Bangkok Doesn't Need a Smart City

Let's cut to the chase. Preserving cultural heritage sounds like an obviously good thing. The hard part is how.

Picture the first option. You walk into a historic-looking quarter built entirely from the ground up to resemble old shophouse alleys, and every unit in it is a luxury store. Now picture the second. You walk into a genuinely preserved alley where people are living at squatter density, in a building nobody has surveyed, and nobody in it knows what happens when a fire starts. My guess is that you do not want either. And yet those are broadly the two directions on offer.

Now we are talking about something real.

Three reasons the yes came so easily when Chatpong asked me to write this. First, my day job is designing policy and implementation protocols for smart cities — cities that use technology to benefit the people living in them, which is a lower bar than it sounds and one we clear less often than we claim. Second, I grew up in the shophouses of central Bangkok. My father worked in the Old Town and I spent my childhood running around his office, so this is not a case study to me. Third, I wrote a doctoral thesis on the globalisation of the shophouse in Shanghai, in the course of which I accidentally talked myself out of believing that gentrification means one thing, and that the one thing is bad.

So: I am not a preservationist. I have no attachment to old things on the grounds that they are old, and I will not argue that Bangkok should be kept as it was for anyone's benefit, least of all a visitor's. I am closer to the opposite. I think technology can make this city safer, cleaner in its air and water, less lethal on its roads, less wasteful with its energy, and more economically viable for more people than it currently is, and I want all of that faster than it is arriving.

What I think we get wrong is how fast people can absorb a new environment. Humans do not adapt to a rebuilt city as quickly as a city can be rebuilt. We need remnants — enough of the familiar left standing to orient by, so the new reads as a change to something rather than a replacement of everything. Continuity is not nostalgia. It is the condition under which change is survivable. That is a futurist's reason to keep things, not a conservationist's, and it leads somewhere different.

Why the bastards work

Chatpong calls these buildings bastards, and the affection in the word is doing real work. A bastard has no pedigree, no lineage anyone recorded, no design theory to justify it — and turns out to be more inventive than the legitimate children. When I first heard him use it I recognised the whole argument immediately, because I have watched the same pattern from the government side of the table for years.

I have seen comprehensive policies that cost a fortune fail completely, while the thing the locals rigged up for themselves works perfectly for the people using it. A city I worked with wanted a problem-reporting system. The citizens wanted it too, and between them they built something that fit what people actually needed — rather than procuring an expensive, policy-backed platform that would have arrived late and fitted nobody.

Long procurement cycles, bureaucratic process, distance from the facts on the ground, and often a plain lack of capacity. That is how both policies and projects fail.

So in my line of work I look for clues from citizens: their pain points, and the ingenious circumstantial solutions they have already built. When we design a digital system, we have to understand where the problem actually is before we can design anything meaningful. That sounds obvious. It is routinely skipped, because understanding the problem takes longer than specifying a platform, and only one of those two things can be put in a procurement document.

A bastard building is that same intelligence, in concrete. Somebody had a problem, no budget and no permission, and solved it anyway. The result is illegible to a planner and completely legible to a neighbour. This essay is about what happens if we stop treating that as a defect.

The arithmetic that eats a street

Cities became the engine of growth, so land inside cities became the asset. Not all land equally — land that already has infrastructure. A plot with a road, a sewer, a power line, a transit station within walking distance is worth many times a plot without, because someone already paid for the expensive part. In Bangkok, the land with infrastructure has an awkward feature: it is already occupied. It is occupied by the capitalist infrastructure of a previous era, which we call the shophouse — a shop on the ground floor, the shopkeeper's house stacked above it, which is exactly what the name says and one of the more honest names in architecture.

There are roughly 400,000 of them left in Bangkok. There were something like 750,000 at the peak, during the boom of the 1960s and 70s.¹ That is the number that should stop you: not how many exist, but the slope. We are a quarter of a million units into a demolition that nobody ever decided to carry out. It is simply what happens when each individual transaction makes sense on its own terms.

