Search
QR Code https://file.pide.org.pk/pdfpideresearch/discourse-2024-3-26-ai-and-educational-elitism.pdf

AI and Educational Elitism

Publication Year : 2024
Author: Bina Shah

It’s up to global South educators to harness AI as a learning tool without damaging scholarship in an already fragile educational environment

Bina Shah

What sci-fi junkies and technology experts have been dreaming about for decades is finally here: artificial intelligence that possibly equals or even surpasses human intelligence. Inevitably, AI is going to disrupt learning and education as we know it. Some educators are excited by this opportunity; others, understandably terrified. This is a critical time for educators to familiarize themselves with AI’s capabilities so that the technology serves the classroom, not the other way around. And the need is most urgent in classrooms outside of the Global North, or the developed world.

Global South educators must take control of AI and its tools, understand how they work, and how their students are using them (as students and young people tend to always be ahead in terms of technology use). And they have to do it faster than the time they took to get used to the Internet in the late 90s and early 2000s, because the tech industry is already working on how to monetize classrooms. Sal Khan of Khans Academy envisions a personalized AI tutor for every student and personalized AI teaching assistant for every teacher in the world. More educational tools will crowd the market; the economic incentive, combined with the way AI can be misused will make a mockery of assessment  — imagine students using AI to write papers and educators using AI tools to grade them. It will replace human interaction with AI-assisted shortcuts, creating an isolation that the pandemic already showed us was harmful to students’ mental and emotional health.

In the Global South and elsewhere, high school and undergraduate students already use AI to write their essays from start to finish, especially in postcolonial and decolonized countries where a European language was imposed on a non-English speaking population. Nadia Agha, Professor and head of the Gender Studies Center in Shah Abdul Latif University, Khairpur, notes that students have always plagiarized the work of others, but are now using ChatGPT for everything – writing proposals, conducting research – without the critical skills needed to evaluate the accuracy or even the logic of the written output.

In Pakistan, the Higher Education Commission treats the use of AI as plagiarism and only recommends the use of plagiarism software to check it, which is a flawed and faulty way of dealing with the problem. Universities around the world are scrambling to formulate sound AI policy, while others simply mandate that it should not be used at all.  But as Agha observes, research candidates are not trained well in critical thinking and academic writing; AI is already negatively influencing Global South researchers and graduate students, who use it to produce substandard research papers and dissertations that fail to meet a global standard of excellence for inclusion in the worldwide academe.

Stephen Lyon, Dean of the Faculty of Arts and Sciences at Aga Khan University in Karachi, recognizes that AI poses a threat to the acquisition of critical thinking and analytical skills. But he is equally confident that AI will level a playing field that has been skewed for many years, for international and working-class undergraduate students who have been marginalized from non-vocational education. He believes AI’s generative writing tools can help undergraduate students revise and rewrite their theses and documents, giving them the thinking space to develop intellectual skills often taken for granted in elite universities.

In Pakistan the research environment has already been “fashioned by incompetent faculty who are improperly trained to supervise students,” according to Muhammed Z. Ahmed, a Pakistani PhD who works with the USDA. It’s the same in other Global South Countries: “Western publishers rate our work lowly, and do not publish our work as easily as they publish from elsewhere,” says Augustine Gitonga, an academic writing editor and librarian at the Aga Khan University in Nairobi. He is not anti-AI, but he prefers the use of AI tools like Grammarly and Writeful to support students’ writing, and help them develop skills, rather than using Generative AI like ChatGPT, which “weakens the actual capacity to do real work.”

Mathew Barber, a historian at the Aga Khan Institute for the Study of Muslim Civilizations and its Center for Digital Humanities, has been deeply engaged in the discussion on AI in education. In his field, the ability of AI to read and interpret large sources of data across languages, text types and systems, is a significant form of knowledge creation that nevertheless requires human oversight. AI can even aid human creativity, resulting in interdisciplinary courses and hybrid subjects for classrooms of the future.

But the dominance of English as the lingua franca of the LLMs, with scant attention paid to developing similar models in non-European languages, poses the danger of increasing the already substantial gap between research in academic rigor and intellectual standards produced in the global North and global South. Educators in the MENA, African, and SWANA regions face the prospect of AI distorting that gap to the point that it can’t ever be bridged.

David Chojnacki, an international educator with 5 decades of experience who works in Near East and South Asian education as a government consultant, believes that as educational AI becomes more skilled, people will be more easily satisfied with its results but less concerned with the quality of learning and scholarship taking place in the classroom. Furthermore, in turning over many academic responsibilities to AI, people will mistake intelligence for wisdom.

A more practical use of AI is in the field of vocational training, where global South students can learn to use AI in order to create the generic texts and content needed in business, government, and commercial settings. But this will also exclude those students from the levels of higher education that produces thinkers, leaders, and artists who will use AI in different ways — as a supplement, not a replacement, for creativity and scholarship.

Mathew Barber, who strongly supports the use of AI in academics, recalls how as a student from a non-elite background he entered Oxford University without the skills to write academic papers. In weekly sessions with his tutors, he was made to write papers that were never marked for a grade. In this manner he learned how to write, how to form arguments and defend a thesis statement, to cite sources correctly. This grueling experience engendered a love for academic discipline and excellence that may not have emerged if AI had been there to do the hard work for him.

In countries like Pakistan, where quantity is already valued over quality, AI threatens to overwhelm the educational field completely. The temptation to turn to ChatGPT to write that research paper or essay, especially in English, is too high right now. And educators who themselves may not have had proper training in academic research and writing might even encourage the use of AI among their students, regardless of its flaws and errors. A future where AI robs Global South students of the chance to develop as critical thinkers and human beings will put them at a permanent disadvantage in the wider world.

All educators must address their own fears about AI by learning everything they can to objectively harness it as a learning tool in the classroom, one that augments scholarship without damaging its long-term human advantages. For Pakistan, and the rest of the global South, AI will force us to change the education system, if not to dismantle it completely. If we learn to focus more on processes — the acquisition of learning, the development of critical thinking skills, academic integrity — rather than products (the graded research paper, the essay, the take-home exam), the disruption might actually work in our favor.