Rooted in the Origin of Species, Charles Darwin makes the argument for adaptation. This, he argues, is a survival mechanism. It is not clear that South Africa is doing enough to survive.
During the pandemic and its aftermath, it became increasingly apparent that the age of artificial intelligence (AI), so often spoken about through a theoretical framework, had arrived. The changes were sweeping.
This was Darwin's very theory in action. As a survival mechanism, the world went online and deployed intelligent technologies to augment human capabilities. Against an extreme context such as the pandemic, resisting this shift would prove to be futile. The breadth and width of these changes have been quite remarkable.
Untold possibilities
AI's transformative nature and its potential merit a lecture in its own right. There are examples and possibilities alike in areas such as healthcare, climate change, and disaster prediction. In healthcare, AI enables personalised medicine, accelerates drug discovery, and improves diagnostics.
In climate science, AI models analyse complex data for better understanding and prediction. Disaster management benefits from AI's early warning and response optimisation. Beyond this, AI is also a catalyst for overcoming traditional barriers.
For instance, in the financial domain, AI solutions promote financial inclusion by offering banking services to underserved populations. In healthcare, AI-driven technologies increase accessibility, especially in remote areas with limited medical infrastructure.
In education, AI facilitates online learning platforms, personalised education experiences, and adaptive tools, thereby democratising access to quality education on a global scale. Youth empowerment is fostered through AI education and training programmes, equipping the younger generation with skills relevant to the digital economy. Gender inclusivity is promoted as AI contributes to equal opportunities in education, employment, and entrepreneurship.
There are untold possibilities. But in our own context, what is the impact of AI – both real and imagined? As South Africa continues to grapple with an economic malaise – one that I must caution has been around well before the impact of the pandemic took hold – we too must adapt.
Of course, the growth and unemployment numbers are often used as markers of this malaise, but it is deeper than this.
Growth is well below 1% and expected to remain stagnant for quite some time, and unemployment is above 30%, and the youth segment is ever-increasing. The census conducted in 2022 and released a few months ago provided further insight into South Africa's development trajectory – notwithstanding the issue of undercounting as 30% of homes and 31% of individuals were not part of the overall data sample.
As the numbers indicated, there are apparent knowledge gaps. Almost one-third, or 31.8%, of individuals aged five years and older attended some kind of educational institution.
Nationally, 86.8% of these individuals attended primary or secondary schools, while a further 5.8% attended tertiary institutions. Only 2.1% of individuals attended Technical Vocational Education and Training (TVET) colleges.
The significant drop in terms of tertiary education is cause for concern. Just last week, it was troubling to learn of companies such as Volkswagen and ArcelorMittal sending out resounding warnings to the South African government about retrenchments.
A recent Investec brief on economic growth suggested that South Africa's distinct lags are attributable to a lack of investment in knowledge, innovation and research and development. Intriguingly, these are the very metrics important in the age of AI.
Striking a balance
This cannot be merely a survival mechanism, but rather, it should be a catalyst, and we must emerge stronger. As I reflected on the topic at hand, I contemplated the very idea of peril. Of course, this term, in relation to our current wave of technology, has been deployed by Klaus Schwab.
As he warns, we must strike a balance between the promise and the peril. And as a person trained in engineering, we must maximise the promise and minimise the peril at the same time. This is multi-objective optimisation and therefore, requires human judgement to determine how much we weigh the promise versus the peril and this is all in the so-called the Pareto frontier.
Of course, by peril the immediate understanding is the potential misuses of AI systems. But what of the peril of not embracing AI at all? In a country such as South Africa, which faces an economic trough, growing poverty and increasing inequality, the danger of not keeping pace with the advancements in AI is profound.
This year has presented an entirely new wave of AI and, with it, novel opportunities. In order to understand this wave, it is imperative to understand what AI is. AI is a technique that essentially makes machines intelligent. While computers traditionally relied on people to tell them what to do and how to react, AI means that machines can learn and make their own decisions.
