In a talk she gave at a BetaNYC conference, Reshma Saujani, founder of Girls Who Code, said, ‘There are always moments in technology where when a new technology came out, we had a lot of fear about the change that was going to come. And our immediate reaction was to basically figure out how we take it away from people’s hands. And the implications of that way that we have traditionally treated technology has wreaked havoc.’ The validity of this statement is evident in data about the global perception of artificial intelligence (AI), specifically Generative AI. In a 2021 survey on the Perceptions of Risk from AI by Lloyd’s Register Foundation, when asked whether they thought AI would help or harm people in their country in the next 20 years, Eastern Africa (51%), Central/Western Africa, and Southern Africa (37% each) had the most pessimistic responses.

According to the same report, the countries most optimistic about AI were leading in AI innovations and investments. These numbers suggest that people's openness to technology makes it easier to explore its potential. When most people strongly oppose a type of technology, it becomes harder for them to adopt it despite its many benefits and opportunities to influence positive growth.

Generative AI is not the first technological innovation to be met with scepticism. Many of what is today considered disruptive technology, such as the Internet, which early users thought would have no use outside of academia, to more recently, cryptocurrencies, which were seen as a fad and platform for illegal financial activities and have even been banned in some countries, are some examples. Despite these initial or still existing reservations, the use cases and benefits of these innovations are the same as those of AI. While it is essential to consider the risks innovation brings, it is also important to ask how institutions leverage them to solve big problems.

The problem with only focusing on the negatives is that while other countries and continents with more optimism towards AI embrace and include it in developing their sectors and economies, Africa remains behind. Even if, eventually, general perceptions begin to change, the continent may still be left playing catch-up or stuck with existing AI technologies built without local contexts in mind. Thus, Africans become customers, yet again, instead of innovators.

While some fears about implementing AI tools are understandable, like the fear of it taking away already scarce jobs or the many unethical ways it has/can be used, it is still important to highlight AI's several benefits and opportunities on the continent. Educating people on how AI could help improve lives and showing areas for possible innovations is important. 

What is Artificial Intelligence and How it Works

In basic terms, artificial intelligence (AI) refers to innovative or advanced technology trained to perform activities previously only possible for humans or the human mind. Such activities include reasoning, problem-solving, learning, interactions, diagnosis, and more.

AI integration and use is as old as the 1950s. The concept can be traced to an old paper by Alan Turing on machines' ability to exhibit intelligent behaviour. The term artificial intelligence was officially coined in 1956 by computer scientist John McCarthy. It has also existed in modern technologies, such as 

• Siri and Alexa on iPhones and Amazon, respectively

• Voice assistants who understand commands and respond with answers or actions

• Chatbots on sites used in place of live support 

• Facial recognition for unlocking devices and identification

• Netflix and personalised recommendations, etc.

The above are long-existing examples of artificial intelligence and show how it has always been present as features in everyday devices.

Although AI technology has recently received significant attention, especially with the advent of OpenAI’s ChatGPT, it has always existed on some level. AI is indeed artificial, meaning it is human-made. It is trained to be as intelligent as possible based on available datasets from which some problems stem. Due to the emphasis on mimicking human intelligence, AI sometimes reflects a computer concept, ‘Garbage in, garbage out’. This means that bias and toxic content generated or exhibited by tools result from the data quality used in training or making this technology. Even though artificial intelligence is not a perfect technology, it still has many benefits and use cases that cannot be ignored.

According to McKinsey’s 2022 State of AI report, the number of organisations globally using some sort of AI technology more than doubled from 20% in 2017 to 50% as of 2022 because they saw positive results in streamlining and optimising processes. This number is only one area, but AI has proven helpful across industries, even in Africa.

Current AI Innovations in Africa: Challenges and Use Cases

Africa is still in the early stages of AI adoption, and several factors prevent it from being one of the top contributors to the global sector. A significant indicator is the distrust of the technology and the unavailability of organised datasets or labelled data. Labelled data refers to data that has been organised and tagged to represent a specific context. Doing this makes it easier to train AI technology on them. For example, images on an app such as Instagram can be classified as unlabelled, except they have a qualifying tag like hashtags that explain what they are or categories they belong to. Data on many African experiences or contexts needs to be updated. Where data for the continent exists, it needs to be more organised and labelled, making it challenging to build contextually relevant AI technology.

