The UAE Bought an AI Future. Now It Needs Builders.

Start with the number that should bother you, because it's the honest one. In the first nine months of 2024, every startup in the UAE combined raised about $380 million. That was down eighteen percent on the year before, a steeper fall than the wider region managed. It is less than what a single well-known AI company in San Francisco now raises in one round.

Now hold that next to the other set of numbers. MGX, a hundred-billion-dollar AI investment vehicle, stood up in 2024 and began writing equity checks into OpenAI's $500 billion Stargate, the largest computing-infrastructure project anyone has attempted. A gigawatt of AI compute is rising out of the Abu Dhabi desert, the first slab of a planned five-gigawatt campus that its builders call, without much exaggeration, the largest of its kind outside the United States. Washington has cleared its most advanced chips for export to the Gulf. Microsoft is fifteen billion dollars in and owns a piece of the country's national AI champion. And the UAE appointed a minister for artificial intelligence back in 2017, before most governments had so much as a department for it.

Those are two portraits of the same country. The first is the most deliberate, best-funded national bet on AI anywhere outside America and China. The second is a startup scene that raises less in a year than a good week on Sand Hill Road. The distance between them is the most interesting thing about the UAE right now, more interesting than either portrait on its own, and reading it correctly is the whole game if you're weighing whether to come.

Most people read that gap as a contradiction, as proof the whole thing is sovereign money with nothing real underneath it. I think it is better understood as a sequence. The UAE did something almost no country does. It bought the hard, capital-intensive half of a technology ecosystem first, and only now is it going looking for the cheap, human half it skipped. It built the cathedral before the congregation arrived.

That is the reverse of how Silicon Valley happened, and the reversal matters. The Valley grew out of people first: the defectors, the hobbyists, the dense mesh of engineers who kept quitting to start things and helping each other do it. Capital and institutions came afterward, drawn to those people. The buildings were the last thing to show up. You can't understand the place without seeing that the scarce, precious input was always talent, and everything else organized itself around that.

The UAE inverted that order. It began with the resource it has in genuine abundance, which is capital plus the political will to deploy it at a speed no democracy can match, and it spent that on the things money can actually buy. Money turns out to buy a surprising amount. It buys compute: G42's Khazna already controls roughly seventy percent of the country's data-center capacity, and the Stargate build will dwarf even that. It buys proximity to the frontier, because when your sovereign fund is an equity backer of OpenAI's infrastructure and Microsoft owns a slice of your national champion, you are not standing on the edge of this technology. It buys policy: a real AI ministry, a Dubai blueprint that installs a chief AI officer in every government department and a regulatory sandbox beneath them. It even bought the domain name. The country owns dub.ai, its own name spelled out in the top-level domain reserved for artificial intelligence. That is a flourish, but an honest one. It tells you how constantly these people think about this.

What money cannot buy is the congregation. For all its compute and capital, the UAE has historically been thinnest in the place that matters most: a grassroots community of people who build.

The boosters tend to skip this part, so I will say it plainly. A builder community is not a co-working floor full of founders or a conference with good catering. It is something more specific, almost ecological: a dense, informal network of people who have shipped real products, stuck around long enough to do it more than once, answer each other's two-a.m. questions, put their exit money into the next cohort, and peel off to start the company that competes with the one they just left. It compounds over time. It takes a decade to grow, and you can't import it wholesale, because the very thing that makes it valuable is that it grew where it stands.

The UAE did not have that, for reasons that are structural rather than accidental. For fifty years its labor market ran on a transient bargain: come, earn well, pay no tax, go home. That is a wonderful deal for a mid-career professional and a poor one for a builder ecosystem, because builders need to believe they can put down roots. The talent the country attracted was pointed at services, trade, finance and real estate, not at sitting in a room for seven years to ship a product that will probably fail. And the engine that actually seeds a builder community, the big homegrown exit that mints a generation of operators-turned-angels, has fired exactly twice in living memory: Souq, sold to Amazon in 2017, and Careem, sold to Uber for $3.1 billion in 2019. Two exits. The region is still waiting on a third at that scale. PayPal alone produced enough of them to remake an industry.

So when someone tells you the UAE is "the next Silicon Valley," a little skepticism is in order, and that falling funding number is your evidence. The capital is real and the buildings are real, but a place does not turn into a builder ecosystem because a sovereign fund wants it to. It becomes one when builders choose it, and until very recently they mostly didn't.

