Avo Solutions
By Nthabiseng Mhlanga, UJ Social Impact Assessment Master’s Candidate
April 19, 2024
Edward Snowden once remarked, “Every person remembers some moment in their life where they witnessed some injustice, big or small, and looked away because the consequences of intervening seemed too intimidating. But there's a limit to the amount of incivility and inequality and inhumanity that each individual can tolerate. I crossed that line. And I'm no longer alone.” This reflection resonates profoundly as we contemplate the role of Artificial Intelligence (AI) in addressing the entrenched inequities in South Africa's higher education system.
As crises like the COVID-19 pandemic forced a shift to remote learning, the conversation has frequently circled around the promise of educational technologies to bridge gaps in access and opportunity. Scholars and researchers argue that AI, a technological marvel capable of optimising tasks and facilitating unprecedented learning experiences, holds the key to democratising education. Pedro et al. (2019) define AI as intelligence exhibited by machines that optimise their likelihood of accomplishing preset objectives, and if implemented thoughtfully and ethically; it has the potential to mitigate existing disparities and promote equity in higher education.
Indeed, AI offers the potential to revolutionise learning experiences and enhance academic outcomes. It promises to democratise education, making learning more accessible irrespective of geographical boundaries. Leading scholars like Professor Tshilidzi Marwala highlight AI's capacity to increase productivity and create job opportunities, ostensibly heralding a new era of educational equity. However, such discussions often overlook the deeply rooted inequities within our education system.
AI's implementation in South African higher education must be examined through critical lens informed by structural injustice theories and Black feminist scholarship. This perspective reveals how AI, despite its advanced capabilities, risks perpetuating historical legacies of structural injustice and inequity. The apartheid-era barriers, compounded by ongoing socio-economic disparities, continue to limit the educational opportunities for marginalised individuals. Marginalised individuals refer to ‘those from low socio-economic backgrounds, rural areas, and individuals with disabilities, who have been historically disempowered and oppressed by influential and discriminatory groups in various forms, including economic inequality, political disenfranchisement, and restricted access to education, healthcare, housing, and employment (Pedro et al. 2019; Messiou 2012; Mowat 2015:457). These barriers are not mere vestiges of the past but active obstacles that AI alone cannot dismantle. Often, discussions about equality and access focus on the idea that enhancing access to technology-based educational tools will reduce inequity in higher education (Hall et al. 2020). This implies that AI could potentially eliminate disparities in higher education. However, AI should not be viewed as a solution for dismantling these injustices, its 'intelligence' is derived from existing biases and prejudices. AI learns from the data and behaviours that society provides. Numerous instances show that AI bots and 'smart' recruitment/selection tools can perpetuate learned racism and discrimination. For example, ChatGPT.
Holstein and Doroudi (2022) assert that AI systems have the potential to reduce achievement gaps by scaling the benefits of personalised tutoring. Yet, this potential clashes with the harsh reality of financial constraints faced by marginalised individuals. The underlying question is not just whether AI can bridge the equity gap, but how marginalised individuals will afford (and access) AI tools amid existing financial, educational and systemic hardships.
The critical issue lies in our failure to acknowledge the systemic and social barriers ingrained in South Africa's socio-economic and political fabric. These barriers manifest as inadequate funding for institutions serving marginalised communities, disparities in the quality of education and limited financial aid for low-income students. Without addressing these foundational inequities, AI's promise remains unfulfilled.
Moreover, the integration of AI requires robust infrastructure, including smartphones, laptops, internet connectivity, and reliable electricity—resources that are scarce for over 50% of South Africans living in poverty https://theconversation.com/how-current-measures-underestimate-the-level-of-poverty-in-south-africa-46704. These communities often lack basic amenities like proper schools, houses, and consistent water supply, further exacerbating the digital divide.
To harness the true potential of AI in higher education, we must first confront and dismantle these systemic barriers. Policymakers, educators, civil society, and critical scholars must collaboratively address how technology, especially AI, can be leveraged to promote equity and access for marginalised groups. Only by creating an equitable foundation can we hope to reap the benefits of AI in education.
In conclusion, while AI holds remarkable promise for transforming higher education, it cannot be viewed as a panacea for deeply entrenched educational inequities. We must first rectify the historical and systemic injustices that impede access to education. Only then can AI serve as a genuine catalyst for educational equity in South Africa.
References Hall, J., Roman, C., Jovel-Arias, C. and Young, C., 2020. Pre-service Teachers Examine Digital Equity Amidst Schools' COVID-19 responses. Journal of Technology and Teacher Education, 28(2), pp.435-442. Holstein, K. and Doroudi, S., 2022. Equity and Artificial Intelligence in Education. In: W. Holmes, R. Huang, F. Miao, eds., The Ethics of Artificial Intelligence in Education. 1st ed. Routledge, pp.151-173. Messiou, K., 2012. Collaborating with Children in Exploring Marginalisation: An approach to inclusive education. International Journal of Inclusive Education, 16(12), pp.1311-1322. Mowat, J.G., 2015. Towards a new conceptualisation of marginalisation. European Educational Research Journal, 14(5), pp.454-476. Pedro, F., Subosa, M., Rivas, A. and Valverde, P., 2019. Artificial Intelligence in Education: Challenges and opportunities for sustainable development.
