In July 2022, Crown Prince Mohammed bin Salman released a promotional video for The Line—a planned linear city stretching 170 kilometers across the Saudi desert, 500 meters tall, 200 meters wide, clad entirely in mirrored glass, housing nine million people, with no cars, no streets, no carbon emissions, and every daily necessity within a five-minute walk. The video was produced with the aesthetic confidence of a prestige television trailer and the engineering specificity of a fever dream. There would be flying taxis. Robot butlers. A sports stadium suspended 350 meters in the air. Vertical farms. Artificial intelligence managing the entire city like a cognitive organism. The prince called it “a civilizational revolution” and said it would be the most livable city on the planet “by far.”

Three and a half years later, construction on The Line is suspended.

The Saudi sovereign wealth fund paused work on September 16, 2025. The NEOM CEO quit. The 2029 Asian Winter Games, which were supposed to be held at Trojena—a ski resort built on manufactured snow in the Saudi mountains, another NEOM subproject—were indefinitely postponed in January 2026. The workforce has been cut by roughly 35 percent. Over a thousand employees have been relocated from the remote construction site to Riyadh. The sovereign wealth fund wrote down $8 billion from the project. Internal evaluations suggest the final cost of The Line alone could approach $9 trillion—not billion, trillion—which is roughly nine times Saudi Arabia’s annual GDP. And the Financial Times reported that MBS himself has now privately accepted that the original vision will be realized as something “far smaller.”

This is, by any objective measure, one of the most spectacular collisions between architectural ambition and physical reality in modern history. And the thing is, it was always going to end up here. The warning signs weren’t subtle. They were load-bearing.

What was actually proposed

The original specifications for The Line read like someone took a megastructure from a science fiction novel and submitted it as an engineering brief without checking whether the laws of physics had signed off. Two parallel mirrored walls, each 500 meters tall—roughly the height of One World Trade Center—extending 170 kilometers in a straight line from the Red Sea coast inland through desert valleys and over mountain terrain, with a 200-meter-wide gap between them containing a multi-layered city stacked vertically: residential, commercial, recreational, and transportation layers, all climate-controlled, all connected by automated transit, all powered by renewable energy.

The planned population density was 260,000 people per square kilometer. For comparison, Manila—the most densely populated city on Earth—has a density of roughly 44,000 per square kilometer. The Line, as designed, would be six times denser than the densest city that currently exists. And it would achieve this density not by building outward, as every city in human history has done, but by building upward and linearly, which creates engineering constraints that compound at every scale: structural loading, wind forces on a 500-meter-tall continuous surface, seismic risk in a region with active fault lines, thermal management in a desert where surface temperatures regularly exceed 50°C, water supply for nine million people in one of the driest regions on Earth, and the logistical challenge of moving millions of tons of construction material to a remote site with no existing infrastructure.

The cost estimate of $500 billion was, in retrospect, the most optimistic number in the history of optimistic numbers. An engineering analysis by Imperial College London noted that constructing The Line to its stated specifications within the proposed timeline would require building at 15,000 times the rate of normal U.K. construction. The volume of The Line’s enclosed space—roughly 17 billion cubic meters—at standard high-rise construction costs of about $1,000 per cubic meter implies a structural cost alone of $17 trillion. You don’t have to be a construction engineer to sense that the math wasn’t done before the video was shot.

What was actually built

The construction that did happen is not nothing—and this is the part that gets lost in the “it’s all a fantasy” narrative. Satellite imagery from late 2024 shows massive earthworks, completed buildings, grid-like infrastructure layouts, and support facilities along the western end of the route near the coast. The 170-kilometer trench that was excavated across the desert is visible from space. Concrete foundations have been poured. Vertical cores—the structural columns that will support the mirrored walls—have been started along the initial 2.4-kilometer “Phase One” section.

Phase One was always going to be the proof of concept: a small section near the Red Sea containing anchor assets—residential units, commercial space, a marina, and the foundations for the stadium—designed to demonstrate the concept and attract further investment. The work on that section is real, it’s substantial, and it represents an enormous expenditure of capital, labor, and engineering effort.

But Phase One is 2.4 kilometers out of 170. That’s 1.4 percent of the total length. And even that section isn’t complete. The cores are partially built. The steelwork for the outrigger beams hasn’t been installed. The mirrored cladding—the visual signature of the entire project—exists only in renders. The timeline for completing even this initial section is unclear, because construction was suspended in September 2025 and as of March 2026, work has not resumed.

