Video Gamer is reader-supported. When you buy through links on our site, we may earn an affiliate commission. Prices subject to change. Learn more
While of course – there is a huge contingent of people out there using ChatGPT to cheat in college and school, there’s also a majority who aren’t using it to cheat. Regardless – many are being accused of cheating by their academic institutions – and the internet is rightly torn in half.
Students are among some of the craftiest and most inventive people out there. Any amount of authority or scrutiny from teachers and lecturers alike will always be met by resistance – whether that’s through clever loopholes or outright cheating. Now, being the cheeky free-thinking souls that they are, students have figured out some of the best ways to use ChatGPT without getting caught in schools, colleges, and universities.
One trick, noted on this Reddit thread, involved ensuring that the document format you’re submitting has a way to track changes and record edit history. If a student then uses ChatGPT to generate content – they can then hand-write the artificial intelligence generated text into the document, while also committing all of the regular spelling errors and logistical mistakes usually made – before then correcting them. If a student is then accused of cheating – they can bring out the edit history as evidence that it was written by an imperfect human. However – we don’t condone this particular ‘trick’ as at the end of the day, students making use of it aren’t really benefitting from it.
One particularly amusing failed attempt at using ChatGPT for homework resulted in a student using it for their Spanish homework. However, the AI generated Spanish translation also included ChatGPT’s infamous catchphrase: “Sorry, but as an AI language model I don’t …”
Of course, this particular student was caught out pretty easily – before their teacher posted it online for the world to see.
Recent updates to Turnitin have given it the ability to detect ChatGPT content, though this has simultaneously been met by opposition from students. For example, AI detection tools are now commonly being used to accuse students of cheating – even when they haven’t.
In particular, a college instructor recently went viral after accusing their entire class of using ChatGPT, and allegedly failing them all. The particular incident is being investigated by the college, and a spokesperson has recently come forward and stated that “No students flunked or were prevented from graduating because of the issue,” according to NBC.
The issue has sparked online discourse on how academic institutions are placing too much trust in ChatGPT detectors. In fact, ZeroGPT were recently found out to have allegedly used ChatGPT to generate their own privacy policy through their own detection software, which the following Reddit thread discusses.
ChatGPT detectors and AI detection tools alike are proving themselves to be inconsistent at times, and it’s clear that a tool flagging content as ‘possibly AI’ is not enough to make a decision from. If teachers are concerned about their students using AI to create content – the technology simply has to improve. Of course, this is a lot easier said than done, and it will take time, but as of right now, AI detectors and teachers are currently fighting a war against ChatGPT and students.
It’s quite ironic, really, when you consider that many of the students accused of using ChatGPT are being found out by teachers using AI tools themselves.
Cover image generated by Dall-E, edited in-house.
