About a year and a half ago, I wrote about my kid’s experience with an AI checker tool that was pre-installed on a school-issued Chromebook. The assignment had been to write an essay about Kurt Vonnegut’s Harrison Bergeron—a story about a dystopian society that enforces “equality” by handicapping anyone who excels—and the AI detection tool flagged the essay as “18% AI written.” The culprit? Using the word “devoid.” When the word was swapped out for “without,” the score magically dropped to 0%.
The irony of being forced to dumb down an essay about a story warning against the forced suppression of excellence was not lost on me. Or on my kid, who spent a frustrating afternoon removing words and testing sentences one at a time, trying to figure out what invisible tripwire the algorithm had set. The lesson the kid absorbed was clear: write less creatively, use simpler vocabulary, and don’t sound too good, because sounding good is now suspicious.
At the time, I worried this was going to become a much bigger problem. That the fear of AI “cheating” would create a culture that actively punished good writing and pushed students toward mediocrity. I was hoping I’d be wrong about that.
Turns out … I was not wrong.
I’m accused of being AI on other sites simply because I construct complex sentences with regularity – and use emdashes.
This predates the ai bubble. There used to be a really common “plagiarism detector” (something like CheckMeIn?] that would generate a “similarity score” with a database of literature. Institutions were welcome to set their own thresholds of what they considered too similar. I hit the threshold multiple times in completely original works by using language that was simply too literary or formal in nature.
Mind I had been accused of plagiarism by teachers prior to those tools for much the same reason based only on vibes, so maybe that was a step up, since students could use it ahead of time.
There was a news story around that time of somebody getting taken through disciplinary action due to getting close to 100% similarity on the tool - eventually to discover that their own essays had Venn included in the database.
This is why I love hearing large businesses complain about all the fake low effort job applications they get with AI as if this wasn’t the inevitable end state of fakeness and inauthentic corporate language they have been pushing for decades.
As if that same job posting wasn’t itself likely written with AI lol
HEY HEY that is different ok!
I’m accused of being AI on other sites
Other sites? Happens here, too. The best answer is troll them by imitating AI.
18% is nowhere near high enough to be throwing around accusations like that. Seems like the teachers don’t know how to interpret the results.
– and use emdashes.
That’s more a matter of 95% of people not even knowing how to type a ‘–’ with their standard keyboard layout.
Word processors (like MS Word) have been doing it automatically since I was in school. Same with double spacing after a period.
Which is great for one application, but two spaces after each period would be hell to edit down to AP Style.
I mean, Ctrl+H and switching two spaces to one is easily doable, but that’s not where I want to start the editing process.
In my experience that is in fact more of a MS Word feature (and very inconsistent at it) than a general word processor feature. But maybe I’m underestimating the impact on “average texts” simply because my use of MS products is far below average.
I doubt the average student is using anything other than Word, unless they are using AI to begin with.
I realized recently that I enjoy reading peoples janky personal messages online. You know it’s a real person. Or at least it probably is.
I hope I’ve been sufficiently janky for you with my posts.
AI checkers seem like a stupid and lazy way to determine if a student used AI to write their paper when the teacher could simply sit down with the student to ask them about the content of their to paper.
AI checkers for text (but the same is true for the ones pretending to spot AI pictures and videos) also don’t work by definition.
The AI tries to make it’s "product’ perfect. It does not have the ability to spot its own mistakes and telltale signs, or it wouldn’t make them in the first place.
So every AI check is actually cheating. In pictures and videos with hidden watermarks, in text with typical clues like the mentioned ‘–’ or vocabulary more prevalent in AI texts that the average human work.
I think that’s a very intentional feature, brought by the techbros who urge kids to skip uni.
I can’t get Idiocracy out of my mind when I read this…
