2,867 karma · joined February 21, 2011
What would your proposal be for fixing what you’ve identified as the underlying problem?
> A lot of solutions require way more code to implement than to describe the necessary properties of.
That's true to an extent, the additional code often define the emergent and undiscovered properties of a system.
However, LLM outputs change with slight re-wording of prompts and with each new model release. I could hand write a test that says if x > 1 make sure y happens, but then what productivity was gained?
Coding effectively is hard AND painful for the majority of folks. And I’d venture to guess they don’t want to be responsible for debugging LLM output either. Just like I don’t want to fix a car, or do my own plumbing.