There is a sharper version of that slope, and it comes from a thesis almost nobody has read. In 2010 a Chulalongkorn master's student named Quin Limp took the 1907 cadastral survey of Bangkok — a 1:1,000 map made to issue land title deeds, which is why it records individual buildings — and counted every shophouse in Samphanthawong district. There were 2,430, in sixteen reconstructable design types, including a run of 63 identical units across Songsawat, Charoen Krung and Yaowarat. Then he counted what was left.²

Each transaction makes sense on its own terms. Hold a plot on a main road with a three-storey building on it, where zoning and the market will support thirty, and the difference between three and thirty is the return. Clearing the low-rise is the shortest path to it. For a long time we called this progress, and for a long time it was — the city needed density, offices, hotel rooms, tax base. Nobody was wrong.

The history is not decoration. Charoen Krung was cut in 1862 because the foreign consuls petitioned the King that they were falling ill with nowhere to ride, and masonry rows followed along it. Rama V drew the line of Song Wat himself in 1892, and ordered Yaowarat through the 1890s, and the Privy Purse built and leased rows to merchant tenants — which means the shophouse was, for a while, an instrument of royal property income. Siam Cement was founded in 1913 and the frame went from load-bearing wall to concrete skeleton somewhere in the decades after. By the 1960s more than half of everything built in Thai cities was shophouses, and by the mid-1970s about seventy per cent of Bangkok lived in one.³

Then at some point people begin to notice that something is no longer the same. The original residents have moved out, because they sold or because they were priced out. Franchises come in, because a franchise can pay rent that a family shop cannot. Prices rise. And the street ends up with a retailscape you could find in any city on earth — the same coffee chain, the same convenience store, the same phone shop, arranged in the same order. Sterile is the word people reach for. I think the more precise word is generic.

You can watch this happen along a single avenue. Sukhumvit 26 became Phrom Phong and its shopping. Sukhumvit 55 became Thonglor. Sukhumvit 63 became Ekkamai and its nightlife. In each case the shophouses went first, and in each case what arrived was more valuable per square metre and less distinguishable from anywhere else. Sukhumvit 71 — Pridi Banomyong — is one of the few that still has its stock, still occupied, still hybrid. It has a Burmese community at the southern end that turned an old theatre block into a fresh market and food stalls. It has massage parlours and karaoke bars and new coffee places in the middle. It has furniture-factory-hardware-shop-motorcycle-garage hybrids at the north. It also has a BTS station and a developer's spreadsheet with its name on it. That is why the studio picked it.

The concession I have to make first

This is not a paper against building things. I want to concede the strongest version of the other argument before I make mine, because an argument that cannot survive its own counter-example is not worth publishing.

A great many shophouses should come down. They were built cheaply and fast by settlers who needed a shop and a bed, not by anyone designing for a sixty-five-year service life. Some were built for perhaps that long and have been used harder than that. Reinforced concrete built to 1960s practice, in a hot wet climate, with decades of informal modification cut through it — some of these buildings are genuinely dangerous, and the people living inside them are the ones in danger.

Never preserve a heritage building at the expense of the humans inside it. If the structure is unsafe, you protect the people first and you argue about the architecture afterwards.

I learned that rule the expensive way. I spent more than a decade studying Shanghai, and my doctoral work was on the lilong — the lane housing that the city spent the 1990s and 2000s deciding what to do with. The preservation movement there produced real victories and at least one recurring failure, which was the building saved while the family inside it kept living without a private toilet, because the listing protected the fabric and said nothing about the person. Do not do that. A building is a means.

So: the ones that are fine

Set aside the dangerous ones. What about the shophouses in decent structural condition, on good land, coming down anyway because the arithmetic says so? There is a sentimental version of this argument and I do not want to write it.

The usual defence of these buildings is the sense of place. I believe in the sense of place. I also know it does not fly alone in a room with a developer, and I would rather win the argument than be right in a way that changes nothing. So let me start with money and carbon and come back to feeling later.

