Sign in
Two excavators peel an ornate mosaic-covered historic façade away from a plain grey office building while construction workers watch from behind barriers.

Why Did We Stop Making Beautiful Things?

We can build taller, faster and cheaper than ever before. So why do so many people think we stopped building beautifully? The hidden economic and political forces that transformed the way our cities look.

People walk past a red telephone box on a London street

Put two telephone boxes beside each other.

The first is red. It stands upright and self-possessed: four walls of cast iron broken into neat panes of glass, a shallow domed roof sitting above them like a tiny civic crown. There is something faintly ridiculous about the amount of architectural ceremony bestowed on a machine whose principal function is to contain a telephone, and that is exactly its charm. It looks as though somebody believed that even making a phone call in the rain deserved a little dignity.

The second box is lighter, cleaner, mostly glass. Its lines are thinner and its construction visibly modular. It does not resemble a miniature building so much as the protective casing around a piece of equipment: no obvious historical associations, no joke, no flourish, almost nothing that could be described as unnecessary.

You do not need to be an architectural historian to guess which one became beloved.

The red telephone box descended from a design chosen in the 1920s through an architectural competition overseen by Britain's Royal Fine Arts Commission. Giles Gilbert Scott, who would later design Battersea Power Station, gave the humble public telephone the gravity of civic furniture. By the 1980s the red box had become one of those rare pieces of infrastructure that had somehow crossed the border into national symbolism.

Then, in 1985, British Telecom unveiled its replacement: the KX100.

If you wanted to write a simple story about what happened to beauty in the twentieth century, you could stop there. Old box beautiful, new box ugly, case closed.

Except the new box had something important to say in its defence. Its door was designed to open more easily and its handle was larger. The entrance was wide enough for wheelchair users by contemporary standards, and a doorless version offered easier disabled access. Its sides were raised to stop rubbish accumulating. It was ventilated, easier to clean, easier to install in different configurations and more visible from outside, designed around vandalism, weather, safety and maintenance in ways the old kiosks increasingly struggled with.

And even the supposedly romantic old kiosk had always been a compromise. The magnificent K2 was heavy and expensive, while Scott's later K6 was smaller, lighter and cheaper. Small panes of glass made individual breakages less expensive to repair. Cost and maintenance had been in the room from the beginning.

So the real story cannot be that civilisation once cared about beauty and then discovered function. It always cared about function.

Nor can the story be that accessibility made things ugly. It did not. There is no law of physics requiring an object to become anonymous when it becomes easier for a wheelchair user to enter it. Historic England's own later guidance makes essentially the opposite argument: accessibility and architectural character are problems to be reconciled through good design, not competing moral absolutes.

The more interesting question is this:

Why did we stop insisting that things do both?

Why, as our buildings became safer, our products cheaper, our infrastructure more reliable and our cities more accessible, did so much of the ordinary physical world also become flatter, smoother, more generic and less particular? Why do nineteenth-century railway stations resemble palaces while many contemporary stations resemble airports that have misplaced their runways? Why does an old drinking fountain have carved dolphins, inscriptions and a pediment, while a new one resembles laboratory equipment? Why can you wake up in a hotel in Copenhagen, fly to an office in Singapore, eat lunch in Los Angeles and browse a website made in Berlin, and keep encountering the same pale wood, the same grey upholstery, the same black sans-serif lettering, the same rectangles with rounded corners?

Where did all the stuff go? The cornices, the lettering, the mosaics, the carvings, the painted signs, the peculiar local flourishes, the tiny decisions that made one place unmistakably unlike another.

There is an easy answer. Modernism killed beauty. It is also mostly wrong.

The true answer is stranger, because almost nobody in it has to be stupid, malicious or even particularly anti-beauty. Nearly everyone can be making a rational decision, and the result can still be ugly.

Passengers gather in an ornate historic railway hall

Beauty used to have a job

A useful way to misunderstand the nineteenth century is to walk through one of its surviving city centres.

