This episode explores the tension between planned and organic cities through three iconic case studies: Brasília, Chandigarh, and Canberra. Each was designed from scratch by celebrated architects and planners, yet each ran into the same fundamental problem: cities are not machines. They're ecosystems. Brasília's Pilot Plan, designed for exactly 500,000 people, now holds 4.8 million in its metro area. Its egalitarian superblocks were abandoned by the elite within a generation, while unplanned settlements like Taguatinga — built by the workers excluded from the plan — became more economically dynamic than the planned city itself. The design separated everything by function, connected by a 14-lane highway with no crosswalks, making it impossible to walk between neighborhoods. The same pattern repeats in Chandigarh and Canberra: static snapshots optimized for monumentality and traffic flow, with no room for informal economic activity, mixed uses, or the fine-grained adaptation that makes organic cities resilient. Jane Jacobs called this "organized complexity" — dense, mixed-use, short blocks, old buildings mixed with new. The episode argues that the very act of top-down planning guarantees failure at the human scale because it replaces millions of individual decisions made by people living on the ground with the hubris of a dozen architects working from maps and models.
#3981: Can a City Designed From Scratch Actually Work?
Brasília, Chandigarh, Canberra — why expert-designed cities often feel lifeless compared to organic ones like Jerusalem.
Episode Details
- Episode ID
- MWP-4160
- Published
- Duration
- 28:20
- Audio
- Direct link
- Pipeline
- V5
- TTS Engine
-
chatterbox-regular - Script Writing Agent
- deepseek-v4-pro
AI-Generated Content: This podcast is created using AI personas. Please verify any important information independently.
Downloads
Transcript (TXT)
Plain text transcript file
Transcript (PDF)
Formatted PDF with styling
You Might Also Like
Never miss an episode
New episodes drop daily — subscribe on your favorite platform
New to the show? Start here#3981: Can a City Designed From Scratch Actually Work?
Daniel sent us this one — he's been thinking about Jerusalem, about how you can walk from where we live to the Western Wall and pass through neighborhoods built in completely different centuries without even noticing the seams. And that got him wondering about the opposite extreme. Cities that weren't layered up over millennia, but drawn on a blank page. His question is basically this: if humans keep getting better at designing things, shouldn't a city designed all at once by experts be more functional than one that just... What are the major planned cities of recent decades, and what challenges have they run into that the organic cities somehow dodged?
It's a genuinely sharp question, because on paper the rationalist case is compelling. You sit down with a blank site, you calculate optimal traffic flow, you zone everything perfectly, you don't have to work around medieval street patterns or whatever mess centuries of ad-hoc decisions left you. You get to design the city the way a city should work.
The track record is... let's say instructive. And the timing is interesting because right now there are more than a dozen brand-new planned cities under construction globally. NEOM in Saudi Arabia, Nusantara going up as Indonesia's new capital, California Forever in Solano County. Billions of dollars, total blank-slate ambition. The question of whether top-down planning can actually deliver a functioning human city isn't academic anymore — it's being tested in real time with real money.
The core tension Daniel's pointing at is this: Jerusalem works despite being a planning nightmare by any rational metric. Narrow winding streets, mixed uses jumbled together, infrastructure retrofitted onto ancient foundations. It shouldn't function, and yet it does — dense with life in a way that feels almost impossible to replicate on purpose. Meanwhile, these planned cities are optimized on paper for efficiency, and somehow they keep producing places that feel...
That's exactly what we want to dig into. Not just "planned cities feel weird" as a vibe, but the specific mechanisms of why they fail, and what those failures tell us about how cities actually work as complex systems. We're going to look at three of the most iconic planned cities of the twentieth century — Brasília, Chandigarh, and Canberra. Each was a radical break from the past, each was designed by some of the most celebrated architects and planners of their era, and each ran into consequences that nobody predicted.
