Comment by oblio

> What you're getting at is the heart of the problem with the LLM hype train though, isn't it?

> "We should have rigorous evaluations of whether or not [thing] works." seems like an incredibly obvious thought.

Heh, I'd rephrase the first part to:

> What you're getting at is the heart of the problem with software development though, isn't it?