The impending AI divide

AI may deepen the digital divide, with poorer nations and communities lagging behind in access and benefits.

Listen to article

The power and potential of AI to change our world is now an intensely debated topic. AI proponents are enthusiastic about the potential of this technology to offer fixes to major global problems, while cynics warn of the need to regulate AI. Less attention has however been paid to how AI will impact the existing digital divide which has led information technologies to become another marker which sustains the divide between the haves and the have-nots.

Leaving aside the potential existential fear that AI may become smarter than humans, and try to exert power over us, it is already becoming clear that those who adopt AI technologies will be at a clear advantage compared to those who remain unable to do so. While the internet savvy around the world have already begun tinkering with generative AI tools like ChatGPT, the fact remains that much of the world is yet to even come online.

A so-called 'digital divide' is apparent not only between different countries but also within them. In the US, for example, a third of the population is said to lack access to the internet at fast enough speeds to even use software programmes like Zoom, which need much less data processing capacities than AI-enabled technologies. In the global south the digital divide is even starker.

Those without internet access lose out on business opportunities and remain unable to avail educational or health care facilities. Recall how the decision by many governments to move schooling online posed a significant challenge for schools and students who do not have access to computers and the Internet.

People like tech-billionaire Bill Gates estimate that there will be significant levels of AI usage by the general population in advanced economies like the US in less than two years, and that even resource-constrained countries will be able to benefit from AI without much of a lag-time. However, this optimistic estimation fails to take note of the existing structural inequalities which will invariably prevent such an even spread of AI.

Nearly 2. 5 billion people are estimated to lack access to the internet. Such people already face impediments accessing health care and educational resources, and this gap may grow as we transition towards greater use of AI. Yet, connecting the unconnected is not a problem that the market mechanism and big technology firms will solve on their own. If this were possible, we would not have such a big gap between those who have access to the internet or not, and the even benefits of market-based and capital-intensive growth would have become more widespread by now.

The US Council on Foreign Relations recently stressed the need for overcoming the digital divide by equipping all individuals with essential AI skills, such as understanding how to find a job using AI, how to use Generative AI responsibly, how to manage small business with new AI tools. However, closing the digital divide also needs building the digital infrastructure needed to power a more equitable and productive workforce. How this can happen without adding to the ecological footprint of energy-intensive AI data centres is still an unanswered question. The environmental impacts for enabling AI adoption are yet not fully understood, and the thirst for critical minerals to power AI will likely worsen exploitation usually found at lower rungs of global supply chains. Consider, for instance, the horrific conditions of unregulated labourers mining cobalt - a vital ingredient in running digital technologies - in Congo. Similar problems will crop up in other countries, with underregulated or unregulated economies, including Pakistan.

Overcoming the digital divide needs to avert top-down approaches and focus on making AI adoption more inclusive and sustainable. However, addressing the concerns and constraints of poorer people has not been the primary focus for big technology firms thus far. Hence, AI is more likely to reproduce or even exacerbate existing disparities around the world, instead of disrupting or transcending them, at least in the foreseeable future.