[Column] Patrick Ndegwa: Homegrown intelligence - How AI in Africa is taking shape
In February 2022, the Economic Commission for Africa (ECA) and the Government of the Republic of Congo inaugurated the African Research Centre on Artificial Intelligence, an initiative committed to researching digital technologies and applications. The centre – the first of its kind on the continent – will facilitate education and skills development for the integration of new AI-driven solutions, promote economic growth, and reinforce continental ownership of digital tools and resources.
The centre marks a turning point for AI on the African continent. While a great deal is being done by big tech entities and international companies, the field is still open to ‘Africanisation’, as in contextualising and implementing it locally to enable and explore our own applications. By keeping up to date on the latest trends, while also understanding how they and the underlying technology work, we can determine how AI on the continent will look and feel.
The emergence of AI mechanics and technologies
During the last decade, the world at large has shaken off pop culture ideas of how AI will manifest. While we may someday have robot assistants that do all the housework or even fully-automated vehicles speeding along our highways, today’s manifestations of the technology are built on fundamental concepts that seem rudimentary but are enabling us to automate, digitise, and strengthen business operations.
Current AI applications are much more dependent on data. By combining datasets and computer science, the scientific field seeks to solve problems by mimicking human-like decision-making based on the information it is given. Enabled by machine learning (ML), the result is automated and relevant to both consumers and enterprises, such as speech recognition, automated customer services, data consolidation, and digital asset management. Cloud computing especially plays a role when it comes to establishing a secure and networked environment in which these AI-enabled systems can manifest.
To that end, the current-day pursuit of AI development is primarily focused on the ability to gather, store, and process data. This requires infrastructure: dedicated and robust networks that can accommodate more and more data essential to enterprise operations.
The application of those technologies
Businesses are under increasing pressure to adopt the technologies of tomorrow. The race is one to acquire systems and solutions that both embody 4IR principles, and empower them to offer products and execute actions that consumers and enterprises will consider essential in the medium-to-long term.
AI plays a central role in these systems and solutions, and companies are embracing it. In a 2021 McKinsey survey, 57% of business respondents in emerging economies reported AI adoption in at least one function, up from 45% in 2020, while past surveys show the most common functions include service operations, product and service development, and marketing and sales.
According to Gartner, small and wide data, used to strengthen AI analytics and produce more detailed and situational results than big data can, is one trend that is driving innovation in AI. By 2025, 70% of organisations will be compelled to shift away from big data to small and wide data – a means to make AI less data dependent and improve analytical processes.
This trend shows that even established concepts such as big data are being surpassed, and it’s essential that companies are equipped with the appropriate resources and capabilities to respond to a changing environment.
If we are to envision AI in Africa, it starts with surveying the landscape, with a focus on digitisation. UNCTAD reports that in several African countries, digital economies serve as one of the main drivers of growth and account for more than 5% of GDP. Coupled with the fact that more than 80% of the continent’s population has a mobile phone subscription, there is and continues to be massive potential for digital services, which also have the potential to be underpinned by innovative AI-driven systems.
Alongside initiatives such as the African Research Centre on Artificial Intelligence, we see real innovation in African startups and enterprises that work with AI to develop new solutions. For example, insurtech platform AiCare uses algorithmic data to produce mobility insights by way of telematic systems, a new way to gather and process data about how Kenyans behave on the roads. Another startup, Ai Kenya, aims to bring industry experts and professionals together to create a unified approach to expanding the technology in the region.
AI in Africa requires the input of stakeholders, including government, Internet service providers, entrepreneurs, data scientists, and all professionals operating within the field to produce new and cutting-edge data-driven African solutions to African problems. It’s only by working together that AI can manifest real change for the continent and all who live on it.