alt Hacker NewsThere are at least different kinds of messes, and the new kind may offer different challenges which may eventually require different (easier? harder? invented-yet?) techniques than the ones we're used-to.
The problem I'm already having with LLMs (in miniature) is code/comments/docs where I don't know whether it's someone's intent or whether its un-reviewed LLM babble. That creates a continual stumbling block for figuring out whether the thing should be the way it is, or whether I'm looking at an unusually-coherent error.
Take the exact same messy pile of bytes but promise: "Everything in this made sense to a human at some point in the process", and that changes how one approaches the problem.
Before code would be intentionally a certain way, done as a known compromise that's less than ideal, or the developer flubbed in some way. Using commit comments and other surrounding details you might be able to identify which in many cases which could guide your next steps. I don't think you can do that with an LLM's output.