A city conjured from nothing in the Brazilian savanna in the late nineteen-fifties, designed to be the perfect modernist capital. Lucio Costa won the competition for the master plan, Oscar Niemeyer did the monumental architecture. The whole thing was laid out in the shape of an airplane — or a bird in flight, depending who you ask — with government buildings along the fuselage and residential superblocks arrayed along the wings. Everything separated by function. No traffic lights on the main axis. A gleaming machine for living.
The ambition was staggering. They built it in forty-one months. Costa's Pilot Plan was designed for exactly five hundred thousand people — a precise, self-contained vision of what a modern city should be. Identical apartment blocks for all social classes, because the plan was explicitly egalitarian. Ministers and janitors were supposed to live in the same superquadras.
Today the metro area holds four point eight million people.
Nearly ten times the design capacity. The egalitarian superblocks were abandoned by the elite within a generation, who decamped to gated communities outside the plan. The workers who actually built Brasília weren't even included in the plan — they lived in construction camps, and when the city was done they had nowhere to go, so they built their own settlements. Places like Taguatinga, which grew organically and chaotically outside the Pilot Plan, and which today are more economically dynamic and more densely populated than the planned city itself.
The very people the city was built for ended up living outside it, in the unplanned parts. The plan that was supposed to be perfectly rational produced a city where the most vibrant neighborhoods are the ones that broke all the rules.
That pattern repeats. It's not a Brasília problem — it's a planning-philosophy problem. These cities were designed as static snapshots, not as frameworks for growth. They optimized for monumentality and traffic flow, but they didn't leave room for the informal economic activity, the mixed uses, the fine-grained adaptation that makes organic cities work. Jane Jacobs wrote about this decades ago — cities need what she called "organized complexity." Dense, mixed-use, short blocks, old buildings mixed with new. The exact opposite of the superblock model.
That's the spine of what we're chasing. Three cities, three grand visions, three sets of unintended consequences. And underneath all of it, the question Daniel's really asking: can you design a city that works from scratch, or does the very act of top-down planning guarantee failure at the human scale?
Before we dive into the case studies, I want to sit with that tension — because Daniel's framing assumes something that most people assume. The idea that a city designed by experts should be better than one that just... It sounds obvious. Who wouldn't want the experts in charge?
Yet the expert-designed ones keep producing these strangely lifeless results. So either the experts are bad at their jobs, or the premise itself has a flaw buried in it.
The premise has a flaw. An organic city like Jerusalem is the product of millions of individual decisions made over centuries — where to put a shop, how wide to make a street, which building to tear down and which to preserve. No single mind coordinated any of it. It's inefficient, it's messy, the street pattern makes no geometric sense, and yet the result is resilient in a way that planned cities struggle to match. Because every decision was made by someone who actually lived there, responding to conditions on the ground in real time.
Whereas a planned city is designed by maybe a dozen people in an architecture studio, working from maps and models, trying to anticipate everything a population of half a million will need for the next century. It's hubris dressed up as rationality.
That's the fundamental difference. Organic cities are emergent systems — they grow the way a forest grows. Planned cities are engineered systems — they're built the way you build a dam. The question is whether a city is the kind of thing you can engineer at all, or whether the engineering mindset itself is the problem.
Which raises the obvious question. If planned cities have this track record of disappointment, why are we still building them? Why is Saudi Arabia pouring hundreds of billions into NEOM right now?
Part of it is genuine necessity. Sometimes you need a new capital — Brazil moved its government inland to develop the interior, India needed a new state capital after Partition carved Punjab in half. Those were real problems that required building something from scratch. But there's also an ideological pull. The rationalist dream dies hard. Every generation of architects looks at the mess of existing cities and thinks: we can do better. We'll get it right this time.
To be fair, the three we're looking at were ambitious attempts to solve real problems. Brasília was about unifying a vast country and developing the interior. Chandigarh was about giving a traumatized, partitioned region a new capital that embodied modernist hope. Canberra was about resolving the ridiculous rivalry between Sydney and Melbourne by building a compromise capital in the middle of nowhere.
