Don Recommends FG Invest in AI Readiness Infrastructure
Oluchi Chibuzor
A university don in the department of Computer Science, Lagos State University, Prof. Oluwatoyin Enikuomehin, has recommended that the federal government invest in building in AI readiness infrastructure.
This was as he urged the Africa as a continent to be intentional in building AI datasets as institutional datasets.
Speaking at the 127th LASU inaugural lecture series, Enikuomehin, said that societies have always been shaped by their dominant technologies.
He said that societies that benefit most from artificial intelligence will not be those that simply adopt it fastest, but those that adopt it most thoughtfully.
“Refusing the comfortable illusion that because a system is digital it is therefore neutral is the beginning of genuine Al readiness. Protect the most vulnerable from algorithmic harm.
“Invest in AI readiness infrastructure, universal broadband connectivity must be treated as public utility. Digital literacy programmes must reach every segment of the population. High-quality African datasets must be collected, curated, and published with open access.
“The algorithmic society is being designed now. The question is not whether Nigeria will be part of it. The question is whether Nigeria will help design it or simply inherit what others have built. Nigeria has both the right and the responsibility to contribute to, not merely consume, the global Al order.
“Every graduate of a Nigerian university should be able to engage critically with Al systems in their professional domain, not merely as users but as commissioners, evaluators, and challengers.
“Universities should begin building the mathematical and physical science foundations for quantum literacy now, through research programmes, postgraduate training, and international partnerships,” he said.
The lecture themed, ‘Thinking Machines And The Future Of Thinking: Cyberphysicalism, The Transformation of Human Thought And The Emergence Of An Artificial Intelligence Driven Algorithmic Society’, posited that AI represents far more than another technological innovation.
According to him, “Societies have always been shaped by their dominant technologies. Agriculture produced agrarian societies. Steam engines produced industrial societies. Computers produced the information society. Artificial Intelligence is now giving birth to something fundamentally different: an Algorithmic Society.
“Throughout this lecture, I have argued that Artificial Intelligence represents far more than any other technological innovation. It extends cognition, reconstructs expertise, reshapes institutions, and transforms the relationship between biological intelligence and computational intelligence. If these propositions are correct, then one important consequence follows naturally: society itself cannot remain unchanged.
“Artificial Intelligence is now initiating another transformation, one unlike all previous revolutions because it does not simply change how we work. It changes how decisions are made. Increasingly, algorithms influence choices that were once made exclusively by human beings. They recommend, predict, prioritise, optimise, classify, generate, and advise. In many situations, they now participate directly in decision making.”
He stressed that Africa accounts for approximately nineteen percent of the global population, yet African language content in the major datasets used to train the world’s most powerful Al systems is estimated at less than one percent of total training data.
“Africa must build AI datasets as Institutional assets. Every clinical interaction, every agricultural observation, every educational intervention within university-affiliated institutions is a potential data point for training Al systems that serve African contexts.
“The proportion of top-tier Al research publications from African institutions is approximately two percent of global output. The proportion of global Al compute infrastructure located on the African continent is less than one percent. These numbers are not merely embarrassing. They are warnings.
“Consider a Nigerian hospital that deploys an Al diagnostic system for malaria, one of the most common and most deadly diseases on the continent. The system was trained predominantly on patient data from Europe and North America, where malaria presentations are rare and often atypical. When deployed in Lagos, the system encounters presentations that are routine to every Nigerian physician but statistically unusual in its training data. It misclassifies. It under-diagnoses. Patients are harmed.”
