Arctic Dementor

If somebody on twitter posts something, and somebody replies "Source?", then it's all "HELP HELP RANDOS IN MY MENTIONS". Twitter users are really status-conscious.

In YouTube comments, it's different. I once saw a YouTube comment that said: "I got a PhD in $TOPIC and the video is wrong." The next comment said: "I am a professor at a university and I teach $TOPIC and the video is correct and that guy is full of shit." – "Liars, both of you. Nobody who has a PhD posts YouTube comments" – "Well, I know I am a full professor, and I believe the other guy has a PhD, I just know he's wrong. One of my students pointed me here and I thought I'd weigh in".

On tumblr though, you have somebody ask "Source?", and then OP gives the source. OP usually doesn't say "I have the credential, shut up", but instead OP gives their reasoning and citations, the exact same way somebody would if their credentials were "I had to write a term paper about this and I skimmed 5 books about this topic". There is no presumption that this is about status. There is no expectation for people to shut up because one of them has a PhD. Nobody is surprised when somebody has done extensive research into some niche topic like the French penal system in the early 19th century.

I think my own expectations of evidence and how people are supposed to behave are forever tainted by the way mainstream journalists used to treat tentative results from psychology and neuroscience, and the fact that you don't actually need a degree in computer science to be a programmer, but you do need theory. Self-taught programmers often mix these up. They say "you don't need a degree", but they imply "you don't need to know about formal languages, database theory, computability, operating systems, or any of that. You can just start and do stuff."

In other fields, the difference between "knowing the theory" and "having a degree" is never hand-waved in the same way, and if it is, you're a crank. In software, cranks are the default.

Also, there’s an expectation here that everyone else here is a grad student, former grad student, or temporarily inconvenienced grad student, so someone asking for a source is more likely to be understood as one aspiring academic asking another for a source, vs a challenge to someone’s authority.

-Cory Doctorow.

People sometimes respond to my criticism of some thing or person X by accusing me of "hate" for X, and I can be confident that is usually false, the emotion I am feeling about X is not hate, because I know what hate feels like from other times when I am feeling it.

Right now, for example, I hate Doctorow. I want him punished. I want him censored. I want him humiliated. I want him hurt.

My first reason for hatred is that Doctorow appears to think the quoted nonsense is a reasonable claim.

on protein bars — undated, evergreen

Brian Maxwell mixed the first batch of PowerBars in his Berkeley kitchen in 1986, using oat bran, milk protein, corn syrup, and a press he'd built out of a cabinet hinge and some plywood — he was a Canadian marathoner who'd bonked at the 26-kilometer mark of the 1977 Commonwealth Games and decided, on the basis of that very specific physiological humiliation, that the existing endurance-nutrition market consisted of orange slices and bananas and that there ought to be something denser. The bar was at this point still a cottage product, sold by Bill Rodgers's running store in Boston and a handful of bike shops on the West Coast, and it would remain so until the early '90s when triathlon culture and the Tour de France's English-language coverage made carbohydrate timing legible to a much larger English-speaking audience that had previously thought a Snickers and a Gatorade was a perfectly adequate pre-race kit.

(Maxwell sold PowerBar to Nestlé in 2000 for somewhere around $375 million, in cash, on the eve of the bar category exploding into mass retail. He died of a heart attack four years later at age 51, which is the kind of biographical detail you cannot work into the marketing copy and which the marketing copy has accordingly never worked in.)

What actually happened in the bar category between 1986 and now is not really a story about athletes or nutrition. It's a story about dairy.

Specifically, whey.

