Either you have to give the AI the points you want to convey, then just put those points in a message. Or you don't have anything to convey, then don't post a message.
I don't see why anyone would want a slopified version of whatever it was that I had to say.
Lots of people lack confidence around their writing, and many people (particularly in tech) are not english native speakers. I can definitely see both of those groups getting use out of AI assistance in writing.
That being said, I sometimes use AI to see if I've missed anything, but the last thing I'll give up to our future AI overloads is writing text, as I enjoy it.
Not to single out OP or anything, but the more we do things on our own, the more likely we are to build our confidence. Relying on something or someone to hold our hand risks slowing down personal growth.
So I probably wouldn't argue against this in all cases, except where someone is just outsourcing all their thought to the model(s), that feels much worse to me.
I am neither (a native speaker), but you learn and gain confidence by doing. Also, I think flattening the beauty of language (however imperfect) into LLM-speak is a huge loss. Everyone sounds the same and boring through an LLM.
In my case I just use AI to fix typos and save people time if I repeated myself in a message.
You’re foaming at the mouth a bit about AI here rather than reading what I said. I don’t see how shortened or fixed grammar is slop. The article argues the opposite.