Daniel sent us this one — he's been thinking about seasteading again, but not the boat-based version. The actual question is: what if you just built an artificial island? Find the shallowest spot just outside any country's exclusive economic zone, dump enough material to create something solid, build a house on it, a beach, something that won't get swallowed by the ocean. Has anyone actually tried this as a housing solution or a micro-state play? And the more grounded version of the question: do coastal countries with space constraints actually use artificial islands for infrastructure and population growth, and how do the engineering challenges work at that scale?
There are really two completely different conversations here, and I want to pull them apart before we dive in. One is the libertarian dream — the "I will build my own country with blackjack and sand dunes" version. The other is what actual nation-states do when they're land-constrained and have serious money to throw at the problem. They share some engineering DNA but the legal and practical realities are night and day.
I'm guessing the nation-state version actually exists.
Oh, it exists on a scale that's hard to wrap your head around. Singapore has expanded its land area by about twenty-five percent since independence in nineteen sixty-five. Twenty-five percent of the entire country is manufactured ground. They've used over a billion cubic meters of sand — which, by the way, created a whole geopolitical mess with Indonesia and Malaysia banning sand exports because Singapore was literally erasing their islands.
A billion cubic meters. That's the kind of number where you just nod and pretend you can visualize it.
It's about four hundred thousand Olympic swimming pools. But here's the thing that connects to the prompt — Singapore isn't building in international waters. They're building within their own territorial sea, usually extending existing land or joining islands together. The engineering is extraordinary but the legal framework is settled. They own the seabed, they own the resulting land, and nobody disputes it.
The "just outside the EEZ" part of the question is where everything gets complicated.
And this is where I want to talk about something most coverage of seasteading completely misses. There's a real-world case study that answers the prompt almost perfectly, and it's called the Republic of Minerva.
That sounds like a skincare brand.
It was nineteen seventy-two. A Nevada real estate millionaire named Michael Oliver — who had already tried and failed to start a libertarian country on a platform near the Bahamas — decided to try the artificial island approach. He found two submerged reefs in the Pacific Ocean, about two hundred sixty miles southwest of Tonga. The Minerva Reefs. They were just below the surface at low tide, which made them perfect — you could build up from them without starting from the deep ocean floor.
Critically, they were outside anyone's territorial waters.
Oliver's group — they called themselves the Ocean Life Research Foundation, which is a beautiful bit of branding for what was essentially a real estate play — they hired a dredging barge from Australia. They sailed out in late nineteen seventy-one, dumped sand and coral onto the reefs, and built two artificial islands. They planted a flag. They declared the Republic of Minerva. They minted coins. They had a constitution and a national anthem.
Of course they minted coins. The true mark of a serious nation is immediately monetizing the merchandise.
They issued a declaration of independence to neighboring countries. They were actively recruiting settlers. Oliver's plan was to build a libertarian utopia with no taxation, no welfare, no economic intervention — basically a country as a business.
Then reality showed up.
Reality showed up in the form of King Taufa'ahau Tupou the Fourth of Tonga. Who looked at this situation and said — and I'm paraphrasing — absolutely not. Tonga claimed the Minerva Reefs as part of its historical territory. In June nineteen seventy-two, King Tupou personally sailed to the reefs with a contingent of convicts and a military band. They tore down the Minervan flag, raised the Tongan flag, and the band played the Tongan national anthem.
Wait — convicts and a military band? That's the most theatrical sovereignty assertion I've ever heard.
A South Pacific Forum meeting later that year formally backed Tonga's claim. The reefs are part of Tonga to this day. And here's the lesson that the prompt is essentially asking about: the engineering worked. They actually did it. They built islands where nothing was before, cheaply and relatively quickly. What failed was the legal layer. You cannot simply declare land yours in the ocean and expect the world to shrug.
Because someone always already thinks they own it.
Or will decide they own it the moment you show interest. The Minerva Reefs were technically outside Tonga's territorial waters, but Tonga argued they were historically part of the Tongan maritime domain. And Tonga had a navy and a seat at the UN. Oliver had a dredging barge and some coins.
The answer to "has anyone tried this" is yes, and the answer to "did it work" is a king showed up with prisoners and a brass band.
Let's talk about why nobody has succeeded at this, even with much better technology than nineteen seventy-two. The UN Convention on the Law of the Sea — UNCLOS — which came into force in nineteen ninety-four, is remarkably clear on this point. Artificial islands, installations, and structures do not possess the status of islands. They have no territorial sea of their own. Their presence does not affect the delimitation of the territorial sea, the exclusive economic zone, or the continental shelf.
