Author : Chioma Nwaodike
Affiliated organisation : Internews
Site of publication : internews.org
Type of publication : commentary/blog
Date of publication : 17 March 2020
Lien vers le document original
The Basics: What is AI and where is it used?
Far from science fiction, AI is deployed in almost all the technologies we use daily. Apple devices, Google’s facial recognition system, Instagram face filters, Facebook news feed, Netflix, and Spotify all draw on machine learning to process and analyze large amounts of data in order to improve and personalize products and services. Governments and public entities also use AI for a wide variety of applications. For example, facial recognition is used by police to track and identify criminals or by customs and border control when scanning passports and cities, and municipalities develop customer service chatbots to help with public service delivery.
The AI Digital Divide in Africa
Big Data: Access and Connectivity
Only 28.2% of people in Africa have access to the internet. As data is the prerequisite asset allowing AI systems to function, questions about connectivity and unequal access to data also must be considered. For instance, in Nigeria data is expensive and internet connectivity variable. These disadvantages developers and AI entrepreneurs. Without reliable core infrastructure, affordable data plans, and easy access to technologies, current digital divides will only be exacerbated with the rise and continued advancement of AI. This is especially the case in areas in which access to the internet is limited to low-bandwidth mobile, as many AI applications require faster internet speeds and better software.
Applications that employ AI are largely built-in major languages such as English and Chinese, and as of now, I have not seen any such application that understands many African languages
With slower and less reliable internet access, African developers operate at a competitive disadvantage. Tech companies based in Europe and the US are thus able to draw materials or data from the developing world, and then “import” them back as finished goods (software). In this way, AI risks becoming an extractive industry that pulls resources out of the country to the benefit of others, designing closed source software that often cannot be used by African developers.
Language and Locally Relevant Content:
There are different gaps that exist both across and within the different countries in Africa, between urban and rural areas, and according to gender and language. The connectivity challenge in the region leads to AI produced for non-African languages, with less local content and African-produced open-source AI technology, and a dearth of resources for local training.
To understand these challenges, let me provide an example. If I ask SIRI in my native language about the weather in Nigeria, it is certain that I’ll not receive a correct answer or that SIRI will even understand this simple question. Applications that employ AI are largely built-in major languages such as English and Chinese, and as of now, I have not seen any such application that understands many African languages. If these barriers are not eliminated, with more emphasis placed on AI development for local languages and local content, AI will aggravate the digital divide and exclude individuals from the benefits of artificial intelligence.
In this way, AI risks becoming an extractive industry that pulls resources out of the country to the benefit of others, designing closed source software that often cannot be used by African developers
Policy Framework
Policies enabling African AI development must be rooted in a human rights perspective and respect international norms in addressing the challenges highlighted above. Any meaningful dialogue around policy requires input and collaboration with a wide variety of stakeholders including international bodies such as the ITU and UN, regional African organizations, technology companies from inside and outside Africa, and civil society groups that focus on protecting African citizens and marginalized populations online. Measures such as fairness, accountability, transparency as well as safety and security should be at the core of any technical, ethical, and regulatory framework.
Data Politics and Access: First and foremost, data protection and privacy must be central to the debate around artificial intelligence in African countries. The African Union, in agreement with State members, the United Nations, and other international bodies should unlock the power of data by developing a standard data governance framework that does not discriminate based on region of the world and takes into account existing global inequities in technology production and consumption.
National policies should include principles relating to the production of both proprietary and open-source software, management of data, and affordable access. An appropriate regulatory framework linking data access and protection can create an enabling environment for the development of artificial intelligence in Africa. Public and private stakeholders should work together to develop common resources, databases, platforms, and tools that are open and encourage African technology development.
Connectivity: Connectivity remains a challenge on the continent, impacting the ability of consumers to access AI-enabled applications and services as well as the ability of African developers to build innovative new technologies by and for Africans. To promote more equitable practices and prevent exploitation, there must be renewed efforts to invest in core infrastructure in order to ensure affordable broadband and adequate power.
Local content and language: To increase investment in local content and African languages, national policies must include principles that encourage consultation of local users when developing AI applications. Drawing on rural churches, schools, and community radio stations when developing local content will create an inclusive environment where content is rooted in local communities and culture.
Public and private stakeholders should work together to develop common resources, databases, platforms, and tools that are open and encourage African technology development
Education: The future of AI-oriented work in developing countries necessitates the inclusion of technology into school curricula as well as the development of linkages with higher education programs. It is necessary for students and teachers to build AI-oriented curricula that integrate technology and technological concepts into a variety of contextual learning (science, art, business, etc.). This will enable youth to acquire essential ICT skills in order to promote responsible technology consumption and build African AI industries.
Connecting high-level education programs on AI to core infrastructure tied to community hubs can also offer opportunities for younger Africans to lead the way in African-produced AI applications. Proper education and knowledge of AI systems will give this next generation the skills required to be productive through the “4th industrial revolution” and reduce the high rate of unemployment in the region.
Les Wathinotes sont soit des résumés de publications sélectionnées par WATHI, conformes aux résumés originaux, soit des versions modifiées des résumés originaux, soit des extraits choisis par WATHI compte tenu de leur pertinence par rapport au thème du Débat. Lorsque les publications et leurs résumés ne sont disponibles qu’en français ou en anglais, WATHI se charge de la traduction des extraits choisis dans l’autre langue. Toutes les Wathinotes renvoient aux publications originales et intégrales qui ne sont pas hébergées par le site de WATHI, et sont destinées à promouvoir la lecture de ces documents, fruit du travail de recherche d’universitaires et d’experts.
The Wathinotes are either original abstracts of publications selected by WATHI, modified original summaries or publication quotes selected for their relevance for the theme of the Debate. When publications and abstracts are only available either in French or in English, the translation is done by WATHI. All the Wathinotes link to the original and integral publications that are not hosted on the WATHI website. WATHI participates to the promotion of these documents that have been written by university professors and experts.