Big Bad Things — Learning to Live Next to the Future

We say no first.

It is one of the most reliable things about us. Set almost anything new in front of a person — a dish, a stranger, a device, a building going up on the corner — and the first response is rarely delight. It is suspicion. We meet the unfamiliar with our guard up, and we call our wariness good judgment.

There is a name for this, and a reason. Psychologists call it neophobia — a mild, built-in fear of the new — and it is not a flaw so much as an inheritance. For most of human history, the unknown thing was as likely to hurt you as help you, so a species that hesitated outlived a species that didn’t. We sit, all of us, somewhere on a spectrum between fear of the new and attraction to it, and we lean, by default, toward caution.

The counterweight is familiarity. In 1968, the psychologist Robert Zajonc demonstrated what he called the mere-exposure effect, which posited that repeated, harmless contact with a thing is enough, all by itself, to move it from strange to safe, from suspect to liked.[1] First impressions are almost always wrong, and almost always sour. The interesting question — the one this piece is about — is what carries us across the gap from that rough first meeting to something like welcome.

Usually it is time. But for the big things, the consequential things, the ones we cannot simply wait out, it is something more deliberate. It is design. Design is exposure engineered — the craft of shortening the distance from no to yes.

Nothing tests this right now like the data center.

The Most Important Building Nobody Wants

The numbers are stark. In a Gallup survey this year, seven in ten Americans said they would oppose an AI data center being built in their local area — nearly half of them strongly.[2] A newer poll put the figure at three-quarters.[3] This is the physical body of artificial intelligence, the most consequential building type of the century, the structure the entire digital economy is now racing to pour concrete for — and, by the numbers, it is the least welcome thing you could propose to build near an American home.

It is not only an American reflex. The fight has gone global — but the world has split into two camps, and the split is telling.

Across much of the wealthy, resource-constrained democratic world, the answer has been to hit the brakes. The Netherlands imposed a national moratorium on new hyperscale campuses and now limits the largest to just two designated sites. When Meta tried to build a mega-facility at Zeewolde, an anti-data-center party swept the local council, and the Dutch Senate blocked the land sale until the company walked away.[4] Singapore froze all new data-center construction for three years over energy, water, and land. It reopened only under some of the strictest conditions on earth — near-perfect efficiency, fully renewable power, liquid cooling, mandatory heat reuse — approving barely a hundred megawatts out of more than thirty bids.[5] Ireland’s grid limits pushed Amazon and Microsoft to scrap planned Dublin builds.[6] Australia is drafting rules that would require new centers to bring in their own renewable energy supply and demonstrate a public benefit, amid mounting objections to their water, noise, and power draw.[7] Even Japan now faces five-to-ten-year waits just to connect to the Tokyo grid, and rising local opposition in its crowded cities.[8]

And then there is the other camp. In India, the same buildings are not resisted but courted. The government’s draft 2025 policy dangles twenty-year tax holidays, full electricity-duty exemptions, single-window approvals, and dedicated “Data Centre Economic Zones.” Investment commitments have already passed a hundred and twenty-six billion dollars and are climbing toward a hundred and eighty, with capacity projected to grow some sevenfold by 2031.[9] Parts of the Gulf and state-directed China are racing in the same open-armed direction.

The line between the two camps is worth noticing, because it is the same line this essay is about. The resistance clusters where land is tight, water is counted, and environmental consciousness runs highest — the affluent democracies that most resemble the United States. The welcome is loudest where the growth imperative still outweighs the ledger of costs, the way it once did everywhere. One camp sees a threat to be contained while the other, a prize to be won.

It would be easy to file all of this under neophobia and move on. It’s a new thing whose unfamiliarity causes a predictable flinch, but give it time, and it all will become familiar. But that explanation falls apart the moment you look at an old photograph.

We Didn’t Always Say No

Here is Los Angeles in the 1920s. Not a refinery on the edge of town — a neighborhood. Wooden oil derricks stand in the front yards, hundreds of them, so thick the Los Angeles Times called them “trees in a forest.” On Signal Hill, developers had subdivided the land for houses, but when a Shell crew hit a gusher in 1921, the would-be homeowners simply built oil wells on their tiny lots instead.[10] Families lived in the gaps between the rigs.

At Huntington Beach, the derricks marched right down to the shoreline, and people spread their towels and sunbathed beneath them, a palisade of oil machinery at their backs and the Pacific at their feet.