Even if you build it, it doesn't generate any legal water around it.
A natural island — even a tiny rock that's above water at high tide — generates a twelve-nautical-mile territorial sea and a two-hundred-nautical-mile EEZ. That's about one hundred sixty-five thousand square miles of ocean under your control. An artificial island generates nothing. It's just a structure in international waters. The coastal state — or in this case, no state — has jurisdiction over it, but it doesn't create new jurisdiction.
Which means your libertarian paradise has no legal basis for excluding anyone from walking up and setting foot on it.
Or fishing next to it, or sailing through it, or setting up their own platform fifty feet away. You can't enforce laws because you have no legal authority to make laws. You're essentially a very elaborate picnic.
A picnic with minted coins.
Now, there's a separate but related approach that some groups have explored, which is building on an existing natural feature that's already above water at high tide. The most famous attempt is Sealand.
The platform off England.
Right — and Sealand is actually a terrible example for the artificial island question because it wasn't artificial island construction. It was a World War Two anti-aircraft platform built by the British government in what were then territorial waters. The UK extended its territorial sea from three to twelve nautical miles in nineteen eighty-seven, which put Sealand inside British waters, but by then the Bates family had been squatting there for two decades and had declared it an independent principality.
The UK just...
The British government largely ignored them except for one moment in nineteen sixty-eight when Roy Bates fired warning shots at a Royal Navy vessel that had entered what he claimed were his territorial waters. He was prosecuted for firearms offenses, but a British court ruled that Sealand was outside UK jurisdiction at the time because it was beyond the three-mile limit. That ruling is often misunderstood as Britain recognizing Sealand's sovereignty. It wasn't. It was a jurisdictional technicality that the UK closed by extending its territorial waters.
Sealand is a legal footnote, not a precedent.
And it's not an artificial island — it's a repurposed military structure. The prompt is asking about actually creating land from the seabed, and for that we need to look at what nation-states do, because they're the only ones who can pull it off legally.
Let's go there then. Singapore, Dubai, the Netherlands — these are the poster children.
They're doing it for completely different reasons with completely different constraints. Singapore's land reclamation is driven by pure existential need. It's a city-state with five and a half million people on an island that was about five hundred eighty square kilometers at independence. Every square meter matters. They've added about one hundred forty square kilometers through reclamation — that's where Changi Airport sits, where much of the port infrastructure lives, where Marina Bay is.
The iconic skyline photos of Singapore — a lot of that is on manufactured ground.
Most of the downtown core. And the engineering has evolved dramatically. Early reclamation was simple — dump sand, let it settle, build on it. But sand is finite, and as I mentioned, regional neighbors cut off exports. Singapore now uses a technique called empoldering, which they adapted from the Dutch. Instead of raising land above sea level by piling up sand, you build a seawall, drain the water from inside it, and create dry land at or slightly below sea level. It uses far less sand and it's more resilient to sea level rise because you're managing water with pumps rather than relying on elevation.
The Dutch approach — the country that's been doing this for centuries.
The Netherlands has about seventeen percent of its total land area below sea level, protected by dikes and pumps. But what's interesting for this conversation is that they've also built entirely new land in the sea. Flevoland, which is a whole province, was created in the twentieth century by draining part of the Zuiderzee. It's about two thousand four hundred square kilometers — roughly the size of Luxembourg — and it's now home to over four hundred thousand people.
Four hundred thousand people on land that was literally ocean a century ago. That's the scale where this stops being a curiosity and becomes actual national infrastructure.
The engineering timeline is worth paying attention to. The Afsluitdijk, the big dam that closed off the Zuiderzee, was completed in nineteen thirty-two. The last of the polders in Flevoland was completed in nineteen sixty-eight. That's a thirty-six year project, and the Netherlands is one of the wealthiest, most technically capable countries on earth with centuries of experience in water management.
When someone reads a seasteading blog post about building their own island paradise, the relevant comparison isn't "can I do this" — it's "the Dutch needed thirty-six years and a national budget.
They were building in shallow, sheltered, inland-adjacent water. The North Sea is not the open Pacific. The engineering challenge of building an artificial island in deep international waters, exposed to open ocean swells and storms, is several orders of magnitude harder. The seabed might be thousands of meters deep. Even at the shallowest points outside EEZs — and most of those are seamounts or mid-ocean ridges that are still hundreds of meters down — you're not dumping sand and hoping it piles up.
You'd need to build a structure from the seafloor up.