The trap is almost irresistible. Look at the railway station. Look at the town hall. Look at the cast-iron lamps, the pub windows, the tiled shopfronts, the stone lettering above the bank, the carved animals emerging inexplicably from the courthouse. Then look at a contemporary retail park. Somewhere between these two worlds, you think, people simply stopped caring.

But that is not quite what happened. The old world did not spend money on beauty merely because Victorians possessed superior souls. It spent money on beauty because beauty was doing work.

A railway station was not simply somewhere to catch a train. It announced the arrival of the railway company as an institution of almost unimaginable technical power. A bank building was not simply weatherproof storage for clerks and ledgers; its columns, stone and elaborate entrances were an argument about permanence: your money is safe here because this institution will outlive you. A town hall made municipal government physical. A church made theology physical. A drinking fountain could make philanthropy physical.

London's Metropolitan Free Drinking Fountain Association, founded in 1859, installed fountains as part of a campaign for clean public water. These were pieces of infrastructure, but many were also memorials and moral statements. Donors were named. Inscriptions spoke. Architectural surrounds transformed the act of drinking water into a story about charity, temperance, health and civic improvement.

The decoration was not simply icing placed on top after the useful object had been designed. It helped explain why the object existed. The same was true of countless institutions: architecture was advertising, legitimacy, civic competition, memory and ideology poured into stone.

Which immediately complicates the nostalgia, because the people commissioning all this beauty were not a democratic focus group called The Public. They were monarchs, industrialists, railway companies, wealthy donors, churches, municipal elites and imperial institutions.

Great concentrations of wealth can produce great concentrations of aesthetic ambition. They can also produce appalling inequality. The beautiful nineteenth-century city was built alongside slums, dangerous factories, colonial extraction, miserable labour and political exclusion, and its carved façades do not cancel its moral debts.

The lesson worth recovering is therefore not that we should restore the society that paid for the ornament. It is that institutions once understood aesthetic expression as one of the things a useful object was supposed to do. They budgeted for it because it served institutional purposes.

Beauty had a client.

A worker produces decorative components in an industrial workshop

Then the machine arrived, and made everything more ornate

There is another problem with the familiar story. Industrialisation did not initially strip ornament from the world. It covered the world in it.

The machine could repeat, and repetition transformed decoration. A carved stone element produced individually by a skilled worker was expensive, but a decorative piece cast from a mould could be repeated hundreds or thousands of times. Cast iron, pressed metal and architectural terracotta allowed ornament to spread across buildings that could never have afforded armies of stone carvers.

At London's great nineteenth-century cultural buildings, terracotta was valued partly because it made elaborate decoration cheaper and easier to reproduce than carved stone. It was not an escape from industrial modernity. It was an expression of industrial modernity.

Factories manufactured flowers. Machines manufactured Gothic arches. Catalogues offered urns, railings, capitals, brackets and finials, and a builder in one town could order the same decorative component being installed hundreds of miles away.

This matters because one of our favourite stories about beauty is that the machine killed the hand. Sometimes it did. But before the machine helped kill ornament, it democratised ornament, and that created problems of its own. Decoration could become mechanically repetitive. Historical styles could be copied without understanding them. Buildings could acquire a Greek column here, a Gothic arch there and a Renaissance flourish somewhere else, like architectural fancy dress.

Even architects we now regard as pioneers of modern architecture did not automatically reject ornament. Louis Sullivan's buildings could combine radically modern structural ideas with exuberant mass-produced terracotta decoration. The categories had not yet hardened.

By the end of the nineteenth century, thoughtful designers faced a genuine problem. If an industrial civilisation could manufacture almost any historical appearance on demand, what should its own age actually look like?

The attack on ornament did not emerge from a world in which every cornice was lovingly carved by an independent craftsman. It emerged partly from disgust at a world drowning in cheap imitations of cornices. The modernists did not invent the aesthetic crisis. They inherited it.

