This resonates with me. LLM output in Spanish also has the tendency to "write like me", as in the linked article.
On that regard, I have an anecdote not from me, but from a student of mine.
One of the hats I wear is that of a seminary professor. I had a student who is now a young pastor, a very bright dude who is well read and is an articulate writer.
Fast forward to 2025. The guy still has time for some Internet theological flamewars. In the latest one, he made (as usual) a well argued, long-form Facebook post, defending his viewpoint on some theological issue against people who have opposite beliefs on that particular question. One of those opponents, a particularly nasty fellow, retorted him with something like "you are cheating, you're just pasting some ChatGPT answer!", and pasted a screenshot of some AI detection tool that said that my student's writing was something like "70% AI Positive". Some other people pointed out that the opponent's writing also seemed like AI, and this opponent admitted that he used AI to "enrich" some of his writing.
And this is infuriating. If that particular opponent had bothered himself to check my student's profile, he would have seen that same kind of "AI writing" going on back to at least 2018, when ChatGPT and the likes were just a speck in Sam Altman's eye. That's just the way my student writes, and he does in this way because the guy actually reads books, he's a bonafide theology nerd. Any resemblance of his writing to a LLM output is coincidence.
In my particular case, this resonated with me because as I said, I also tend to write in a way that would resemble LLM output, with certain ways to structure paragraphs, liberal use of ordered and unordered lists, etc. Again, this is infuriating. First because people tend to assume one is unable to write at a certain level without cheating with AI; and second, because now everybody and their cousin can mimic something that took many of us years to master and believe they no longer need to do the hard work of learning to express themselves on an even remotely articulate way. Oh well, welcome to this brave new world...
On that regard, I have an anecdote not from me, but from a student of mine.
One of the hats I wear is that of a seminary professor. I had a student who is now a young pastor, a very bright dude who is well read and is an articulate writer.
"It is a truth universally acknowledged" (with apologies to Jane Austen) that theological polemics can sometimes be ugly. Well, I don't have time for that, but my student had the impetus (and naiveté) of youth, and he stepped into several ones during these years. He made Facebook posts which were authentic essays, well argued, with balanced prose which got better as the years passed by, and treating opponents graciously while firmly standing his own ground. He did so while he was a seminary student, and also after graduation. He would argue a point very well.
Fast forward to 2025. The guy still has time for some Internet theological flamewars. In the latest one, he made (as usual) a well argued, long-form Facebook post, defending his viewpoint on some theological issue against people who have opposite beliefs on that particular question. One of those opponents, a particularly nasty fellow, retorted him with something like "you are cheating, you're just pasting some ChatGPT answer!", and pasted a screenshot of some AI detection tool that said that my student's writing was something like "70% AI Positive". Some other people pointed out that the opponent's writing also seemed like AI, and this opponent admitted that he used AI to "enrich" some of his writing.
And this is infuriating. If that particular opponent had bothered himself to check my student's profile, he would have seen that same kind of "AI writing" going on back to at least 2018, when ChatGPT and the likes were just a speck in Sam Altman's eye. That's just the way my student writes, and he does in this way because the guy actually reads books, he's a bonafide theology nerd. Any resemblance of his writing to a LLM output is coincidence.
In my particular case, this resonated with me because as I said, I also tend to write in a way that would resemble LLM output, with certain ways to structure paragraphs, liberal use of ordered and unordered lists, etc. Again, this is infuriating. First because people tend to assume one is unable to write at a certain level without cheating with AI; and second, because now everybody and their cousin can mimic something that took many of us years to master and believe they no longer need to do the hard work of learning to express themselves on an even remotely articulate way. Oh well, welcome to this brave new world...