Education as an academic discipline has often been criticised for not being adaptive quickly enough for the fast-changing world. Now, as universities grapple with making their curricula relevant for the jobs of the future, another challenge has emerged.

The disruption by ChatGPT has cast further doubts on the relevance of the traditional approach to teaching and assessing students. ChatGPT is an AI language model capable of following instructions in a prompt and providing detailed human-like responses, much like a chatbot. As it can respond, synthesise knowledge and produce essays, bypassing the detection of software that detects copy and pasting, it has the potential to circumvent learning while producing good enough assignments.

Education is different from the industries or the business world. In business, goods or services (both are products) are often the only thing that matters. A process is often only as useful as its role in bringing the product to the market more efficiently or effectively. This is one reason business tends to incorporate technology faster: it can use any process that helps. In education, however, the main focus is often the process itself. Learning is a process, not a product by itself. Here, learning achievement or a degree (the main products in education) are often expected to arise naturally from the process.

ChatGPT disrupts this assumed association between process and product by producing an assessment (output) without adequate learning (process). When this happens, education as a knowledge or skill acquisition process is diminished and might even face the possibility of becoming defunct. This disruption will be experienced both in the schooling and higher education sectors.

However, a greater impact will be felt in higher education, which has incorporated technologies at a faster rate. More specifically, it is in the area of student assessments that ChatGPT will cause the biggest disruption. If current educational models do not adapt, and adapt swiftly, a disturbing possibility (admittedly extreme) exists that marks or degrees will be awarded for entirely artificial intelligence- (AI) generated assessment. In other words, education’s setup, functions, fundamental assumptions and way forward are under threat.

Many educators’ first instinct, facing such disruption, is to banish the use of ChatGPT in educational assessments. Framed as a means to cheat, ChatGPT is treated with suspicion. Our contention is that this approach will result in a disconnect between education and the real world it prepares students for. One would not be able to banish ChatGPT in the world the students operate in after they graduate, so to cut ChatGPT out of learning and assessment contexts would be like experiments run only in the lab, removed from the authentic and context-responsive principles many educationists advocate for.

In this sense, ChatGPT will force us educators to reconceptualise assessment. For example, what exactly do we want to assess our students for? What is the best way to assess the students for whatever objectives we have planned for our lessons? What assessment can’t be automated by ChatGPT? These considerations will fundamentally change how assessments are designed.

A new baseline

To achieve this, we propose an integrative approach. This approach would incorporate ChatGPT into teaching and assessment, from a new baseline in which students’ augmented abilities are taken into account. The assessment would move on from an evaluation of skills that can be outsourced to ChatGPT. 

In other words, the objective becomes to train students to get the most out of ChatGPT while nurturing their own thinking and writing skills. This should include, for example, using ChatGPT to help understand and digest concepts, provide a snapshot synthesis for a certain topic, and improve the usage of words or the coherence of writing.

The challenge educators face is then to up their game to nurture and assess skills that cannot be easily simulated by ChatGPT. Simple regurgitation of content is no longer meaningful in an age in which a question can be fed as a prompt to AI and a passable response generated with a click. For assessment to remain meaningful, it will need to evaluate students’ augmented ability given their interaction with ChatGPT.

In addition, we propose an open-minded approach that embraces the challenge ChatGPT presents to us as a teachable moment. Moving away from punitive measures, a developmental and adaptive educational stance, in our opinion, will be much more helpful. This is also related to another priority in education: instilling ethical practice. Deliberate discussions about the language model’s ethical use can free students and educators from the stalemate the argument has reached, caught between a natural desire to tap into ChatGPT as a resource and a more conservative condemnation and restriction of its use.

Given that there is no way to fully predict the extent to which ChatGPT will disrupt the educational landscape, this open-minded approach can take the form of asking more questions and deliberating on one’s reasoning. For example, is using the tool to aid the thinking process (eg, through prompts like “What is the connection between this and that?”) ethical? 

What about the practice of entering an assignment question, then copying, pasting and submitting its response after paraphrasing? Or asking ChatGPT to read a few articles, extract a summary and synthesise the differences and similarities among the articles to assist in understanding the main debate on a particular topic? Or using it to assist with the structure, cohesion and logical progression of a document? 

Further discussions and deliberations around these issues open up the space for reflection, which is more useful than simply banishing this powerful, galvanic and useful tool. DM

Ku Ye is an associate professor at the Department of Education Leadership and Management, Faculty of Education, University of Johannesburg. Dr Karen McCarthy is a lecturer at the Department of Childhood Education in the same faculty.