Now look at Cleveland. Along the Flats, families lived in the shadow of steel mills so enormous and so ceaseless that the neighborhoods beside them breathed furnace smoke around the clock. The Cuyahoga River that ran past the mills was so thick with oil and industrial waste that it caught fire — not once, as legend has it, but at least a dozen times across the century. For most of that century, nobody thought it was worth a headline.[11]

Look long enough at those pictures and the easy story inverts. These were not timid people flinching at the unfamiliar. They lived, willingly, inside the loudest, dirtiest, most frankly monstrous machinery their era could produce. If earlier generations were braver about big bad things than we are, the data center’s problem isn’t that people fear the new. It’s that people have changed their minds about what they’ll tolerate.

But “braver” is the wrong word, and getting it right is the whole point. They weren’t fearless. They were making a different bargain, on different information.

The bargain was jobs. The derrick in the yard and the mill down the street paid the mortgage. The working-class neighborhoods that hosted them were their fiercest defenders at first, because the machine was the paycheck. And the smoke that came with it was not read the way we read it now. For the better part of a hundred years, industrial smoke was a sign of prosperity — plumes over a skyline meant work, wages, and a town that was going somewhere. Clean air read as unemployment; the smell of sulfur was, quite literally, the smell of money.[12]

The information changed slowly, and then all at once. In 1948, a temperature inversion trapped industrial smog over Donora, Pennsylvania, for five days. Consequently, it killed twenty people, sickened seven thousand, and became the first crack in the idea that smoke was harmless prosperity.[13] Rachel Carson’s Silent Spring followed in 1962, helping spark a public consciousness around environmentalism. The Cuyahoga burned again in 1969, and this time, somehow, the country finally looked. The Environmental Protection Agency and the Clean Air and Clean Water Acts arrived around 1970.[14] Over a single generation, we grew a new sense — call it environmental consciousness — an acquired ability to see the hidden costs that earlier eyes slid right past. We learned to price the water, the air, and the long tail of harm.

That new sense is exactly what the data center now walks into. This is the thing the industry has not understood: our “no” is no longer just the old animal flinch at the unfamiliar. It is a more informed refusal — a public that has learned to see what a big building quietly takes at real cost. We are not the people on the Signal Hill lots. We can see the water. We can see the grid. We have become smarter, and smarter publics are harder to sell.

Everything Taken, Nothing Given

That informed refusal has three specific objects, and the data center has designed for none of them.

The first is energy no one can read. A hyperscale facility draws power at the scale of a city, but that scale is invisible and abstract until it surfaces, monthly and metered, on a utility bill. Grid strain is the one harm of planetary computation that arrives pre-attributed to every household, and it reduces to a single sentence: the data center raised my bill. No one designed the energy story to be legible, and no one designed its benefit to be shared. All the public reliably receives is the cost.

The second is water no one can spare. An average data center can use up to five million gallons a day to cool itself.[15] This resonates in an era when we have finally learned to count and associate value to water. Here is the curiosity buried in it: the fix already exists. Closed-loop cooling recycles its water and cuts consumption by fifty to seventy percent. Rainwater harvesting is already running at Microsoft sites in the Netherlands and Ireland.[16] Knowing this opens a genuinely provocative thought — if a data center were designed to catch and recycle its own water, its natural home would be a water-rich climate, the wet places, not the deserts we keep building them in. Designed honestly for water, the building would redraw its own map. Water isn’t the only thing that decides where these go, of course — the giant training campuses that teach the models can sit almost anywhere, which is exactly why they should sit where the water is, while the inference centers that answer your queries have to stay near the cities they serve.[17] Which only sharpens the point. The buildings we could put anywhere, we are putting in the driest places we could have chosen, and the ones we must put beside people are precisely the ones that need to behave like good neighbors.

The third is jobs that never arrive — the deepest break with the derrick and the mill, because it is the missing bargain. A large facility runs on a skeleton crew of twenty to thirty permanent staff per hundred megawatts. In Virginia, the densest data-center ground on earth, the industry generates roughly one permanent job for every thirteen million dollars invested, compared with about a hundred and thirty-seven thousand dollars per job on average across every other industry and corresponding facility elsewhere.[18] The building is almost entirely equipment. There is nearly no one inside for the community to become. It is the monster next door that doesn’t even employ the town.

Add it up, and the refusal stops looking like fear at all. The data center consumes power, water, and land at civic scale and returns, so far, almost nothing legible, beautiful, or shared. Of course the answer is no when asked about one being conceived in your own backyard.