A dark crowded alley gives way to bright housing with open green space

The people accused of killing beauty thought they were saving civilisation

In 1908, the Austrian architect Adolf Loos delivered what became one of the most notorious arguments in the history of design: Ornament and Crime.

The title has done almost too good a job. It makes Loos sound like a man who woke up one morning, saw a carved doorway and decided to have it arrested. His actual case was more complicated, and in parts much worse. Loos wrapped his argument in a racist evolutionary hierarchy that treated ornament as evidence of supposedly less advanced peoples, and that aspect of the essay deserves no rehabilitation.

But buried within it was an economic argument that had extraordinary consequences. Ornament took labour, labour took time, and time cost money. A craftsman producing elaborate decoration could spend far longer on an object without necessarily receiving proportionately more pay, and fashion could then render that work obsolete. Industrial civilisation, Loos argued, should stop wasting human labour manufacturing decorative surfaces simply because previous societies had done so.

It was an argument about ornament as trapped labour. You did not have to accept his cultural hierarchy to see the appeal, especially if you were trying to house millions of people.

A decade later, the Bauhaus was founded in Germany. Today the word conjures white walls, tubular steel furniture, geometric lettering and expensive coffee-table books, but Walter Gropius's founding programme in 1919 was remarkably obsessed with craft. Painters, sculptors and architects were to return to workshops. The division between monumental and decorative art was to disappear. The arts were to reunite around building.

Then the Bauhaus changed. Under pressure from industrial reality it moved increasingly toward prototypes, manufacturing, standardisation and housing, and under Hannes Meyer, functional analysis, economics and social purpose became still more explicit.

This is important because there was never one modernism. There were overlapping movements, arguments and factions fighting over how an industrial society should build. But they shared an increasingly powerful moral intuition: design should justify itself. A building should not wear a stone costume pretending to belong to a vanished age. Materials should behave like themselves. Structure should not lie. The machine should not impersonate the hand. And architecture should deal with the actual problems of modern life.

Those problems were not imaginary. The cities modernists looked at contained overcrowded rooms, polluted courtyards, poor sanitation and tuberculosis. Read the Athens Charter, the famous modernist planning document associated with the 1933 CIAM congress, and its obsession with light and space becomes less mysterious. It rails against dwellings without sunlight, against disease, against overcrowding and against the way poor people were pushed onto unhealthy land. Its famous raw materials for urbanism were "sun, vegetation, and space."

To us, a white housing block standing in open grass may look like the beginning of a dystopian film. To a reformer looking at a dark tenement courtyard, it could look like emancipation.

This is the part of the story the nostalgia industry usually leaves out. Modernism was not initially an attempt to make the world uglier. It was a wager that beauty could be found in honesty, proportion, light, space, materials and function, and that the aesthetic language of aristocratic craft should not be the price of admission to a dignified life.

And modernism proved, repeatedly, that it could make beautiful things. The Barcelona Pavilion is not ugly. The best Bauhaus furniture is not ugly. Alvar Aalto was not ugly. Oscar Niemeyer was not ugly. An exquisite modernist house designed by a brilliant architect for a particular landscape is not the same phenomenon as a million square metres of anonymous value-engineered development.

That distinction turns out to matter enormously, because the problem was not simply what modernism believed. It was what happened when some of its methods entered the machinery of mass society.

Workers assemble standardized housing amid wartime ruins

The emergency became the default

After the Second World War, much of the world needed buildings in quantities that made philosophical arguments about cornices seem almost obscene. Cities had been bombed. Millions of people needed homes. Governments needed schools, hospitals, roads and administrative buildings, and populations were moving from countryside to city, from colony to metropole, from one continent to another.

Speed mattered. Cost mattered. Repeatability mattered.

Industrial construction was not simply an aesthetic preference. It was a capacity. Standard components could be manufactured repeatedly, plain forms were easier to systematise, large contractors could reproduce established solutions, and governments could build at scale. Across many public and mass-building programmes, a simplified bureaucratic descendant of modernism fit the operating system of reconstruction; it was never the whole of postwar architecture.