Despite this and other critical AI challenges in Africa, such as insufficient infrastructure and skills, some emerging and existing AI technologies are creating positive impacts in their industries. Currently, these innovations are spread mainly across finance and health. 

Here are a few:

• M-Pesa in Kenya uses AI for fraud detection and evaluation of client data, which helps provide personalised solutions, including qualifying customers for micro-loans based on their history.

• Kudi, now known as Nomba, started as a chatbot integration on social media apps. It responds to people’s financial needs and helps them with tasks, from buying airtime to paying bills quickly via text-based messaging. This model worked well, especially for those to whom online banking was inaccessible.

• Ubenwa, a Nigerian AI-powered software, is helping parents diagnose their babies with neurological and respiratory conditions like asphyxia, which is a leading cause of infant mortality, using just their cry sounds.

• Google Ultrasounds is a portable handheld AI solution that can read ultrasound images and provide information. This is important, especially in rural areas and villages with little to no easy access to hospitals and trained sonographers. Jay Patel, Director of Technology at Jacaranda Health, a Google partner and healthcare provider focused on maternal health in Sub-Saharan Africa, believes in AI technology's power to “significantly improve maternal and newborn health and push the boundaries of what is possible in healthcare.”

• A program using AI was used in a Mozambican prison to detect inmates with tuberculosis within five minutes of the scan and quickly eradicate the spread of the disease.

Growth Opportunities for AI in Africa

According to a Statista report, the African AI sector has the potential of reaching $6.9 billion in revenue by 2024. Yet adoption, innovation, research, and even investments are still slow across the continent. The continent is largely underrepresented globally except for four countries, Egypt, South Africa, Nigeria, and Kenya, which contribute most to AI research and development. However, these countries rank very low globally regarding their AI capacity. According to a 2023 report by the Global AI Index of the 62 countries with a level of investment in AI, the earlier mentioned African countries rank at 52, 55, 61, and 62, respectively.

The rankings were based on seven sub-pillars: Talent, Infrastructure, Operating Environment, Research, Development, Government Strategy, and Commercial.

Africa has unique problems that AI can solve while providing employment opportunities and increasing national growth for its countries. The following identified growth opportunities for AI on the continent have been divided into innovation areas and policies. Both are needed to improve the sector's overall performance and are covered in the seven sub-pillars mentioned earlier.

Innovation Areas

There are many innovation areas, including agriculture, where adopting AI tools can improve food production, soil quality, etc. We will dwell on the following two as they form a good foundation for many others.

• African Languages and Natural Language Processing (NLPs):

This component of AI gives computers and technology the ability to understand, interpret, and generate human language. This type of AI technology is typically trained on limited datasets, making them a perfect area for African languages that are grossly underrepresented in AI and tech. 

While Google Translate provides translations for 25 African languages, more is needed. This gap is a crucial innovation area when you realise that English is only the third most spoken language on the continent. Five of the 13 countries in the world where less than ten percent of the population speak English are African countries. According to a Telegraph mapping, the countries are; The Gambia, Malawi, Algeria, Uganda, and Tanzania. The Gambia, a popular tourist destination, has fewer than three percent of English speakers.

NLP tools built to aggregate African languages can preserve the culture. As well as making the internet inclusive and accessible to non-English speakers, it could ease access to healthcare, e-commerce, emergency services, news, and education. In the agriculture sector, NLPs can provide a means for rural farmers to access knowledge and assistance easily in their local languages.  This is a more friendly way to adopt digitalisation while working with researchers like the farmers in Senegal, who use WhatsApp voice notes to collaborate. It can also provide a means of communication for tourists, immigrants, and those looking to learn these languages.

• Data and Research Centers:

Africa needs more companies invested in data and AI research to start filling the existing gaps. According to data, Sub-Saharan Africa accounts for only 1.06% of the total AI journal publications globally, while East Asia and North America account for 42.87% and 22.70%, respectively.

The importance of data and research cannot be overstated in AI development, especially in Africa, where many opportunities and use cases abound. From Google’s AI research centres in Ghana and Kenya to organisations like Lelapa AI in South Africa, The Masakhane Research Foundation, and even Data Scientists Network (DSN) in Nigeria, all are committed to building Africa’s AI talents. DSN also has a mandate to ‘raise one million AI talents in Nigeria.’ Functional research centres encourage skill acquisition, problem discovery, and innovation.