Faced with the backlash, the industry has produced two answers which create two opposite ways of seeing the opportunity. One is to build bigger — the arms race for the largest facility ever poured and gigawatt campuses drawing the power of a mid-sized city. The other, championed by NVIDIA, is to put a data center everywhere. Small AI compute nodes bolted onto new homes, a server farm in every garage.[19] But bigger only makes the thing more monstrous, and everywhere only makes it inescapable. Both change the box’s size or number. Neither option changes its manners.

We have engineered the data center to the absolute limit and designed it almost not at all — which means we’ve given it every reason to fail its first impression and no help whatsoever in adopting it as something we wish to live with.

The good news is that we know how to help. We have done it dozens of times.

A Short History of Learning to Live With Things

The reassuring truth, underneath the neophobia, is that almost everything you now cannot live without was refused at first but evolved beyond that refusal because someone designed its welcome. The pattern is always the same three beats: radically new and resisted → substantial design work → familiar, then necessary. The design almost always does one of two things. Either it makes a frightening machine feel like a room you already trust, or it makes an invisible force visible and safe.

Consider the elevator. A box on a rope, and the animal terror of the rope giving way. That terror was earned. In the elevator’s earlier days, before any safety standards existed, a single metropolitan area like Chicago could average forty elevator deaths a year.[20] To overcome that distrust of an unfamiliar thing, Elisha Otis did something brilliant in 1854. He performed the solution. He rode an open platform up at the Crystal Palace and ordered the cable cut with an axe so that the crowd could watch the brake catch.

The second correction to this beastly new thing is the quiet one. The Victorian passenger elevator was dressed as a parlor, with ornate wood paneling, mirrors, plush upholstered benches, and a uniformed attendant navigating you along the ride. The moving box was disguised as a refined drawing room, so you could stand calmly inside it while it did something no room had ever done. We still run the same trick years after the elevator’s invention, but we’ve swapped the velvet for brushed steel and piped in soft music to smooth the nerves for the now-familiar journey. Fear has been engineered so far out of the object that American elevators now make eighteen billion trips a year, with roughly one injury per million rides — among the safest journeys a person can take.[21]

The elevator has a twin in the sky. The airplane’s whole history is this same instinct, maturing. When flight was new, it was also genuinely lethal. The early US airmail service killed so many of its pilots, roughly one in seven, that the aviators themselves called it “the suicide club.”[22] To counter that reality and find some comfort in the experience, early cabins in the 1920s and ’30s were furnished like flying living rooms, with wicker armchairs, table lamps, curtains, and wood trim. This strategy for making the unfamiliar feel familiar was borrowed straight from railway saloons and ocean liners. Some cabins were fitted with berths and privacy curtains, so you could put on pajamas and sleep as though you’d never left home.

By the 1970s, reassurance had become seduction aboard the Boeing 747. A curved staircase led up to a cocktail lounge in the sky, with club chairs and a piano bar. The cabin was costumed no longer as a room you would merely tolerate, but as one you would never want to leave. Today the move has gone somewhere stranger. The aesthetic is no longer a costume on the machine. It is science.

On Qantas’s ultra-long-haul Project Sunrise flights, the cabin lighting is tuned to your biology. Tapping into your diurnal psyche, twelve scenes developed with the University of Sydney’s Charles Perkins Centre simulate sunrises and sunsets across the ceiling, nudging your circadian clock to blunt the jet lag of a nonstop run from Sydney to London.[23] Over the century, and with better engineering, flight itself went from that suicide club to roughly one fatality per ten billion passenger-miles, by some measures tens of thousands of times safer than it once was.[24] And it did so even as the design worked ever harder to make the cabin feel like anything but a machine. The parlor became a lounge, which in time became a piece of applied chronobiology. What began as decoration wrapped around a fear has become design wrapped around the body itself — aesthetics fused with science at last.