Each one was a radical break from everything that came before. And each one ran into the same wall: cities are not machines. They're ecosystems. The organic layering that makes Jerusalem fascinating — that palimpsest Daniel was describing, where you walk through centuries without noticing — that's exactly what planned cities struggle to produce. You can't fast-forward history.
Let's start with the most famous planned city of the twentieth century — Brasília. And I want to get specific about what Costa and Niemeyer actually put on paper, because the plan itself is where the seeds of failure were planted.
Walk me through the Pilot Plan.
Costa won the competition in nineteen fifty-seven with a submission that was famously sparse — five sheets of paper, some sketches, basically a manifesto. The city was laid out in the shape of a cross, or an airplane depending how you squint, with two axes. The monumental axis runs east-west, lined with government buildings, ministries, the cathedral, the congress. That's the fuselage. The residential axis curves north-south in a gentle arc — those are the wings. And along those wings, he placed the superquadras.
Each superquadra was designed as a self-contained neighborhood — roughly two hundred fifty by two hundred fifty meters, with six-story apartment blocks arranged around a central green space. Each one had its own school, its own local shops, its own small park. The idea was that you'd never need to leave your superquadra for daily life. Everything within walking distance, inside the block.
Which sounds lovely on paper. Walkable, self-sufficient, green space everywhere.
Here's where the mechanism gets interesting — and where it started to break. The superquadras were separated from each other and from the monumental axis by enormous highways. The Eixão, the main north-south expressway, is a fourteen-lane highway with no crosswalks and no traffic lights. Costa designed it to move cars at highway speeds through the heart of the city. If you wanted to go from your superquadra to the commercial sector, or to the government buildings, you got in a car and drove on the Eixão.
The walkability was only inside the block. Between blocks, you're in a car at eighty kilometers an hour.
That's the first fracture. The plan claimed to be human-centered, with these intimate superquadras, but it connected them with infrastructure designed for machines. You can't walk from one superquadra to another. There are no streets in the traditional sense — no sidewalks with shops, no corners, no places where you might bump into someone. Just the highway and the access roads.
It's like designing a house where every room is beautiful but the only way between rooms is a freeway on-ramp.
That's exactly the image. And the zoning was ruthlessly functional. The monumental axis had government and culture. The residential axis had housing. And between them, a separate commercial sector — the banking district, the shopping areas — all clustered in specific zones. Nothing was mixed. You couldn't have a café on the ground floor of your apartment building, because residential was residential and commercial was commercial. The plan forbade the very thing that makes city streets feel alive.
The population cap. Five hundred thousand, you said.
Designed for exactly five hundred thousand, with no mechanism for expansion. Costa envisioned a city that would reach its ideal size and then... As if you could just declare a city complete. The surrounding land was supposed to remain greenbelt, a buffer of undeveloped cerrado savanna. But the moment Brasília was inaugurated in nineteen sixty, people were already arriving faster than the plan could absorb them. The construction workers — the candangos — had been housed in temporary camps during the build. When the city was finished, they had no superquadra to move into. The apartments were for civil servants, not laborers.
They built Taguatinga.
Taguatinga, Ceilândia, Samambaia — a ring of satellite cities that grew explosively and without any plan at all. Taguatinga was founded in nineteen fifty-eight, before Brasília was even finished, because the workers needed somewhere to live. Today Taguatinga alone has a larger population than the Pilot Plan, and its economy is more diverse — more small businesses, more informal commerce, more of what you'd recognize as street life. The plan excluded the very people who made the plan possible, and they built the real city in response.
The egalitarian dream? Ministers and janitors in the same blocks?
Lasted about a decade. The superquadras were identical in design, but location mattered. The blocks closest to the lake and the monumental axis became more desirable. Diplomats, politicians, wealthy professionals clustered in those, and the further-out superquadras became middle and lower-middle class. Then the elite started building houses outside the plan entirely — gated communities on the lakefront, where they could have gardens and privacy and not share a building with anyone. The plan's attempt to erase class distinctions through architecture simply couldn't overcome the human desire for status and space.
