Turnitin, utilized by all eight universities in New Zealand, has taken a step forward to prevent cheating through the activation of software able to detect the use of AI, such as ChatGPT. The software is said to have an accuracy rate of 98%, and it has been enabled for customers in New Zealand.
The introduction of an update to help counter the use of digital learning tools for cheating has been welcomed by academics. Still, they said its efficacy would be short-lived and ineffective against students familiar with more advanced techniques.
James Thorley, the regional vice-president for the Asia Pacific of Turnitin, announced that the software could assess what proportion of composition was authored by Artificial Intelligence and accordingly pinpoint the problem areas.
AI-generated work can be difficult to determine compared to other cheating forms. Thornton noted that this makes such work harder to identify.
“Each time that ChatGPT or any other AI writing tool generates anything it is unique so no two creations will ever be the same, so you cannot detect it in the same way that you detect copy-paste plagiarism from whatever source”
“You’re really looking for this difference between how an AI writer would write and how a human would write but obviously that’s a lot more complex than simply matching the same text.”
Thorley declared that even after modifying AI-generated content to look like original material, Turnitin could still effectively detect plagiarism. He claimed that surveys conducted in the United States revealed a quarter to half of the college students had used AI for different tasks.
Thorley continues to say:
“Everyone’s actually very excited about the potential of these kinds of tools. Detecting doesn’t necessarily mean that there’s misconduct happening, that it’s wrong. It’s all about getting a bigger picture of what’s going on and starting conversations.”
Catherine Moran from the University of Canterbury stated that due to AI, the University is overhauling its assessments and anticipates Turnitin will be utilized for assignments that are yet to make modifications.
Understanding The Short-Term Impact Of TURNITIN Updates
Simon McCallum, a senior lecturer in software engineering at Victoria University, has shared his view that the update to Turnitin would assist in uncovering AI misappropriation; however, he believes its effectiveness will be short-term since most students tend to rely on rudimentary methods.
Simon McCallum says:
“There’s a whole bunch of these other AI tools that when you layer them on top of each other make any of those detections much less certain and the problem we have is that if you’re going to ping someone for plagiarism you have to be certain that they break the rules not just have a hint that it’s probably an AI that wrote this,”
ChatGPT and similar “large language model” programs are being utilized heavily by students, according to a statement from Dr. McCallum. He further revealed that these programs offer an effective means of aiding them in their studies.
Simon McCallum continues to say:
“Students are using GPT a lot. Some of them are getting very good at it and the ones who are good at prompt-engineering, creating good input to the AI, are able to generate very realistic work that looks very much like their own work in much shorter time than they would if they were having to write it themselves,”
According to him, students should gain familiarity with AI since it will be commonplace in many future jobs. However, his concern is that some are utilizing it without putting in any critical thought.
Simon McCallum went on to say:
“I’m worried when people are abdicating thinking. They’re not using their mind or making the effort to learn,”
“We might need to consider every assignment a group assignment because we have to think that every student has access to a group member, an AI, that can do some of the task for them and so then you have to say, well instead of treating it like your own work, in the group how did you contribute.”
McCallum suggested that the entire education system should shut down to give teachers and lecturers enough time to figure out how best to use tools like ChatGPT and similar ones, thus underscoring their importance.
Simon McCallum adds to say:
“This is a pandemic-level event. It’s the kind of thing which is going to fundamentally change a lot of what we do. In the pandemic we suddenly had to go online and we suddenly had to teach in a whole new way and this is the same level of change that’s needed,” he said.
“Let’s just close schools for a month, five, six weeks, and just spend that time doing professional development with all the teachers and with the university lecturers and say all of you need to get up to speed with this because it has fundamentally changed education.”
Why Universities Should Reevaluate The Use Of Essays As An Admission Requirement
Universities New Zealand executive director Chris Whelan believes that universities and schools are not required to shut down; however, they should consider assessing students differently.
Chris Whelan says:
“Best practice internationally is to recognise that it’s out there and to redesign assessments around it. So in some cases encourage assignments to be set using generative AI for a first draft and then to provide an assignment around what’s good or bad about [it]. But in other instances it’s to set assignments that generative AI can’t replicate,”
Paul Geertsema, a senior lecturer at the University of Auckland, proposed that universities reshape their usage of essays as a mode of assessment. Whelan commented that the redevelopment of Turnitin holds great potential.
Chris Whelan continues to say:
“We need to reassess why we are giving people essays to write. The truth is that ChatGPT and similar tools can write beautiful essays on most topics and we need to revisit what is that we want people to be able to do,”
“This is a very powerful tool and society can really benefit from its use and we should do whatever we can to encourage its adoption in industry and in the workplace so we are able to compete effectively and generate wealth for our society.”
Vaughan Couillault, head of the Secondary Principals Association, believes it would be hard for teenagers to foolishly use AI to produce essays, as their teachers had a sufficient understanding of them, enabling them to differentiate between the usual quality of assignments and exceptional ones.
Rather than relying on homework alone, teachers could ask their students to show them the first drafts of their work. Assigning work for completion in class may also be a rewarding way for students to grow and learn.
It is important to note that the development of such software also raises concerns about privacy and data protection. Developers of these tools must ensure that the software is used ethically and does not violate the rights of individuals. Moreover, regulations and monitoring tools must keep pace as AI technology evolves.
Overall, developing software to detect the use of AI systems is an important step toward responsible AI use. By promoting transparency and accountability, we can ensure that AI technology is used for the benefit of society while mitigating its potential risks.