Whey is the liquid that runs off when you make cheese — milk has casein and whey proteins in roughly an 80/20 ratio, and the cheesemaking process drops the casein out of suspension to make the curd while leaving the whey behind as a sour off-yellow byproduct that for most of the 20th century the American dairy industry had no idea what to do with. They fed some of it to pigs. They dumped a lot of it in rivers, where it caused massive oxygen-depletion fish kills because whey is essentially sugar water from the perspective of pond ecology. The EPA started writing whey-disposal regulations in the 1970s under the Clean Water Act, and suddenly every cheese producer in Wisconsin had a compliance problem that translated directly into a cost line on the balance sheet — the stuff you used to dump now had to be hauled or treated, and that meant somebody at the Wisconsin Milk Marketing Board was going to spend the next twenty years figuring out how to sell it instead.

This is the substrate. Everything else is paint.

The path from "we have to pay to get rid of this" to "we are selling this for forty dollars a pound in a plastic tub at GNC" is a thirty-year project conducted by, basically, the Wisconsin Milk Marketing Board, the National Dairy Council, a handful of food-science PhDs at land-grant universities (Wisconsin-Madison especially), and the international bodybuilding subculture, which provided both the demand and the cultural framework. The bodybuilders are key here — Joe Weider's magazine empire (Muscle & Fitness, Flex, Shape, and the IFBB itself, which Weider also basically ran) had spent the '60s and '70s building a readership of guys who would believe nearly anything about protein, who treated dietary supplements as the line item between "natural" gains and "real" gains, and who were already buying powdered milk protein blends from mail-order ads in the back of the magazines.

The dairy industry had a waste product. The bodybuilders had a credulous, captive, methodologically incurious market. The cross-coupling reaction was inevitable.

It just needed catalyst.

The catalyst was the development of cross-flow microfiltration and ion exchange techniques in the 1980s and early '90s that let processors fractionate whey into specific isolates — whey protein concentrate at 80%, whey protein isolate at 90%, hydrolyzed whey at higher purities still — each of which could be sold at successively higher price points to successively more credulous customers, each with a story attached about absorption rate or amino acid profile or post-workout window. None of the stories were exactly false. None of them mapped cleanly onto the actual hypertrophy outcomes either. They were good enough to sell the powder, and once you had the powder, the bar was just a delivery format.

Met-Rx is the connecting tissue between the supplement market and the food market. A. Scott Connelly, an anesthesiologist, developed Met-Rx in the late '80s for clinical nutritional support — it was originally aimed at hospital patients who couldn't eat solid food and needed dense calorie loads, which is the same problem space that produces every meal-replacement milkshake in every nursing home in the country. Then Connelly noticed that the bodybuilders were buying it in bulk, because the bodybuilders had figured out before the company did that a hospital-grade meal replacement was structurally identical to what they wanted, which was a way to get 600 calories and 60 grams of protein into themselves without having to eat another goddamn chicken breast.

(There is something almost poetic, in a strictly materialist sense, about a nutrition product originally designed for people too sick to eat being absorbed by a subculture trying to get too big to live in a normal-sized doorway. The product spec is the same. The customer base is unrecognizable. This happens constantly in American consumer goods and almost nobody notices.)

By the late '90s the bar wars are fully underway. PowerBar is the endurance bar. Met-Rx is the bodybuilder bar. Balance Bar (40-30-30 macros, Zone Diet alignment, founded 1993) is the diet-fad bar. Clif Bar — founded 1992 by Gary Erickson, named after his father Clifford, the founding legend is the "epiphany ride" in the Sierras where Gary realized mid-ride that he could make a better bar than what he was eating, which, fine, but Erickson had also been a baker for years and was already operating a commercial kitchen, so the kitchen-table founder myth papers over the fact that Clif Bar was a professional food operation from the first day it sold a unit — is the granola-coded prosumer bar. The category fragments into demographics and then the demographics fragment into ever finer demographics, and at every point the underlying ingredient stack is some version of: oats, corn syrup or brown rice syrup, soy protein isolate, whey concentrate, a vegetable oil, and flavorings. The differentiator is the story. The substrate is dairy byproduct and cheap carbohydrate.

Then comes the Quest era.

