So I had put 'prompt engineering' in my "fake stuff" mental bucket. Not that's not a real thing, but rather that it seems like pseudo-knowledge.
But [1] made me rethink that: LLM are a reflection of knowledge that does exist. And if you can find ways to reliably mine it, then the method of mining conveys information about the structuring of that information. (That paper doesn't say that at all, it was merely the catalyst for that realization.)
In other words, there's a quite direct mapping with what is identified in [1] and the activities one does in requirements gathering, requirements analysis, specification creation and specification refinement.
@JacquesC2 I'm not sure I fully grok your point, but it sounds like you may find https://benlevinstein.substack.com/p/whats-going-on-under-the-hood-of#%C2%A7representing-truth interesting as well, if you haven't already seen it
