Can AI Be Nigerian-Smart?

The first time I tried an AI-powered voice assistant, I was convinced that AI might be built for the world, but it’s definitely not ready for Nigeria—or Nigerians. As a Nigerian, you’re expected to be two things: book smart and street smart. In fact, being street smart—knowing when to speak, how to speak, and what […]

The first time I tried an AI-powered voice assistant, I was convinced that AI might be built for the world, but it’s definitely not ready for Nigeria—or Nigerians.

As a Nigerian, you’re expected to be two things: book smart and street smart. In fact, being street smart—knowing when to speak, how to speak, and what to say—is sometimes more important than being book smart. But AI? AI doesn’t get it. It studies culture, claims to understand humans, but somehow, Nigerians didn’t quite make the cut.

I asked AI to pronounce my name—Lilian Ogazi. The robotic voice, crisp and efficient, butchered it beyond recognition. “Lai-lee-an Oh-gah-zee,” it said, with the confidence of someone who had no business being confident. That was the moment I knew—AI is smart, but it’s not Nigerian-smart.

AI vs. Nigerian Context: A Struggle

AI is changing how we work, think, and write. In newsrooms, it’s generating headlines, fact-checking, transcribing interviews, and even suggesting story angles. But no matter how advanced it gets, it doesn’t understand us.

It doesn’t know that “How far?” means different things depending on tone. It doesn’t understand that “I dey come” could mean I’m actually coming, I’m still at home chilling, or I’m never coming at all. It struggles with language, culture, and context—three things deeply embedded in Nigerian journalism.

As journalists, we navigate indigenous languages, Pidgin, code-switching, and expressions that don’t translate neatly into English. In fact, some things can only be truly expressed in Pidgin. But AI? AI is still stuck on robotic translations, and that’s a problem.

AI Still Needs Nigerian Supervision

Despite its flaws, AI isn’t useless in the newsroom. It helps fight misinformation, enhances headline generation, and—most importantly—saves time. But for all its intelligence, AI still needs human supervision.

It doesn’t know that “Oya na” is different from “Oya now.”
It doesn’t know that “E go beta” is definitely not “Ego butter.”
It doesn’t know that when a Nigerian aunty says “My dear,” it’s not always affection—it might just be the start of a long, unsolicited lecture.

Maybe one day, AI will evolve enough to read between Nigerian lines. Maybe it will understand that when my mother calls my full name, it’s time to run. Maybe, just maybe, AI will go beyond capturing our words—and finally capture our essence.

But until then, I’ll keep laughing at its mistakes and remembering that some things—our language, our culture, our way of telling stories—are still best left to us.