or any practical tips.
Thanks!