Why it stalled

The proximate cause is money. Oil prices have been softer than Saudi Arabia’s budget requires, the sovereign wealth fund’s portfolio has underperformed, and the combined cost of Vision 2030’s megaprojects—of which NEOM is only one—has exceeded the kingdom’s ability to fund them simultaneously. When you’re also building the world’s tallest skyscraper in Jeddah, a massive entertainment city called Qiddiya with a Six Flags theme park, and dozens of other giga-scale developments, something has to give. The Line, being the most expensive and most speculative of the bunch, was the logical candidate for a strategic pause.

But the deeper cause is that the project was, from inception, designed backward. The vision came first. The engineering came second. The constraints came last—or, more accurately, never fully arrived. This is the pattern that every failed megaproject follows, from Fordlandia to the Concorde to the Superconducting Super Collider: you start with an inspiring image of the finished product, work backward to figure out how to build it, discover that the physics or the economics or both don’t cooperate, and spend years trying to close the gap between what you announced and what you can deliver before quietly scaling back and declaring the scaled-back version was the plan all along.

The specific engineering problems with The Line have been catalogued exhaustively by independent analysts. Pouring concrete at scale in a remote desert environment requires perfect consistency over long durations—too fast and the ingredients separate, too slow and it sets unevenly. The mirrored glass exterior would create a solar death ray effect, concentrating reflected sunlight onto the ground between the walls at temperatures that could melt asphalt. The structural loads on a 500-meter-tall continuous wall extending for 170 kilometers—including wind loading, thermal expansion, and seismic forces—exceed anything that’s been built anywhere on Earth. The water supply for nine million people in the Tabuk desert would require the largest desalination infrastructure ever constructed, in a location with no existing water infrastructure. Each of these problems is solvable in isolation. Together, at this scale, in this timeline, in this location, they compose something approaching impossibility.

What happens now

According to the most recent reporting, architects have been tasked with figuring out how to repurpose the infrastructure that’s already been built—the trench, the foundations, the cores—into something deliverable. The leading candidates appear to be a much shorter initial city section near the coast (the 2.4-kilometer Phase One, potentially extended to 5 kilometers) at a reduced height, with the remaining earthworks potentially repurposed for industrial use, including AI data centers. The coastal location is considered an asset—saltwater for cooling, proximity to shipping lanes, existing port infrastructure from the Oxagon industrial zone nearby.

The full 170-kilometer vision has been deferred to a “multi-decade timeline,” with 2045 cited as a possible completion date—though at current pace, independent analysts have projected full realization could stretch into the 2070s or 2080s, if it happens at all. The Mukaab, a massive cuboid building planned for another Vision 2030 project in Riyadh, has already been cancelled outright. Trojena’s Asian Winter Games have been indefinitely postponed. The broader NEOM ecosystem is being triaged: some components (Oxagon, the industrial port) appear viable; others (Trojena, Sindalah luxury island) are in critical condition.

What it actually tells us

The Line is not a story about Saudi Arabia being uniquely delusional. It’s a story about what happens when a planned city is designed as a marketing asset rather than an engineering project—when the render is more important than the spec sheet, when the announcement timeline drives the construction timeline rather than the other way around, and when the person commissioning the project has the authority to override every engineer in the room who’s trying to explain why the physics don’t work.

The historical record on planned cities built from scratch is not encouraging even under far less ambitious parameters. Brasília works but is widely considered sterile. Naypyidaw, Myanmar’s purpose-built capital, is a ghost town. Masdar City in Abu Dhabi, billed as the world’s first zero-carbon city in 2006, has been quietly scaled back to a small neighborhood. Songdo in South Korea, designed as a ubiquitous computing city, is roughly half-occupied a decade after opening. The pattern is consistent: planned cities that succeed tend to be modest in scope and flexible in design. Planned cities that lead with a grand vision and a promotional video tend to become very expensive lessons in the difference between rendering and reality.

The Line may still produce something useful. A 2.4-kilometer coastal development with advanced infrastructure and renewable energy systems would be a significant achievement, even if it bears almost no resemblance to the mirrored canyon city in the original video. But that’s the downgrade that reality imposes on ambition when the constraints weren’t named before the plan was drawn, and it’s a pattern as old as city-building itself.

We cover NEOM, The Line, and the full history of utopian megaprojects—from Fordlandia to Auroville to the kibbutz movement—across our Utopian Societies course. The pattern of visionary ambition meeting structural reality is the through-line of the entire course, and The Line is its most expensive modern example.

