Ask HN: Improving as mid-level dev with forced use of LLMs

I’m interested in how do you grow as a developer when the usage of LLMs is basically forced on you?

Background: 8-ish years of computational physics & chemistry research (scipy) and switched over to industry backend dev (.Net) 2 years ago. Goal is to grow in backend dev.

So far I’ve been resisting them as much as possible in favor of old school rtfm, docs, and ddg/so, so I’d actually learn, but the org is insisting on us doing as much as possible copilot stuff (with tracking usage metrics), disregarding any learning of how stuff works angle.

Asking around, it seems that the interviews for other jobs (and jobs themselves) are also inreasingly LLM use focused.

Any suggestions?