Quest Nutrition launches in 2010 with a bar built around a very specific food-chemistry trick — isomalto-oligosaccharide (IMO) syrup, a soluble fiber that tastes sweet, used in place of most of the conventional sugars, which lets them advertise twenty grams of protein and one gram of sugar on a bar that still tastes like a candy bar. The IMO situation is itself a structural rhyme worth pausing on, because the FDA classifies soluble fiber as fiber for labeling purposes, and Quest was using it as a sugar substitute, which meant the nutrition label read like a miracle and the actual product was a chewy mass that tasted nothing like the chicken breast and broccoli the bodybuilders had been eating for decades. Quest sold a billion bars in five years. Got bought by a holding company in 2019 for around a billion dollars. Then in 2018 the FDA revised its fiber rules and IMO mostly got disqualified from the fiber count, and the entire formulation had to be reworked, and most consumers never noticed because most consumers are not actually reading the labels they think they're reading.

A parallel current.

(The amino-spiking lawsuits. Several major protein supplement brands got hit with class actions in the mid-2010s — MusclePharm, Muscle Milk, various store brands — because the FDA-mandated protein measurement is based on nitrogen content, and you can goose the nitrogen content by adding cheap nitrogen-bearing amino acids like glycine or taurine that the body doesn't use for muscle protein synthesis the way it uses complete proteins. A bar advertising 20 grams of "protein" might be, mechanistically, 14 grams of whey and 6 grams of glycine spike, and the label is technically accurate and the consumer is technically getting screwed. The lawsuits mostly settled. The amino spiking mostly continues. The FDA does not have the testing budget to police every protein bar on the shelf and the manufacturers know this.)

Then the market pivots on aesthetics.

RXBar is the next move and probably the most interesting one structurally. Founded 2013 in Chicago by Peter Rahal and Jared Smith, working out of Rahal's parents' basement, the pitch was the inverse of every supplement-coded predecessor — radical label transparency, "3 egg whites, 6 almonds, 4 cashews, 2 dates, No B.S.," the ingredient list right on the front of the package in big sans-serif type. They were not selling protein per se, they were selling the absence of the supplement aesthetic. The customer was somebody who had been to a CrossFit gym, had been told whey was bad and seed oils were bad and ultraprocessed food was bad, and wanted a bar that did not look like it had been formulated by a chemist in an industrial park in Carlsbad. RXBar sold to Kellogg's in 2017 for $600 million, four years after launch.

The Kellogg's acquisition is the punchline.

Kellogg's is the original industrial breakfast company, the one that invented the cold cereal category at Battle Creek a hundred and twenty years ago by way of John Harvey Kellogg's Seventh-day Adventist anti-masturbation program (the original purpose of corn flakes was to suppress the libido of the patients at the sanitarium, a fact the W. K. Kellogg Company has never quite scrubbed from the historical record), and the whole arc from corn flakes to RXBar to whatever comes next is the same company selling the same demographic — the urban professional who wants to feel like they're eating well — a product that tracks the prevailing nutrition anxiety of the decade. In 1900 it was constipation and the threat of meat. In 1980 it was cholesterol. In 2010 it was sugar. In 2020 it's seed oils and ultraprocessed food. The wrapper changes. The shelf does not.

Battle Creek and Carlsbad are the same city.

The category sits in an interesting place now, because the protein-bar attack surface has expanded sideways into the Kind bar / Larabar / nut-fruit-energy-snack space and downward into the candy aisle, where Hershey and Mars have both launched protein versions of their existing properties — Snickers Hi Protein, Reese's Hi Protein, the M&M's protein bar — which represents the candy companies looking at the bar category and recognizing it as, essentially, candy with a permission structure. The permission structure being: I am eating this for performance reasons. I am eating this because I went to the gym. I am eating this because it has macros.

The macros, of course, are not really the point. The point is that an entire generation of food consumers has been trained, by the supplement-magazine subculture and its corporate descendants, to accept a candy-coded product as a meal-coded product if the back of the label has the right numbers on it.

