In general, AI is as impactful as any technology can be. However, for those of us who like to enjoy the process (the act of gathering the info, structuring them and editing them), whether it be writing code, or writing in general, there's a joy in the act of building - straight up finished pieces of work handed over to you, robs you of this space to think and formulate your own views.
I do agree on the "existential crisis" part of it. At work, every time I see someone sign-off of something AI generated without much edits, I feel this fear that we're on to slippery slope where there's no turning back.
How do you do that post-mortem? I can agree, let's say if the problem is 'pick top k elements', and you already know what needs to be done, you can hone in on what went wrong. But in cases where you don't know how to come up with the steps, how do you go about that?
I do agree on the "existential crisis" part of it. At work, every time I see someone sign-off of something AI generated without much edits, I feel this fear that we're on to slippery slope where there's no turning back.