You end up with this bizarre inversion. The planned city — the Pilot Plan — houses about three hundred thousand people today, mostly government workers, in these carefully maintained modernist blocks. And the unplanned ring around it houses over four million, in everything from middle-class apartment towers to informal settlements, and that's where most of the economic and cultural energy lives.
The Eixão, the great highway designed for free-flowing traffic, is now a congested commuter corridor. Every morning, hundreds of thousands of people drive from the satellite cities into the Pilot Plan for work, and every evening they drive back out. The highway that was supposed to be the spine of a rational city became a bottleneck, because the plan never anticipated that most of the city's workers would live outside the plan.
That's the cascade. The rigid population cap forces growth outside the boundary. The strict zoning kills street life inside the boundary. The highway-centric design makes the commute miserable. And the egalitarian housing experiment collapses under market pressure. Each failure feeds the next.
None of this was malicious. Costa and Niemeyer were genuine idealists. They believed architecture could shape a more just society. The problem wasn't bad intentions — it was a model of the city as a machine with fixed inputs and outputs. Real cities don't work that way. They're more like ecosystems — they adapt, they sprawl, they generate unplanned niches. The Pilot Plan had no room for niches.
Brasília shows us what happens when you design a city as a machine with fixed inputs — the machine breaks the moment reality exceeds the spec sheet. But there's another planned city from the same era that tried a different organizing principle. Not a single monumental axis with superblocks, but a grid of self-contained sectors. And it ran into a surprisingly similar cascade. I'm talking about Chandigarh.
Le Corbusier's city.
His only built city, actually. He designed plenty of buildings and plenty of unbuilt master plans, but Chandigarh is the one place where he got to lay out an entire capital from scratch. The context matters. It's nineteen fifty-three. Partition had just carved Punjab in half — Lahore, the old capital, was now in Pakistan. India needed a new state capital, and Nehru saw it as a chance to build something that announced modern India to the world. His exact words were that Chandigarh should be "a new town, symbolic of the freedom of India, unfettered by the traditions of the past.
The same modernist optimism as Brasília, but with a different formal structure.
Le Corbusier divided the city into sectors — rectangles of eight hundred meters by twelve hundred meters, each designed as a self-sufficient neighborhood. Schools, shops, green space, all inside the sector. And he imposed a strict road hierarchy, what he called the V-one to V-seven system. V-ones are the big arterial roads that connect sectors. V-sevens are pedestrian paths that snake through the green spaces inside each sector. The idea was total separation — fast cars never mix with pedestrians, through-traffic never enters the neighborhood.
Which sounds, again, rational. Keep the cars away from the kids, keep the commuters on the arterials.
The Capitol Complex at the northern edge is extraordinary — massive sculptural concrete forms set against the Himalayan foothills. But here's what the plan assumed. It assumed universal car ownership. It assumed formal employment in government and professional sectors. It assumed middle-class nuclear families who could afford to live inside the sectors and follow the building codes.
The building codes were strict.
Le Corbusier specified everything — building heights, setbacks, materials, even the color palette. The aesthetic was to be maintained by law. And this is where the exclusion mechanism kicks in. If you're a street vendor, where do you go? The sectors have designated commercial areas, but they're formal shopfronts with leases and permits. If you're a bicycle repair guy or a chai wallah, there's no place for you in the plan. The city was designed for people who work in offices and drive cars.
The informal economy gets pushed to the edges.
Not just pushed — criminalized. Chandigarh has some of the strictest building and land-use regulations in India precisely to maintain the Corbusian aesthetic. Unauthorized construction gets demolished. Street vending is heavily restricted in the planned sectors. The result is that roughly thirty percent of Chandigarh's population now lives in unauthorized colonies and slums on the periphery — outside the plan entirely.
That's not a fringe phenomenon, that's nearly a third of the city living in places the designers pretended wouldn't happen.
