Will our writing all converge to a generic AI style?
Let's take a quick detour into LaTeX. Back in the late '80s, before LaTeX was the standard, there was TeX—a system with no default formatting, which meant everyone had their own unique style for papers. Then LaTeX arrived, and suddenly all our papers looked polished and professional. But the catch was, LaTeX only got you about 80% of the way there. The original manual even mentioned that you needed to add some extra TeX commands to really finish the job. Most of us didn’t bother, though, and soon all our papers had that same uniform look. It was efficient, and it was good enough. Microsoft Word ended up doing the same thing for everyone else—you could tweak it to be different, sure, but most people didn’t. It turns out most of us are just fine with "good enough."
I generally don't like to use large language models to write for me. But I do use them to refine my work, to catch errors or suggest improvements. The thing is, AI likes to take what I write and make it sound smoother, and I often think, "Wow, that really does sound better." But here’s the tricky part: it might be better, but it’s not mine. It’s the AI's voice. And still, if the stakes aren’t too high, sometimes I let the AI’s version slide. That’s the start of a slippery slope. Before you know it, we’re all letting AI make our writing a bit more generic, a bit more uniform. And eventually, we end up writing to match the AI’s preferred style.
For this blog post, I didn’t resist at all. Could you tell this is ChatGPT's style?
This was my original post that I fed into ChatGPT.
ReplyDeleteWill our writing all converge to generic AI style?
Let me take a detour to LaTeX. Before LaTeX came out in the late 80s there was TeX, which had no default formatting. So everyone would have their own paper style. Then LaTeX came out and all our papers looked great, as long as you didn't look too closely at the content. The original LaTeX manual said that LaTeX only got you 80% of the way to formatting the paper and you needed to add TeX commands to get the rest. But no one bothered doing so, and all our papers looked the same. Microsoft Word had the same effect for everyone else. You can make some effort for your writing look different, but most people don't care.
I don't like to use large-language models to write for me. But I do use them to look over my writing to get fixes and suggestions. But AI likes to rewrite what I write and often I'm like "wow, that sounds so much better". But I need to resist, because it's not my style, rather it's the style of the AI program. But if what I'm writing isn't that important, maybe I don't bother to rewrite the rewrite and I go down a slippery slope. If others follow we end up all our writing looking alike. And then we start tailoring our writing towards the AI style.
I didn't resist at all for this blog post. Could you tell this is ChatGPT style?
I honestly find your version much better than AI's one.
ReplyDeleteYes, the first paragraph sounded obviously like it was AI generated or had an AI pass over it. The second paragraph was less clear to me -- the only word choice that seemed suspect was "I generally" which is unnecessary and "But here's the tricky part" which is so generic.
ReplyDeleteThen again, if the topic of the paragraphs wasn't AI itself, maybe I would be less keyed into looking for this sort of bland language.
you should attach, in your system prompt, a past post of yours, and ask ChatGPT to write the new thing in that style - so it's not just the content that's yours but also the style.
ReplyDeleteFor even more fun, attach a past post of Gasarch's and ask it to write the above post in that style :)
Simple solution. Conferences ban the use of LLMs to write or rewrite your work.
ReplyDelete1) Hard to enforce
Delete2) To what end?