Researcher Develops AI Learning Framework and Platform for African Diaspora Trust Networks

Across the world, technology platforms are designed to solve problems at scale. Yet some of the most persistent problems remain overlooked. Some systems quietly harm the people they are meant to serve, while entire communities operate trust networks outside the digital infrastructure that dominates today’s economy. That is where Oluseyi Olaolu Olukola believes technology needs […]

Across the world, technology platforms are designed to solve problems at scale. Yet some of the most persistent problems remain overlooked. Some systems quietly harm the people they are meant to serve, while entire communities operate trust networks outside the digital infrastructure that dominates today’s economy. That is where Oluseyi Olaolu Olukola believes technology needs to change.

Olukola, a doctoral researcher at the University of Southern Mississippi, is currently working on two projects that address very different problems. One exists in academic research, inside artificial intelligence systems used in education. The other is a platform he is building independently for the African diaspora. What connects them is a single idea. The way a system is designed determines who it ultimately serves and who it may unintentionally harm.

One of the problems Olukola’s research examines is rarely discussed openly within the artificial intelligence industry. Intelligent systems can appear to be functioning perfectly while quietly producing harmful outcomes that nobody measures.

In AI-powered educational systems, for example, algorithms often optimize for signals such as engagement rates, completion metrics, or test performance. When those signals are misaligned with real learning outcomes, the system does not crash or generate errors. Instead, it continues operating normally while gradually degrading the learning experience of the student depending on it. Every metric on the dashboard may still look successful.

Olukola’s doctoral research targets exactly this gap. Rather than adjusting datasets or fine-tuning models, his work focuses on redesigning how decision-making occurs inside these systems. The goal is to introduce architectural constraints that prevent certain types of harmful optimization from happening in the first place.

“The question is not whether the AI is performing well,” Olukola said. “The question is whether ‘performing well’ means what we think it means and who pays the price when it does not.”

Researchers familiar with early discussions of the framework describe the work as part of a growing effort to rethink how artificial intelligence interacts with human learning environments.

While that research unfolds inside academia, Olukola is also building something entirely separate outside the university.

The project is a digital platform designed for a problem that millions of people in the African diaspora encounter daily but that few technology companies have tried to solve properly.

Across North America, Europe and other regions, diaspora communities operate extensive informal economies built on trust and reputation. Electricians, developers, tutors, tailors, builders and accountants regularly find work through community referrals rather than through global digital marketplaces.

The talent exists. The trust networks exist.

What does not exist is a platform designed around how those communities actually function.

Most existing marketplaces were designed from assumptions about how work is discovered, rated and verified within Western gig economies. Those models often fail to capture the relational trust systems that diaspora communities rely on.

Olukola’s platform attempts to translate those trust systems into digital infrastructure. Instead of replacing community networks with anonymous ratings or algorithmic rankings, the platform is being designed around a principle that many technology companies have overlooked. Trust that is controlled and validated by the community itself.

“The trust is already there,” Olukola said. “What is missing is technology that respects how that trust works instead of trying to replace it.”

What makes Olukola’s work notable is not simply that he is working on two different ideas at once. It is that both projects are driven by the same underlying argument.

In artificial intelligence, poorly designed systems can quietly harm students while appearing to perform well.

In digital marketplaces, poorly designed platforms can ignore the social structures communities already rely on and replace them with systems that serve different interests.

In both cases, the architecture of the system determines the outcome.

One project applies that principle to the design of AI systems used in education. The other applies it to digital infrastructure for one of the world’s largest but least formally recognized economic networks. If either effort succeeds, the implications extend far beyond a single research paper or product launch.

Because the question both projects ask is the same. Not simply what technology can do, but whose interests it ultimately serves.Top of Form

Bottom of Form