In September 2026, newsfeeds lit up with a story that seemed perfectly heartwarming on its surface: truck driver Joe Macken spent twenty-one years in his free time gluing together a model of New York from balsa wood, cardboard, and glue — all five boroughs, 320 sections, around 800,000 buildings. His daughter convinced him to post a video on TikTok, the clip racked up millions of views, and the Museum of the City of New York opened an exhibition called "He Built This City" — the handmade city standing steps away from the permanent exhibition about the real New York itself.
A story designed for "aww, how sweet" stubbornly refuses to stay sweet on second glance. Back in first grade, Macken went on a field trip to the Queens Museum and saw the Panorama of the City of New York: a 1964 model covering nearly nine hundred square meters, inhabited by 895,000 buildings. "I walked around it and thought: this is what I want to do," he recalled decades later. The man they called a solitary DIY-er turned out to be the last link in a century-long chain: he wasn't just building a city — he was building an answer to the model he'd seen as a child.
And that's when the non-obvious question emerges. Not "why does a weirdo need a city model," but why has a civilization that has satellite imagery, CAD files of every building, and Google Earth on phones built the same city out of physical material at least four times over a hundred years — each time for completely different reasons?
Start with the Panorama itself, because its fate breaks the usual notion of a "museum model." In 1961, Robert Moses — the man who built New York for thirty years and who, according to biographer Robert Caro, was "the greatest builder America had ever known" — commissioned Raymond Lester's firm to model all five boroughs for the 1964 World's Fair. More than a hundred people worked for four years. The cost is known to the penny: $672,662.69.
But the main thing isn't the size, it's a clause in the contract. It states in black and white: the model must be designed "as a comprehensive planning and research tool for use after the World's Fair closes." This is not a souvenir. This is — a snapshot of the city officially designated for work. They made the model modular: 273 sections on metal legs that could be carried away, updated, replaced. And indeed: in 1967, 1968, 1969, and 1974, Lester's firm incorporated new buildings into the model, and these updates through 1974 were paid for not by the museum but by Moses's bridge monopoly — the Triborough Bridge and Tunnel Authority — meaning the device lived de facto within the city's production infrastructure.
The scale of this planning-mind machine is easy to underestimate. One inch on the model equals a hundred feet in reality, a person the size of a grain of sand, the Empire State Building fifteen inches. The Panorama contained more than 830,000 structures (the museum's official page lists 854,635 at the premiere, and after the 1992 update — 895,000; the spread in sources honestly shows how hard it is to even count a city). For standard construction, Lester invented 190 standardized forms plus 24 combinable ones; for 25,000 landmarks — skyscrapers, museums, stations — each was made by hand. Windows weren't painted: phosphorescent paint through stencils, because the model could do "night" — the day-night cycle lasts twelve minutes, and in darkness the city glows with windows and parks glow green.
And yet — a detail that captures the entire twentieth century: 3,172 colored bulbs are sewn into the model. Each marks a municipal building — police precinct, school, hospital, courthouse, library, subway station, electrical substation. Moses wasn't building a model of houses; he was building a showcase of municipal socialism — "look how much the city does for you." The model was an argument.
The contract with Lester allowed deviation from reality of no more than one percent. On paper this sounds innocent, until you learn what was done to achieve it. Joseph Ivancic, president of Lester Associates, recalled: "Everyone thought we'd call the city and they'd give us a book listing every building in the five boroughs. No such thing existed." A unified catalog of New York City buildings in 1961 did not exist in nature. They had to build it themselves: 109 vertical and 5,000 oblique aerial photographs, thousands of photos of individual structures, 19 topographic maps from the Geological Survey, 35 Sanborn maps with municipal facilities, city tax maps with every lot marked with its footprint and number of floors.
Thirty-five bridges in the model were made from original engineering drawings: the design was photographically transferred onto brass foil and acid-etched — the finest cables of the Verrazano and George Washington bridges exist on the model as actual copies of their load-bearing geometry. The terrain was sculpted from then-novel polyurethane foam using topographic maps. The signature in the bushes on the Jamaica Bay islands — "Bill," "Ed," "N.P." (different sources read it differently) — was left by workers ensuring themselves immortality at 1:1200 scale.
