I’ve been coding with AI for about 6 months, and I think my coding skills have started to atrophy. Or is this a placebo effect?
Yes, as AI coding technologies have developed, people’s coding proficiency has declined. They have become increasingly reliant on AI auto-generating code instead of relying on their own memories. As a result, there is less and less code that can be memorized. Thus causing a vicious cycle.
imho it depends if you review and correct the suggested code or hurry up and push-push-push to production whatever it suggests . in my case I rarely get correct suggestions so I re-read it and argue with it in my mind , then sometimes use a bit of that and sometimes not. and usually when the suggestion is good, I’m like “wow nailed it! no freakin way!” ahah
I felt like I’m getting more practice at design patterns such that I can recognize and ask the LLM to refactor this in this design pattern or code principle.
If not, I’d ask which design pattern would be perfect in this situation and I get to learn more about it.
It’s also useful for writing tests
The exact opposite. I can forget about syntax bullshit and completely focus on high-level design. If by ‘coding skills’ you mean my ability to spot an errant semicolon, maybe it has decreased.
But if you mean my ability to write clean code fast, it has improved by a LOT, and quickly.
I been having fun creating mini apps and getting the code and pasting it in. I rarely have been adding code myself and therefore no longer taking advantage of copilot/cursor ++ auto complete
It is just so much more fun to me having a code money and focusing on the bigger picture