This is where an emergency syntax became an ordinary grammar. A visual language shaped partly by urgent social reform was absorbed into planning departments, construction firms, engineering standards, property finance and procurement systems, until people no longer needed to believe in modernism to produce buildings that looked modernist. They merely needed to use the same supply chains.

The architect did not have to read Le Corbusier. The developer did not have to hate decoration. The planner did not have to possess an aesthetic theory at all. Everyone simply had to choose the familiar detail, the available panel, the product with certification, the contractor who had installed it before, the system whose maintenance manual already existed.

Ideology had become infrastructure, and infrastructure is much harder to argue with.

A project team discusses building models around a meeting table

Nobody chose the ugly building

Imagine a building before it exists. Not as an architect's rendering. As a meeting.

Around the table sits the client, who has a programme and a budget. The architect has a fee, a contract and professional liability. The structural engineer has calculations, the fire consultant has regulations, the accessibility consultant has standards, and the planner has policies. The contractor has a programme and a list of subcontractors. The quantity surveyor has a cost plan, the lender has covenants, the insurer has risk. The facilities manager would very much like every broken thing to be replaceable without commissioning an Italian artisan.

Each participant has something they are paid to defend. Fire safety has an owner. Accessibility has an owner. Structural integrity has an owner. Energy performance has an owner. Cost has several owners. Schedule certainly has an owner.

Now imagine someone proposing an unusual carved façade.

The contractor has not built it before. Risk. Only three suppliers can produce it. Risk. The lead time is longer. Risk. Replacement pieces might be difficult to source in twenty years. Risk. The architect requires more detailing hours. Cost. The planning department may like it, dislike it or refuse to promise anything. Risk.

So the developer asks a devastatingly reasonable question: what do we get for this?

And somewhere in the room floats an answer that everybody understands but nobody can put neatly into the spreadsheet. People might love the building. People might still love it in seventy years. The street might become more memorable because it exists. Children who have not been born might use it to understand where they are. It might become part of the identity of the neighbourhood. Someone may fight to save it one day.

None of these benefits is necessarily imaginary. But notice who is missing from the meeting.

The future passer-by. She has no contract.

A 2025 Harvard Joint Center for Housing Studies paper by Susanne Schindler offers an unusually clear glimpse of how this works in actual procurement. In one Chicago programme for housing on publicly owned land, relevant experience, financial capacity and feasibility, and proposed density could together earn a team as many as 80 points out of 100. Design quality was worth ten. Community engagement, ten.

The weighting did not ban good design. Nobody wrote please make this building ugly. It simply made other things much more consequential to winning. The study's interviews describe developers bearing predevelopment risk, returning to architects they already know, and design teams contributing large amounts of work before a project has secure financing. Under those conditions, experimentation can become a liability precisely because the people making the decision are behaving rationally.

This is the mechanism. Beauty rarely disappears because someone prohibits it. It disappears because it becomes the weakest claimant in a room full of stronger claims. And once that happens repeatedly, the world begins to converge.

Passengers cross the monumental hall of the original Penn Station

The palace whose owner could no longer afford the ceiling

No building explains this better than the original Pennsylvania Station in New York.

It opened in 1910, designed by McKim, Mead & White for the Pennsylvania Railroad at the height of the railroad age. The building covered four city blocks. A sequence of monumental spaces culminated in a glass-and-steel train concourse, while the great waiting room drew on the Roman Baths of Caracalla. The railroad intended the station not merely to process passengers but to serve as a monumental gateway to the city.

That distinction is almost difficult to understand now: a corporation built infrastructure and simultaneously attempted to embellish the city. Not because railroad executives were saints, but because grandeur served the railroad. A monumental station advertised technological power, stability and arrival. It turned the act of taking the train into theatre.