Before money or carbon, though: there is a legal fact I did not know until I went reading, and it is the strangest and strongest thing in this essay. On a great many shophouse plots, you are not allowed to rebuild what you tear down. Setback rules require a building over two storeys to stand six metres back from the centreline of a road narrower than ten metres.⁴ An old shophouse already occupies its entire plot, right up to the pavement, because it was built before that rule existed. Demolish it and the replacement has to step back — onto land that, on a twelve-metre-deep plot, is most of the site. So for a whole class of buildings, reuse is not the sentimental option and not even the economical one. It is the only lawful thing you can do with the plot, and the alternative is to hold an empty lot.

Reuse is not the sentimental option on these plots. It is the only lawful one, and most owners do not know it.

Every time we tear something down and build something new, we spend carbon. Not operational carbon — the electricity the building will use, which everyone now measures — but the carbon already spent making the thing that exists, thrown away, plus the carbon of demolishing it, plus the carbon of making its replacement. The industry calls this embodied carbon and has only recently started taking it seriously.

The number that matters most is where embodied carbon sits inside a building. Roughly 55% of it is in the substructure and superstructure — the foundations and the frame.⁵ That is the part a retrofit keeps and a demolition throws in a truck. So before any argument about character or community, a reuse project starts with more than half the carbon already banked, and a demolition starts by writing it off and then spending again.

The classic study here is The Greenest Building, from the Preservation Green Lab in 2011. It compared reuse against new construction across six building types and four American cities on a seventy-five-year horizon, and found that it takes between ten and eighty years for a new energy-efficient building to overcome, through better performance, the climate impact of its own construction. For most types and climates, twenty to thirty years.⁶ If your city has a 2050 target, a building you demolish in 2026 and replace with something efficient may not break even until after the target has passed.

That same study also found a case where reuse lost. A warehouse converted to apartments performed worse than a comparable new building, because the conversion was materially heavy — they replaced so much that they spent more than they saved. I include this because it is the boundary of my own argument and I would rather draw it myself. Reuse is not automatically greener. A retrofit that guts a building to its frame and then rebuilds everything inside it in aluminium and glass is a demolition with better manners.

What I actually cannot tell you

I went looking for the evidence on this and it is not there.

There is no embodied-carbon study of the Bangkok shophouse. None. There is good Thai work on standard low-rise houses, and a life-cycle assessment of one Thai high-rise using national emission factors, and neither is a shophouse. Every coefficient in the model on this page is imported from British or American benchmarks. Concrete is concrete and the direction of the answer is not in doubt, but Thai cement, Thai construction practice and the Thai grid are all different, and anyone who tells you they know the Bangkok number to two significant figures is guessing.

There is no institutional benchmark anywhere for whole-building deep retrofit. The GLA, RICS, RIBA and LETI all publish new-build benchmarks; none publishes the retrofit equivalent. Which means every reuse-versus-rebuild comparison in circulation, including mine, has a well-measured demolition side and an estimated reuse side. That asymmetry quietly favours knocking things down, in every assessment that does not say so out loud.

The most useful thing anyone could fund off the back of this studio is not another render. It is a life-cycle assessment of one ordinary Bangkok shophouse.

On cost the evidence openly contradicts itself. Gensler, from more than 1,300 office-to-residential assessments, puts conversion around 30% cheaper than new construction.¹⁰ CBRE, the following year, puts conversion at 250 to 650 US dollars a square foot against roughly 320 for new office — often more expensive. Both are shown on this page because the honest answer is that it depends on the building, and a consultancy that pretends otherwise is selling something.

The research is already done

That is the gap on carbon. On almost everything else, the opposite is true, and it is the more annoying problem.

There are at least fifty-seven graduate theses on Bangkok shophouses and historic districts sitting in the repositories of Chulalongkorn, Silpakorn, Thammasat, KMUTT, and half a dozen universities abroad. People have surveyed the stock, measured the units, interviewed the owners, mapped the 1907 cadastre, costed the renovations, and traced the legal obstacles in detail. Master's students have been solving this problem, one district at a time, for forty years.