The basic idea behind AI is to see whether we can give computers some of the decision-making abilities that we as humans have. There are three broad types of AI – prediction machines, clustering machines and generative machines. Prediction machines, such as Artificial Neural Networks (ANNs) or Convolution Neural Networks (CNNs), are designed to forecast future outcomes or make predictions based on historical data.
Clustering machines, such as k-Means, are used for grouping similar data points together based on certain characteristics, and ultimately identifying patterns within data that might not be immediately apparent to humans.
Generative machines, including Chat GPT and generative adversarial networks (GANs), which have definitely seen a surge in recent months, are capable of creating new content, such as images, text, or even music, that resembles human-created content based on algorithms that learn patterns from existing data. This goes beyond text generation.For example, synthetic data is created by algorithms that reproduce some structural and statistical properties of real-world data.
AI as a leapfrog?
It can address challenges such as data scarcity, privacy, and bias issues and raise concerns about data quality, security, and ethical implications. Understanding how the technology works is important as AI gains pace. Over the next five years, we will see a much more rapid adoption of this technology.
As an intriguing OECD report puts it: "While AI augmentation can amplify human potential by improving decision-making, multitasking, and problem-solving abilities, it also poses risks such as decreased self-reliance, reduced critical thinking, and potential loss of certain skills." This statement can be interpreted in various ways.
AI promises heightened efficiency across industries, personalised experiences, advanced medical diagnostics, and innovative solutions to global challenges. However, it also presents concerns such as job displacement leading to economic inequality, biases in AI systems perpetuating social disparities, security and privacy challenges, ethical considerations due to opaque decision-making processes, the potential for social isolation and dependency on technology, and new and pervasive weaponry.
Striking the right balance between the advantages of AI and mitigating its potential negative consequences will be essential for its integration. There is an argument to be made for AI as a means to leapfrog – or bypass traditional stages of development, particularly in relation to economic growth, poverty eradication and social inequality.
In recent years, South Africa has made significant strides in embracing AI. The 4IR Commission proposed eight recommendations, and these can be seen within the context of AI:
Educate South Africa on AI;
Establish the National AI Institute;
Use AI to reindustrialise South Africa;
Develop a Data Institute;
Incentivise the adoption of AI;
Build AI infrastructure;
Educate lawmakers on AI;
Develop implementation capacity.
South Africa has made some strides in its 4IR journey, with notable achievements in various areas.
In terms of infrastructure development, South Africa is investing in critical infrastructure, such as broadband connectivity, data centres, and high-performance computing facilities.
On skills development, initiatives like the Youth Employment and Skills Development Initiative (YESD) and the National Artisan Development Institute (NADI) are focused on developing the skills needed for the 4IR workforce. On innovation and entrepreneurship, South Africa is fostering an innovation ecosystem, supporting startups and SMEs in developing 4IR-driven solutions. South Africa has established the National AI Institute jointly hosted by the University of Johannesburg (UJ) and the Tshwane University of Technology (TUT).
Added complexities
With a burgeoning tech startup scene and a growing number of AI-focused research institutions, the country has started to tap into AI's transformative capabilities. From healthcare to agriculture to finance and education, AI applications are already making an impact, promising more efficient processes and better decision-making. Yet, there are added complexities.
The country must be conscious of the inequities and inequalities that persistently prevail. Social justice demands that we ensure that advancements occur on all fronts without abandoning the citizens of this country and ensuring that access to the enormous benefits of AI is provided.
What is required is responsible utilisation of AI to counter any negative effects, especially in a country with stagnant growth. AI certainly has the potential to catalyse economic growth in South Africa.
Artificial intelligence in South Africa comes with special dilemmas – plus the usual risks
By automating repetitive tasks, AI can enhance productivity across various industries, leading to cost savings and improved competitiveness. Moreover, developing and deploying AI solutions can create jobs in areas such as AI development, data science, and AI ethics, fostering a new wave of skilled employment opportunities. Our education systems need to be primed to deal with these trends in the job market, where new skills of agility, responsiveness, technical savvy, curiosity and others are embedded in the curriculum.
