Positive impact: AI can serve as a tireless tutor available 24/7 that adapts to each student’s pace and learning style and could even offer
explanations in their native language. Photo: File
In the Tragedy of King Lear, through the character Gloucester, William Shakespeare wrote: “So distribution should undo excess and each man have enough.”
Written over four centuries ago, these words resonate with haunting clarity today. Inequality has long been a marker for humankind. It has been a persistent shadow across eras and epochs. While we often think of our journey through innovation, this has not necessarily been to the benefit of all.
We have inevitably marched forward but we have not sufficiently allowed all of humanity to enjoy the formidable markers of our journey. In the digital age, inequality takes a different bend. The digital divide denotes a gap in access but it is actually far more insidious than mere connectivity. Beyond access to technologies, it encompasses divisions in access to skills, digital literacy and opportunity.
The World Economic Forum (WEF) asserts that there are three divides: the AI revolution risks deepening global inequality by widening the gaps among companies, workers and countries. These divides create a chasm between communities that can shape their digital futures and those for whom technology remains an impenetrable barrier. The impact is astounding. AI deepens many of our existing divisions, from amplifying biases to concentrating power and even threatening to leave entire populations behind.
Yet, as artificial intelligence (AI) deepens many of our divisions, we must look to its promise as a solution through the lens of societal impact. This broad digital shift unfolding before us calls us to empower humanity. This may seem at odds with many narratives around AI. We are surrounded by headlines proclaiming superintelligence, by dystopian warnings of job displacement and by utopian promises of effortless abundance.
We hear too often about AI replacing doctors, eliminating teachers or surpassing human creativity, thus making us wholly irrelevant. Perhaps our error lies in viewing these structures and systems solely through their technical capabilities. When we shift our lens from technical capability to societal impact, an entirely different landscape emerges. We begin to see AI as a tool for democratising access to knowledge rather than a replacement for human intelligence. Herein lies an opportunity to liberate people and create space for meaningful contribution.
Let me be concrete about what this means in practice. Consider education. In many parts of the world, quality education remains inaccessible. In South Africa, for example, millions of young people are excluded not because they are unwilling but because structural barriers prevent them from participating meaningfully in these opportunities.
These barriers include underfunded schools, limited digital infrastructure, socio-economic exclusion and even the legacy of apartheid spatial planning. For those of us at the helm of higher education institutes, it has been apparent that AI offers tangible ways for us to close this access gap. For instance, AI can serve as a tireless tutor available 24/7 that adapts to each student’s pace and learning style and could even offer explanations in their native language. Critically, of course, this AI tutor should not replace the human teacher.
The goal is not automated education but rather amplified human teaching that reaches more students more effectively. AI could also be used to reach a wider audience. For example, we know that developing more massive open online courses (MOOCs) could be one solution to access challenges, funding constraints and skills gaps. It is by removing geographical, financial and physical barriers that online education provides opportunities to those who might otherwise be excluded from traditional academic pathways.
Or consider healthcare. So many die from preventable diseases because they lack access to medical expertise. AI diagnostic tools, which can be run on basic smartphones, can analyse symptoms and suggest treatments. This brings specialist-level knowledge to remote clinics staffed by community health workers. AI-powered ophthalmology tools can detect diabetic retinopathy from a simple retinal scan, enabling early intervention in communities that may never have access to an ophthalmologist.
Algorithms trained on chest X-rays can identify tuberculosis with accuracy comparable to expert radiologists, which is critical in regions where TB remains a leading killer but diagnostic infrastructure is scarce. Voice-based symptom checkers can conduct preliminary health assessments in local languages, triaging patients and helping community health workers prioritise urgent cases. Much like my example of education, the goal is not to replace the doctor but to extend their reach. We must go further. For instance, we have seen promising initiatives where AI is being deployed to predict disease outbreaks by analysing patterns in social media alongside environmental factors. This gives health systems precious time to prepare and respond.
Language models can now provide health guidance via text message in dozens of African languages. These innovations work because they are designed with local contexts in mind. They account for intermittent electricity and limited internet connectivity. They respect local health practices while complementing them with evidence-based medicine. They are built in partnership with the communities they serve, rather than being imposed from outside.
Or consider small-scale farmers, who represent the majority of the world’s food producers. These farmers lack access to agronomists, soil testing laboratories, market information, and weather forecasting services. They farm with knowledge passed down through generations, which is increasingly insufficient as climate change disrupts rainfall patterns and introduces new pests.
The consequences are devastating and risk threatening food security for entire nations. AI can analyse soil conditions and weather patterns while providing personalised recommendations in local languages through simple voice interfaces. These systems can alert a farmer in a rural area that their maize crop shows early signs of fall armyworm infestation and recommend organic pest control methods before the damage becomes catastrophic.
They can advise farmers on optimal planting dates based on hyperlocal weather predictions while accounting for the specific microclimate of their particular plot. They can also inform farmers about fair market prices in real-time. This serves as a way to democratise knowledge that was once available only to large industrial farms. It must be noted that the role of AI is not to tell the farmer that their ancestral knowledge is wrong. But it is a tool to enhance that knowledge with information they couldn’t previously access. The result is a powerful synthesis of indigenous knowledge and technological innovation, which creates agricultural practices that are both productive and sustainable.
From my vantage point in higher education, I recognise that we are witnessing new frontiers of knowledge in terms of what we know and who has access to this knowledge. This democratisation of expertise is perhaps the most revolutionary aspect of inclusive AI.
Empowering humanity in the age of AI means increasing the capacity for participation and understanding. In this regard, innovation serves the collective well-being rather than narrow interest. To bridge the digital divide, we must cultivate inclusive AI innovations by creating systems that are accessible and equitable.
Inclusive AI begins with representation. The data that feeds our systems must reflect the plurality of the human condition across languages, genders, racial dynamics, geographies and cultures. Here, communities shift from passive subjects of technological experimentation to active co-creators in the design and deployment of AI systems. This, of course, requires investment not only in infrastructure but in digital literacy, education and capacity building.
By equipping people with the skills to critically engage with technology, we move from mere access to meaningful empowerment. Bridging the divide also calls for ethical frameworks that prioritise transparency and fairness. Governments, private actors and civil society must collaborate to ensure that the benefits of AI are distributed equitably. This could involve encouraging open-source innovations, supporting localised AI solutions tailored to specific cultural and linguistic contexts, or promoting cross-sectoral partnerships that democratise both knowledge and opportunities.
If designed inclusively, AI can reimagine what equity means in the 21st century. Crucially, this reimagining cannot be confined to universities and traditional knowledge institutions. Higher education, of course, plays a vital role but we must resist the temptation to position ourselves as the sole architects of this transformation. We must do away with any notions of ivory towers.
To refer back to some of my examples today, this means recognising a subsistence farmer’s knowledge of their land or a community health worker’s understanding of local health practices. This is how we move closer to Gloucester’s long-echoed vision of a world where “distribution should undo excess”.
Letlhokwa George Mpedi is the vice-chancellor and principal of the University of Johannesburg.