In 1960, the United States Navy purchased a Pacific white-sided dolphin. Not to train it. Not to weaponize it. To study how it swam. The researchers at Point Mugu, California, wanted to understand the dolphin’s hydrodynamic efficiency—how it moved through water with so little drag—because they thought the answer might help them build faster torpedoes. The torpedoes never got faster. But somewhere in the process of studying the dolphin’s body, someone noticed that the animal was extraordinarily intelligent, easily trainable, and capable of operating untethered in open ocean without swimming away. By 1963, the Navy Marine Mammal Program was formally established. By 1965, a bottlenose dolphin named Tuffy was carrying tools and messages to aquanauts living 200 feet underwater in the SEALAB II habitat off La Jolla. By 1967, the program was classified. It stayed classified for over two decades.

The program was declassified in the early 1990s, and what emerged was not what the conspiracy theories had predicted. There were no laser-equipped attack dolphins. No kamikaze cetaceans strapped with explosives. No underwater assassins trained to kill enemy divers with poisoned darts—a rumor that surfaces approximately once per hurricane season and has never been true. What the Navy had actually built was something considerably less cinematic and considerably more interesting: a program that exploits the fact that dolphins have a biological sonar system so sophisticated that no technology humans have built can match it, and that this sonar can be directed—through years of positive-reinforcement training—toward finding mines, detecting swimmers, and recovering objects on the ocean floor. The dolphins aren’t weapons. They’re sensors. And they’re better sensors than anything the defense industry has managed to engineer.

What dolphins actually do for the Navy

The Navy Marine Mammal Program, based at Naval Base Point Loma in San Diego, currently operates with roughly 120 marine mammals—primarily bottlenose dolphins and California sea lions—organized into five operational teams, each designated by a “Mark” number. The teams are trained for specific mission profiles, and the division of labor between species is based on biology, not preference.

Dolphins handle mine detection. Their biological sonar—echolocation—works by emitting clicks from a structure in their forehead called the melon, then processing the returning echoes to build a three-dimensional acoustic picture of their environment. The resolution of dolphin echolocation is extraordinary. A trained Navy dolphin can detect a mine buried in seafloor sediment, distinguish it from surrounding debris, and mark its location with a transponder—all in murky water where human divers can barely see their own hands and sonar equipment returns a useless mess of false positives. The reason the Navy hasn’t replaced dolphins with autonomous underwater vehicles is not sentimentality or tradition. It’s that the dolphins are genuinely better at this task than any machine the Navy has tested. The program director, Dr. Mark Xitco, put it directly in a 2024 interview: the animals are natural hunters, and all the Navy does is change what they’re hunting for.

The Mark 7 team is the primary mine countermeasure unit. These dolphins locate mines in shipping channels and coastal waters and mark them for explosive ordnance disposal teams to neutralize. They’ve been deployed operationally—during the Iraq War in 2003, Navy dolphins cleared mines from the port of Umm Qasr, enabling humanitarian aid ships to dock. That wasn’t a training exercise. That was a real minefield in a real combat zone, and the dolphins found mines that conventional minesweeping equipment had missed.

California sea lions handle a different set of tasks. They lack echolocation but have exceptional underwater directional hearing and low-light vision, which makes them ideal for swimmer detection and object recovery. The Mark 5 team trains sea lions to detect and intercept unauthorized divers approaching Navy ships or harbor facilities. In a 2011 demonstration, a Navy sea lion successfully located and tagged a Navy SEAL attempting to infiltrate a harbor—five times in a row. The sea lion attaches a clamp connected to a line onto the swimmer’s leg, and surface personnel reel them in. The swimmer generally doesn’t know the sea lion is there until it’s too late, because the sea lion operates silently and approaches from below in dark water where human visibility is near zero.

The Mark 4 team uses dolphins for swimmer detection as well, providing force protection for ships at anchor and in port. After the USS Cole was attacked by a suicide boat in Yemen in 2000, the Navy significantly expanded its marine mammal force protection capabilities. Dolphins can detect an approaching swimmer at distances far greater than any underwater sensor system and can do it in harbor environments full of acoustic clutter—boat engines, currents, pier structures—that confound artificial sonar.

How training works (and why they don’t leave)

Navy dolphins train for five to seven years before operational deployment. Sea lions train for two to five years. All training uses positive reinforcement exclusively—fish, toys, tactile interaction, verbal praise. The Navy’s stated policy prohibits aversive training techniques, and the program’s veterinary staff includes nearly 20 veterinarians providing round-the-clock care.

The part that surprises most people: the dolphins work untethered in open ocean. There is no leash, no fence, no barrier between a Navy dolphin on a mine-detection mission and the entire Pacific. They can leave whenever they want. Over the decades of the program, a few have. Almost all stay. Dr. Xitco’s explanation is practical rather than sentimental—the dolphins are comfortable, well-fed, socially stimulated, and have built trust relationships with their handlers over years of cooperative work. If the relationship weren’t working, the animal would simply not come back, and the Navy would be out one very expensive training investment.