Which is essentially offshore platform construction. And that technology does exist — the oil and gas industry has been doing it for decades. But a deep-water platform costs billions of dollars and requires constant maintenance. The Perdido platform in the Gulf of Mexico, operated by Shell, sits in about two thousand four hundred fifty meters of water. It cost something like three billion dollars to build and install. That's for a single industrial facility, not a livable community with a beach.
The shallow-spot strategy is doing a lot of work in the prompt. Find the shallowest point just outside an EEZ. But the ocean doesn't cooperate — the continental shelf drops off, and by the time you're outside the two-hundred-mile EEZ you're typically in deep water.
There are exceptions, and this is where the geography gets interesting. The prompt mentions finding "the shallowest point" and there are some genuinely intriguing candidates. The Rockall Bank in the North Atlantic, for example — Rockall itself is a tiny granite islet that the UK claims, but the surrounding bank has depths of around one hundred to two hundred meters. That's shallow enough to build on, but it's well within claimed EEZs. The Mascarene Plateau in the Indian Ocean has large areas of shallow water, but again, it's within the EEZs of Mauritius and Seychelles.
What about the actual high seas? The areas beyond any national jurisdiction?
That's the problem — almost all of the shallow areas of the world ocean are on continental shelves, and continental shelves are almost entirely within EEZs. UNCLOS allows a coastal state to claim an extended continental shelf beyond two hundred nautical miles if it can demonstrate that the seabed is a natural prolongation of its land territory. There's a whole UN commission that adjudicates these claims. The point is, the legal and geological maps overlap almost perfectly. If it's shallow enough to build on affordably, someone almost certainly claims it.
The Venn diagram of "shallow enough to build" and "nobody owns it" is basically two circles in different rooms.
Touching at maybe a few points. There are some submerged features in the South China Sea that are in disputed waters — but "disputed" is not the same as "unowned," and building an artificial island there is how you start a military confrontation, not a housing development.
Which brings us to China's artificial island program, which is the elephant in the room for this entire conversation.
It's the most significant artificial island construction program in human history. Since twenty thirteen, China has built over three thousand two hundred acres of artificial land on reefs and atolls in the Spratly and Paracel Islands in the South China Sea. They've turned submerged features into fully militarized islands with airstrips, harbors, radar installations, and living quarters.
Three thousand two hundred acres. That's about five square miles of brand-new land.
They did it through massive dredging operations. They used cutter-suction dredgers — enormous ships that literally chew up the seabed and pump the resulting slurry through pipelines to the construction site. The sand and coral rubble is piled up, compacted, and stabilized. Some of these features were barely above water at low tide before China started work. Now they have three-thousand-meter runways.
This is the part of the prompt about whether countries actually use this for infrastructure and population. China's answer is: yes, but it's military infrastructure, and the population is soldiers.
The international reaction has been — to put it mildly — negative. The Permanent Court of Arbitration in The Hague ruled in twenty sixteen that China's activities violated the Philippines' sovereign rights and that China had no legal basis to claim historic rights to resources within the nine-dash line. China rejected the ruling entirely and continued building.
Artificial island construction in disputed waters is a geopolitical flashpoint, not a housing innovation.
Let's step back to the engineering question, because this is where it gets fascinating. How do you actually build an island that won't get washed away? The naive approach — the "pile up sand and hope" method — has a fundamental problem called liquefaction.
Which sounds terrifying.
When loose, water-saturated sand is subjected to seismic shaking or even just repeated wave loading, the grains can lose contact with each other. The sand essentially turns into a liquid. Anything built on it sinks. This is why land reclamation projects spend years on ground improvement before any construction begins.
What does ground improvement actually mean?
Dynamic compaction — you drop enormous weights from cranes repeatedly onto the ground to force settlement and densification. Vibro-compaction — you insert a vibrating probe into the ground that rearranges the sand grains into a denser configuration. Stone columns — you drill holes and fill them with gravel to create drainage paths and reinforcement. For large projects, you might do all of these in sequence, and the process can take years before the land is buildable.
That's in sheltered coastal waters. If you're in the open ocean, you're also dealing with wave erosion, storm surges, and the fact that the sea really wants your island to not exist.
Wave energy is the relentless enemy here. The significant wave height in the North Atlantic can exceed ten meters in winter storms. The force that water exerts when it hits a structure is staggering — we're talking tons per square meter. Any artificial island needs an armored perimeter, usually with large rock or concrete units called riprap or dolosse — those interlocking concrete shapes you see on breakwaters — to dissipate wave energy before it reaches the core of the island.
Those things that look like giant concrete jacks.