None of it reaches the room where a demolition is decided. It is in Thai, behind a repository login, in a PDF with a broken link, indexed under a title nobody would search for. I read four of them properly for this essay and each one contained something I would have otherwise had to guess at — the 87% loss in Samphanthawong, the stair as a measured cause of vacancy rather than an aesthetic complaint, the setback rule that makes rebuilding illegal, the survey finding that owners will hold a building empty for years rather than let it below their price.

I have put the whole bibliography on this site with every working link I could find, because the cheapest research contribution available to this studio is not new fieldwork. It is reading what is already on the shelf and telling somebody.

Four things Shanghai taught me

A word about why I am the one saying this. Chatpong made his students bring a case study from their own city before they were allowed an opinion about Bangkok, which I think is the single best decision in the whole studio. I should hold myself to it. I was born in Bangkok and grew up in it; I then spent a decade in Cambridge, Massachusetts, taking degrees and working at MIT and Harvard. So Bangkok is not my case study, it is my city. My case study is Shanghai, where I spent years on the lilong — the lane housing — and specifically on gentrification, as an anthropologist rather than an architect.

What I found there does not sit comfortably on either side of the usual argument. Gentrification is not one process with one villain. It is several things happening at once — some of them genuinely bad, some of them straightforwardly good for people who had nothing, and some of them simply the price of a neighbourhood becoming somewhere people want to be. The romantic left-wing reading, that any change to a poor neighbourhood is a theft from its residents, did not survive my fieldwork. Neither did the developer's reading. What I watched was more specific: the government picked which structures were worthy of preservation, which quietly made every unworthy structure available for immediate bulldozing, and the residents of the worthy ones went on living exactly as badly as before, because the renovation had gone into the façade.

Affordability is a by-product of diversity, an ingredient in almost all great cities. It is not something you can protect by freezing a place.

I should say plainly that I am not a Jane Jacobs partisan, and I think her influence has done real damage to how we argue about cities. The move she legitimised — a neighbourhood declaring itself finished and defending its own texture against newcomers — turns out to be indistinguishable from ordinary property protectionism once it is in the hands of people who already own the block. Ed Glaeser is right about this. The economic case for diversity is not preserving Greenwich Village for whoever bought in before it got expensive. It is housing more people, close to work, in a city that needs their labour and owes them a place to live. Anyone who cites Jacobs at me while opposing the housing that would let a delivery rider live within an hour of his job is not making an argument about cities. He is making an argument about his own street.

So when I say keep the shophouses, understand what I am not saying. I am not saying freeze Sukhumvit 71. I am saying that these buildings are, right now, several hundred thousand units of housing and workspace sitting on serviced land next to transit, and that the fastest way to get more people living well in this city is to make them usable again — not to spend twenty years replacing them with a smaller number of more expensive units and calling the difference progress.

If reuse is the right answer, the question becomes how, and this is where the Shanghai years are worth something. Four rules, in the order I would apply them.

First, preserve in groups, never one at a time. A single saved building surrounded by towers is a specimen. The thing worth keeping is not the object, it is the connectedness — the fact that the people in it have known each other for thirty years and that this is legible in how the street works. Save a block and you keep a social fabric. Save one façade and you keep a photograph.

Second, protect the people before the building. Already said. It is the rule I would enforce hardest.

Third, bring new functions in, but learn from the old one before you do. These buildings worked for decades without air conditioning: high ceilings, cross ventilation, a light well, a deep plan that stays cool at the back. If your renovation makes the building dependent on mechanical cooling, you have taken a low-energy structure and given it a permanent bill. And there is a specific move available here that we keep missing — the house part of the shophouse is still a house. Staff could live above the shop they work in, which is what the typology was invented for, and which would take a commute off the road and a rent burden off a working household in the same gesture.

Fourth, and this is the one people find counterintuitive: do not keep the façade and fill it with luxury retail. My objection is not ideological. It is that there is not enough demand. Scarce heritage frontage attracts expensive products because expensive products need to borrow authenticity, and in China nearly every repurposed building ends up housing whatever can pay the rent, which is luxury. Then the whole district is luxury, and a district of luxury has one kind of person in it at one time of day. A shophouse that draws people in — ideally around the clock, ideally people who live nearby — is worth more to the street and, over a long enough lease, worth more to the owner.

