Balancing the AI scales: Does the potential for AI in Africa outweigh the current risk?
Vodacom is a pioneer of social, digital and economic transformation, in the markets in which it operates, through technological innovation and is now leading the discussion and practical application of the evolving use and governance of Artificial Intelligence and big data analytics to drive a step change. In a recent report by McKinsey & Company, AI-enabled advancements indicate that Africa could contribute up to as much as US$1.2 trillion to the global economy by 2030, representing a 5.6% increase in Africa’s GDP. Although Africa tends to lag behind the developed countries in the adoption of emerging digital technologies, there is significant potential for AI in Africa both from an economic and humanitarian perspective.
The magnitude of this potential is largely due to the diversity of benefits AI offers especially to assist with the attainment of the United Nations Sustainable Development Goals. For example, AI was used by academics at the University of Johannesburg to predict the first peaks of the Covid-19 pandemic in South Africa. Public health policy makers and officials were able to use this information to quantify the impact of government interventions and plan ahead. Similarly, Vodacom used AI and big data analytics to assist government in curbing the spread of Covid-19 after a superspreader event in the Free State, predicting the movements of attendees and preventing further spread, using movement modelling. This intervention was a game changer at the onset of the pandemic.
Major holiday periods are a prime example where we leverage AI and big data analytics to assist with mobile telephony network planning and managing increased congestion when people migrate from one province to the next, to ensure we continue to provide acceptable standards of quality of service. While these major holiday periods are happy times for most, the number of incidents on roads tends to increase due to drunk driving, fatigue, and reckless driving. Using AI and big data analytics, government officials may be able to identify problem areas better, placing officers in strategic locations to manage the circumstances which lead to these fatalities.
AI and big data analytics also have the potential to completely revolutionise how we address access to education and educational outcomes by identifying learners who are struggling or likely to drop out early. This would give educators the chance to intervene, providing additional academic and emotional help.
As AI evolves, we are seeing an explosion of generative AI tools such as GPT-4, AlphaCode, Bard, Claude, ChatGPT, just to mention a few. AI’s ability to generate text, images, and other media has been both a blessing and a curse for education, media, and entertainment. In the entertainment world, people have started using AI to generate lyrics and soundtracks with minimal effort. For example, the late rapper Ricky Rick, who sadly took his own life, was able to pass on a strong positive message about mental health from beyond the grave through AI. Ricky’s family, working with his industry partners, used AI to create a song using words from his social media posts. This development and creativity is phenomenal but also raises a number of ethical and legal issues, including regarding intellectual property rights.
Generative AI is a fantastic tool that can be used effectively for research and proofreading. Although generative AI can complete and/or complement some human brain functions, it still lacks the ability to identify disinformation and struggles to manage copyright requirements when pulling from a source and therefore, circumspection is required when using these tools.
Data privacy and confidentiality are other areas of concern in the use of AI and big data analytics. Due to its open-source nature, any individual who puts their data into a generative AI platform risks that information being stolen or compromised. This, again, is an issue that has arisen due to lack of comprehensive regulation and has made some people distrust anything that makes use of AI. Another area of concern is bias (including unconscious bias) in the use of AI that perpetuates racial and gender stereotypes through biased profiling.
Several countries will be going to the polls globally in 2024 and AI and big data analytics may play a central role in elections. Used responsibly, AI and big data analytics may enhance administrative processes in elections by processing vast election data which is managed by electoral bodies and how politicians disseminate their campaign messages to their constituencies. However, in the absence of regulation, the useability of generative AI during this period has been pulled into question. Its vulnerability to manipulation means many are asking questions about the potential of AI driving perceptions through disinformation and perpetuating biased opinions that do not reflect the views of the electorate.
Due to lack of effective regulation on AI and big data analytics globally, issues of privacy, confidentiality, copyright, ethics, and prejudice are subject of major debates. The European Parliament seems to be on course to finalise the regulation of AI. Furthermore, several governments around the world are debating ways to manage the use of AI either through broad national legislation or sector-specific regulation. It is prudent that each country encourages the responsible and ethical use of AI, based on its cultural, social and economic nuances to promote amongst others, fairness, safety, transparency, equity and equality.
As a champion for the responsible use of AI and big data analytics, Vodacom strives to manage its use of these tools through self-regulation as we embed them in our operations within the company and in our offerings to customers. Recently, following guidelines we use on big data analytics, the company has developed guidelines on the use of generative AI to promote innovation and to manage the above-mentioned concerns around ethics, prejudice, copyright, confidentiality and privacy.
As advocates of technological innovation, we are aware of the pitfalls in the use of AI and big data analytics and the need to manage these tools with circumspection, and therefore, the concerns should not outweigh the benefits- instead, debates on effective regulation, whether self- or government sponsored are critical to find long lasting solutions that will not stifle innovation.