There is a woman I know who still grinds her millet by hand. Not because she cannot afford a machine. Not because the mill down the road is too far away. She does it because, she says, the grain speaks to her through the stone. She knows when it is ready not by a timer, but by the sound, the resistance, and the fine dust that rises like breath.

ALSO READ: Adapting to AI: How workers aged 50+ are navigating change

She is, by every modern definition, a Luddite. And she is, by every measure that matters, one of the wisest people I have ever met.

We have been using that word incorrectly for two hundred years.

ALSO READ: Africa must not be left behind, Kagame tells AI Summit

The original Luddites were not afraid of machines.

The original Luddites were not afraid of machines. They were skilled craftsmen – weavers, framework knitters, and textile workers – who understood their trade so completely that they could see exactly what the machine would do to it. They were not reacting out of ignorance. They were reacting out of knowledge: deep, embodied, generational knowledge.

ALSO READ: Africa’s AI adoption stuck in pilot phase, new report finds

They smashed the looms not because they did not understand them, but because they understood them all too well. We remember them as fools. History often does that to people who are right too early.

ALSO READ: How AI is reshaping the workplace

Now we find ourselves in another machine moment, possibly the most significant one our species has ever created. Artificial intelligence has arrived. It is in the recruitment software screening your CV before a human ever sees it. It is in the content moderation deciding what speech is acceptable. It is in the credit scoring determining whether you get a loan, in the predictive policing models deciding which neighbourhood gets watched. In some cities, it is writing the news. In some hospitals, it is reading medical scans.

And most of us are jumping in. Enthusiastically. Uncritically. As if enthusiasm were the same thing as understanding.

Here is what the original Luddites knew that we have forgotten: you do not have to oppose something in order to understand it first. Resistance and study are not opposites. In fact, the deepest resistance, the kind that actually changes outcomes is always built on knowledge. You cannot push back against what you cannot name. You cannot negotiate with what you have not examined.

The African context makes this more urgent, not less. Many of the AI models being deployed across this continent were not trained on our languages, our contexts, our histories. They were trained largely on data from the Global North which means they carry embedded assumptions about what is normal, what is dangerous, what is progress. When a hiring algorithm trained on American corporate culture evaluates a CV from Kigali, it is not neutral. When a content moderation model flags Swahili slang as suspicious, it is not neutral. Neutrality was never on the table. It was never offered. To use these tools without studying them is not progress. It is a very elegant form of surrender.

This is not an argument against AI. It is an argument for the woman with the grinding stone for her instinct to know the thing in her hands before she lets it shape her life. Study the model. Understand what it was trained on. Ask who built it and for whom. Ask what it optimises for and what it quietly deprioritises. Ask what happens to the knowledge that does not fit inside it.

The Luddites were not anti-future. They were pro-dignity. They wanted to move into the new world as full human beings, not as inputs.

That distinction is still worth defending.

The wisest people in the room are not always the ones moving fastest. Sometimes they are the ones who have stopped, hand on stone, listening for what the grain is trying to say.

The writer is a communications specialist, writer, and strategist.