The most frightening machines of their age, the elevator and the airplane, became bearable through an identical initial design move: wrap the unfamiliar in the clothing of a room you already love. The automobile got the same treatment from the other direction. Early cars arrived panicking horses and drawing laws that made a man walk sixty yards ahead of them waving a red flag. That concern drove the first pivot in the car’s design. The solution was to disguise the machine as the thing it replaced: the “horseless carriage,” a buggy with the horse subtracted. One inventor, Uriah Smith, went further. In 1899, he patented a car with a wooden horse’s head bolted to the front so that real horses wouldn’t bolt.[25]

Today the automobile has its own language, and the carriage it once imitated survives only in the word we still use to name it. It has grown far safer, too. Its death rate per mile is down roughly 93 percent from where it stood in 1923.[26]

We end where the elevator pointed all along: up. The skyscraper and the elevator are a matched pair. One made the tall building usable, and the tall building made the elevator necessary. Two big bad things that each solved the other’s problem. But the tower had a fear of its own, and it answered in a different language. People were certain it would topple. So the early skyscrapers reassured through sheer visible heft. Chicago’s Monadnock Building sits on masonry walls six feet thick at the base, a wall that is as much reassurance as structure.[27] That confidence didn’t arrive overnight. It took fifty years of steady innovation, and tower after tower rising and standing, before a wary public stopped bracing for collapse and the skyscraper settled into the familiar. Only then could the walls come down entirely. We now wrap our tallest buildings in nothing but glass — the fear so fully gone that we make the structure disappear.

These have all been big things — towers, aircraft, and the machines we ride. But the same pattern scales down as easily as it scales up, right into the house, where we are least forgiving of anything new. The smallest, most domestic objects fought the very same fight, through the very same three beats.

At first, the refrigerator was genuinely lethal. Early electric models ran on toxic, flammable coolants like ammonia, sulfur dioxide, and methyl chloride, and leaks poisoned families in their sleep. In 1929, a methyl-chloride leak from a hospital’s refrigeration system in Cleveland ignited and killed more than a hundred people. Frightened homeowners dragged their refrigerators out into the backyard. The pivot was chemical and theatrical. A new nonflammable coolant emerged, and chemist Thomas Midgley, in 1930, inhaled it and used his breath to blow out a candle to prove it was safe.[28] Better styling quieted the noisy “monitor-top.” Today it hums, unremarked, at the center of every kitchen, opened a dozen times a day without a thought.

The microwave landed in the 1970s straight into Cold War radiation panic. It was a science-sounding box that cooked with invisible rays. This is the one case here where the fear ran well ahead of the facts. Unlike the exploding refrigerator, the danger was mostly imagined. So reassurance came in two forms. First was a visible scaffolding of trust: federal radiation standards in 1970, FDA certification in 1971, and years of magazine articles patiently explaining how the thing actually worked. Then came the styling. The earliest units looked like laboratory equipment rendered in all chrome, clunky dials, and heft. Through the 1980s, the industry wrapped them in faux wood grain to match the kitchen cabinets, and by the end of the decade the box began to vanish altogether, with the option to neatly tuck the device over the range or build it flush into the millwork.[29] An invisible danger was first made legible, then made to disappear. Today the fear is so thoroughly gone for some that it is hard to believe it was ever there.

Our story compresses most vividly in the thing in your pocket, because the phone had to be re-humanized at every single leap. The first telephones frightened people with the strangeness of a disembodied voice, and with rumors that they damaged your hearing. Design answered on two fronts. The object evolved from an awkward box into the ergonomic handset, the comforting cradle, and the twirled, knotted cord: for the better part of a century, the most familiar object in the house. The behavior was engineered too. The earliest phone books printed etiquette pages instructing people how to conduct themselves on the line, right down to the greeting. Bell wanted us to answer with “ahoy.” Edison pushed “hello.” Hello won, was printed into the directories and carried by the operators, and we have said it ever since.[30]

Making that familiar phone mobile demanded fresh design thinking all over again. The object had to learn to be held on a public street and carried in a pocket. It took years of reinvention to get there, through the brick, the flip, and the slab-with-a-keypad, each shape bringing the device a little closer to something that felt ordinary in the hand. Then, in 2007, the iPhone arrived and threw out the whole vocabulary. It was mocked on arrival for having no keyboard and given “no chance.” It answered with a touchscreen that simply worked, dressed in leather and felt, and the click of a shutter that wasn’t there. That costume it quietly shed, in flat design, once we’d internalized the trick.[31] Today it is the featureless glass slab we cannot put down. And yet the old corded phone held on longer than anyone remembers: mobiles did not become the primary phone in most American homes until 2016.[32]

The pattern runs back centuries, proving none of this is new about us. The table fork was condemned in the eleventh century as devilish vanity. When a Byzantine princess dared to eat with one, a churchman blamed her early death on the sin of it. Yet through slow refinement and the spread of dining culture, the fork became the most unremarkable object on any table on earth.[33] The umbrella took even longer, and it took sheer persistence. When Jonas Hanway carried one through the streets of 1750s London, where a man with an umbrella was thought soft, he was jeered and pelted with rubbish. There was no clever redesign. Only one stubborn man’s public exposure held it up for thirty years, until the thing turned ordinary.[34] Everything ordinary was once objectionable.

