Graduate uses ChatGPT to write a university essay – that gets a passing grade

A graduate who used a powerful artificial intelligence ‘bot’ to write a college essay successfully fooled a professor, who gave the report a passing grade.

Pieter Snepvangers used the controversial ChatGPT AI to write an essay as part of an experiment to see if cheaters could use the software for their courses.

He told the technician to prepare a complex 2,000-word article on social policy, which he did in 20 minutes.

Pieter then asked a professor at one of the major Russell Group universities to mark him and give him his assessment and was surprised when the tutor said he would have been given a mark of 53, a passing grade of 2:2.

The professor called the essay “suspicious” but said it was closer to the work of a “lazy, lazy” student than an AI, admitting: “You definitely can’t cheat your way to a first-class degree, but you can do trap. their way to a 2:2.’

Pieter says he was surprised to find a college professor admitting that students could “cheat” to get a passing grade.

Since its launch three months ago, ChatGPT has concerned schools and universities around the world.

The software allows users to ask any question and receive an AI-generated response in seconds that mimics the style and syntax of a human response.

Students in the US have been banned from using the software in schools and UK universities are ‘fighting’ to review how they can detect its use.

Writing for the student news website The boardPieter said: ‘I found a fairly prestigious Russell Group university and asked one of their professors if I could take his senior social policy assessment to see if ChatGPT could really work.

‘I wanted to know what grade I could get and whether or not he would detect that the essay was written by a bot.

“Under the premise of being a third-year social policy student completing a 2,000-word essay worth 75 percent of a unit, I went to work.”

Pieter started by simply asking the software the essay question and asked for 2000 words with references.

However, the technology, which was created by the Elon Musk-founded firm OpenAI, only managed to return 365 words at first, just 15 percent of the requested number.

The graduate decided to take a different approach and asked the bot ten separate questions, all related to the essay question, eventually managing to get 3,500 words out of the AI.

Pieter came up with the experiment and asked a university professor if he could tell if the essay was written by an AI.

He then took the best paragraphs the software had given him and copied them down in an order that “looked like the structure of an essay.”

He did not change or rewrite any of the words, and his essay was complete in 20 minutes.

He said: ‘In total, 20 minutes to produce an essay that is supposed to demonstrate 12 weeks of learning.

‘Nothing bad. I nervously sent it to my professor and waited for the verdict.

Once ticked off, Pieter was surprised to discover that even though the software hadn’t delivered a top-notch rating, it had still achieved a 2:2 pass rate.

Asked if it was obvious the article was written by a robot, the speaker didn’t “think it would have been very clear” but said it was a bit “suspicious.”

His comments continued: ‘Basically, there is no reference to this essay. It is very general. It does not go into details of anything. It is not very theoretical or conceptually advanced.

‘This could be a student who has attended classes and engaged with the topic of the unit. The content of the essay could be someone who has been in my classes. It wasn’t the most terrible in terms of content.

A snippet of the essay written by the controversial artificial intelligence software, which a university professor said was “suspicious” but reminiscent of essays written by “lazy and hesitant” students.

The only item where ChatGPT completely failed was the lack of references in the text, however, the professor said that if a student “had slipped in something that seemed plausible”, the essay would receive a grade of 53.

They also said that if Pieter had simply added references from the module reading list, “it might even have made it to 50”.

The speaker even admitted that of the essays he had marked so far, a shocking 12 percent of them showed signs of having been written with AI software.

Pieter said: ‘The truth is that software doesn’t give you the answer all at once. You will have to structure your answers in a more coherent order.

‘But I spent ten minutes doing this and got a 53; It wouldn’t have taken much longer to add some reading list references and bump it up to a 2:2 max.

‘ChatGPT is only three months old. You wouldn’t bet against me writing a 2:1-worthy essay in another three months.