Ask HN: How can I grow as an engineer in the AI era

I'm a junior developer, with less than 1 yoe. I have to use AI at work because deadlines are tight, and if I try to solve something by myself, it would take too much time. Work has to be done like yesterday. I'm trying not to use AI when I fix some bugs because they don't have such tight deadlines, but I feel so behind when it takes too long.

As you can probably guess, AI is destroying me on different levels. It often creates much better solutions than I do because I don't know this large codebase well. It can write code without forgetting to add something somewhere. But sometimes I can see that the suggested solution is bad, and I argue with a soulless machine about it.

I don't know much. My solutions are too straightforward and stupid. I can't make architectural and design decisions, or the ones I make are simply bad. I can't sit down and describe how I'm going to solve something because I don't know what might come up when I start. And I'm trying not to feed slop to PR reviewers, and I truly learn something when they suggest something to me.

I want to be a good engineer, and I thought that if I started working, it would come to me naturally. But now I'm in this LLM loop where everyone is expected to use AI, yet it gives me nothing in terms of real experience, and people don't want to teach anyone anymore.

I don't know what to do. I want to become a good engineer even if this profession won't exist, and now my life and my family depend on my income, so I can't just decide "It's not for me".

What should I do? Spend my free time trying to understand the codebase without AI? Try to create something myself without AI? Improve my problem-solving skills, learn new things from books, practice implementation?