Let me paint you a picture.
It's 2025. Humanity has built machines that can write poetry, diagnose cancer, pass the bar exam, and generate a suspiciously convincing image of a cat in a business suit. We have artificial intelligence. We are, by every metric, living in the future.
And yet Africa, a continent of 1.4 billion people sitting on top of the largest reserves of cobalt, lithium, coltan, and manganese on the planet, minerals that are the AI revolution, is largely watching this future happen to it. From the outside. Through a window. That it cleaned.
Take a moment. Let that land.
The Vanity Card Nobody Prints
Here's what the press releases won't tell you:
The GPU in your laptop? Cobalt from the DRC. The battery in the server farm running the model that just wrote someone's cover letter? Lithium from Zimbabwe. The coltan in every piece of electronics that makes "the cloud" more than a metaphor? Mined, largely, by African hands, for African wages, to build systems that will then be sold back to Africa at American prices.
If irony were a mineral, Africa would be the world's largest exporter of that too.
The tech giants, bless their hoodie-wearing, kombucha-drinking hearts, have discovered that Africa is a magnificent place to find three things: raw materials, cheap labor, and an emerging consumer market hungry enough to absorb whatever product they decide to launch here next. What Africa is apparently not, according to the org charts of Silicon Valley, Seattle, and Shenzhen, is a place where intelligence originates.
Which is fascinating. Because the continent produced mathematics, astronomy, architecture, and philosophy while Europe was still working out the whole "walls on buildings" concept.
The Talent Is Here. So Where's the Product?
I've met them. You've met them. The Kenyan developer who could out-code half of FAANG on a bad internet day. The Nigerian data scientist doing groundbreaking NLP work, with a salary that wouldn't cover a San Francisco studio apartment's electricity bill. The South African ML engineer consulting for a London firm, solving their problems remotely, invisibly, without a single line in their annual report crediting the continent that produced her.
We are not short of talent. We are not short of ideas. We are not short of hunger.
What we are short of is the infrastructure of audacity, the venture capital that says yes to an African founder without asking if they have a "Western co-founder," the government policy that treats AI development as a national priority rather than a paragraph in a speech, the universities that train engineers in the morning and fund their startups by afternoon, the collective self-belief that says: we do not have to wait for permission to build the future.
A Brief, Uncomfortable History Lesson (Stay With Me)
Colonialism, at its core, was a remarkably elegant business model: extract the resources, control the processing, sell back the finished goods at markup. It ran for centuries. We called it history. We wrote Truth and Reconciliation reports. We moved on.
Except, and here's where it gets awkward at the dinner table, the model didn't move on. It just got a rebrand. Now the extraction is data. The processing happens in data centers in Virginia and Amsterdam. And the finished product is an AI assistant that doesn't understand Swahili idioms, can't read Amharic handwriting, and will confidently give you healthcare advice calibrated to a Western body in a Western context.
Digital colonialism isn't a conspiracy theory. It's a business strategy. And unlike the original version, it doesn't even need ships.
So What's the Actual Problem?
Let's be precise, because vague outrage is everyone's least favorite newsletter.
1. Capital doesn't follow talent here, yet. African AI startups raised a fraction of what comparable companies raise in comparable markets. Not because the ideas are worse. Because the check-writers are in rooms where Africa is a charity case, not a business opportunity.
2. Policy is playing catch-up while the game changes every six months. Most African governments are still writing the AI regulation framework while the technology they're regulating already has three new versions out. We need policy that's built with technologists, not handed down by committees who think "neural network" is a medical term.
3. We're training for other people's problems. When the brightest minds on the continent are optimizing for jobs at multinationals, they're solving San Francisco's problems with Lagos' brain. That's not a personal failing. That's a structural incentive we built. We can unbuild it.
4. The data gap is real and it matters. AI learns from data. If the data is predominantly English, Western, and WEIRD (Western, Educated, Industrialized, Rich, Democratic), then the AI is predominantly those things too. Building AI that actually works for African contexts requires African data, collected by Africans, labeled by Africans, for problems Africans actually have.
5. We keep asking for a seat at the table instead of building the table. Seats can be taken away. Tables are harder to repossess.
What Would Winning Look Like?
Funny you should ask.
Winning looks like a 19-year-old in Kampala training a model on local language data and selling it to a telecom. Winning looks like a government procurement policy that defaults to locally-built AI solutions. Winning looks like a pan-African compute infrastructure so our researchers stop begging for AWS credits like digital Oliver Twists. Winning looks like the cobalt that leaves the DRC coming back as a patent, not a product someone else built with it.
Winning looks like us, building the thing.
Not consulting on the thing. Not implementing someone else's version of the thing. Not writing Medium posts about the thing while the thing gets built elsewhere.
Africa has the minerals. Africa has the talent. Africa has the youngest population on the planet, a demographic that, in tech terms, is called a market when you're selling to them and human capital when you're employing them and untapped potential when you're writing thought leadership about them.
What Africa does not yet have, in sufficient quantity, is the institutional confidence to stop asking for a seat at the table and start asking a different question entirely:
Why are we eating at their table at all?
The ingredients were ours to begin with.
If this made you uncomfortable, good. If it made you want to build something, better.
— The Driver