OLORI LOLADE SIYONBOLA
A Brief History of Artificial Intelligence in Africa
Artificial Intelligence has transformed our world irreversibly, but the most dramatic impact of this technology remains to be felt. Whether in the AI-for-profit world or AI-for-good, artificial intelligence will continue to transform the way we live at breakneck speed. Unfortunately, most of us don’t understand it fully, nor do we recognize the full scope of its power, which is why the risks associated with this technology are staggering, as are the potential opportunities for AI to improve the well-being of the majority of the world’s population. The role of Africa – her people, her Diaspora, her cultures – in the development of artificial intelligence has long been understated. But a new generation of radically visionary computer scientists is making it impossible to continue to erase Blackness from the AI landscape.
Due to the increased recognition of the work of Black AI professionals like Timnit Gebru, Joy Buolamwini and Meredith Broussard and films like The Social Dilemma – there’s a major push currently to make AI more equitable and just. The bias coded into artificial intelligence programming by the overwhelmingly white male mass of AI developers across the industry has real-life implications for BIPOC, which can ultimately mean life and death, particularly with policing and in the case of self-driving cars. Buolamwini’s 2020 film, Coded Bias, highlights these disparities.
Re-Framing the Tech Narrative
Google, arguably the largest AI company on earth, has come under fire recently for purging itself of leading AI ethicists, those who advise on its AI policy. Following the major scandal surrounding their firing of Ethiopian-born Eritrean AI scientist, Timnit Gebru, a number of other ethicists were unjustly forced out of the company, signalling a rogue era in Silicon Valley. In sync with the African tech startups exploding in popularity and economic power, a few organizations are working to dislodge white privilege in tech by increasing opportunities for Black AI engineers. Black in AI (https://blackinai.github.io/#/ ), co-founded by Gebru, is the leading platform for Black professionals in the field of AI. Alliance4ai, founded by Alex Tsado, takes this work a bit deeper by cultivating networks of young AI professionals in Africa and in the Diaspora and providing them with access to AI education rooted in African history and cultural pride.
Teaching their members about Africa’s innovative past makes a huge difference in Alliance4ai’s retention of talent and in the strength of their AI networks. Most people of African descent entering technology professions struggle to overcome the intimidation of an industry that appears to be owned and run by prototypical white males, with the occasional Asian representative.
Most of us have internalized the overarching Western media narratives painting white male technologists as innovators, and everyone else as beneficiaries of their “genius”.
Rarely is credit given to the African origins of the technology on which AI is based or to the people of African descent (PoAD) responsible for advancing AI throughout the world. This piece, the first of our series on Artificial Intelligence, will break from this tradition.
Artificial Intelligence Technology Began in Africa
As long as we’ve had divination systems in Africa, we’ve had what is now known as artificial intelligence. There is a long history of scholarship linking the spiritual technology of African divination systems with the digital technology of computer science, both having a basis in advanced mathematics. Our divination systems have always been based on data and binary combinations, which is exactly what Google is: an amalgamation of an insane amount of data being queried by different algorithmic combinations. In the Ifa divination system for instance, a binary code (algorithmic combination) is used to determine which Odus (books of knowledge) must be accessed for a particular inquirer.
Ron Eglash’s 1999 book, Modern Computing & Indigenous Design provides a foundation for studying the algorithmic systems across African architecture, textile design, divination, hair design and so much more. On divination specifically, a 2013 article by André Croucamp refers to the practice as a form of information technology that organizes and accesses information. Also in 2013, a paper by Alamu, Aworinde and Isharufe referred to the Ifa divination system of the Yoruba as “an ancient binary computer system which has successfully linked the probability of numbers with the complexities of the human condition and the ever complex flux of events”. The Ifa divination system relies on the 256 Odu Ifa which are believed to represent all the possible scenarios that humans can experience. This is similar to the maximum 256 combinations believed to be possible in binary mathematics in computer science.
In the Portal to African Intelligence, we’ll deepen our exposition on the intertwined history of African Intelligence and Artificial Intelligence, and further show how modern-day AI is built on African Intelligence.
Why We Need the History
With the above foundation, it stands to reason that digital technology is not something that was introduced to Africa by Westerners, but rather that there was a direct relationship between the two worlds that produced all the technology that we have today. Note that none of our devices would work without the Coltan mined in the Congo under inhumane conditions, that IBM has been working with engineers in Nigeria since at least the 1970s, and note too that a Nigerian named Philip Emeagwali invented the internet.
Understanding the African origins of AI and computer science should help Africans, African Americans and the larger African Diaspora take on, with urgency, our role in reclaiming this technology for the benefit of mankind – in pursuit of Sustainable Development Goals – rather than for the profit of a select few. It is only through this reclamation that we can have more influence on the tools we use, which are directing the fate of our world.
Our generation must come to understand the potential of AI to either Liberate us fully or re-enslave us permanently.
The AI Visionaries of Africa & her Diaspora
While Western AI ecosystems ignore Black talent, Africans in places with even the least internet penetration are developing AI solutions that increase food security, prevent deadly diseases, and reduce crime. Companies like Tambua Health and Agrix Tech have attracted Western eyes and dollars due to their utilization of deep tech to solve universal human problems. But they didn’t get there by themselves.