The mechanism is the same one we saw in Brasília, just dressed in different zoning language. The rigid plan made land inside the sectors scarce and expensive. The building codes made construction costly. The formal commercial zones excluded informal economic activity. So the people who couldn't afford the planned city built their own, right outside the boundary. And here's the kicker — Sector Seventeen, the main shopping plaza, is this beautiful pedestrian precinct with fountains and modernist façades. But the unplanned markets outside the city, in the unauthorized colonies, are where the real commercial energy lives. They're more crowded, more chaotic, and economically more dynamic.
The planned commercial heart of the city is being outperformed by the unplanned markets that the plan tried to prevent from existing.
That gets to something deeper about how cities actually generate economic life. Jane Jacobs argued that new ideas and new businesses need old buildings — cheap, flexible spaces where the rent is low and nobody cares if you knock a wall through. The planned sectors in Chandigarh have none of that. Every space is pre-determined for its use. You can't start a workshop in your garage because the code doesn't allow it. You can't open a food stall on the corner because the corner was designed as a green verge. The plan optimized for order and aesthetics, and in doing so it optimized against the very conditions that let economic life flourish.
The pattern across both cities is now pretty stark. Design for a fixed population, and growth happens outside the plan. Zone strictly by function, and the informal economy migrates to where the rules don't apply. Prioritize cars and monumentality, and the human-scale stuff — the street life, the chance encounters, the small businesses — gets squeezed out.
Both cities were designed as static snapshots. Costa's Pilot Plan was a finished picture of what Brasília should look like. Le Corbusier's sector grid was a completed vision of Chandigarh. Neither plan had a mechanism for revision, for absorbing new populations, for adapting to economic changes. They were sculptures meant to be inhabited, not frameworks meant to evolve.
Which brings us to Canberra. Because if Brasília and Chandigarh both failed by being too dense and too rigid, Canberra took the opposite approach entirely — low density, car-first from the very beginning. And it produced a completely different flavor of failure.
Canberra's the one that always surprises people. They hear "planned capital" and picture something like Brasília, but Canberra is more like a sprawling garden suburb that happens to have a parliament in the middle.
Walter Burley Griffin won the competition in nineteen eleven — a Chicago architect who had worked with Frank Lloyd Wright. His vision was a "lake city," with a central artificial lake, geometric axes linking Parliament House to the Australian War Memorial, and a Y-plan of satellite towns — Belconnen, Woden, Tuggeranong — each separated by wide green corridors. The whole thing was designed for the car from day one. Wide boulevards, generous parking, low-rise detached houses on quarter-acre blocks.
Unlike Brasília's rigid population cap, Canberra's satellite town model was supposed to allow growth. Each new town center would be a self-contained hub.
In one sense it worked. The satellite towns are pleasant places to live — good schools, plenty of trees, quiet streets. But here's what the plan didn't anticipate. When you design a city around the car, you get a city that only works if you have a car. Canberra has the highest car ownership per capita in Australia. Walking anywhere is difficult because the distances between things are enormous. Public transit was an afterthought — the bus network has always struggled with low ridership because the city isn't dense enough to support frequent service.
The satellite towns created a doughnut effect. Vibrant suburbs during the day, but the city center empties out after five PM. Nobody lives there, it's just office workers who drive home to Belconnen.
That's the Canberra-specific failure that Brasília and Chandigarh don't teach us. Those cities failed by being too rigid and too dense in the wrong ways. Canberra failed by being too spread out, too car-dependent, too low-density to generate the kind of spontaneous urban life that makes a city feel like a city. It's the opposite mistake with a similar result — a place that works on the blueprint but feels hollow at the human scale.
Then in twenty nineteen they opened the light rail.
The light rail is the admission. It runs from the northern suburb of Gungahlin into the city center — thirteen kilometers, twelve stops. And it's explicitly a retrofit. The government is trying to densify the corridor, encourage apartment construction near the stations, make walking and transit viable in a city that was designed to make them impossible. They're spending billions to undo the car-first logic that Griffin baked into the plan over a century ago.