(A side-story I keep wanting to write is the military piece — the original D-ration chocolate bar that Hershey developed in 1937 for the Army Quartermaster Corps, the formal specification of which was "must taste slightly worse than a boiled potato" so soldiers wouldn't eat them outside of emergencies, which is essentially the inverse of the modern protein bar's design constraint, which is "must taste enough like a candy bar that the customer doesn't notice they're eating dairy waste-stream powder." The D-ration was engineered against consumption. The Quest bar is engineered for it. The technology is roughly similar. The cultural use is opposite.)

What you have on the shelf at the gas station in 2026 is a category that began as a Berkeley marathoner's kitchen experiment, was rescued from cottage-industry obscurity by the bodybuilding magazine empire, was financed in its industrial phase by the dairy industry's need to monetize a regulated waste stream, was refined by food chemists exploiting a series of FDA labeling loopholes, was rebranded for the post-CrossFit clean-eating market, was acquired by the industrial breakfast giants, and is now being closed-loop reabsorbed into the candy aisle it spent forty years pretending it wasn't.

wtf does "Battle Creek and Carlsbad are the same city." mean? This is a post about power bars. Are you hallucinating, old man?

Two American cities, three thousand miles apart, founded within four years of each other in the 1880s, both named after European spa towns, both organized in their first incarnation around the proposition that mineral water plus dietary discipline plus correct living habits would heal the white American middle class of its various complaints.

Both wellness projects then went broke or sputtered, the founders aged out or died, and both cities had to figure out what to do next. What they did next, in both cases, was become company towns built around a single oddball product category that nobody had been planning for, that emerged out of the ruins of the wellness project, and that turned the city into the global production capital of a thing most Americans use without ever thinking about where it comes from.

Battle Creek, Michigan, makes most of the breakfast cereal eaten on Earth.

Carlsbad, California, makes most of the golf clubs swung on Earth.

That sounds glib but the numbers more or less hold up. Three of the four largest golf-equipment manufacturers on the planet (Callaway, TaylorMade, Cobra) are headquartered within two miles of each other in Carlsbad, with the fourth (Titleist) maintaining a substantial presence there even though its corporate HQ is in Fairhaven, Massachusetts. The city economic-development office counts 116 firms in its "sports innovation cluster." Locally the area is known as Titanium Valley, which is the kind of nickname only somebody on the chamber of commerce can produce with a straight face but which is also, as a description of the underlying industrial reality, accurate.

Battle Creek is the headquarters of Kellogg's (now WK Kellogg, recently acquired by Ferrero for $3.1 billion in October 2025 after the Kellogg's split into two companies) and Post Cereals, plus a Ralston-Purina cereal plant of long standing, plus various smaller operations. "Cereal City" has been the official municipal nickname for over a century. The Cereal Festival is held the second Saturday in June and features what's billed as the world's longest breakfast table.

The structural parallel is what's interesting.

Both cities started as the same thing. In 1882, a Rhode Island sea captain named John Frazier was drilling for water on his farm in coastal San Diego County and hit two artesian springs whose mineral content turned out to be chemically near-identical to the famous spa waters of Karlsbad, Bohemia (today Karlovy Vary, in the Czech Republic).

A group of investors led by a German immigrant named Gerhard Schutte, who had grown up near the real Karlsbad and knew how the European spa trade worked, bought 160 hectares from Frazier, built a four-story hotel and bathing facility, and started promoting the place as "Carlsbad", anglicized spelling, same product. By 1887 the Santa Fe Railroad was making stops at the Carlsbad depot to let passengers drink water from Frazier's 51-foot well tower. The pamphlets they distributed nationally compared their water favorably to the 30,000 annual visitors of the Bohemian spa and asked, in italics, "is this not as near Eden as any the world has ever seen?"

That was Carlsbad's first business model. People with money were supposed to come from somewhere else, drink the water, take the baths, get well, and tell their friends.