They interlock when stacked and create voids that absorb wave energy. A single dolos can weigh twenty tons or more. You need thousands of them for even a modest breakwater. The cost adds up fast.
Between ground improvement, perimeter armoring, and the actual fill material, we're talking about an engineering project that makes a skyscraper look like a garden shed.
We haven't even talked about utilities. Fresh water, electricity, sewage — on a natural island, these are challenges. On an artificial island in the open ocean, they're existential problems. Fresh water either has to be shipped in, which is ruinously expensive, or produced through desalination, which requires enormous amounts of energy. Electricity either comes from a cable from shore — impossible in international waters — or from on-site generation, which means fuel shipments or renewable systems with battery storage.
Solar and wind would work, at least.
They would, but you need redundancy and storage for calm, dark periods. And salt spray corrodes everything. Maintenance costs on offshore structures are typically three to five times what they would be on land. Everything rusts faster, breaks faster, and costs more to fix.
The engineering is possible but punishingly expensive, the legal framework is hostile to the micro-state idea, and the shallow spots outside EEZs barely exist. This is not a promising picture for the libertarian island dream.
Let me push back on my own pessimism here, because there is a version of this that's interesting and might actually happen, just not in the form the prompt is imagining.
Not boats, not anchored platforms — purpose-built floating urban districts that exist within a country's territorial waters but create new space where land reclamation isn't feasible. The difference is that these are explicitly under the jurisdiction of an existing state, so the legal problems vanish. And the engineering, while still challenging, is actually more tractable than building an island from the seabed up.
Because you're not fighting the depth of the ocean.
You're not fighting the depth at all. A floating structure moored in deep water only needs to anchor to the seabed, not build up from it. The mooring system handles wave motion. The structure rises and falls with the tide naturally. And you can build modularly — add more platforms as needed, reconfigure the layout, even relocate if conditions change.
Has anyone actually built one?
There are a few serious projects. The Maldives Floating City is under development — a joint venture between the Maldivian government and a Dutch architectural firm called Waterstudio. It's designed to house about twenty thousand people on a series of floating platforms in a lagoon near Male. The structures are tethered to the lagoon floor and connected by floating walkways. Construction started in twenty twenty-two and the first units are supposed to be completed around now, in twenty twenty-six.
The Maldives has a pretty strong incentive to figure this out, given that their entire country is basically at sea level.
The highest natural point in the Maldives is about two point four meters above sea level. They're the canary in the coal mine for sea level rise. But they're also a functioning sovereign state with a clear legal framework, so they can do what the Minerva people couldn't — actually create livable jurisdictions.
There's the Busan project, right? In South Korea?
This is the one backed by UN-Habitat. The design was unveiled in twenty twenty-two — three interconnected floating platforms totaling about six hectares, designed to house twelve thousand people. It's meant to be a prototype for floating cities as a climate adaptation strategy. The platforms are hexagonal, which apparently distributes stress better and allows for modular expansion.
So they're building a floating game board.
A very expensive floating game board. The projected cost for the Busan prototype is over six hundred million dollars. For twelve thousand residents. That's about fifty thousand dollars per person just in construction costs, before you even think about operations and maintenance.
Which is not actually insane compared to land prices in major coastal cities.
That's the argument floating city advocates make. In Hong Kong or Monaco or Manhattan, fifty thousand dollars per person for new land would be a bargain. But those are the most extreme real estate markets on earth. For most places, traditional land reclamation is still far cheaper.
The floating city concept is a solution for places that are both extremely dense and extremely rich — or existentially threatened by sea level rise.
Which is a small but growing club. Kiribati, Tuvalu, the Marshall Islands — these are countries that might not exist in a century if sea level projections hold. They're looking at floating infrastructure not as a luxury but as a survival strategy. And they have something the seasteading movement never had — actual sovereignty.
That's the thread running through all of this. Sovereignty is the thing you can't build from sand.
It's what the seasteading movement fundamentally misunderstood. They thought sovereignty was a technical problem — if you build the platform and declare independence, you've created a country. But sovereignty is a social and political construct. It exists because other states agree that it exists. Tonga showed up at Minerva with a king and a band and that was the end of it, because nobody was willing to go to war over a sandbar built by a Nevada real estate developer.
The band was really the finishing touch. You can't argue with a military band.
It's the acoustic assertion of sovereignty. But let me circle back to something in the prompt that we haven't fully addressed — the more practical question about whether countries with space constraints actually look at artificial islands as viable infrastructure.
Right, beyond the flashy examples. Is this a real tool in the urban planning toolkit, or is it a prestige project for petrostates?