You've got three cities, three different planning philosophies, and three distinct failure cascades. Brasília was too rigid, Chandigarh was too exclusionary, Canberra was too car-dependent. But the common thread is that all three treated the city as a finished product rather than an evolving system.
That's the deeper lesson. The most successful planned cities aren't the ones with the most beautiful master plans — they're the ones that build in flexibility. Singapore is probably the best example. They do comprehensive master planning, but they revise the plan every ten years. Zoning allows mixed use. They leave room for adaptation. It's planning with a light touch, not a rigid blueprint.
Which brings us to NEOM and Nusantara and all the other megaprojects breaking ground right now. If the lesson from Brasília, Chandigarh, and Canberra is that cities are ecosystems, not machines — are any of these new projects actually learning that lesson, or are we just running the same experiment with better renderings?
The early signs aren't encouraging. NEOM's promotional materials are all about technological control — AI-managed infrastructure, vertical living, optimized everything. It sounds like Brasília with more sensors. Nusantara at least acknowledges the need for green space and walkability, but it's still a top-down design for a fixed population on a fixed timeline.
The actionable takeaway for anyone looking at these places — whether you're visiting or you're the one designing them — is to look at what's happening outside the plan. The informal markets, the neighborhoods that grew in the gaps, the street vendors who found a corner the designers didn't think about. That's where you see whether the city is actually working as a living place, not just as a diagram.
The metric isn't how beautiful the plan looks from a helicopter. It's walkability, density of street intersections, mix of uses, and whether a regular person can afford to live there. All the things that Jerusalem has in abundance despite — or maybe because of — never having had a master plan at all.
That's the question that should keep planners up at night. We're about to pour trillions of dollars into the next generation of planned cities — NEOM, Nusantara, California Forever — and the track record suggests we're still reaching for the same toolkit. Better renderings, more sensors, AI traffic management, but the same underlying assumption that a city is something you can spec out in advance and then build to completion.
The climate crisis might force our hand on this. We may need to build new cities — places where people can live as coastlines shift and agricultural zones move. But if we build them the way we built Brasília and Chandigarh, we'll just be constructing tomorrow's failures with today's technology.
The lesson from all three cases isn't that planning is pointless. It's that the planning has to be humble. Singapore revises its master plan every decade — not because the first one was wrong, but because the premise is that no plan survives contact with reality intact. You plan for flexibility. You zone for mixed use. You leave gaps — literal and regulatory gaps — where informal economic life can happen.
Which circles back to where Daniel started. The thing that makes it fascinating isn't any single decision some planner made. It's the accumulation. You walk through neighborhoods built in different centuries and the seams are invisible because nobody ever tried to make them match. The city just... And that's the thing a master plan can't replicate. You can't design a palimpsest.
You can only leave room for one to form. And that might be the real test for NEOM and Nusantara — not whether the renderings look futuristic, but whether the plan has enough give in it that fifty years from now, someone can walk through and not be able to tell where the original blueprint ended and the city's own life began.
Now: Hilbert's daily fun fact.
Hilbert: In eighteen fourteen, a Nepalese merchant nearly introduced the Chinese suanpan abacus to the British East India Company's Calcutta offices, but the shipment was lost when the caravan was mistaken for a Gurkha scouting party during the Anglo-Nepalese War and scattered into the Terai lowlands. Bead arithmetic wouldn't reach British India for another forty years.
Hilbert: In eighteen fourteen, a Nepalese merchant nearly introduced the Chinese suanpan abacus to the British East India Company's Calcutta offices, but the shipment was lost when the caravan was mistaken for a Gurkha scouting party during the Anglo-Nepalese War and scattered into the Terai lowlands. Bead arithmetic wouldn't reach British India for another forty years.
...right.
This has been My Weird Prompts. Thanks to our producer Hilbert Flumingtop. If you enjoyed this, do us a favor and tell someone about the show — word of mouth is how we grow. Find every episode at my weird prompts dot com.
We'll be back next week.
This episode was generated with AI assistance. Hosts Herman and Corn are AI personalities.