The state of AI in South Africa

A look at some of the most recent developments in artificial intelligence in South Africa, the key challenges facing the sector and the long-term prospects for growth.

As part of the ongoing process of rebuilding after decades of political unrest and economic uncertainty, companies across South Africa are engaged in the development of a nascent artificial intelligence (AI) sector. So, what have been some of the most recent developments in the AI sector in the country? What is the long-term potential for South Africa to emerge as a global centre of expertise in AI development?  What strengths and weaknesses does the country currently possess in this area? And how best might IT professionals contribute to the ongoing development of the country as a global hub? 

Growing sector

In recent years, a growing number of AI companies have emerged in South Africa - and there are signs that the country is gradually building up a solid base of expertise in the technology.  One interesting example is Cape Town-based AI Software and Solutions company Cortex Logic, which works across a range of industry sectors to augment and automate legacy processes, ultimately transforming them into intelligent systems via its AI Engine.  As Jacques Ludik, Founder and Managing Director at the company, and also President of the Machine Intelligence Institute of Africa (MIIA), explains, the Engine works by solving ‘strategic and operationally relevant problems by unlocking the value offered by the fourth industrial revolution and mobilizing data science, Internet of Things (IoT), Big Data and analytics'.

Another example is AI start-up DataProphet, a machine learning-based solution that analyses and interprets a company's data using a prescriptive AI system - and enables production teams to eliminate defects and scrap, and minimise downtime by prescribing the optimal parameter control limits and set points.  As Isaac Matsa, Country Sales Manager at DataProphet, explains, the company now boasts a team of 40, typically made up of postgraduates in mathematics, statistics, computer science or engineering.

"It's our vision to be the global leader in AI for manufacturing in Industry 4.0.," he says.

In pursuit of this aim, the company has recently raised funding from Knife Capital, giving it what Matsa describes as the ‘runway to expand operations internationally' and a significant boost to its ‘growth and expansion efforts into Europe, North America and South America'.

Other notable examples include Sandton-based outfit Xineoh, which has created a platform that is capable of predicting consumer behaviour with AI, and CLEVVA, a technology company specialising in decision navigation - helping to guide staff, customers and digital workers (or bots) through desired decision-making journeys so they ‘reach the right outcomes in a consistent, compliant and context-relevant way'.

‘Huge Potential'

When it comes to technological developments, Matsa believes that South Africa has been ‘overlooked' for some time - although he is also keen to stress that Cape Town, sometimes referred to as ‘Silicon Cape', is recognised as Africa's tech hub. 

"There is huge potential in this market and a lot of highly skilled talent in South Africa, especially data scientists and computer engineers.  When you compare this globally, resources are also much more cost effective than in other markets," he says.

"One of the things that drive the founders of DataProphet is to show that we can compete globally, at a Silicon Valley level, with local talent," he adds.

Despite this potential Matsa admits that the country still needs to address a number of weaknesses - including the fact that investing in South Africa is ‘complicated for an international investor from a legal standpoint'.  Moreover, in order to achieve ‘global leadership', he reveals that his company needs to sell its services to large manufacturing markets - all of which are ‘based overseas and are typically in high-cost environments'.

"Raising capital in South African currency while growth costs are in Dollars or Euros is challenging.  For a South African company to grow in this increasingly competitive environment, it is important to raise foreign capital," he says.

Elsewhere, Vian Chinner, CEO at Xineoh, also offers a rather guarded assessment, and argues that the long-term potential for South Africa to emerge as a global centre of expertise in AI development is ‘minimal', for several reasons.  To begin with, he believes that few South Africans possess the ‘standards and work-ethic required to compete with Silicon Valley giants'.

"This may be due to a lack of exposure to these competitors, or any other reason. The problem is not that we do not have the capability to get there, but there will need to be a radical shift towards productivity and excellence if South Africa is to be competitive in this," he says.

Chinner also stresses that examples of so-called ‘angel investors' are thin on the ground in South Africa, primarily because individuals ‘are taxed at a very high comparative rate, leaving the upper middle class with much less money to invest'. In addition, Chinner points out that every employee in a South African start-up must ‘perform like a superhuman because the competition is so brutal'.

"Ten to twenty companies are usually pursuing an idea that just became possible. In the end, there will be only one incumbent remaining. In the USA you give an employee not pulling his weight two weeks' notice. In South Africa, it's a massive process to get rid of an errant employee, and coders earn a lot of money. If you do it wrong and get a CCMA judgement against you, it will be in the R2 million range, and it can bankrupt the start-up," he says.

"Our education system is [also] not up to standard in the fields that I know, namely computer science and mathematical modelling.  It's possible to graduate most of our universities without having touched Python, which is the industry standard for rapid prototyping of machine learning and AI models," he adds.

Thinking Differently

In Ludik's view, there are numerous opportunities for IT professionals and AI companies to contribute to the ongoing development of building a global centre of AI expertise in South Africa - and he believes that one of the most powerful ways to make such contributions is by delivering ‘impactful domain and application-specific African solutions to some of Africa's unique problems'.

"Some of these solutions, such as those in fintech, healthcare, agriculture, and education, can also be scaled across the African continent and to other parts of the developing world. Other contributions include educating the market on AI and its applications," he says.

He also points to the key role of organisations like the Machine Intelligence Institute of Africa (MIIA), that aims to ‘empower the African community to help build a strong and innovative AI and Data Science community in Africa' and ‘help transform Africa by networking together the critical mass of resources, promote and sponsor learning activities, and strengthen scientific and technological excellence, mentoring and collaboration on the continent'.  MIIA has also recently initiated an to maintain a comprehensive country-level review of AI and machine learning resources, people and projects in Africa.

Meanwhile, Ryan Falkenberg, Co-CEO and Co-Founder at CLEVVA, argues that South African IT professionals have an opportunity to ‘think differently about AI and automation' - chiefly because developing countries are ‘specialising in forms of AI that exclude human involvement, and assume an Internet of Things'.

"Within Developing Countries, this is in many cases not practical in the short to medium term. There are too many breaks in the digital eco-system, and we still rely heavily on humans to fill these gaps. A specialisation on both-and solutions, where humans and digital workers are seen more as partners than competitors, could see us differentiate globally," he says.

"In other words, if our starting premise is to include human capital in our automation thinking, we will develop innovative solutions that realise this outcome," he says.

Ultimately, Chinner argues that IT professionals, and the private sector in general, should ‘stay positive and focus on the things it does have control over' and ‘foster a culture of excellence and raise the standards of production in South Africa'.  He also argues that the RSA government can ‘completely change the playing field' by addressing four key points: making tax rates more internationally competitive, reforming labour legislation - with cities or provinces designing their own labour laws, improved foreign exchange control and making university funding ‘contingent on closer collaboration with the private sector on curriculums'.

Meanwhile, Matsa points out that, with investment into Industry 4.0 becoming a focus for many economies, there is now a demand for new ideas.

"The manufacturing sector has experienced a huge shift in the fourth industrial revolution and through AI, we have helped organisations achieve the optimal state of production. AI enables a substantial improvement to traditional analytics available to manufacturers.

"It would be great to partner with IT professionals at the forefront of innovation through machine learning, governments and industry bodies globally to share best practices. This could be through industry events, working groups - or even networking through online forums to achieve what we did not know was possible," he says.