Kenya, Africa hurdles in artificial intelligence race

Kenya is still miles behind in positioning itself for the takeoff of Artificial Intelligence (AI) and other advanced technologies in Africa.

Out of Africa's 55 states, only three are well prepared for the uptake of the technology, despite a global shift towards automation, according to a report released recently by Oxford Insights

The survey, Government AI Readiness Index 2020, ranks Mauritius, Egypt, and South Africa the highest on the continent for their early adoption of AI subsets such as Machine Learning, Deep Learning, Computer Vision, Neural Networks, Robotics and Natural Language Processing.

Out of 172 nations that were involved in the research, Kenya came at position 71 in a ranking index dominated by developed economies.

Mauritius is at position 45, followed by Egypt at position 56 while South Africa is ranked 59.

Other African countries that are showing commendable progress in spreading the awareness of the technology are Seychelles (68th), Tunisia (69th) and Rwanda which closes the number of nations from the continent in the top 100 list at position 87.

The survey paints the picture of a continent still struggling to recognise emerging technologies as a key driver in the global seismic shift of industries towards the Fourth Industrial Revolution where less manual work, cost and time are becoming the defining factors for every successful economy.

"These results should not come as a surprise, as African countries have historically lagged the rest of the world in technological developments. Despite the different generations of technology that African countries have 'leapfrogged,' they still face the persistent challenge of catching up, as the pace of technological change outstrips their leaping abilities," states the report.

The top five countries in the world are the United States, the United Kingdom, Finland, Germany and Sweden, underpinning a continued dominance by Western powers in the Fourth Industrial Revolution (4IR) arena.

Sub-Saharan Africa has the lowest score, followed by Latin America and the Caribbean, and South and Central Asia.

The major hindrance to the implementation of AI projects across the continent has been the lack of digitised data upon which AI algorithms are built.

Kenya, for instance, which has a government AI taskforce, has in the past years struggled to convert data stacked in traditional formats in both public and private data silos, despite a higher penetration of digital products in the past decade compared to neighbouring countries.

Mauritius remains the only country in Sub-Saharan Africa to publish an AI strategy, doing so in 2018 with a long term plan to support its Blue economy through a maritime Internet of Things (IoT).

Egypt unveiled its national AI Strategy in 2019, with the core aim of training personnel and supporting research and development of the technology.

South Africa is yet to do the same, but the government, just like Kenya has set up a presidential commission on the 4IR to come up with a proper plan for AI implementation.

But while Africa has made efforts to ease access of data, it is largely not in machine-readable formats such as XLS, CSV, JSON and XML. This is the format used to create algorithms for self-driving cars, weather forecasting, crucial life-saving medical gears, robots and automation of business operations.

However, the survey reveals that at least 86 African companies in 17 countries are using 4IR technologies in the agricultural sector by 2020, AI being pivotal.

Over the last decade, Kenya's total value of investment in AI is estimated at Sh13 billion, South Africa at Sh165.8 billion and Nigeria at Sh60.3 billion, according to Microsoft's Artificial Intelligence in Middle East and Africa 2019 Outlook report.

Though these figures indicate optimism for AI growth in Africa, the global AI market was Sh15.8 trillion in 2020, according to the International Data Corporation (IDC), with projections of global revenues surpassing Sh30 trillion by 2024, with a five-year compound annual growth rate of 17.1 percent. The continent is thus contributing and benefitting the least from AI.

Founder of Harare-based technology and energy company Econet, Strive Masiyiwa says the trillions in the global AI market will be split by the United States and China, with both nations presumably taking 70 percent, with the current trend.

"Of the remaining 30 percent, the African bloc should strive to get at least 10 percent. Let's create our own version of Silicon Valley, our own platforms whilst including everyone," he remarks.

Over 50 African countries still lack AI strategies, but the African Union's AI Working Group met for the first time in 2019 to develop a continental approach to AI and exchange expertise between countries.

"Collaboration in this manner could help countries develop AI strategies, identify other regulatory and governance issues, and learn from regional best practice," the report states.

In Africa, the report adds, limited preparation of appropriate regulatory and ethical frameworks is evident, and governments themselves generally have low use of emerging technologies.

Rwanda expects to have a data protection policy in place by 2020, expecting it to indirectly address key AI-related governance issues. Kenya's Data Protection Act of 2019 only takes care of data privacy, totally leaving out AI, with the two becoming increasingly inseparable.

The academic and entrepreneurial communities are particularly active. A mapping of emerging AI hotspots in the Global South identified 148 players in nine sub-Saharan African countries, mostly in academia.