The elephant

I have to name the thing everyone in Bangkok is already thinking about.

Xintiandi in Shanghai — Benjamin Wood's project, which I researched for years — took a lilong block, kept the shells, and made a high-end dining and retail district. It was enormously successful and it set the template for two decades of Chinese heritage-led development. Bangkok now has its own version at Woeng Nakhon Kasem, where a developer bought the old town block and rebuilt it. I want to be fair about this: it is better than a tower. The height limit near the Grand Palace meant they could not have gone tall anyway, but they could have built something sealed and inward-facing, and they did not. It is small-grained and it faces the street.

My question is not whether it should exist. My question is how many a city can have. Because the value of that model comes entirely from its scarcity — it is an experience, and an experience has to be unusual. One of them in a city is a destination. Three is a theme. Five and you have Disneyland: a fantasy, priced for people who can afford the ticket, populated by characters they were sold in advance. How many Disneylands can a city support? One. Perhaps.

So if there are 400,000 shophouses and the Xintiandi model can absorb, generously, a few blocks, then the model is not the answer to the problem. It is the answer to a much smaller problem that happens to be very profitable. The remaining 399,000 need something else, and that something else has to work at ordinary rents, for ordinary businesses, without a masterplan.

What the students found

Which brings me to the studio, and why this particular set of drawings matters more than its author probably intends.

Chatpong sent twelve students to Sukhumvit 71 and did something I would not have thought to do. He did not fly them in to be experts in three days. He made them first document a local best practice from their own city — Fuzhou, Qingdao, Chengdu, Chaoshan — and only then look at Bangkok. The effect is that the Bangkok observation arrives through a comparison rather than through a fresh pair of astonished eyes, and astonished eyes are shallow. Anyone who has watched a foreign consultant discover street food knows exactly how shallow.

What came back is a set of proposals that all share a structure, and I do not think the students coordinated it. Each one takes a system Bangkok already runs on, which currently has no address, and gives it a building. Win motorcycle taxis, which have no base, get a ramp and a place to sleep. The fresh market that a convenience store replaced gets its ground floor back, at three hundred stalls, on a structural grid that turns out to be the same four metres as the market it replaces. Street food gets a serviced wall with water and drainage and power, bookable by the slot. Nightlife gets a buffer strip that is a hairdresser at noon and a bar at midnight, so the street stops dying for twelve hours a day. Ritual gets a canopy and a calendar.

None of these are technology projects. And — this is the part I would not want a reader to miss — none of them is really a repurposing project either. Adaptive reuse is the technique, not the ambition. What each of these students is actually drawing is a piece of social and economic infrastructure: a mega-complex for micro-mobility, a hybrid marketplace, a building that trades day-shift for night-shift. They are not intervening in a building for the sake of intervening in a building. They are proposing an ecosystem that a neighbourhood can run on, and keep running on, long after the drawing is filed.

That distinction matters commercially as well as intellectually. A repurposing project is a one-off transaction with a fixed return. An ecosystem project builds the thing the neighbourhood needs in order to keep being worth investing in — which is the difference between selling a building once and owning a position in a place that keeps appreciating because people want to be there.

The smart city we already have

Bangkok does not have a smart city deficit. It has a legitimacy deficit. The systems are running. They are simply not written down, not measured, and not permitted.

Consider what the win motorcycle network actually is. It is a demand-responsive, last-mile transit system, with a distributed rank structure, community-negotiated pricing, and a labour pool that self-organises by neighbourhood. It emerged in the 1980s because the sois run a kilometre deep off the main roads and nothing else would serve them, and because women walking home late were being robbed and the neighbours with motorbikes started giving lifts. Any transit authority on earth would be proud of having designed it. Nobody designed it. It has no allocated space, no shelter, no formal recognition, and it grosses millions of baht a year while parking illegally on the corner it serves.