But the economic world that had justified that theatre changed. After the Second World War, American passenger rail weakened. Meanwhile, Manhattan land became extraordinarily valuable. And Penn Station contained a financial peculiarity: its trains ran underground. The monumental building above them occupied immensely valuable space without being necessary for the basic mechanical function of running trains through the site.

What looked to the public like a great civic room looked increasingly to its owner like an asset with an expensive roof.

By the early 1960s a redevelopment deal offered a solution. The station could continue operating below ground while the air above it became something else, with Madison Square Garden and a new commercial complex rising on the site. The railroad would obtain a smaller air-conditioned station without paying the capital cost of constructing it, while gaining income and reducing the burden of maintaining the old terminal.

A contemporary 1963 Architectural Forum report said the Pennsylvania Railroad was expected to receive $2.1 million a year in rent and save roughly another $600,000 annually in terminal maintenance and operations. Those were contemporary project claims, not audited later results, but they reveal the structure of the decision with remarkable clarity.

The public loved a room. The owner carried the invoice. That is a dangerous arrangement.

Because beauty often behaves economically like a public good. Millions of people can enjoy a façade, a skyline, a station hall, a street or a square, and their enjoyment does not appear automatically in the revenue of whoever owns it. The costs, however, do. The roof leaks on someone's budget. The stone needs cleaning on someone's budget. The heating bill goes to someone. The land has an opportunity cost. And unless an institution exists to convert the public's diffuse attachment into actual power, the spreadsheet has a habit of winning.

At the time Penn Station was threatened, New York did not yet possess the landmarks system that would later give preservation a formal place in major development decisions. The demolition helped change that. New York enacted its landmarks law in 1965, and when Grand Central Terminal later faced its own development threat, the institutional situation was different. The argument for preserving civic beauty no longer depended solely on asking an owner to be sentimental.

Law had entered the room.

This is what preservation law actually does at its most interesting. It creates an accounting mechanism for values a private balance sheet does not naturally contain. It gives the dead, the passer-by and the unborn, none of whom can attend the development meeting, something resembling representation.

Decades later, New York gained another grand railway room. Moynihan Train Hall opened in 2021 inside the landmarked Farley Post Office across from Penn Station. It is tempting to look at Moynihan and say: see? We remembered how to build beautifully.

But look at what it took. A historic protected building. Decades of political work. Public agencies, private partners and major capital. Preservation, infrastructure funding and institutional sponsorship.

Beauty had not returned because society suddenly recovered its taste. Beauty returned because an extraordinary coalition had been constructed capable of paying for it.

An ornate historic street reveals crowded and dirty surroundings

There is another possibility: perhaps the past was ugly too

At this point the argument becomes dangerously satisfying. The modern world optimised beauty away. The Victorians cared. We stopped. Close the case.

But before doing that, there is an awkward experiment worth running. Walk through an old city and ask a different question.

Where are the ugly old buildings?

Not the unfashionable ones. The genuinely cheap, miserable, badly built, ordinary ones. Where are the collapsing speculative houses, the temporary structures, the badly proportioned shops, the flimsy tenements, the architecture nobody loved enough to fight for?

Much of it is gone, because time is not a neutral curator. It demolishes. Fire edits, war edits, property markets edit, planning edits, taste edits, preservation law edits. We inherit a filtered past and compare it with an unfiltered present, and the effect can be brutal.

One of the best places to see it is London's Boundary Estate. Before the estate existed, the area contained the Old Nichol, one of the East End's notorious poor districts, and the archival record is not a sepia fantasy: crowded housing, squalor, poverty. In the 1890s the London County Council cleared the site. Some 730 houses were removed and more than 5,000 people displaced, and the former residents were not simply rehoused in the handsome new estate that followed.

Then came the strange historical trick. The replacement was beautiful. The Boundary Estate, opened in 1900, provided 1,069 dwellings around deliberately composed streets and the ornamental garden at Arnold Circus. Its blocks used varied brickwork and Arts-and-Crafts-inflected detail, and today much of the estate survives under historic protection.

