We Built a Mirror and Called It God

In 2024, a Nigerian woman applied for a job.

She prepared. She researched the company. She understood the role.

Then she sat down for her interview — and discovered she was not speaking to a human being. She was being evaluated by an AI system. The machine asked questions, assessed her answers, and decided whether she was qualified.

She had no way to know what criteria it used.

No way to appeal the decision.

No Nigerian regulatory body she could turn to if the evaluation was wrong or biased.

That woman was me.

And the most unsettling part was not the rejection.

It was the silence. Nobody — not the company, not the government, not any institution — was required to explain how the machine made its judgment about my future.

The oldest story in the world

Human beings have always done this.When our ancestors encountered something they could not explain — lightning, disease, drought — they did not leave it unexplained. They filled the gap with power. They built gods from their fear of the unknown.This is not weakness. It is how the human mind works. We cannot tolerate a vacuum of understanding. So we assign authority to whatever occupies that vacuum first.AI occupied the vacuum first.A parent who does not know how to raise their child asks AI. A student who does not know how to study asks AI. A government that does not know how to make policy asks AI. A teenager who does not know how to make friends asks AI.None of them stopped to ask what AI actually is.And the institutions that should have told them — governments, regulators, schools — did not. Because they were too busy being afraid of it themselves. Or too busy selling it.

---

What AI actually is

A math equation doing multiplication at scale whose coefficients were tuned to minimize a loss on observed data.

I know that sounds reductive. But it is the most honest description available.

AI does not think. It does not feel. It does not understand. It models patterns in human language and data — everything we have ever written, said, and recorded — and returns the most statistically likely response.

When you ask AI how to raise your child it is not giving you wisdom. It is giving you the average of everything humanity has ever written about raising children — compressed, fluent, and delivered with extraordinary confidence.

A mirror does not know your face. It only reflects what you put in front of it.

AI is a very confident mirror. We built it. Then we called it God.

The institutional failure

This did not happen by accident.The people who build AI systems understand exactly what they are. They also understand what Herbert Simon identified in 1971: a wealth of information creates a poverty of attention. Give people enough information, fast enough, in a convincing enough voice, and they will stop asking where it comes from.The business model of AI depends on you not asking.Every subscription you pay, every query you send, every decision you outsource — that is revenue. That is the signal that tells investors the $725 billion capex cycle is rational. Your dependency is the product being sold.And governments — particularly in Africa, particularly in Nigeria — have provided almost no protection.Nigeria’s National Information Technology Development Agency has released AI guidelines. But guidelines are not law. There is no binding regulation. No accountability mechanism. No transparency requirement for AI systems making decisions about Nigerian citizens.Which means when an AI system evaluates a Nigerian job seeker, denies a Nigerian loan application, or filters a Nigerian student’s university submission — nobody is required to explain how it made that decision. Nobody is accountable if it was wrong. Nobody can be challenged.The silence I experienced in that interview is not a personal anecdote. It is policy. It is the sound of a government that has not yet decided its citizens deserve protection from the systems shaping their futures.

The thing nobody is saying

AI did not make itself powerful.

We made it powerful by outsourcing our judgment to it before we understood what it was.

The parent who asks AI how to raise their child is not lazy. They are operating in a system that gave them AI before it gave them the support, the community, or the institutional infrastructure that used to answer that question.

The teenager who uses AI as a friend is not broken. They are lonely in a culture that replaced community with performance — where you must be visible, record-breaking, and productive to matter — and AI is the only thing that never judges them for being ordinary.

The government that uses AI for policy is not incompetent. They are operating in a world moving faster than democratic institutions were designed to handle.

The fear is real. The dependency is real. The danger is real.

But the source of the danger is not the technology.

It is the gap between what AI is and what we have been told it is. That gap is where the power lives. And that gap is maintained — deliberately, profitably — by the people who benefit from your confusion.

What intentional understanding looks like

You do not need to become an engineer to be safe in an AI world.

You need to ask six questions. Every time. About every AI system that touches your life.

Who built this?
Who controls it?
What data does it use?
What decisions can it make about me?
Who benefits from those decisions?
Can I challenge them?

These are not technical questions. They are citizenship questions. They are the same questions you would ask about a bank, a court, a government agency, or any institution with power over your life.

AI is an institution now. It deserves the same scrutiny.

The shift is not from fear to trust. It is from fear to understanding. You do not have to trust AI. You have to know what it is — a tool built by humans, trained on human patterns, deployed by organisations with specific incentives — so that you can use it on your own terms instead of being used by it on theirs.

Keep it simple. Keep it human. Follow the decision chain.

Who decides?
Who benefits?
Who is accountable?
Can you intervene?

Three signals worth watching

Whether Nigeria’s NITDA moves from voluntary AI guidelines to binding regulation with enforcement mechanisms before the end of 2026. The African Union’s Continental AI Strategy has been adopted — the question is whether member states implement it or treat it as another document.

Whether the EU AI Act’s high-risk category provisions — which require transparency, human oversight, and the right to explanation for AI systems making consequential decisions — begin to influence AI deployments outside Europe, particularly in Africa where tech companies operate with far less accountability.

Whether any major AI company operating in Nigeria voluntarily publishes the criteria its AI recruitment and evaluation systems use — or whether that transparency requires a law to force it.

The mirror does not become less powerful when you name it.

But you become harder to fool.

— Victoria, Axis Brief