One distinction decides which of these corrections last. Disguise fades. Dignity endures. The purely cosmetic fixes buy a first impression and age badly, because everyone eventually sees the costume: the car wearing a wooden horse’s head, the microwave hiding under fake wood grain, the iPhone dressed in stitched leather it soon threw away, the fake 5G tree that fools no one.[35] The corrections that hold did something harder. Otis demonstrated his own brake. The skyscraper carried its reassurance in six honest feet of stone. Midgley drank his own coolant. None of them hid the new thing. They solved it in the open, and let people watch.

And here is the part that matters most for the data center. Familiarity is not the goal of this work. It is the reward, and it comes slowly, sometimes across a century. What design needs to do in this moment is braver. It needs to let people stand in front of something genuinely unprecedented and not turn away. The brake did not make the elevator ordinary. It made an extraordinary thing bearable long enough for time to do the rest.

That is the real assignment now. Not to disguise the new as the old. Not even, yet, to make it familiar. It is to do the substantial design work that lets us meet the unprecedented honestly, and to trust the years to turn it into a landmark.

What the History Asks of the Data Center

What does that honest work look like for a data center? Start with what it is not. The industry has already reached for the costume — the timber cladding, the visitor gallery, the landscaped berm, the shaggy plant-draped building doing its biophilic impression of a hillside through borrowed jargon. That is the fake 5G tree at the scale of a city block, putting makeup on a problem the design never touches. It will age the way every disguise ages.

The honest move is the one the microwave made, only bigger. An invisible danger made legible instead of hidden. At kitchen scale that meant standards and patient explanation. At civic scale it means letting the building’s enormous metabolism — its heat and water and load — become something a neighbor can see, and better yet, use. Pipe the waste heat into the homes and pools around it. Turn the water from a hidden intake into a visible civic asset, and design the cooling to catch and recycle what it can, which, followed honestly, would push these buildings toward the wet places and away from the deserts we keep choosing. Give the community a real stake, since the jobs won’t. And do it at every size, because the node on the house and the campus on the ridge both need manners.

None of that is disguise. It is the opposite. It is a building confident enough to show you what it is.

A Type With Almost No Precedent

Here is the part the fear hides, and it may be the most hopeful thing in the whole story. The data center is one of the very few genuinely new building types of our lifetime. Almost everything else we design is an inheritance. The house, the office, the school, and the stadium all arrive carrying centuries of accumulated form, and the work is mostly correction through the questioning of habits the type stopped examining generations ago. The data center carries almost none of that weight. It has no settled face. It has no civic role. It has no worked-out relationship to the street it sits on or the rural setting it commands.

We keep treating that blankness as a problem to cover up, when it is the rarest gift the profession is ever handed. The chance to author a building type almost from scratch. It looks like a windowless shed not because it must, but because no one has yet been curious enough to ask what else it could be.

The First No Is Not the Last Word

We opened on a human reflex, and we will end on a discipline. The rough first impression is a stage, not a verdict. Everything we now depend on once stood on the wrong side of our suspicion. The elevator. The airplane. The telephone. The fork beside your plate tonight. Each of them crossed over because someone was curious enough to design the crossing.

The data center meets a public that has grown a new and hard-won sense, and that raises the price of its welcome. We are not the family on the Signal Hill lot, content to raise children in a forest of derricks because the derrick paid. We have learned to see the water and the power and the carbon, and we will not be talked out of what we can see. That is not an obstacle to design around. It is the clearest brief the profession has been handed in a generation. Not how big. Not how many. But, how welcome.

That brief will not be answered by an engineer making the box bigger, or a marketer making it cuter. It will be answered by curiosity — the same posture this studio brings to every typology it meets. The willingness to stop at the thing everyone else hurries past and ask what it is, what it takes, what it could give, and what it might yet become.

Fear designs a bunker. Greed designs a bigger bunker. Only curiosity designs a building worth living beside.

Every big bad thing was, at first, exactly that — and every one was eventually taught its manners. The data center is still waiting — not for a better disguise, but for someone curious enough to actually design it.