In January 2023, a team of scientists from the University of Plymouth Marjohn published the research paper Chatting and Cheating: Ensuring Academic Integrity in the Era of ChatGPT. The article raised the problems of using neural networks in writing scientific papers, but it turned out to be written by ChatGPT, according to The Guardian.
The article was published in the journal Innovations in Education and Teaching International. According to Professor Debbie Cotton of the University of Plymouth Marjona, who is listed as the lead author of the publication, her team warned the editors of the journal about the use of a chatbot when writing the material. However, neither reviewers nor readers knew anything about this. The four researchers who reviewed the publication rated it as human-written.
“We wanted to show that ChatGPT writes at a very high level,” Cotton explained. “It's an arms race. Technology is improving very quickly and it will be difficult for universities to overtake it.”
According to Thomas Lancaster of Imperial College London, there is a growing panic in universities due to the rampant use of chatbots for writing essays and scientific articles.
“If the work is only a written document, it is incredibly difficult to prove that it was written by a machine, because the quality of the text is often very good,” he noted and added that the vocabulary and literacy of bots is often better than that of students.
According to him, teachers can still find signs that the student wrote the work using ChatGPT. The subtlety lies in the fact that the bot does not quite understand the bibliographic references properly and often uses “suspicious” references or simply invents them. However, in this case, students simply need to check their work and add real sources of information.