And here's what this engineering means in historical context: all of this was done in an era when a unified digital twin of the city didn't even exist as a concept. The Panorama wasn't a sentimental toy — it was the most accurate unified model of New York that existed on the planet. Urban-resolution satellite imagery wouldn't appear until the nineties, and until then a complete "snapshot" of the city existed in exactly one copy — in Queens, made of foam and balsa.
Now the juicy part. If a model is a snapshot of a city, then anyone who places their project on it at the same scale gets an honest answer about how that project will change the city. That's exactly how the Panorama once functioned as a weapon — in the most apolitical hands possible.
In the late 1980s, Donald Trump proposed Television City on Manhattan's west side — a megacomplex with the world's tallest building. A group of citizens calling themselves Westpride commissioned a model of the project at the Panorama's scale and physically installed it on the city model. The result worked instantly: "it became obvious the project was frighteningly gigantic," writes American Heritage, and the tower was never built. No slides, no rhetoric, no infographics: the visual truth of weight — concrete, impossible to dispute in a public hearing.
This is the main engineering ethic of physical models that digital can't replicate: scale is a contract. Every building has the same coefficient to reality, and therefore comparison is honest by construction, not by choice of angle. A render cheats with lighting and camera position; half a dozen expert reports cheat with assumptions; a polyurethane block placed next to another polyurethane block cannot cheat. Moses knew this and therefore built his showcase of roads at this scale; Westpride knew this and therefore bought a model of the enemy. One tool — two opposite masters, and it served both honestly.
By the way, Moses himself got an irony from the model he didn't ask for. When the model was fully updated for 1994, a New York Times Magazine critic looked at the result and wrote: "The good news is the model looks just like the real city. The bad news is it looks just like the real city, because thirty years of other people building looks like six months of Moses working." The snapshot is honest even to the one who commissioned it.
The showcase-model lineage predates the Panorama by a quarter century, and within it are two figures separated by an abyss. The first — Norman Bel Geddes, industrial designer who built "Futurama" for General Motors at the 1939 World's Fair: a model of America's future covering an acre — half a million buildings, a million trees, fifty thousand cars, of which ten thousand continuously drove on a fourteen-lane highway. A visitor sat in one of 552 plush chairs and for eighteen minutes "flew" over this America to a narrator's voice: the journey started in daylight, passed through night, and ended at dawn — a metaphor for "new era." A poll of a thousand people leaving the fair gave GM's exhibit 39.4 points versus 8.5 for second place. Business press wrote about thirty thousand people a day standing in line.
"Futurama" wasn't a forecast but a price list: it sold the country the idea of federal highways — and by admission of the US Department of Transportation directly influenced the laws that gave rise to the Interstate Highway System. The model changed the planet's landscape, though it only existed for two seasons. After the fair closed, like most fair props, it was turned over for war needs — a metaphor more precise than any you could invent: the showcase of tomorrow's world went to be melted down for today's world.
The second figure — Albert Speer, and his model was in spirit a twin brother to "Futurama." In 1938–1939, Berlin housed a physical model of "World Capital Germania": a north-south axis, a triumphal arch whose opening would entirely contain the Parisian one, and a domed "Great Hall" 290 meters high, whose oculus — 46 meters — would contain the dome of St. Peter's together with the Pantheon rotunda. "As world capital, Berlin will be comparable only to Ancient Egypt, Babylon, and Rome!" — so Hitler described the city that stood before him in scale.
The same genre, the same mechanics — a city built to scale to transmit to a person the experience of the future. The difference is that for "Futurama" the future was advertising's consumable material, while for "Germania" it was a promise tied to the date "1950." In 1943 work stopped; the concrete cylinder driven into Berlin soil to test whether it could hold the arch settled 18 centimeters in three years against six allowable — and stands to this day, the project's most honest monument: the model turned out more accurate than the propaganda. The earth physically wasn't going to hold the future drawn for it at 1:1000 scale.
Both 1939 stories reduce to one thing: in the twentieth century, the city model first became not a copy of what exists but a device for experiencing what doesn't exist. And this device was used with equal enthusiasm by a car salesman and a totalitarian empire. The tool is neutral; the master determines honesty.