Government Policies

“There remains this idea in some places that AI is evil…people still don’t know if AI is secure and if their data is safe,” Mehdi Sayegh, a co-founder of AI Connect Africa, told when asked about the state of AI on the continent.

Unsurprisingly, the top 10 countries leading in AI research and technology have solid policies and laws protecting data rights and use. Meanwhile, in Africa, there is still little legislation on this. A  2020 paper on AI deployments in Africa reported that only 17 out of the 55 African Union (AU) states have enacted comprehensive data protection and privacy legislation. So far, only Kenya has an AI-related Bill, the Kenya Robotics and Artificial Intelligence Society Bill, 2023. However, many have claimed it seems like an extortionary means for taxing new AI companies rather than a bill protecting people’s rights, data, and privacy.

For the integration of artificial intelligence to progress on the continent, policies must encourage an innovative environment and be more protective of data, privacy, and even labour, both foreign and local. Government policies are an essential growth opportunity factor that, if done well, could attract investments and improve innovations on the continent.

The African market has much more catch-up to do than its counterparts in other continents. The potential for Artificial Intelligence in Africa is numerous across sectors, from policies to innovations, education, investments, etc. However, achieving these potentials will be difficult and take longer if people do not see or understand how this technology can improve their lives.

References

The Lloyd’s Register Foundation World Risk Poll: “A Digital World: Perceptions of risk from AI and misuse of personal data”. [https://wrp.lrfoundation.org.uk/LRF_2021_report_a-digtial-world-ai-a nd-personal-data_online_version.pdf].

Paxful University: “List of countries where Bitcoin is banned”. [https://paxful.com/university/bitcoin-banned-countries-list/].

McKinsey: “The state of AI in 2022 - and a half decade in review”. [https://www.mckinsey.com/capabilities/quantumblack/our-insights/the-state-of-ai-in-2022-and-a-half-decade-in-review].

Founders Factory Africa: “In An AI Powered World, Africa’s Data Dearth More Apparent Than Ever”: [https://www.foundersfactory.africa/blog/ai-powered-world-africas-data-dearth-more-apparent].

Ade-Ibijola, A., Okonkwo, C. (2023). Artificial Intelligence in Africa: Emerging Challenges. In: Eke, D.O., Wakunuma, K., Akintoye, S. (eds) Responsible AI in Africa. Social and Cultural Studies of Robots and AI. Palgrave Macmillan, Cham. [https://doi.org/10.1007/978-3-031-08215-3_5].

Medical Xpress: “AI use in Mozambique jails spawns new hope in TB fight”. [https://medicalxpress.com/news/2023-11-ai-mozambique-spawns-tb.html].

The Global AI Index: [https://www.tortoisemedia.com/intelligence/global-ai/].

Statista: “Artificial Intelligence - Africa”. [https://www.statista.com/outlook/tmo/artificial-intelligence/africa#market-size].

World Data: [https://www.worlddata.info/].

The Telegraph: “Mapped: Where to go if you can’t be bothered to learn the language”. [https://www.telegraph.co.uk/travel/maps-and-graphics/mapped-english-speaking-countries/].

Techopedia: “Top 10 Countries Leading in AI Research and Technology in 2024”. [https://www.techopedia.com/top-10-countries-leading-in-ai-research-technology].

Rest of World: “WhatsApp voice notes are revolutionising farming in Senegal”. [https://restofworld.org/2023/whatsapp-voice-notes-farming-senegal/].

African Business: “ChatGPT and the future of African AI”. [https://african.business/2023/01/technology-information/chat-gtp-and-the-future-of-african-ai].

African Business: “AI: the African opportunity”. [https://african.business/2024/04/technology-information/ai-the-african-opportunity].

Data Guidance: “The Kenya Robotics and Artificial Intelligence Society Bill, 2023”. [https://www.dataguidance.com/sites/default/files/the_kenya_robotics_and_artificial_intelligence_society_bill_2023.docx.pdf].

TechCabal: “Kenya’s AI and robotics bill proposes a $6,250 fine for unregistered entities”. [https://techcabal.com/2024/02/08/kenya-robotics-and-artificial-intelligence-society-bill/].