My column this week is the fourth part of my recent paper for the NTU-SBF Centre for African Studies at Nanyang Business School, Singapore, where I am a research fellow. References are in the original article.
On-the-ground AI research to bridge knowledge gap & incorporate African factors
Artificial intelligence (AI) is the ability of a digital computer or computer-controlled robot to perform tasks commonly associated with intelligent beings. AI research falls into one of three categories: Strong AI, applied AI or cognitive simulation. Strong AI aims to produce machines with human-like thinking abilities. Applied AI or “advanced information processing” seeks to build “smart” systems for lucrative or life-saving vocations like medical diagnosis or stock-trading. Cognitive simulation involves using computers to test human cognition theories.
Global tech firms IBM and Google have established AI research labs in Africa; the former in Nairobi, Kenya and Johannesburg, South Africa and the latter in Accra, Ghana. They are largely applied AI research ventures. These AI initiatives mitigate fears about the likelihood of Western-designed AI systems excluding African preferences and cultures. They also help ensure African researchers participate in global AI efforts; which were hitherto all based outside Africa.
Some African AI researchers abroad are seizing the opportunity to return to their homeland. And a next generation of African AI researchers would also now be able to avail themselves of the knowledge and experience without having to venture afar. African governments should be mindful of the ethical motivations of these AI research ventures by global tech firms, however.
Challenges and constraints
Infrastructure remains inadequate, data costs are high, and internet penetration of below 40 percent remains relatively low, logistics can be nightmarish, many Africans remain relatively digitally illiterate, and the regulatory environment is uncertain or even antagonistic.
Online marketplaces have been filling the digital skills gap with foreign labour. To manage regulatory and financial risk, major online marketplaces are not only registered outside of Africa but are also tapping foreign capital markets for funding. Both online suppliers and consumers have genuine cause for concern. For example, African governments have been shutting down the internet during elections and protests. The ease with which governments are able to do so exposes the fragility of the internet on the continent. Part of the appeal of the internet is its supposedly “borderless” nature. However, if governments can stop the usage of the internet on a whim, its infrastructural utility diminishes.
Perhaps the only distinction is that such behaviour is not limited to African countries; Russia is building a sovereign internet wall, for instance. Besides, even when the shutdowns are not motivated by government action, the fact that most Africans use the internet via their mobile phones, through networks that are relatively fragile, means that they easily lose access to the internet from time to time due to fault-induced shutdowns.
External private actors have also been able to shutdown at least one African country’s internet infrastructure; in Liberia, for instance. Still, many are able to defeat these shutdowns. And legal institutions have surprisingly proven to be resilient in keeping government overreach in check in some African countries.
In any case, calls are being made for some form of global regulation of internet activities. And despite the obvious drawbacks of overregulation and censorship, some African governments are undaunted. They are increasingly looking to tax online activity or have some form of regulatory oversight over the internet in their jurisdictions. For example, Uganda has imposed a social media tax. More recently, Uganda announced plans to nationalise its internet data exchange service, raising fears of inefficiency and likely regulatory overreach in the aftermath.
Conclusion
The internet is a net job creator. Cheaper smartphones and mobile data, a mobile money boom, and burgeoning e-commerce on the African continent are spurring incremental economic growth and creating jobs. Digital IDs, smart and balanced regulation, and certainly much cheaper smartphones and internet would be key elements to achieving these goals on scale. African governments should ensure all their citizens have digital IDs, can make payments digitally, and in collaboration with global tech giants, are able to acquire low-cost smartphones and use the internet cheaply. In the aftermath, they would all be able to participate in online marketplaces as suppliers or consumers.
There is already an example of how private enterprise and governments can accelerate the process. In India, a largely cash-based society, an entrepreneur is already leveraging on the country’s now well-established digital identity system to include all Indians in the digital economy. Recognising high data costs are a constraint for poor people, the entrepreneur is offering data services on his mobile telecommunication network almost for free and almost certainly at a loss in the interim.
His strategy is clearly to formalise all Indians, irrespective of class, in a digital ecosystem that enables them conduct almost all their daily activities, from socialising to buying groceries, online. While his ambition is not an ideal scenario, in light of obvious antitrust issues, it exemplifies the potential of the internet and how the digital ecosystem it creates would spur incremental economic activities and create jobs.
African governments have a good example in this Indian case on how the internet opportunity can be operationalised on the one hand, and the potential monopolistic drawbacks on the other. In tandem, government policies should prioritise human capital and infrastructural development. Entrepreneurs should also be able to secure financing easily. These would only be possible if African governments create a conducive business environment in their respective jurisdictions. The job creation thesis of the internet rests on these foundations.
Rafiq Raji
