Within a matter of weeks, ChatGPT has given the education sector much to worry about, especially around how they should plan and redesign academic assessments. A primary concern with ChatGPT is how it can lead to increased cheating. The sector has been dealing with contract cheating for an extensive period of time so this is another variable in contract cheating rather than something that is completely new. Contract cheating is a form of academic dishonesty in which students outsource or contract their academic milestones such as assignments, exams or thesis to a third party. Keeping in mind that contract cheating has previously involved engaging human writers for essays, and the sector has previously risen to the challenge. The sector is now fighting cheating against an imperfect AI entity.
Amid intellectual threats and potential for cheating, some academic institutions and state governments have banned ChatGPT (such as New York public schools, New South Wales, Queensland, Victoria and Tasmania in Australia). Others have begun to redesign exams and assessments with reports of some Australian universities already considering reverting to traditional pen and paper testing methods. Besides intellectual threats, ChatGPT is also feared to lead to the deskilling of students and future workforce. Its spoon-feeding approach is seen as having the potential to harm human creativity.
Why banning ChatGPT is not a solution?
While ChatGPT is facing criticism for potentially ending the tradition of homework and undermining the overall integrity of academia, the “for ChatGPT” educators view it as a valuable teaching tool. They argue that banning ChatGPT is not a feasible solution for several reasons. Firstly, students can always access ChatGPT through their personal devices, making a ban on school-owned devices ineffective. Secondly, countermeasures such as GPTZero are unreliable. GPTZero claims to identify text written by ChatGPT, but testing has revealed that it can be easily fooled by changing a few words, or rearranging sentences using another AI system. In fact, ChatGPT’s own detector didn’t perform well at this stage. Until detectors have advanced to a relatively mature state, it may be worth reconsidering if depending on these tools is an optimal use of teachers' time and resources.
Proponents of ChatGPT argue that it should not only be embraced, but also utilized for better learning and teaching opportunities. They suggest that ChatGPT can function as an after-hours tutor for students, providing clear explanations of complex information and offering personalized learning based on the student's cognitive abilities and intelligence quotient. ChatGPT can also reduce the time spent on research and prevent students from repeating work that has already been done. Pro-ChatGPT educators and administrators propose various ways to incorporate it for the improvement of the education sector. For instance, South Australia's education department permits the use of ChatGPT for students, excluding graded work.
What ChatGPT and other generative AI tools will bring to education is likely to be disruptive and ChatGPT and DALL-E (another application by OpenAI to generate digital images from natural language description) are just one of the firsts. If AI in education is inevitable, then our focus should be on educators and policy makers to create a suitable learning environment where AI is embraced and higher order learning in students is achieved. For example, rather than fearing ChatGPT's ability to facilitate learning, we should embrace it as a tool that can stimulate critical thinking and creativity. Depriving students of innovative learning technologies will not enable us to compete in a rapidly changing technological environment that exists in the real-world. It is evident that ChatGPT and those new technologies that follow will be here to stay and is likely to become an integral part of our lives, much like other technologies. Now that it has arrived, ignoring it won’t be a solution for educators and education policy makers.
The broader questions around ChatGPT in the education context centre around the challenges that the sector will face, including questions around the ethical aspects of AI in education, how AI will lead to a rethink of the way educators teach and learn, and the policies that will guide the sector to ensure that advances in AI can be embraced effectively much like how we have embraced video conference tools for online teaching during the pandemic.
It's worthwhile to keep in mind that ChatGPT is a relatively new technology. It was released to the world in November 2022 so there may be more hype than an objective assessment of the capabilities and limitations of ChatGPT in the education space. For educators, taking stock and understanding what the technology is from a policy and learning outcome perspective will be crucial so the sector does not develop a knee-jerk response.
We do feel confident that the sector will be able to manage any negative aspects of ChatGPT. The less known variable instead is how ChatGPT can be embraced for the ‘good’ of teaching and learning. This is a question that will need much more discourse and exploration both by educators and policy makers.
Watch the Hub’s contribution to the discussion as reported on Ten News: https://www.linkedin.com/feed/update/urn:li:activity:7027062350566330368
Professor Kok-Leong Ong
(Director, Enterprise AI and Data Analytics Hub - School of Accounting, Information Systems and Supply Chain, RMIT College of Business and Law)
Dr Samar Fatima (Enterprise AI and Data Analytics Hub - School of Accounting, Information Systems and Supply Chain)