It's both, and the distinction matters. Dubai's Palm Jumeirah and the World Islands are the prestige examples — built with oil money, marketed as luxury real estate, and environmentally catastrophic in ways that are still being understood. The Palm disrupted coastal currents, caused erosion on nearby beaches, and buried coral reefs under millions of tons of sand.
The World Islands are the ones shaped like a map of the world that you can see from space.
Most of them are empty. Construction started in two thousand three, the global financial crisis hit in two thousand eight, and the project essentially stalled. Some islands have been developed, many are just sandbanks slowly eroding back into the Gulf. It's a cautionary tale about building artificial islands as speculative real estate rather than as necessary infrastructure.
Versus Singapore, where it's not speculative — it's literally the only direction they can grow.
Even Singapore is hitting limits. The sand supply issue I mentioned earlier is a genuine strategic vulnerability. They've had to get creative — using recycled construction waste, exploring deep seabed mining for fill material, and as I mentioned, adopting Dutch-style poldering to reduce sand requirements. The Tuas Port project, which will be the world's largest container terminal when fully completed, is being built partly on reclaimed land and partly using caisson technology — enormous concrete boxes that are floated into position, sunk, and filled.
That's the technology that built a lot of harbor infrastructure in the nineteenth and twentieth centuries.
It's having a renaissance because it's less sand-intensive than traditional reclamation. You prefabricate the structures on land, float them out, sink them precisely into position, and you've got a stable foundation. The Mulberry harbors used in the Normandy invasion in nineteen forty-four were essentially the same principle deployed at urgent speed.
The engineering toolkit is actually pretty mature. It's the application that's constrained — by cost, by law, by environmental impact, by the sheer difficulty of the open ocean.
I think that's the core answer to the prompt. Yes, the engineering exists. Yes, countries use it — extensively, in some cases. But the version where an individual or a small group builds an artificial island in international waters as a housing solution or a micro-state? That's not an engineering problem, it's a category error. You're asking a technical question about what is fundamentally a political and legal impossibility.
Even if you solved the legal layer — even if you found a shallow spot nobody claims and built your island and somehow got enough states to recognize it — you'd still be running a continuous engineering project against the ocean, which is the most patient and well-funded adversary on earth.
The ocean has infinite time and infinite energy. Your island has a maintenance budget.
That's the four-word version of this entire episode. The ocean has a maintenance budget.
Let me add a footnote that's encouraging. The floating city projects in the Maldives and Busan are not just vanity projects. They're serious attempts to solve a real problem — how do you house people when coastal land is disappearing? And they're doing it within existing legal frameworks, with sovereign backing, using engineering that's been validated in offshore oil and gas for decades. That's the version of this that might actually scale.
The seasteading dream migrates from international waters to territorial waters, from libertarian micro-states to climate adaptation infrastructure, and from private islands to public housing.
Which is exactly the opposite of what the seasteading movement wanted, but it's the version that might actually work. The Oceanix project in Busan is specifically designed to be affordable housing, not luxury condos. The Maldives project is about keeping a nation's population housed as the land literally disappears beneath them.
There's something almost poetic about that. The libertarian dream of escaping government ends up being co-opted by governments as a tool for keeping their citizens alive.
The ultimate failure mode of seasteading wasn't engineering or even law — it was the assumption that you could build a community without politics. Politics is just the name for how groups of humans make decisions about shared resources. You can't escape it by moving to a platform. You just get a smaller, more intense version of it.
If your platform is in international waters with no legal framework, the politics arrives in the form of whoever has the biggest boat.
King Tupou understood that perfectly.
Now: Hilbert's daily fun fact.
Hilbert: In the Ottoman Empire, the office of the Chief Astrologer — the Müneccimbaşı — was a formally budgeted administrative position that managed a staff of subordinate astrologers, maintained an official observatory, and issued legally binding calendars that determined when government offices opened and closed. The position was so bureaucratically entrenched that when a seventeenth-century Müneccimbaşı named Hüseyin Efendi was executed for a failed prediction, his office continued operating with an interim appointee within three days.
The bureaucracy that survives its own bureaucrat's execution is the most Ottoman thing I've ever heard.
Legally binding calendars. There's a phrase that's going to sit with me.
This has been My Weird Prompts. Thanks to our producer Hilbert Flumingtop. You can find us at myweirdprompts dot com or wherever you get your podcasts. If you enjoyed the show, leave us a review — it helps.
Until next time, keep your islands above the waterline and your astrologers on payroll.