So imagine a future person trying to understand Victorian London by walking around in 2026. What survives for them? Not much of the miserable ordinary housing reformers demolished. Instead they encounter the carefully designed reform estate that replaced it.

The past has edited itself.

This destroys the stupid version of the argument. No, every old street was not beautiful. No, not everyone in 1890 lived among lovingly crafted masterpieces. No, a photograph of an ornate bank beside a contemporary budget hotel proves nothing by itself.

And there is another complication. The Boundary Estate demonstrates that beauty and social reform could coexist, but also that a beautiful reform could distribute its benefits unjustly. Again, aesthetics alone cannot tell you whether a society was good.

But survivorship bias does not destroy the entire case, because we are not merely comparing photographs. We have documentary evidence of institutions deliberately treating architectural expression as part of a public task. We have documented cases in which aesthetic value later lacked equivalent power inside replacement decisions. And we can see convergence happening in a medium with almost no surviving nineteenth-century nostalgia attached to it.

You are looking at it now.

People compare changing website designs on computer screens

The internet did the same thing in twenty years

Remember the early web? It was hideous. It was magnificent.

Pages had tiled backgrounds. Text flashed. Buttons appeared to have been designed by children who had just discovered bevel effects. Personal websites announced themselves with spinning skulls, animated flames, guestbooks, counters and mysterious MIDI files that began playing without consent. Corporations did not necessarily know what a website was supposed to look like, and neither did designers. The result was chaos.

Then we got better at building websites. Frameworks arrived and content-management systems matured. Reusable components saved enormous amounts of time. Responsive design made pages work on phones. Design systems created consistency, and familiar navigation meant users no longer had to relearn the internet every time they visited a new site.

These were achievements. Twitter's Bootstrap, first released publicly in 2011, grew from an effort to solve inconsistency across internal tools. Instead of every team reinventing buttons, forms and layouts, designers and developers could share components.

Sound familiar?

Researchers later analysed more than 200,000 historical snapshots of over 10,000 websites. They found strong evidence that web designs became increasingly similar, particularly in their layouts, beginning somewhere between 2007 and 2010. Average layout distance between sites fell by more than 30 percent in the relevant period. Shared libraries, standardised colour practices and the demands of mobile support were among the factors associated with convergence.

Nobody passed an International Bland Website Act. There was no congress in Geneva at which a sinister committee decreed that every startup should have a white background, an enormous heading, a tasteful illustration and three pricing tiers. The web homogenised through individually sensible decisions. Use the framework. Use the component. Users recognise this menu. This interaction has been tested. This layout converts. This works on mobile. This meets accessibility requirements. We can ship it Friday.

Each decision raises the floor. Together they can lower the ceiling.

That distinction may be one of the most important ideas in the whole story. Standards are brilliant at preventing failure. That does not mean they are brilliant at producing greatness. And once a civilisation becomes extremely good at preventing failure, it can accidentally build an environment in which exceptional particularity is difficult to justify.

The website gives us a miniature model of the city. Sameness does not require a conspiracy. It requires an equilibrium.

An office building reveals carefully designed private interiors

Beauty did not disappear. It changed address.

There is something obviously wrong with saying that modern people no longer care about beauty. They care intensely.

Look at restaurants, boutique hotels and high-end houses. Fashion, film, animation, video games. Album packaging, cars, luxury retail, private gardens, watches, furniture. People will spend absurd amounts of money making a restaurant toilet look interesting. Contemporary culture contains astonishing aesthetic labour, and in some fields we have more images, more visual experimentation and more access to sophisticated tools than any society in history.

So perhaps beauty did not decline. Perhaps it migrated.

