Is Data Science, Big Data, AI and Machine Learning blowing up in South Africa?
Economically advanced nations are leading a frenzied obsession with artificial intelligence and massive data sets, but has the furore affected South Africa yet?
We spoke to Geoff Nitschke about the state of South Africa’s development in data-centred and artificial intelligence technologies. Geoff is a senior lecturer in Computer Science at the University of Cape Town. Here are the key insights:
South Africa is lagging behind major economic powers, but…
Unsurprisingly, South Africa is years behind superpowers of technological development in data science and AI-related fields. In the US, you can study data science at countless universities and colleges. All industries are led by major companies that are poaching top data science talent, and opening up exploratory development departments with the expressed goal of bringing the special powers of AI, and machine learning in particular, to bear on their business practices. This is true of markets in other major economies across Europe and Asia, to varying degrees. It is now common to see predictions that China will dwarf America’s big data and AI scene, since it has enormous data pools and its government is aggressively stimulating growth in this area. And across the world, people are cutting out formal training altogether by using online resources like algorithms, tutorials and data sets, to up-skill themselves, since centres of higher learning are struggling to keep up with rapid developments in these technologies. The demands from businesses are outstripping the abilities of many universities and colleges to provide, but online resources are giving people the tools they need to fill roles in often newly-created capacities within companies. South Africa is nowhere near this level of market sophistication, but Nitschke sees plenty of evidence that a catch-up has begun in the country.
FinTech is getting interested, and interesting
The money in South Africa is following the international trends. Large banks like ABSA and Nedbank, and insurance companies are expanding and refocusing their research divisions to explore data science opportunities, according to Nitschke. Investment management companies such as Allan Gray are doing the same. Other industries are also looking to data science, but Finance broadly contains the stand-out examples of businesses looking to use the likes of machine learning to better understand their clients, and their own business models, in order to provide the best and most profitable products and services to their customers. This naturally leads to a rapidly rising demand for skilled data scientists, programmers and the like.
South African society continues to have a very traditional view of education. Since so few people have opportunities to study at universities, and there is such a dire skills shortage across all industries, the common and reasonable belief is that formal education is the undeniable guarantee of a successful career, which it is. This leads talented students to universities to pursue traditional, long-existing vocations like medicine, accounting and engineering. Newer kinds of work are not trusted or valued in the same way, so computer science and data science have not yet come to rival the older, more illustrious paths of study. Many universities don’t have the tools or personnel to offer computer science courses anyway. General lack of affordable access to computer hardware and software, and the internet, for most South Africans has meant very low rates of computer literacy.
This is a problem for South African industries that are yearning for skilled young programmers to incorporate into their research divisions. In the first world, higher education networks are meeting the demand because so many universities are providing computer science and data science. The alternative in the first world is people up-skilling themselves through online materials. It is rare to find South Africans doing this, because so much of the population has no access to the internet, very little education, and very little computer literacy either.
All of this means that industries that are trying to create the kinds of developments in data science and AI use overseas are struggling to find the workers they need to do so.
Industry is pushing the universities to change in South Africa
Nitschke says that the excitement overseas, and the demand from major companies in South Africa, has driven universities to data science and AI research and teaching. He says that data science is still quite new in many of the academic programs in the country, but there are cases of some universities incorporating data science into their offering. Nitschke mentions that the University of Pretoria offers data science at undergraduate level and also offers some Masters topics. The University of Cape Town is starting some data science-related projects through Nitschke and his colleagues.
This is vital progress because many South African companies, like those in Finance, have enormous data sets, covering millions of customers, that can be crunched to uncover knowledge that ought drastically improve the effectiveness of business.
Start-Up Culture is small, but crucial, and growing
Nitschke gives one significant reason why some South African industries have been slow to embrace the seemingly obvious opportunities that data science and AI work can provide: market dominance.
The vast majority of the country’s industries are ruled by a few huge, all-powerful businesses. We have a few large banks, a few large cellular service providers, and few large retailers etc, and all of them hold enough of the market to feel fairly comfortable with the ways they have been doing business. As a consequence, many big companies don’t feel a pressing need to do exploratory work that costs resources, even if it may have a big pay-off in the long run. They are happy with what they are doing now, and don’t feel under threat from new business models, even though they are aware that they may be threatened in future.
This is why Nitschke believes the burgeoning start-up scenes, especially those that are expanding in Cape Town and Johannesburg, are so important. Start-ups typically make value by creating and adopting new and ingenious products, business models and business strategies. Against behemoths with far greater resources, this is the only leverage with which they can compete. Start-ups are more likely to give employees the agency and power to incorporate the latest data science work into their companies. This forces competitors to take notice and respond to newer, better practices, and thus strengthens industries as a whole.