In the age of LLMs this seems like a waste of time to even think about constraining and formalizing languages. Why should human users restrict themselves in their expression if LLM systems can easily point out any remaining ambiguities?
What they seem to be doing at the moment is riffing on the ambiguities. If they point them out, then the outcome will be .. less ambiguous language which will tend to be like ACE.
Gimme the ball is idiomatic. How do you know it means Give me, and not Furble the ball? We're taught Give me, we say Gimme. Both have a role.
I'm interested to see an LLM do that. Technically, nearly every sentence is ambiguous, but that clearly doesn't bother us. An automated system that can point out sentences humans are likely to interpret in multiple ways would be interesting. However, no such thing seems to exist.
I recently made an LLM-based tool to identify inconsistencies with the INCOSE guide to writing requirements (identifying vague language, for example). Also have had some success translating those outputs to more structured language like LTL. I think it is doable
When working as a tech writer, I was mildly obsessed with this. Nothing drives users away from your documentation like a lack of clarity that signals that (1) you're on your own now, pal, and (2) the writer had zero clue of the potential for chaos in interpreting what he is writing.
You make a valid point, LLMs intuitively change the need for such formalized languages. It may not be a total waste of time, but I fail to see how this slight exaggeration justifies how your comment is currently stamped into the ground, likely by the current livid rage currently directed at LLMs in the software developer community.
I get it, people don't like to have their lives disrupted, and I sympathize, just as I sympathized when secretaries, bookstore owners and others were forced to find other ways to make a living. At least in our case we still have (or can find) a job, it has just changed, to some extent even considerably.