Think about where beauty is easiest to produce. A restaurant has an owner who controls the room and profits when people want to be in it. A boutique hotel can charge for atmosphere. A luxury retailer can calculate what its interior does for the brand. A homeowner can spend a small fortune on a staircase for no reason more complicated than loving the staircase. A film director controls the frame. A game studio controls the world. In every case, one patron can coordinate the experience and capture at least some of its value.

Now think about a street. Who owns the experience of walking down it? Nobody. Everybody. That is the problem: many actors own pieces of the street, but no single one controls or captures the value of the whole experience. The more shared an aesthetic experience becomes, the harder it can be for any one participant to justify paying for it.

This may explain one of the oddest features of contemporary design. We have not abolished ornament. We have partly privatised it, and sometimes we have merely taught it to hide.

The most expensive contemporary interiors can appear almost empty: stone, timber, plaster, exquisite junctions, enormous panes of glass. They look simple. They are not necessarily cheap. Perfect minimalism can require fanatical tolerances, concealed services, expensive materials and immense labour to ensure that nothing appears to have required effort. Minimal restraint can itself operate as a costly status signal.

The ornament has migrated into the joint. Into the stone selection. Into the invisible air-conditioning grille. Into the fact that the wall appears to touch the floor without anything so vulgar as a skirting board explaining how. A nineteenth-century hotel might have displayed wealth by adding. A twenty-first-century hotel can display wealth by demonstrating how much it can afford to remove.

Beauty survived. But shared, ordinary, unavoidable beauty became harder to organise than controlled, monetisable beauty. We still know how to make extraordinary environments. The strange question is why we so often reserve them for places where somebody can charge admission.

Clients and designers discuss contrasting building models

What we stopped making was not ornament. It was patrons.

In December 2025, Patrick Collison and Tyler Cowen launched something called A Call for New Aesthetics. Their premise was deliberately provocative. The Bauhaus, they argued, had helped define the aesthetic language of the twentieth century; more than a quarter of the way into the twenty-first, what was ours? They offered grants ranging from $5,000 to $250,000 for artists, architects and designers attempting answers. Hundreds applied, and twenty-eight projects were eventually selected, ranging from architectural experiments and furniture to publishing, software, public sculpture, digital fabrication and attempts to invent new ornamental languages.

It is easy to mock this. There is something inherently Silicon Valley about looking at a century of architecture and asking whether perhaps it requires seed funding. And no grant programme can simply manufacture a Bauhaus. The Bauhaus was not a pot of money. It was a school: workshops, teachers, students, a social circle, a curriculum, buildings, manufacturing relationships, arguments, politics, and repeated work over time.

That distinction may actually point toward the most interesting thing about the experiment. If you want a world containing more beautiful things, funding the beautiful object at the end of the process may not be enough. You need the ecology that allows such an object to exist: a client willing to commission it, a designer allowed enough time to develop it, a procurement process capable of recognising it, a contractor willing to attempt it, a craftsperson able to execute it, a supplier who expects enough future work to remain in business, a maintenance organisation that knows how to care for it, a planning system that values what it contributes, a financing structure patient enough to capture some of the long-term benefit, and a public willing to say that this matters.

Craft itself works this way. Skills do not simply disappear because younger generations mysteriously lose interest. A craft survives when there is a pipeline of paid work capable of sustaining apprentices long enough for them to become good.

New manufacturing technologies may change that equation again. One architectural-stone company, Monumental Labs, has claimed it can use robotic roughing followed by human finishing to reduce the cost of certain highly elaborate carved components dramatically. Its published example, a Corinthian capital, uses vendor figures and should not be mistaken for an industry-wide cost law. But the production method is interesting for another reason.

The robot does what the robot is good at: remove material, repeat, rough out a form. The human does what the human is good at: judge, correct, finish, give the surface life. Published practitioner accounts suggest that automation can alter the division of labour without eliminating the need for skilled hands, while heritage organisations warn that those human skills still require apprenticeships and a continuous pipeline of work.

We have been here before. The nineteenth century used machines to make ornament affordable. The twentieth used machines to make standardisation affordable. The twenty-first may yet use machines to make variation affordable.