Against this backdrop, Joe Macken's story stops being sweet and becomes precise. He's the antithesis of everything above, and simultaneously their legitimate heir. The Panorama had a budget, a hundred people, and four years; "Futurama" had GM's budget and staff; Macken had evenings after his shift, balsa, cardboard, glue, and twenty-one years. When his house ran out of space, he rented a warehouse. The result — a model covering 125 square meters, 320 sections, each roughly a square mile of city.
He made a city that can't be confused with any economic genre. This isn't a city-planning tool: it doesn't update and doesn't claim to be current. This isn't propaganda: it has no client and no agenda. It's not even a museum replica. His model is the only one in this series that answers the question "what does the city mean to you personally." A delivery driver living a hundred fifty miles from New York needed to keep the city with him — so he built it. "I wanted to keep it with me," he explains, "well, I figured I'd better build it."
Curator Elizabeth Sherman recalled when staff first saw the model: "We all stood there and screamed: look, our museum! There's the Met, there's the Guggenheim." This is the key to the effect: the handmade model triggers in a person an act of recognition — you are physically inside someone's memory of the city. Digital maps don't do this not because they're worse made but because they have nothing to do it with: they don't have a hand through which the city passed as effort. Twenty-one years isn't the model's cost. It's its story.
One question remained: why all this in an era when you can "fly through" a city on your phone? The answer has already been given — by the museum itself.
In July 2023, the Queens Museum together with Pratt Institute, CyArk, and Bloomberg Philanthropies launched Panorama Digital Twin — a digital twin of the physical model. The logic here is painfully beautiful: the physical model, frozen on January 1, 1992, is too expensive to update to stay current; the digital one is too cheap and "light" to be loved. So they work in tandem: the physical holds craft, tactility, and the fact "someone assembled this city by hand," the digital holds currency, search, and memory — on the digital twin anyone can "pin their story" to a specific building.
Meanwhile the physical model lives quite the commercial life: since 2009 the museum has sold "adoption" of buildings — for a donation starting at fifty dollars, any structure in the model can be "taken under care" for a year. The Twin Towers of the WTC, added to the model in the 1992 update, still stand on it: the museum decided not to remove them until the site's redevelopment is complete. The model honestly remained "a city across different times": conditions as of January 1, 1992, with several icons from 2009 and later spot-glued on top.
And there's a fourth, living sector of this genre where physical models have no intention of dying. Hamburg's Miniatur Wunderland — 1,694 square meters, 16 kilometers of track, 292,110 figures, more than 400 permanent employees — opened its Monaco section in April 2024: 70 square meters, eleven years of development, over five million euros, 22 meters of Formula 1 track with up to 14 cars racing continuously, controlled by Hall-effect magnetic fields, under 39 kilometers of cable and a hundred thousand lines of code. You can watch the race on "television" screens — cameras 29 millimeters in size are built into the track.
The meaning of this evolution is precise: the model has stopped being a snapshot of a city and become a performance of a city. Not "here's what it looks like" but "here's how it lives right now" — with traffic jams, weather, schedules, and train breakdowns. Come to Hamburg in a month — the Monaco race will run differently: the software is explicitly programmed so each race is unique. The Panorama was called the most honest snapshot of New York; Wunderland has no snapshots at all — only live broadcast.
The physical city model is a technology with a characteristic trajectory: born as a showcase of power, matured as a working tool, aged as a museum exhibit and died with the status of snapshot — but was reborn as memory and as performance. Yet its deep function hasn't changed once in a hundred years. The model is a way to make a person see the city as a whole, not just their street. In 1939 this was used to sell a road or an empire. In 1964 — to show residents the scale of their own infrastructure. In the late eighties — to stop a tower that lawyers couldn't stop. In 2026 — so a person stands over a city made of balsa with binoculars and quietly recognizes their own.
Digital doesn't cancel this, because digital has a different task: it makes the city computable. The physical model makes the city — tangible. These are different interfaces to the same reality, and in 2026 they've finally diverged onto different floors of the same museum: the twin on your phone, the original in the hall where all 800,000 buildings were assembled by one person, about whom not a line would have been written if not for his daughter and TikTok. A good test of any technology: what does it lose if the last pair of hands maintaining it disappears. For the digital twin the answer is — an administrator. For the balsa city — a daughter, a warehouse, and twenty-one years of someone else's evenings. That, honestly, is the entire difference between data and memory.