Grieving the loss of details

I am similarly overwhelmed with a lot of these emotions. There are a lot of different opinions, and it's hard to know what's right and wrong, but one thing I do know is that I don't enjoy it, not one bit. Last year was the first time I've considered leaving this field

I feel for the author. I feel similar about having to implement something to understand it (maybe not as obsessively) and enjoy the lower-level stuff. It especially hurts if you are getting/want to get paid for this kind of work and it's no longer possible.

Planning for the future is extra hard right now. For myself, I've considered switching careers but no idea what kind of work I could even do. I feel like the future of all white collar type work is currently uncertain.

I feel this. FWIW, I think there is still room for people who obsess over understanding details right now, but I can't claim to know what the future looks like or how economically viable that path will continue to be. Yes, if you point an LLM at a specific problem it will often solve it. But they also introduce those problems, and someone has to know enough to know where to point them.

I work mostly on optimization. Currently I'm surrounded at work by people who are rushing to do more than they can understand with the help of LLMs, and stubbornly refuse to collaborate with anyone else because they have so much faith that the LLM will handle it. In practice, it results in a lot of slop and a reduction of quality. For my own part, I've experimented with them but I still find myself mostly writing code by hand, and finding other uses for LLMs, like as a reviewer or in ideation. I still understand everything I do. I think this is the right balance for me because I'm expected to make things better, not produce a volume of features.

I don't know how long I'll be able to keep this up but I feel pretty good about my current position. If I were laid off, I'm not sure I could convince someone to give me a similar role.