It opened for general use at the end of November 2022 and has already sparked much discussion in the higher education and research space. This is ChatGPT, accessible (mostly) for free on its website and other interfaces.
The app is an example of artificial intelligence (AI) areas including natural language processing (NLP) and soft computing. It performs incredible feats of analysis in response to a wide range of prompts.
It can, for example, give a reasonably good response to prompts to write articles or classify complicated data, in a matter of seconds. This has caught the attention of academics and is the latest perceived bête noire by college and university professors.
In the higher education sector, the take-home exam format gained increased usage during the Covid-19 lockdowns and has been retained across many institutions. In the wake of ChatGPT, the knee-jerk response has been to advocate a reversion to the traditional, face-to-face, proctored exam across the board.
However, the exam is only the culmination of the learning done in a module. Part of the educational process in the humanities and social sciences is teaching students to conduct research and write original essays.
On this front, many are concerned about what they see as the diminished ability to ensure authorship of presentations made in essays, the cornerstone of a humanities and social sciences education and the primary tool by which students understand, apply, and synthesize complex concepts are tested.
“How sure can we really be that a student actually wrote the presentation they made and that it wasn’t produced by this cutting-edge AI?” seems to be the main concern.
As with all major innovations, two important aspects have emerged. One sees the AI tool as a threat to education, the other thinks it is a welcome development for the sector.
I fall into the last field. ChatGPT is a good development, irreversible, and its negative aspects can be contained. My brief “experiments” with the tool have not concerned me with this piece of AI, at least in its current version.
Two fundamental reasons are at the root of my optimism. First, and most important, the tool has limitations, thanks in part to the complexity of people as individuals and as a society. Second, and stemming from the first reason, there is no substitute for intellectual engagement. Let’s explore each of these reasons in turn.
First, in my experience, ChatGPT can only give so many answers to the same basic question. Educators can ask questions through the system and identify all possible answers. This is an attractive company and, in fact, possibly a nice one.
In essence, it allows the teacher to see as many possible iterations of possible answers to the assessment questions before students submit them. The professor can then send these responses to their university-mandated plagiarism detection software, with Turnitin being the most widely used.
This alleviates the most common concerns about ChatGPT: that it is not connected to the Internet in real time, and that Turnitin cannot detect plagiarism from ChatGPT, since almost all the answers it can give are unique due to the AI being based on soft computing. .
Soft computing in this case refers to the ability of an AI processor to flexibly “understand” language in the same way that a human would. In practice, it means that ChatGPT can still give responses even when there are typos, which would only cause an error message from a typical algorithm-based AI.
Human nature will also kick in. Students, like all humans, are natural game theorists. That is, they are able to think not only about their actions but also about those of others, in relation to their own interests. Given the task of writing articles on a limited range of topics, many, to the extent that they are rational agents, will refrain from relying on the tool to write their paper, as they would be aware that their classmates would be using the same tool. software.
Teachers should design their curricula and assessments in such a way that the assessments are designed to extract principles and techniques and not just content. They must, like seasoned experts, “get ahead” of their students. This can only help us refine our tools and assumptions about our disciplines and what we teach.
The second reason for my optimism is that AI has limitations that encourage engagement. While it can do impressive things like classify, organize, and comment on complex topics at lightning speeds, AI typically provides short and extremely perfunctory answers to questions, even classic social science prompts like “Write me a topic about Theory X and how”. it can be applied to XYZ, using historical and contemporary examples.”
Faced with such a short, generic, and inadequate answer (which, as noted above, the teacher could initially have them review themselves and put it in a plagiarism detector), the would-be dishonest student has no choice but to investigate further, giving further indications and processing requests.
Through this futile exercise, they are likely to end up doing what they should have been doing in the beginning: conducting honest and in-depth research. Also, the tool cannot cite within the text and does not have access to works published since 2021. This requires the student to engage in the literature and identify relevant sources.
ChatGPT is also just an average student at best. In an article by two psychology professors at Temple University in the United States, they describe how, in response to a question posed to them by an honors level class in their department, AI “cannot distinguish [a] ‘classic’ article in a field that must be cited from any other article reviewing the same content. The bot also tends to keep referencing the same sources over and over again.”
My own experiments with ChatGPT have yielded numerous AI self-contradictions when the questions have been about specialized facts that are nonetheless common knowledge among specialists.
This means that it cannot be used, at least (ironically) not without considerable innovation, to produce dissertations and doctoral theses at the graduate level. Well-trained academics would be able to spot their surface-level shots, among many other mechanisms that exist at that level of education.
Instead of going backwards and withdrawing into the familiar, academics should embrace this new source of change and show dynamism in the way they approach teaching and learning. To do otherwise would be to poorly prepare our students for the complex and ever-changing world that awaits them, and would therefore render the industry increasingly archaic and irrelevant.
In fact, as the philosopher Hericlutus stated over 2,500 years ago, change is the only constant.
Bhaso Ndzendze is a Senior and Associate Professor in the Department of Politics and International Relations at the University of Johannesburg, as well as in its 4IR and Digital Policy Research Unit. His books include Artificial intelligence and theories of international relations (2023) and Artificial Intelligence and Emerging Technologies in International Relations (2021).
Opinions expressed are those of the author and do not necessarily reflect official policy or the position of the mail and guardian.
