People using GitHub Copilot are learning how to code while they review and tweak the AI suggestions. As an outcome, they are becoming better coders in the process.
At this time, I would not agree. That's because AI is literally knocking at the door for many use-cases and we haven't evolved with it yet(especially education). That will likely take another decade at this rate.
I believe the same reliance on Stack Overflow or Google for the most basic answers make people lazy and more impatient. This cheats people out of the process where you typically have these insights for wisdom. The AI also will not get "better" because it's capped at our collective understanding. Which fewer and fewer people will even contribute towards due to new AI-assisted habits.
Only time will tell if it makes people "better". There's no doubt it will be helpful and useful in many settings, especially in education. But this over-reliance makes you only as good as your tool. Which is what my original comment was trying to say.
Hmm if i was to make an argument to "disagree" which I feel reasonably compelled to do - using GitHub copilot to generate (somewhat working) code and reviewing it is akin to watching videos about learning to play the violin. Great for familiarity and knowing where to find said videos, terrible for competence, mastery or originality!
> terrible for competence, mastery or originality!
Would you say the same for finding a solution to your problem in StackOverflow? My argument is that you will always learn from adapting to your situation. And on copilot I think it's the same thing, but now you are finding your solution way faster.
It's not like in the videos example case where you are passively watching something. No, you read the suggested code, execute your code, look for bugs, test it etc. Then you move on to the next step of your coding task or whatever.
Yes I agree with you there. But just the act of consuming (which it seems like what the OP was trying to do without actively engaging in writing/creating on a regular basis) wont make you (imo) better unless you go and build something with it. Reading the dragon book (or SO explanations) is fantastic but if you want to be a PL expert building a toy compiler (and a few iterations) can do wonders!
I believe the same reliance on Stack Overflow or Google for the most basic answers make people lazy and more impatient. This cheats people out of the process where you typically have these insights for wisdom. The AI also will not get "better" because it's capped at our collective understanding. Which fewer and fewer people will even contribute towards due to new AI-assisted habits.
Only time will tell if it makes people "better". There's no doubt it will be helpful and useful in many settings, especially in education. But this over-reliance makes you only as good as your tool. Which is what my original comment was trying to say.