The age of AI is here, and so is cheating on exams and assignments with chatGPT. While colleges and teachers are still figuring out ways to check "AI plagiarism," conflicts arising from "AI cheating" are already in Indian courts.
Across the seven seas, an Australian university has come up with a unique solution - The age-old oral tradition of viva voce, which means "word of mouth" in Latin.
Dr Chris Della Vedova, a professor of biotechnology at the University of South Australia, first encountered the problem after classrooms went online during the pandemic. "With digital exams, we didn't really know if anybody knew anything, which made it hard to assess," he said, speaking to The Guardian. "It was rare to fail students unless they didn't complete them [exams]."
With methods like multiple choice questions and writing essays becoming ineffective due to the availability of AI chatbots throughout the internet at the drop of a hat, Dr Chris and his team decided they needed to change things up. They first tried holding one-on-one 20-minute "conversations" where the examiner would ask questions from a random set of questions based on the curriculum taught.
Based on how students would respond, the examiner asked further follow-up questions to test their depth of understanding of a concept.
"Often you'll ask the first question and get a good memorised answer, but we want to make sure they understand what they're saying, so follow-ups allow gauging how solid their understanding is," Vevoda added.
But this approach has its downfalls as well. Spoken communication across different communities is not the same, and some introverts might get fewer marks based purely on their personality.
The idea of viva voce still exists in Indian educational institutions. Usually, at a higher education level, including Ph. D.s, there is an assessment where one has to "defend" one's dissertation in front of a panel. However, implementing the same at a bachelor or school level is a logistical nightmare since there are not as many teachers compared to a higher number of students to facilitate one-on-one examinations.
There is also the issue of subjectivity with the approach, as judging a conversation for marks may not be consistent as there are many factors involved in the process and bias can adversely impact assessment.