I feel like the fundamentals of debugging sort of transcend the details of programming.
At least for me, I find the art of debugging is often about isolating variables and learning about a system. I think it's the most fun when you can tell that the problem is approachable, but hairy. Like, don't task me to debug bank software written in COBOL and enjoy it, but give me something reasonable and I'm going to methodically work my way through the problem by isolating parts of the system and testing them for variance.
So when we have this AI buddy helping write, debug, whatever... It still presents a system with variables that you can presumably isolate for success. If our AI partnership is to be successful, then there should be a capacity to either work with them or on them to isolate variables and solve problems. How this is done needs to evolve, but I really feel like the process is somewhat fundamental to any of these systems.
I feel like the fundamentals of debugging sort of transcend the details of programming.
At least for me, I find the art of debugging is often about isolating variables and learning about a system. I think it's the most fun when you can tell that the problem is approachable, but hairy. Like, don't task me to debug bank software written in COBOL and enjoy it, but give me something reasonable and I'm going to methodically work my way through the problem by isolating parts of the system and testing them for variance.
So when we have this AI buddy helping write, debug, whatever... It still presents a system with variables that you can presumably isolate for success. If our AI partnership is to be successful, then there should be a capacity to either work with them or on them to isolate variables and solve problems. How this is done needs to evolve, but I really feel like the process is somewhat fundamental to any of these systems.
Agreed. Good points!