Street food is the same story. It is a distributed food-security system that feeds the working population of a metropolis at a price no formal restaurant sector could match, and it began on boats — vendors paddling between houses when the canals were the streets. When Bangkok filled the canals in and turned to face the road, the vendors moved onto the pavement, which is where they still are, without electricity, water, waste collection, or a legal place to put a cart when they finish at two in the morning.

A smart city, as it is usually sold, would install sensors to measure the congestion these vendors cause. I am proposing the opposite: that the technology's job is to give these systems the status the city has been withholding, and that this is both cheaper and more useful than anything in the standard catalogue.

This is where I part company with the model currently being sold across this region. The Singapore template — sensors first, platform second, citizens as data sources — works in Singapore because Singapore built the formal city first and the data layer second. Bangkok did it the other way round. The systems are already running; what is missing is not the technology but the permission. Buying the Singapore product for a Southeast Asian city is therefore not merely expensive. It is a category error: you pay to instrument a formal city you do not have, while the informal one that actually moves your workforce stays illegal, and therefore stays unmeasurable.

What that looks like when you build it

So here is what I have been doing for the last six months, which is going rogue.

I built a national flood monitoring system. Looking at the historical procurement record, a system of that specification would have gone out to tender at somewhere between ten and twenty million baht and arrived, optimistically, after the water had come and gone four times. I built it in about two weeks. It ingests around two million data points, runs longitudinal analysis of how floods actually propagate, and seven municipalities have asked what it costs. It costs nothing. It runs on a server in my house, which occasionally goes down because my mother thinks I am using too much data.

The corridor screen on this site is the same story at smaller scale: an afternoon of code against eleven thousand building footprints, rotating calipers to find each one's short side and long side, to see which shapes read as a shophouse from geometry alone. Building that took longer to write this paragraph explaining it than it took to run. The two weeks I actually spent went into checking the output against buildings I could verify by eye, because a screening tool that is fast and wrong is worse than no tool. That is the part that does not compress: the technology is cheap, and the discipline of trusting your own result is not.

So when someone asks what a bottom-up smart city would cost, my answer is that the cost is not the obstacle and never was. A system that legitimises the win network — that lets a rider register, that publishes where the ranks are, that gives a neighbourhood a way to see its own service level — is a fortnight of work and a laptop. What has been missing is not capability. It is the decision that these systems are worth building for.

Three systems I would build for the shophouse

For this particular building type, three things are missing, and none of them is hard.

The first is a register. Nobody knows where Bangkok's shophouses are. There is no layer, no count, no condition survey — a 400,000-unit asset class with no inventory. The map on this page is a first attempt: every building along Sukhumvit 71 screened by geometry for the shophouse signature, a narrow frontage and a deep plan repeating along a street. It found 2,193 candidates in the study area, 683 of them within 120 metres of the road itself. The median candidate has a 4.5-metre frontage and a 12.6-metre depth, which lands within a decimetre of what the students measured by hand. That is not proof any individual building is a shophouse. It is proof that the population can be found at city scale, by anyone, from open data, in an afternoon.

The second is tenure. Chatpong's collaborator asked me a question I have not stopped thinking about: could you see which shophouses are occupied, which are for sale, which are for rent — from both ends, so a developer can find stock and a family can find a home? That system does not exist and I cannot build it from open data, because tenure is not published anywhere. What I can do is build the frame and leave the fields visibly empty, which is what I have done here. An empty field with a known shape is a specification. It tells you exactly what survey to commission.

One commercial number is public, though, and it happens to be the decisive one. The Treasury appraises land, and those figures are open. Around Sukhumvit 71 the appraised range runs from about 47,500 baht per square metre on an inner soi to roughly 150,000 on the nearest documented main-road band — which puts somewhere between 2.7 and 8.5 million baht of ground under a single shophouse footprint. Bear in mind that appraised value is the base for transfer tax and sits below what a plot actually trades for. So the number is a floor, and it is already several times what the building on top of it is insured for. The demolition argument was never really about the building.

