> Written by a human.
That for some reason uses em dashes and writes in a voice that at times I find hard to distinguish from AI.
Man, I'm tired. Are people just lying? Am I just seeing things? Some mystery third option? Is it meta commentary?
Everything is poisoned.
I suppose it feels incorrect regardless of actual AI use, because it's still the LinkedIn thought leader template with relevant current issue.
Which is interesting, because it is so meta.
It has it all. It has the SpongeBob meme for relatability, it has the vague call to action (mindfulness, lmao) at the end. Ugh. Man.
>That for some reason uses em dashes
You'd be surprised that there are folks on this planet who love em dashes. I'm one of them and I used to write a lot with em dashes, but stopped using it altogether in the past few years because of AI.
Yes but if I am aware enough of the current landscape on the web to put a "written by a human" disclaimer, I am also aware enough of the fact that current em dash perception rightfully isn't very good.
So exactly what you said. You've stopped because you know how it will be perceived.
There is something not checking out with that blogpost is what I'm saying. Things do not feel organic. Which can be AI, but also can be lots of other things, but regardless of that, it smells.
___
Googling the author tells me that they perhaps might just be trying a bit too hard to be taken seriously. Oh well. But anyway
Smell is there. Intent is unclear
Using em-dash is as much a sign that a writer (1) knows how to produce documents properly and (2) has a good grasp of the English language as it a sign of LLM usage.
I went into my text editor, then used a find-replace tool to replace “--“ and “---“ with the appropriate dashes I copied from a character map website. Manually: with my good ‘ole hands mouse and keyboard. I realise that some grammar can just seem like LLM slop, that’s kind of what they have been designed to output. This is why I went out of my way to add that disclaimer at the end.
I enjoy using em and en dashes for punctuation. They provide a nice break that’s not quite a comma, of which I already have way too many, because I tend to overthink grammar.
I’m sorry my writing style is not appealing to you, but don’t accuse me of publishing AI slop, that’s a shitty thing to do.
Nah, man. It's deterministic. If the input sets flags, then output will be "this set flags for me".
No need to make it emotional like that.
Besides, I have simply told you that these aspects set flags for me. Others might not. They might instead just discard it with zero actionable feedback.
I saw on your bsky account that AI is some sort of holy war for you. It is not for me. I just don't want to read stuff that feels inauthentic :D
The thing with such disclaimers like "written by a human" is that a) people can just lie and b) labels are redundant. Either the content speaks for itself, or it speaks in a way where a label doesn't change anything about it.
Especially on platforms full of hustling business frauds like HN, you do not want to base your judgement of something on their self-declarations. You completely ignore any provided guidance or other metadata on how you're supposed to understand stuff and instead parse it as it is. (Of course, context still matters. Always does.)
Anyway. I'm not trying to be an antagonist here.
I have to resist the urge to troll my friends by writing something intionally in the style of AI, just because I find the "AI Style" to be so ugly and annoying. I don't want to cause any more psychic irritation to my friends and family though so I don't do it.
LLMS are here for >3y, enough time to shape the thought processes and language of the society exposed to its output.