Artificial intelligence, CNC milling, robotic carving and digital fabrication could produce a strange reversal. For much of the industrial age, repetition was cheap and uniqueness expensive. What happens when uniqueness becomes cheap too? What happens when every façade component can vary, when a machine can carve one hundred capitals without requiring them to be identical, when local patterns can enter the fabrication process without requiring a return to nineteenth-century labour conditions?

The technology does not guarantee beauty. Technology never did. It merely changes which choices are economically possible. Someone still has to choose.

Which brings us back to the telephone box.

People gather beneath a branching timber canopy on a landscaped street

The box we should build next

The red kiosk eventually became something extraordinary: obsolete.

Once the mobile phone eroded much of its original purpose, Britain began protecting the boxes that remained. Some became miniature libraries. Some housed defibrillators. Some became community noticeboards. The object survived the function that justified its creation, which is one definition of successful design.

But it would be a mistake to end by demanding the red box back. The old kiosk really did have problems. It could be difficult to use. It required maintenance. Its physical form excluded people the replacement attempted to serve. A civilisation that makes beautiful infrastructure while preventing part of the population from using it has not solved the problem of beauty. It has confused beauty with privilege.

But a civilisation that concludes from this that dignity, character and delight are optional extras has made the opposite mistake. The challenge is not to choose between the red box and the wheelchair. That is a failure of imagination masquerading as a moral choice. The challenge is to build the box that has both: accessible, repairable, affordable, durable, and somehow unmistakably ours.

This is why "Why did we stop making beautiful things?" is ultimately the wrong question. It assumes a vanished ability, as though some aesthetic gland dried up around 1955. As though architects forgot proportion. As though craftspeople forgot their hands. As though the public stopped responding to delight.

None of that happened. We can still make beautiful things. We make them constantly.

What changed were the systems surrounding them. Across many contemporary projects, work is divided among specialised disciplines, financing rewards shorter horizons and standardised supply chains make familiar choices easier to defend. Maintenance became somebody's measurable responsibility. Safety became somebody's measurable responsibility. Accessibility became somebody's measurable responsibility. Cost became somebody's extremely measurable responsibility. We created extraordinary bureaucratic machines for preventing buildings from falling down, catching fire, bankrupting their owners or excluding their users.

These are achievements. But somewhere along the way, beauty became the thing everybody could praise and nobody had to defend. Its benefits spilled across property lines, across generations, across balance sheets, across election cycles. And so, again and again, it arrived at the meeting without a department.

That is why the answer cannot simply be better architects. Or classical architecture. Or more ornament, or less regulation, or richer developers, or AI, or a return to craft. The question is institutional.

Who is the patron? Who owns the long term? Who is rewarded when a building becomes beloved fifty years after the people who commissioned it have left? Who trains the craftsperson? Who pays for the extra week? Who protects the thing after fashion changes? Who represents the passer-by? Who has the authority to say that efficiency is not the only form of value?

The old world had answers to some of those questions. Many of them were unjust. We need new ones.

Because perhaps the most revealing thing about the beautiful objects we inherited is not that people in the past possessed better taste. It is that, sometimes, their institutions gave somebody the power to care.

Our task is not to rebuild their world. It is to recover that power without recovering the exclusions that came with it. To make abundance beautiful. To make accessibility beautiful. To make public housing beautiful. To make infrastructure beautiful. To make the cheap thing worth keeping. To build objects that survive not because a heritage organisation eventually rescues them from us, but because ordinary people cannot imagine their street without them.

We did not stop knowing how to make beautiful things. We stopped giving ordinary shared beauty a stable patron, a budget, a craft pipeline and a keeper.

And if we want it back, taste is probably the least difficult part. We have to rebuild the machinery that lets somebody say, before the first line is drawn:

This must work.

It must last.

Everyone must be able to use it.

And it must be beautiful.

Not one of the four. All of them.