While fun and helpful (as outlined in this handy Chat GPT overview) the wild, chaotic phase of Chat GPT has jumped out of the trenches and into the war on academic integrity. Deans and university professors are now debating how to properly counter the plagiarism enabled by this new technology. It’s a big issue and not going anywhere.
If you’ve tried it out, you know Chat GPT can whip up results in seconds, and if a hungover freshman is tasked with turning in an analysis on how the Panopticon relates to postmodern sexuality, you’d be foolish to assume they wouldn’t call upon Chat GPT for backup. Plagiarists now have the keys to the ultimate super weapon. So what new futuristic technology will have to be created to counter it?
THE FALSE FIX:
Plagiarism has always been an issue an academia and, for years, Turnitin.com was the professor’s best antidote. The only problem is, Turnitin.com identifies info copied from the internet. Chat GPT doesn’t copy from the internet and its sources are untraceable. So Turnitin.com isn’t a fix.
22-year-old Princeton Computer Science student, Edward Tian saw an opportunity and created GPT Zero, an app that basically deciphers whether something was likely to have been written by AI. Open AI (makers of Chat GPT) then created there own deciphering app, Hugging Face. But here’s the twisted bit. Even if there’s a 99% similarity match to an AI, there’s no actual way (as of right now) to prove it was written by an AI. Remember, the source is untraceable. It’s merely suspicious.
For this reason, students still use Chat GPT to fly through work. In an anonymous poll published by The Stanford Daily 17% of Stanford students reported using Chat GPT on their finals, albeit mostly for brainstorming. Some have taken to the internet to share their not-so-discreet cheating dilemmas.
Some users were not so sympathetic to the cause:
On other threads, some users offer insight into the ultimate strategy with AI cheating:
But even if AI detection software can’t sift through cheeky edits, and even if it can’t be proven entirely, some professors can just tell when something is fake. With a very powerful tool of their own — the human brain.
Darren Hick, a philosophy professor from South Carolina, was one of the first professors to report catching a student cheating using the new technology. In a long Facebook post he detailed some of the red flags he noticed:
“The first indicator that I was dealing with A.I. is that, despite the syntactic coherence of the essay, it made no sense to someone familiar with the material. ChatGPT also sucks at citing, another flag. This is good news for upper-level courses in philosophy, where the material is pretty complex and obscure. But for freshman-level classes, this is a game-changer.”
So what’s the answer?
THE REAL FIX
Some users on the internet have thrown out some far-fetched theories about what might save academia.
Perhaps these theories are actually right on the money, as many professors have returned to the old, ancient systems of 1995.
In that same Facebook post by Professor Hick, he suggested: “In the future, I expect I’m going to institute a policy stating that if I believe material submitted by a student was produced by A.I., I will throw it out and give the student an impromptu oral exam on the same material.”
Additionally, in that same article from Stanford Daily, a Senior Lecturer in Computer Science revealed that Chat GPT had become so prevalent in efforts to cheat, his class has now returned to pencil-and-paper exams. OG analog stuff.
It’s crazy to think that we’ve advanced so far technologically that the seemingly brilliant counter maneuver, at least for now, is to unplug and return to a simpler time. Effectively, the old school could become the new school, with students being forced to present work in person. Would that be anti-progress or a renaissance? Only time will tell. But for now, it’s the only real solution to Chat GPT on campus.