I noticed this when I was last grading papers. Seeing a typo or grammar nit starts to feel reassuring— a sign that you are reading a student's actual opinions, not LLM generated ones.
Of course, it is far from that simple! This piece suggests we are shortly to enter a weird contest in which everyone tries to signal their writing is human:
"Chicago-based Sean Chou, 54, co-founder of multiple tech startups including an AI company last year, uses artificial intelligence to draft LinkedIn posts. But he said he’ll replace em dashes with two smaller dashes, hoping they’ll look more handmade.“It’s like my artisanal craftsmanship,” Chou said.
He tries to rein in overtly bold statements, too. “[Large language models] get their content from TED Talk transcripts and Reddit opinions, so it has a self- selection bias there, it tends to sound very confident,” he said."
When I read comments or posts that are obviously AI generated or assisted, my hackles rise, and I don't feel like engaging with them. The tells are usually so obvious, I can't understand why the writers don't tweak them. It's not this, it's that. Here's what no one is saying about this. You can so easily make the effort to edit out these tendencies!
I would have the same reaction if my boss sent me an AI generated email, or my professor used AI to grade my work. It is insulting and demeaning on some level. The person made no effort to write the thing, yet expects you to make effort to read it.at
it.at the same time, this is far from simple. AI writing is now so pervasive that it is hard to avoid the style infecting your own. And, adding "artisanal touches" won't help if you used AI to craft the overall arguments, because part of the problem is thinking is IDEAS that converge to the median.This is already like microplastics. In everything. Impossible to extract. Sort of depressing.
wsj.com/tech/ai/writers…