Follow the Money
By Oluwaseyi Ayodeji, Published on oluwaseyiayodeji.com | Sovereign Stack Newsletter
Sovereign Stack — Series preamble: Money in the AI ecosystem
My first job out of engineering school wasn't in engineering at all. It was in Lagos, at PricewaterhouseCoopers, auditing the books of companies that were either already publicly traded or trying to get there. I was twenty-something and mostly quiet in those rooms, but I was watching something I didn't have language for yet: the companies with real access to capital markets behaved differently than the ones without it. Not smarter. Not more disciplined. Differently. They took bets the cash-constrained ones couldn't afford to take, and a year later, those bets had usually become the thing the company was known for. The ones running on overdraft facilities and founder savings were still explaining, why last year's good idea never left the pitch deck.
I didn't understand yet that I was watching the actual mechanism of innovation, not its byproduct. The audit taught me to read a balance sheet. What I actually learned was that ideas don't build companies. Financing does, and it decides which ideas even get the chance to try.
I think about that a lot now, watching the same dynamic play out at planetary scale in artificial intelligence. Everyone is talking about models, benchmarks, and who has the smartest researchers. Almost no one is talking about the thing that actually decides whose ideas ship: who is writing the checks, on what terms, and why.
That is what this series is about. Not the AI race as a contest of intelligence. The AI race as a contest of capital, and what that means for a continent that is mostly watching this particular check-writing from the outside.
Money is not fuel. It's oxygen.
We tend to talk about capital like it's fuel: useful, but secondary to the engine itself. That framing is wrong, and it matters that it's wrong. Fuel makes a good engine go faster. Oxygen determines whether combustion happens at all. Without it, the best engine in the world is just an expensive sculpture.
Capital works the same way for innovation. It is not a multiplier applied after a good idea exists. It is the precondition that decides which ideas get the chance to become real in the first place, which ones get shelved indefinitely, and which ones never get attempted because the people with them already know no check is coming. If you want to know where a society's innovation is actually headed, five years out, don't start with its patents or its universities. Start with where its capital is flowing right now, because capital moves ahead of the outcomes it produces. It is a leading indicator, not a lagging one.
That is the lens for this whole series. And it is why the current moment in AI is worth studying as closely as the technology itself.
What the money is actually doing right now
Start with venture capital, because it is the most visible lane and the easiest to misread. Global venture funding hit a record $510 billion in the first half of 2026, according to Crunchbase, surpassing the entire $440 billion invested across all of 2025. On the surface, that looks like a broad-based boom. It is not. OpenAI and Anthropic alone accounted for $217 billion of that total, 43 percent of all global startup funding in six months. Two companies. Not two sectors, not two countries. Two companies.
And it is not evenly distributed geographically either. The United States pulled in $250 billion of the $297 billion invested globally in the first quarter of 2026 alone, 81 percent of the world's venture capital, up from 55 percent of the total just a year earlier. That is not a market maturing. That is a market concentrating, fast, around a small number of American labs building frontier models.
There is a second lane of money that gets almost no attention next to the funding headlines, and it might be the more important one. Building the actual physical infrastructure behind AI, the data centers, the power plants, the cooling systems, is not primarily a venture capital story anymore. McKinsey estimates global data center spending could reach $7 trillion by 2030, and increasingly that money is coming from debt, not equity: private credit funds have gone from near zero exposure to AI infrastructure a few years ago to more than $200 billion in outstanding loans today, with Morgan Stanley projecting another $800 billion over the next two years alone. This is the quiet engine room of the AI buildout, running largely outside the headlines that cover funding rounds. We will spend an entire piece in this series inside that engine room, because I think most people, myself included until recently, do not understand how much of the AI boom is actually financed on borrowed money.
Then there is China, which is not running either of these playbooks. Rather than compete dollar for dollar on venture terms, the Chinese state is deploying patient, direct capital at the points it considers strategic. The country's semiconductor investment vehicle, known as the Big Fund, launched a third phase last year with 344 billion yuan, roughly $47.5 billion, larger than the first two phases combined, and a separate National AI Industry Investment Fund launched in February 2026 with an initial 60 billion yuan, about $8.2 billion, aimed squarely at early-stage AI projects. Government-linked investors in China went from backing fewer than 10 AI deals a year before 2018 to more than 140 in 2025.
But even that picture is more complicated than "the state funds everything." When DeepSeek closed its first-ever external funding round this year, boosting its valuation toward $50 billion, most observers expected the Big Fund to lead it. Instead, founder Liang Wenfeng put in roughly 20 billion yuan of his own money, nearly half the round himself, specifically to keep control of the company's direction. Even inside China's state-capital model, there is a fight over who actually steers innovation once the money shows up. That tension is worth understanding closely, and it is the subject of the third piece in this series.
The question I actually want answered
Here is where this gets personal, and where I want to be honest about what I don't yet know the full answer to.
If capital is the leading indicator of where innovation goes next, what does it mean that Africa, home to 18 percent of the world's population, accounts for less than 1 percent of global data center capacity? And what does it mean that African startup equity funding fell 37 percent year over year in early 2026, at the exact moment global venture capital was hitting record highs, because many of the same US investors who used to write African checks were busy chasing AI mega-rounds instead?
I don't think the honest answer is simply "Africa needs more investment." That's the version of this argument everyone has already heard, and it hasn't moved much in a decade. The harder, more specific question is this: how much of what currently funds innovation in Africa is actually African capital, and how much of it depends on the same foreign investors who are now visibly distracted by a bigger, more concentrated opportunity elsewhere? If the answer is "mostly the latter," then Africa's innovation pipeline was never as independent as it looked, and the current AI capital boom just exposed how thin that foundation was.
I'm not going to answer that fully here. That's deliberate. Over the next three weeks, this series is going to walk through each of the three lanes I've just described, venture equity, infrastructure and project finance, and state capital, and hold each one up against Africa's specific position. Not to conclude that Africa should copy Silicon Valley or copy Beijing. Both systems have real weaknesses baked into how they work, weaknesses this series will name directly. The goal is to study both closely enough to borrow what works, learn from what breaks, and use that to think through what a financing framework built for Africa's actual constraints, not America's or China's, could look like.
What's ahead
Next week, we start with venture equity: what the concentration in US funding actually means, who gets left out when two companies absorb 43 percent of global startup capital, and what that says about who should be funding African AI ventures if the traditional playbook is visibly distracted.
After that, infrastructure and project finance: the trillion-dollar debt story running quietly behind the AI buildout, and what it would take to build something similar, but sustainable, on African soil.
And finally, state capital: what China's model gets right, what it gets wrong, and whether there's a version of patient, strategic capital that could work for a continent of 54 different fiscal realities instead of one centralized state.
Money moves first. Everything else, the products, the breakthroughs, the headlines, follows it. I want us to start paying attention to the part everyone else skips past.
Follow the money. That's where this series begins.
