3 minute read
Artifical Impact
ChatGPT changes old routine
Jared Carrillo | Staff Reporter
Whether it’s writing notes on your arm with a Sharpie or sneaking a phone during the test when the teacher doesn’t notice, cheating is everywhere you turn.
And it seems to be getting easier.
Now with a simple prompt, a student can have his or her entire essay written by an A.I. bot. It seems like a bulletproof system, but millions of users haven’t considered the repercussions.
ChatGPT is the saving grace to a procrastinating English student, but a true danger to all who use it. During a test trial to see how ChatGPT works, you’ll see how sketchy it is. The server requires an account, and you must give an email, password and phone number. But soon after you enter your email, you receive an alert saying that ChatGPT’s email is dangerous and has been used to steal user’s personal information.
Even with a warning, there will still be a crowd that rolls the dice with this shady service. For them, the free ride will end shortly. In the long run, they’ll have to use their writing skills, which they never trained, on a pen and paper assignment.
In fact, Universities in Australia are already switching to pen and paper assessments in order to fight against the cheating buzz. While America already uses pen and paper assessments for AP Exams and SAT essays, it won’t be long before pen and paper makes its way back into every class in America.
Detection tools have been released already. They don’t do the job to perfection, but these tools are supposed to be able to detect usage of A.I. bots in essays. An A.I.’s essay is 49% artificial and 51% human written, according to thenextweb.com. These detections will scan the text for specific patterns to uncover use of A.I. New detection tools are being developed, perfected and released every day; TurnItIn, for one, is working on a new detection tool.
Even ChatGPT — the “all hail the king” of cheating high school students — released a detection tool of its own to combat “improper use” of its server. This tool can detect if a student’s writing was A.I. with 95% accuracy, according to Stanford.
English Department chair Dr. Paul DiGeorgio works with teachers on methods to prevent students from using A.I. essays. Teachers are familiarizing themselves with the ChatGPT server, and explicitly teaching students what constitutes plagiarism.
“If students have more confidence as writers, there’s no need to use ChatGPT as a crutch,” DiGeorgio said.
While A.I. will be here forever, this trick isn’t here to stay and when it’s gone for good, users will be left blindsided to the alternative: when the teacher gives them a blank sheet of paper, a pencil and an hour to complete their essay, ChatGPT will not be there to help. There is an argument to be made of the proper use of A.I. to benefit students instead of tempt them.