Many of the African AI innovators on the Continent have been empowered by their Western counterparts, People of African Descent who refused to accept the Western narrative that effectively erased Africa and Africans from the AI landscape. These include some of the aforementioned top talent from Google, Facebook, NVIDIA, IBM and many other leading AI companies. They’ve laid an impermeable, highly intelligent foundation for the development of a modern AI ecosystem in Africa, and continue to shape the future of AI globally through their thought leadership and the institutions they are developing. For the above mentioned historical data about these trailblazers, I’m indebted to Alex Tsado, founder of Alliance4ai, who first got me excited about AI in Africa, and schooled me on the ecosystem.
The Africans of AI’s Past & Present
When IBM research scientist Aisha Walcott headed to Kenya in 2013, she became the first in a long string of high profile Western-trained talent sent to develop the AI infrastructure on the Continent. Aisha is a Clark-Atlanta and MIT grad whose most notable work in Kenya focused on using deep learning to address heavily congested traffic in Nairobi.
Her team “developed innovative intelligent transportation systems data capture methods and analytical tools, to provide computational understanding about the local driving and infrastructure context.” Meaning they used machine learning to study traffic patterns so that they could predict, and thus reduce, traffic. They were ultimately able to reduce commuter time by 10%. Aisha is still at IBM Africa as the research manager for AI Science and Engineering team, working on Water-AI-Healthcare innovations.
In 2014 Tunisia, Karim Beguir and Zohra Slim founded Instadeep, an enterprise AI startup using advanced machine learning techniques including deep reinforcement learning to bring AI to applications within an enterprise environment. Their mission is to democratize machine learning in Africa and they were one of the first ever African-founded AI companies to publish original AI research at NeurIPS, the largest AI-related conference in the world.
They’ve since become a major player on the AI world stage, with their platform accelerating business growth exponentially for their clients. Instadeep has partnerships with AI leaders like NVIDIA and Intel; and with universities like Oxford, MIT and University of Michigan. Their high-growth culture and visionary mission has caused many to liken them to the Google of Africa–they’re one of the most competitive employers and a coveted place to work for many youth on the Continent. (Source)
The next significant event in AI’s modern history in Africa was the founding of Black in AI by Moustapha Cisse, Timnit Gebru and Rediet Abebe in 2017. Cisse, a Senegal native, was an AI research scientist at Facebook at the time. Gebru and Abebe, who is Ethiopian, were at Microsoft at the time. Gebru has a BS, MS and PhD in Electrical Engineering from Stanford, and Abebe a BA and MS from Harvard, with a PhD from Cornell. They founded Black in AI to increase Black representation in the field, and subsequently founded Deep Learning Indaba the same year–adding Ulrich Paquet and others from DeepMind to the team–to strengthen Machine Learning and Artificial Intelligence in Africa. Deep Learning Indaba is Africa’s biggest conference for AI researchers, with equal representation from the US and Africa. (Source)
This is also the same time period in which Alex Tsado joined NVIDIA’s cloud computing department to establish their GPU platform–which enables AI–on public cloud platforms, and supported the companies building the future of AI on NVIDIA’s platform. This is when he noticed the glaring erasure of African contributions to AI, and harm towards its people introduced by the exclusionary nature of the industry.
In 2018, Stanford released their annual AI index with zero data on Africa, other than a mention of the Deep Learning Indaba founded by the Black in AI group. The exclusion of Africa from the mainstream AI conversation angered Alex, and propelled him to establish the Alliance for Africa’s Intelligence (Alliance4ai) as a center for information and network cultivation alongside expert Data Scientists Nicholas Litombe (Microsoft) and Linet Kwamboka (Mozilla) and Nuclear Scientist Nomso Kana.
Also in 2018, Moustapha Cisse was tapped to open the Google AI lab in Ghana to “provide developers with the necessary research needed to build products that can solve problems that Africa faces” (CNN)
2020 was the first time African AI researchers had a voice at the UN’s AI for Good Summit, which is the largest platform in the world for strategizing on how to use AI for the benefit of the greater good. The African presence was largely secured and coordinated by the Alliance4ai team in partnership with South Africa’s AI Expo Africa’s AI Media Group.
Thanks to these brave early adopters and thought leaders, we’re seeing a shift in focus for artificial intelligence to be used for the benefit of the overwhelming majority; and we’re seeing an exponential increase in AI policy development on the Continent.
Since 2018 especially, there has been an explosion in the development of AI communities, and startups, and hackathons.
Toward a Liberated AI-powered Future
The most valuable product on the market today is not Bitcoin or Tesla stock. It’s not even gold. The most valuable commodity of our generation continues to be our data. And the race to own the world’s data is on. We have entire governments (especially China) investing heavily in machine learning and AI, even outside of their country, to facilitate the distribution of their ideas and drive markets in their favor. We have leading AI companies like Google, Facebook and Twitter investing in AI in Africa for the same reasons, as well as what scholars are calling Digital Colonialism. Africa’s innovators and leaders must note that the greatest value of AI and big data is in its ability to rapidly improve governance, end food insecurity and increase access to low cost medical care.
this article first appeared in noirpress 26 July 2021