The third is the one that would actually unlock reuse, and it came out of the same conversation. The obstacle to adapting a shophouse is not space, it is circulation. The stair is at the front, on the ground floor, so whoever rents the shop controls access to everything above it. Rent the ground floor to a coffee chain and you have land-locked three floors. Solve vertical circulation and you can let a building in pieces: this floor to a workshop, that floor to a family, the roof to something at night.

I assumed this was an architect's observation until I read the research, and it turns out to be a measured cause. Nabila Imam's 2021 study of forty shophouse owners lists the placement of the stair among the root causes of upper-floor vacancy⁷ — tenants will not take the upper floors because the stair runs through the middle of the building and there is no separate way up. One owner's answer, recorded verbatim in her appendix, is simply: no stairs outside. Attachai Luangamornlert reached the same conclusion from the drawings six years earlier, identifying the staircase as the fixed element that governs everything a shophouse can become.

And here the law closes the circle in a way that would be funny if it were not so expensive. Ministerial Regulation No. 11 defines the removal of a reinforced-concrete stair as a demolition, not a modification.⁸ The one change that would unlock the upper floors of several hundred thousand buildings is legally classified as knocking the building down, which triggers full compliance with current code, which the building cannot meet, because it is old. So the stair stays, the upper floors stay dark, and eventually somebody decides the whole thing is urban trash.

It is a solvable geometry problem — there are a limited number of stair configurations for a four-metre bay, and generating the viable ones for a specific building is exactly the sort of thing worth pointing a machine at. Nobody wants to sit and think about stairs. But the binding constraint is not the geometry. It is a definition in a 1985 regulation, and Thailand has already shown it knows how to write the exception: a 2016 ministerial regulation let existing buildings become small hotels under relaxed standards, grandfathering setbacks and parking to the rules in force when the building was first permitted.⁹ It worked. It was extended three times. It appears to have been allowed to lapse.

Reuse for what, exactly

Everything above is an argument that these buildings should stay. It is not yet an argument for what goes in them, and a preservation case with no programme behind it is how you end up with the luxury-retail outcome by default. So let me answer it with the thing I actually do all day.

I built every system described in this essay on a laptop. Not in an office and not in a lab — carried to a place, opened, worked from. There are thousands of people doing this in Bangkok right now, and the number is climbing steeply, because the tools got cheap enough that a person with an idea and a rented model can build something that would have needed a funded team five years ago. This is a real and growing economy and it has an unusual property: it is almost entirely indifferent to floor plate, prestige and column grid, and almost entirely dependent on where it feels good to sit for nine hours.

Which turns out to be the inner city. Not the business district — the inner city, the mildly chaotic part, the part that is dense and mixed and loud and has four kinds of food within sixty seconds of the door. Sensory richness is not a lifestyle preference for this kind of work; it is an input. And the shophouse is almost comically well-suited to supplying it: a small floor plate that suits a team of three, a shop below that keeps the street alive, a house above that someone can actually live in, and a door onto a pavement where things are happening. A tower cannot manufacture that, and the several hundred thousand units of it that already exist are currently being valued at the price of the dirt beneath them.

The most valuable thing a shophouse has is the one thing no new building can be given: a street outside that is already interesting.

That is one programme. There are others, and the studio found several — a fresh market where a convenience store won, a base for the riders who have none, a vocational school for skills that currently have no classroom, day-and-night use of the same room so a street stops dying at noon. The point is not that I have picked the right one. It is that the question 'reuse for what?' has more good answers than anyone is currently pricing, and that the developer's default answer — clear it, build luxury — is the one that satisfies the fewest of them.

Why this is a business, not a charity

This studio has a sponsor, and I would rather address that directly than pretend a research book is disinterested.

Southeast Asia is where the interesting urban problems are, and Bangkok is the best laboratory in it. Not because it is the richest or the best governed — it is neither. Because it is unusually safe, unusually open, and unusually tolerant. It is a city where being gay or trans is ordinary, where every religion has its building and they are within a kilometre of each other, where tourists and residents use the same streets, and where a new kind of business can start without asking permission first. Those conditions are not decoration. They are why bottom-up systems can emerge here and cannot emerge in cities that are stricter, and they are the reason the research is worth doing in this city rather than a tidier one.

