[mahoho]

Abstraction works because you are able to treat an abstraction as a black box and concern yourself only with its input and output. A segment of code written by an LLM is qualitatively a very different thing; it's more like an open box of crap that you have to inspect and put together yourself, which requires knowledge of the contents, which requires experiencing the drudgery.

[shadowgovt]

I rarely encounter abstractions in the wild that are as nicely "sealed" as the definition implies. Looking at an open box of crap and understanding why it's doing something other than what the author (or you) intend is a valuable skill.

(No idea if this new model of "Ask ChatGPT or Copilot to synthesize a solution and then tune that solution" provides a solid opportunity to improve that skill yet, however).