So I'm currently using LLM-lite tools such as ChatGPT and simple copilot completions. These tools obviously provide utility. I have yet to play with more advanced tooling or models besides these. I'm currently experiencing absolutely no pressure at my job to "up my game," and my hobby programming is mostly done in very niche languages (contribute little to my resume whether or not I finish them).
Given this situation, I've decided to stop using LLM tooling altogether, with the assumptions that 1) the tooling in 2+ years will be completely different anyways so I'm not loosing by not familiarizing myself with the current "hot" tools, and 2) if I ever need to get up to speed with the latest LLM tooling, it is something that one should pretty easily get up to speed, especially with improved models (people have seemed to master the current state of affairs in matters of months, so in the future it should be even easier). I know some people will reply that I'm just wasting time when I could automate stuff, but in reply, I find coding tasks fun (I know, old fashioned) and I feel I otherwise have a good work-life-balance.
Do these assumptions hold up? Are there any perspectives I'm missing?