The dolphins are bred within the program—the Navy has bred its own dolphins exclusively since 1989 and hasn’t acquired any from the wild since. Sea lions are obtained from rescue organizations, marine parks, or occasionally the wild, and only neutered males participate in the program. The animals live in 30-by-30-foot enclosures in San Diego Bay with underwater gates connecting them, so they can socialize freely. When not training or deployed, they swim, play, and do what dolphins do—which, based on available evidence, includes annoying each other and showing off to visitors.

The program’s veterinary and behavioral research output is substantial. Navy marine mammal scientists have published over 1,500 peer-reviewed papers on dolphin and sea lion physiology, cognition, acoustics, and health. A 2023 New York Times feature explored the program’s research on dolphin aging through a 57-year-old Navy dolphin named Blue—a data point in a longitudinal health dataset that no aquarium or wild population study can match, because the Navy has been monitoring individual animals’ blood chemistry, hearing, cardiac function, and body composition continuously for decades. The program essentially invented the protocols for voluntary veterinary participation in marine mammals—training dolphins to present body parts for examination, hold still for ultrasounds, and voluntarily give blood samples—and those techniques have since become standard practice across the entire zoological community.

The ethical debate

The animal welfare controversy around the program is real, ongoing, and not entirely one-sided. Critics—most prominently Ric O’Barry of the Dolphin Project, and various animal rights organizations—argue that confining highly intelligent, socially complex animals for military purposes is inherently unethical regardless of how well they’re treated, that deployment to combat zones puts them in danger they can’t consent to, and that advancing technology should have made the program obsolete by now.

The Navy’s counterargument is that the animals are treated to a standard that exceeds most marine parks, that no dolphin has ever been trained for attack missions (a claim the Navy has maintained consistently and which no credible evidence has contradicted), and that the echolocation capability remains genuinely irreplaceable. The “technology should replace them” argument has been made for thirty years, and the technology keeps not replacing them—autonomous underwater vehicles are getting better, but in cluttered coastal environments with variable sediment, biological sonar still wins.

The honest assessment is that both sides have legitimate points and neither is fully satisfying. The dolphins are well cared for by any measurable standard. They’re also confined, deployed to environments they wouldn’t naturally inhabit, and serving a purpose that has nothing to do with their own interests. Whether that’s acceptable depends on where you draw the line on using intelligent animals as instruments of human policy, and reasonable people draw that line in different places.

The part nobody talks about: what the dolphins teach us

The less discussed dimension of the program is what it’s contributed to our understanding of dolphin cognition and sensory biology. Working with the same individual dolphins over lifetimes—some Navy dolphins have been in the program for 30 or 40 years—has produced data on echolocation processing, hearing sensitivity, cognitive decline with age, social communication, and problem-solving that simply doesn’t exist elsewhere. The longitudinal health records alone are one of the most complete datasets on bottlenose dolphin physiology ever assembled.

The echolocation research has particular implications for engineering. Dolphin sonar can distinguish between objects of nearly identical size and shape based on material composition—they can tell the difference between a hollow aluminum cylinder and a solid one at distance, in murky water, by processing acoustic returns that differ by microseconds. Understanding how the dolphin auditory cortex achieves this has been a research objective for decades, and the answers have implications for synthetic sonar design, medical ultrasound, and underwater communications.

There’s also the hearing research, which has a conservation dimension. The Navy funds extensive study of how anthropogenic ocean noise—from ship engines, sonar systems, and underwater construction—affects marine mammal hearing. This research, conducted in part on Navy dolphins whose hearing baselines have been tracked for years, provides some of the best data available on noise-induced hearing damage in cetaceans. The irony that the Navy is simultaneously one of the largest sources of ocean noise pollution and one of the leading funders of research on its effects is not lost on the marine biology community.

The program occupies a strange space—militarily operational, scientifically productive, ethically contested, and genuinely unmatched in what it’s revealed about dolphin intelligence and sensory capability. Whether you think it should exist depends on how you weigh those dimensions against each other. What’s not debatable is that it has produced more primary data on bottlenose dolphin cognition, health, and echolocation than any other single institution in history, and that the animals it trains can find a mine in a harbor that a billion dollars’ worth of autonomous systems still can’t.

We cover the Navy Marine Mammal Program—alongside carrier pigeons, mine-detecting rats, war horses, and every other animal that’s been drafted into human conflicts—across our Animal Heroes course. If the echolocation-versus-synthetic-sonar question got you, or the 57-year-old dolphin named Blue, that’s where the full story lives.