Now the numbers, because a research book that ends on a feeling is a research book nobody acts on.

Start with the asset. Four hundred thousand units of infrastructure-rich, transit-adjacent, structurally-serviceable floor area — on the order of 20 million square metres of existing structure at the studio's measured dimensions, held by owners who mostly believe it is worth only the dirt beneath it.

The land under a single shophouse footprint appraises between roughly ฿2.7 and ฿8.5 million around Sukhumvit 71, and appraisal sits below market. The building standing on that land is carried at close to nothing.

Now the three things that move that number, all of which are consultancy work and none of which anyone is currently selling. One: on a great many of these plots the setback rule means the replacement building is smaller than the one you demolished, so the demolition case is worse than the spreadsheet says and nobody has told the owner. Two: retrofit starts with 55% of the embodied carbon already banked, which is a real balance-sheet item the moment a client has a disclosure obligation — and a demolition today may not break even on carbon until after 2050. Three: the binding constraint on reuse is not cost but circulation and permitting, and both are solvable by design rather than by capital.

This is not urban trash. It is the largest under-priced asset class in the city, and it is under-priced because nobody has built the tools to value it.

The firm that builds those tools — an inventory, a condition and tenure layer, a circulation solver, a carbon position for each plot — is the firm every developer, lender and city agency in this region eventually has to hire. That is a service line, not a project fee. And the studio has just spent a semester demonstrating what the buildings become once you have it.

That is the opportunity. It is not a preservation argument dressed up in commercial language. It is a commercial argument that happens to preserve things.

The short version

If you are the one deciding what happens to one of these buildings, four things are true regardless of which building it is.

First: the land is worth more than the building, and that gap is no longer a feeling. It is an appraisal, a setback calculation, a carbon position — numbers a spreadsheet can hold.

Second: on a great many plots, reuse is not the sentimental choice. It is the only lawful one, because a demolished building cannot legally be rebuilt to the footprint it had.

Third: a retrofit starts more than half paid for, in carbon a demolition throws away — and a demolition may not break even on that debt inside a normal investment horizon.

Fourth: the thing actually blocking reuse is not money. It is a stair, and a 1985 definition that calls moving it a demolition. Regulation made that problem, and regulation can unmake it.

Everything else in this book is what becomes possible once those four things are taken seriously — a market instead of a mausoleum, several hundred thousand buildings doing something other than waiting to be worth more empty.

What I still walk to

I go back to the old town most weeks — temples, food stalls, cafés, bookshops, the same loop, which after four decades is less research than habit.

It has changed. It is quieter than the street I grew up on. There is less food on the pavement and fewer people who have been there forever, partly because of regulation and partly because the people themselves changed. Gentrified is the available word and it is roughly accurate. But it is still there, and it is still working, and I would argue — without being able to prove it, which I should admit — that the old town of Bangkok is the most alive old town in the world. It has never been given World Heritage status. I have come to think that is part of why. Nobody ever froze it. It was never worth enough to anyone to stop it from continuing to be a place where people live and sell things and argue about the price of fish.

The shophouse is the same proposition at the scale of one building. It was never precious. It was built cheap and fast for people who needed a shop and a bed, it was modified by everyone who ever lived in it, and the modifications are the evidence that it worked. Chatpong calls these buildings bastards, and the affection in that is doing real work: a bastard has no pedigree, no lineage anyone recorded, and turns out to be more inventive than the legitimate children.

I do not know how many we will keep. I know the number is falling and that nobody has decided it should. And I have noticed that every time I explain the win motorcycle system to someone from a city that does not have one, they are impressed — genuinely, not politely — and then they ask who runs it, and I have to say that nobody does, and that officially it is not there at all.

That is the gap I would like us to close. Not by building something clever on top of the city. By writing down what the city already built, and admitting that it counts.