Analysis: ChatGPT is set to disrupt the post-pandemic dominance of written assessments

Post-pandemic assessment and education as a whole are at odds with user accessibility to artificial intelligence

The onset of the COVID-19 pandemic meant that traditional modes of assessment were no longer compatible with distance learning. Academic officials were not prepared nor understood how to effectively deliver education solely online to all students. An inability to administer live monitored exams gave preference to fully cited written assignments as a measure to counteract cheating and academic dishonesty.

In the years following the end of the COVID-19 lockdowns, many schools and universities continued with the trend of written assignments rather than returning to pre-pandemic standards. Although many courses reinstated in-person assessments, across every department in College, the large-scale shift away from examinations during the pandemic made a lasting impression. A large proportion of modules have not returned the associated assessments to exactly what they were previously, with some continuing with no examinations, others with mixed written assignments and examinations, and even some with take-home examinations.

Enter ChatGPT.

Post-pandemic assessment and education as a whole are at odds with user accessibility to artificial intelligence (AI). Software like ChatGPT allow internet users unlimited access to generative AI. Academic institutions must now grapple with yet another imminent issue: what does quality education in the age of AI look like?

Artificial intelligence: What has changed?

AI is far from a new technology, being in fact a prominent field of computer science since the 1950s. Without even realising, AI has long been a facet of peoples’ everyday lives. A study conducted in 2017 by Pegasystems Inc., a world-leading software company, found that while 84% of people were using some form of AI in their daily lives, only 34% were actually aware of what they were doing.

Technologies like Apple’s Siri, Amazon’s Alexa, face-identification, Google Maps and countless others frequently used are all forms of AI. So what has changed? Why is AI such a contentious and feared topic now?

ChatGPT, as previously stated, has been at the forefront of global media for over a year. At the most basic level, as a form of generative AI, the platform allows the user to prompt a question or command and, in return, the software provides a response. ChatGPT quickly gained traction from its inception and almost immediately conversations began to circulate regarding its implications in education.

Not only do generative AI softwares operate in a call-and-response format, but they have the capacity to do much more. For example, a student may receive a multiple-choice homework assignment and input the questions from it into ChatGPT, generating the answers they require; however, it does not stop there. ChatGPT can also generate complete essays and research papers, even providing a bibliography of sources.

In terms of academia, the abilities of generative AI tools are argued to pose a threat to the current model of education. Students are able to easily access a resource that has the potential to complete their assignments, with little to no input of their own. Due to the unreliable nature of AI plagiarism detection softwares, students have been able to pass these works off as their own. The large-scale shift in assessments brought about by the pandemic are at risk of being reversed, with significant emphasis being placed on reinstating in-person examinations.

Even further, the insurgence of generative AI has posed questions regarding the future of education as a whole. In an era where students no longer have to conduct research for their studies, let alone even produce their own work, what does this mean for the quality of education students will receive? Concerns regarding this subject and AI plagiarism have prompted academic institutions around the world to take the initiative, setting guidelines or outright bans against the technology.

Trinity’s response to AI

College has been extremely vocal throughout the past year regarding the use of AI in coursework. As can be expected, using generative AI like ChatGPT to complete assignments is deemed academic dishonesty and therefore banned. The caveat is that apart from plagiarism, each department in College sets out individual guidelines regarding the extent to which students may utilise generative AI as a resource in their studies and assignments. The overarching uncertainty surrounding how to manage AI is reflected within College itself.

Nonetheless, Provost Linda Doyle has committed to engaging with AI holistically as universities come to terms with its impact. In July Doyle said: “I have to find a way for our university to fully engage with this space, to become skilled enough to use and challenge these tools, to have the levels of cybersecurity needed to be resilient in a world in which they are widespread.” It appears that in the future, at least in Trinity, AI will likely be embraced rather than resisted.

A spokesperson for College told Trinity News that while long-term changes may be needed to teaching and learning, which is currently under review, its “current focus is on supporting the adjustment of teaching and assessment practices” in response to generative AI tools.

“Accordingly, we are working to support colleagues to evaluate the pedagogical opportunities presented by GenAI tools and adjust assessments to safeguard their validity and academic integrity”, they said.

“Recognizing that GenAI tools present different opportunities and challenges for different disciplines, we are emphasising that all module and course coordinators should make clear in what way (if at all) these tools may be used in the completion of assessments and how this use should be acknowledged”, they continued.

More and more schools and universities are recognising that resisting AI’s infiltration in education is unrealistic. While not outright endorsing its use, College understands that generative AI is here to stay.

The negative consequences it poses are discussed at length by researchers and academic officials, however, many believe that if wielded properly, AI’s power can positively transform education. It can support teachers and students in its ability to quickly collect data from a plethora of sources and produce in-depth explanations.

Over the coming years, students will more than likely see changes in the manner in which their courses are taught and examined. It is difficult to predict in exactly what ways these changes will manifest, other than perhaps a reversal of the post-pandemic assessment trend. There is much about generative AI that is not yet understood, but there appears to be little doubt that  it has propelled the world into a new era of education.