Artificial intelligence (AI) has become one of the key technologies globally, and at Digital Nigeria, it was one of the main topics of the three-day conference. A panel on Day 1 that included Paycode’s Ayanda Luthuli discussed how African countries could use AI in payments, digital identity, data management, and more.
Financial inclusivity is one of the key goals of African governments, and fintech startups are leading the campaign to bank the marginalized. However, as Ayanda revealed, they face a great challenge with identity management, especially in rural areas.
AI is transforming the sector and making tasks like identity confirmation easy, fast, and cheap.
“I see AI scaling payments,” Luthuli, who is the chief technology officer at Paycode, told the audience at the Abuja International Conference Center.
For Michael Collins, the CEO of AI and machine learning firm Periculum, it’s all about automation. One of the biggest cost burdens for African firms is the cumbersome manual processes. AI allows users to automate several processes and leapfrog many legacy systems that drag entrepreneurs down with unnecessary costs.
While there was a lot of talk about AI disruption, Daniel Doll-Steinberg believes that it will be a transformation instead. Based in London, Doll-Steinberg is the co-founder of EdenBase, an investment company whose portfolio includes 3D firm Threedium.
“I think disruption will be felt in the Western world where we have enormous constraints in how we develop and implement these technologies.”
He dismissed the ‘hype’ around AI, describing it as genuine excitement at the possibilities that the technology offers.
Data is the foundation of AI
“You can’t talk about AI without talking about data,” stated David Edoja, the CEO of Analytics Intelligence.
Edoja revealed that having worked with African companies and governments for four years, a major challenge has been poor data collection and storage, or a lack of it altogether. For the continent to fully leverage AI, there needs to be a change in how data is collected and managed, he believes.
“Only 0.04% of our digital data is [freely] accessible—the rest is either unstructured or hidden,” Doll-Steinberg said.
With AI, how data is structured and presented is also critical. Several studies have proven that AI suffers from bias, but it’s only because the data set used to train the model was skewed against some groups of people.
With bias in AI being a major challenge, how do we ensure inclusivity as the technology explodes?
“The key, now that the tools of AI are available, is to democratize AI,” says Ayanda.
Collins added that the bias in AI is nothing new; Africans face a much bigger challenge interacting with voice assistants like Siri as the data used to train these systems is predominantly from Westerners.
The solution, he says, is building technologies “for us, by us.” This way, Africans can have inclusivity ingrained from the go, and the systems can accommodate the nuances of African users.
“We need more companies that are building for our people—people that talk like us and look like us. This is the only way we will have an ecosystem that creates inclusivity.”
To achieve this, Analytics Intelligence is launching an AI lab and incubator supporting African AI developers, Edoja announced. The company is also engaging in massive data collection for Afro-centric AI training.
Watch: AI truly is not generative, it’s synthetic
New to blockchain? Check out CoinGeek’s Blockchain for Beginners section, the ultimate resource guide to learn more about blockchain technology.