The graduates of the future need a new wave of skills to ensure that they are able to adapt to the turbulence of our new waves and ride these with confidence. This also extends to the workforce, who will witness monumental changes in their work and might need reskilling of a new form and kind.
South Africa must prioritise inclusive education as a transformative force, recognising the public good derived from investments in education. STEM education, especially in underserved communities, is crucial while ensuring equal value for humanities and the arts. Providing AI education empowers individuals for future job markets, reducing economic disparities.
Addressing the AI talent brain drain by enhancing economic opportunities, investing in research infrastructure, promoting AI education, creating a supportive ecosystem, and leveraging the diaspora for knowledge transfer is also imperative.
Our policies also need to be reframed in a manner that speaks to these advancements but also speaks to an ethical approach. While AI enables new technologies that improve efficiency and productivity, it may also lead to increased inequalities among and within countries, and our policies and strategies need to speak to these possibilities, ensuring that there are no gaps in transparency, accountability, safety and ethical standards, which could hinder the development and sustainability of AI.
What is required is regional collaboration to ensure that policy is based on harmonisation, inclusivity and best practices. Moreover, we need to create an enabling environment for open data to boost African AI exchange, innovation and to create markets along with the establishment of AI data as a public asset and a push towards open public sector data to reduce entry barriers and promote AI innovation.
Ethical framwork needed
Then, there needs to be greater investment into AI. This calls for a focus on AI centres that focus on localised applications of AI. Solid investments in facilities and spending for research to facilitate the consolidation of researchers and professors is pivotal to creating an AI-driven economy.
Additionally, there is a need to build AI-specific infrastructure that integrates with existing economic and social infrastructure. We need to look at the generation and delivery of energy, the extension and improvement of water infrastructure and health and educational infrastructure to create a coherent and comprehensive infrastructure network.
It is necessary to leverage AI to address societal challenges, such as improving healthcare quality, enhancing agriculture, and personalising education experiences. Collaboration with chatbots and analytics can ensure real-time student support and extraordinary learning interventions.
These are but a few tangible steps we can take. Importantly, however, we cannot simply forge ahead. Tristan Harris, co-founder and executive director at the Center for Humane Technology, warns that without considering an ethical framework in the development of AI, we risk creating a "digital Frankenstein".
There is a call for us to develop rigorous policies to respond to our shifting context. It has been argued that there was a need to close the gap between policymaking and research in order to achieve better innovation outputs. An important facet of this shift is ethical AI. As AI becomes more integrated into our society, there are several ethical considerations.
Tied to this are concerns about privacy and cybersecurity, which will become increasingly challenging as AI capabilities grow. New safeguards and technologies will be necessary to address these concerns. Striking the balance between these forces, both apparent and emerging, requires us to understand the challenges.
Fairness, transparency, and accountability crucial
Collaboration between AI systems and human oversight will have to converge to ensure the responsible and ethical development and deployment of AI technologies. Many AI systems are trained on biased data, leading to discriminatory outcomes. Addressing bias and ensuring fairness in AI algorithms is an ongoing challenge.
We have to ensure that these systems are inclusive and based on equitable structures and systems. Admittedly, finding the right balance between innovation and regulation to ensure public safety and the ethical use of AI is a continuous struggle.
In this regard, it is necessary to develop governance structures that can adapt to the evolving landscape of AI technology. Ensuring fairness, transparency, and accountability in AI systems will be crucial. The automation of tasks through AI and robotics has the potential to displace certain jobs, leading to unemployment and economic inequality. We have to prepare the workforce to adapt and respond to these changes.
Ultimately, the governance of AI must take into account security concerns, ensure accountability and be guided by regulatory and oversight bodies. If we are to make advancements with AI, we must do so with guiding structures and systems.
There is pressure on us to actively pursue multi-disciplinary as well as multi-sectoral partnerships to foster and promote research in our quest for development with social justice underpinning our agenda.
To return to the thought of Darwin – we must, of course, adapt but we must do so with knowledge. Darwin also said: "Ignorance more frequently begets confidence than does knowledge: it is those who know little, not those who know much, who so positively assert that this or that problem will never be solved by science."