I think that is changing now, and for reasons that are sound rather than wishful. Three things are happening at the same time.

The first is that the bargain itself changed. From 2019 onward the UAE rewrote its residency rules: the ten-year golden visa, no employer sponsor required, opened to founders, to AI and tech specialists, and eventually to freelancers. That can read like immigration trivia, but it is the whole ballgame. It converts earn and leave into build and stay, and a builder community is made up entirely of people who chose to stay. You get no roots without permanence, and for the first time the country is offering permanence.

The second is that capital and talent are finally pooling in the same place. The years since 2023 pulled in a real wave of founders: crypto first, then AI, then simply people worn down by the cost and the visa lottery of the United States. The regional money backs this up, with the honest caveats attached. 2025 was a record year for MENA startups, somewhere around $7.5 billion raised. But more than half of that was debt, and most of the equity went into fintech rather than frontier tech. This is not a deep-tech founder boom yet. It is, though, the first time the money and the people have been thick enough, in the same square mile, for a flywheel to start turning at all.

The third reason is the one that genuinely makes me optimistic, and it is specific to this moment. AI is shrinking the size of the team you need to build a real company. A startup hub used to require a deep engineering labor market, thousands of senior engineers within reach, because software took a lot of hands to build. That is exactly the requirement now being automated away. A serious product can come from three people. So the one resource the UAE was structurally short of, a thick supply of engineers, is the resource the technology itself is making less scarce, and it is happening at the very moment the country has cornered the two inputs that are becoming the real bottleneck: compute and capital.

That is the whole thesis. For the entire history of startups, the scarce input was people who build, and capital ran after them. AI inverts the relationship. Compute and capital are becoming the gating constraints on what can be built, while talent, multiplied by models, needs far less density to be productive. Whether by design or by luck, the UAE positioned itself for precisely the world now arriving. It over-bought the inputs that are becoming scarce and under-built the one that is becoming abundant.

So what should you do with this, if you are a builder somewhere else, reading and wondering?

Start with the honest version. The UAE is not, today, the best place on earth to start a company. If you need a deep bench of senior engineers to hire from on day one, it isn't here yet. If your company lives or dies on a thick angel network of people who have already built what you're building, that network exists but it is thin. And if you want the accidental serendipity of a city where everyone in the coffee shop is shipping something, San Francisco still wins, and it isn't close.

But that is the wrong comparison and the wrong question. The question is not whether the UAE is Silicon Valley yet. It is where the gap between what a place has and what it is about to have runs widest, and whether you are early to it. By that test the UAE is one of the most interesting bets going. You get frontier compute. You get sovereign-scale capital that is openly hunting for places to put itself. You get a government that has written into national strategy that a fifth of its non-oil economy should come from AI by 2031, and that moves at a speed which genuinely startles newcomers. You get zero personal income tax and a visa that lets you stay. And you get a builder community still small enough that you can help shape it rather than fight for air in one that has already set. That is the underrated part. In a mature ecosystem you are a commodity. In a forming one you can be a founder of the ecosystem itself.

The case against is just as clear, and you should weigh it with equal care. The whole project is top-down, and top-down ecosystems share a failure mode: capital pools into mega-projects and national champions while the scrappy early-stage layer goes hungry. That falling funding number is a real warning rather than noise. There is a version of the next five years in which the UAE builds the most magnificent compute infrastructure on the planet and then watches it get used by everyone except a homegrown generation of founders, because the cathedral never did fill with a congregation. That outcome is entirely possible. Anyone who tells you it isn't is selling something.

I don't think it is the likely one. Here is the picture I would leave you with. The UAE has spent a decade and a fortune on the part that is genuinely hard: securing the compute, the capital, the chips, the policy and the seat at the frontier table. It did all of that before it had the builders to justify any of it. That looks like hubris right up to the point where you notice it is the correct order of operations for this particular decade. The expensive part is finished. What is left is the cheap part, the human part, and that is the part finally moving in the country's favor. Permanence is on offer. Capital and talent are thickening in the same place. And the technology of the moment happens to reward the exact resources this country spent years hoarding.

A city sends you a message. For fifty years this one's message was simple: come here and make money. You can feel it changing. What it is starting to say instead, if you listen, is come here and build the thing. The new message is not yet as loud as it will be. But it is audible for the first time, and if you are the sort of person who would rather help build the room than take a seat in a finished one, the next few years here will be worth watching from the inside.

The easy part is done. The interesting part is just starting.