Determined to use her skills to fight inequality, South African computer scientist Raesetje Sefala set to work building algorithms that mark poverty hotspots – developing datasets that she hopes will help target aid, new housing or clinics.
From crop analysis to medical diagnostics, artificial intelligence (AI) is used worldwide for essential tasks.
But Sepala, pictured, and a growing number of other African developers are pioneers tackling the continent’s specific challenges.
Vanir-exodus knowledge is vital for designing AI-driven solutions that work, Sefala said.
“If you don’t have people with different experiences doing the research, it’s easy to interpret the data in a way that will marginalize others,” said the 26-year-old from her home in Johannesburg.
ALSO READ: Looking to the future
Africa is the world’s fastest-growing continent and tech experts say young homegrown AI developers play a vital role in designing applications to address local problems.
“Bringing Africa out of poverty requires innovation and this could be revolutionary as Africans themselves are doing things for Africa,” said Cina Lawson, Togo’s minister of digital economy and transformation.
“We have to use advanced solutions to our problems because you won’t solve problems in 2022 with methods from 20 years ago.”
Digital rights groups warn about the use of AI in surveillance and the risk of discrimination, but Sefala said it could also be used to “serve the people behind the data points”.
She mapped every suburb and town in SA and combined this dataset with satellite data and machine learning algorithms to record the growth of these neighborhoods over time.
Refining the algorithms based on her experiences made the data collected more accurate.
“These kinds of decisions determine who you alienate or include when you build your AI models,” said Sefala, the first AI researcher at the Distributed AI Research Institute, a community-driven research group.
As Covid spread around the world in early 2020, government officials in Togo realized that urgent action was needed to support informal workers, who make up about 80% of the country’s workforce, Lawson said.
“If you decide that everyone will stay home, that means this person will not eat that day, it’s that simple,” she said.
In 10 days, the government built a mobile payment platform – called Novissi – to distribute cash to the vulnerable.
The government partnered with the think tank Innovations for Poverty Action and the University of California, Berkeley, to create a poverty map of Togo using satellite imagery.
Using algorithms backed by GiveDirectly, a non-profit organization that uses AI to distribute money transfers, recipients earning less than $1.25 (about R19) per day and living in the poorest districts were identified for an instant money transfer.
“We texted them that if you need financial assistance, you need to register,” Lawson said, adding that beneficiary consent and data privacy were prioritized. The program reached 920 000 beneficiaries.
“Machine learning has the advantage of reaching so many people in a very short time and providing help when people need it most,” said Caroline Teti, Kenya-based GiveDirectly executive.
With the aim of stimulating the discussion about AI in Africacomputer scientists Benjamin Rosman and Ulrich Paquet co-founded the Deep Learning Indaba, a week-long gathering that began in 2017 with other colleagues in South Africa.
“You used to go to the best AI conferences and there was no representation from Africa, both in terms of newspapers and people, so it’s all about finding cost-effective ways to build community,” Paquet said in a statement. video call.
In 2019, 27 smaller indabas – IndabaXs – were rolled out, with as many as 300 participants at some events.
One was IndabaX Uganda, where founder Bruno Ssekiwere said participants were sharing information about using AI for social issues, such as improving agriculture and treating malaria.
Another result, from SA, was Masakhane, an organization that uses open-source machine learning to translate African languages not normally found in online programs.
On their site, the founders talk about Ubuntu’s SA philosophy, a term that generally means “humanity,” as part of their organization’s values.
“This philosophy calls for collaboration and participation and community,” reads their site.
Now Sefala has built a dataset of SA suburbs and townships, she plans to work with domain experts and communities to refine it, deepen inequality research and improve its algorithms.
“Making datasets readily available opens the door to new mechanisms and techniques for policy-making around desegregation, housing and access to economic opportunities,” she said.
African AI leaders say building more complete data sets will also help address biases ingrained in algorithms. “Imagine Novissi being rolled out in Benin, Burkina Faso, Ghana, Ivory Coast … then the algorithm will be trained to understand poverty in West Africa,” Lawson said.
“If ever there are ways to combat bias in technology, it’s by increasing diverse data sets… we need to contribute more.”
But that will require more funding and wider access to computer science education and technology in general, Sefala said.
Despite such obstacles, Lawson said that “technology will be the savior of Africa”.
“Let’s use the most advanced and apply it right away, or we’ll never get out of poverty,” she said. “It really is that simple.”
This article was written by Kim Harrisberg