Meanwhile in 1866, actually somewhat earlier, but the institution that matters got established in its identifiable form in the 1870s, a group of Seventh-day Adventists in Battle Creek, Michigan, founded what they initially called the Western Health Reform Institute, which was renamed the Battle Creek Sanitarium when Dr. John Harvey Kellogg took over as director in 1876.

Kellogg was a Seventh-day Adventist physician, a vegetarian zealot, and an inventor of moral-hygienic dietary systems that drew, at the Sanitarium's peak, basically the entire American celebrity class of the late nineteenth and early twentieth century, Henry Ford, Amelia Earhart, J.C. Penney, Warren Harding, Sojourner Truth in her old age. The treatment regime involved hydrotherapy (cold baths, hot baths, water enemas, water everything), exercise, fresh air, sunshine, total abstinence from meat, alcohol, tobacco, and sex, plus a vegetarian diet that Dr. Kellogg developed in the Sanitarium kitchens by experimenting with grain preparations that the patients could digest more easily than the heavy Civil War-era meat-and-potatoes American diet they'd been raised on.

Both wellness projects sold the same product. The product was: come here, where the water is special, where the air is clean, where the people who run the place understand what's wrong with how you're living, and we will fix you.

The product was not breakfast cereal. The product was not golf equipment.

In both cases, the company town we now associate with the city is an accident, a byproduct of the wellness project that the wellness project didn't anticipate and didn't initially want.

The Carlsbad spa business cratered first. The original Carlsbad Hotel burned down in 1896 under suspicious circumstances (some thought arson, no charges filed), the Southern California real-estate bubble burst right afterward, the railroad stopped running special excursion trains, and by the early 1900s the town was a quiet stretch of agricultural land, strawberries, avocados, eventually a lot of tomatoes, with one struggling mineral-springs operation and a few hundred residents.

A second luxury hotel opened across the street in 1930, drew some Hollywood traffic for a couple of decades, and then got taken over by the California Division of Forestry and used as office space. Carlsbad was nothing in particular for most of the twentieth century. It incorporated as a city in 1952, mostly to keep from being annexed by Oceanside, with a population of about 3,000.

And then it sat there for forty years.

What happened next is that Ely Callaway, a textile executive who'd had a successful second career running a Napa Valley winery, bought a small golf-club company in Temecula in 1982, moved it to Carlsbad in the early 1990s, and proceeded to invent the modern oversized driver. The 1991 Big Bertha was the first all-stainless-steel driver with a substantially enlarged club head, and it sold spectacularly.

Callaway needed engineers. Carlsbad had cheap land, year-round weather suitable for outdoor club testing, and was an easy commute from the engineering schools in San Diego County. Other golf companies started showing up because Callaway was showing up, Cobra in the mid-90s, TaylorMade not long after, eventually most of the industry's component suppliers (Fujikura, Aldila, Mitsubishi, KBS, these are the shaft makers, the most technically demanding part of a modern golf club) within a few blocks of each other.

By 2010 there were over a hundred companies in the cluster.

The Battle Creek story has a similar shape but ran a little earlier. Dr. Kellogg's brother, Will Keith Kellogg ("W.K." to history, "WK" to his employees and the Foundation that still bears his name), worked at the Sanitarium as his older brother's assistant for years, including helping develop the various flaked-grain recipes the patients ate.

In 1894 the brothers accidentally produced the first version of what would eventually become Corn Flakes, they'd left a batch of boiled wheat out overnight, it went stale, they rolled it through their machines anyway and discovered it flaked. The Sanitarium served the flakes to patients. C.W. Post, a chronically ill failed Texas businessman who'd checked in as a patient in 1891, watched the Sanitarium's food operation closely, left in 1895, opened his own competing health spa down the street called La Vita Inn, and started selling his version of the Sanitarium's grain-based coffee substitute under the brand name Postum.

Post got rich. Within two years he was selling Postum nationally. In 1897 he introduced Grape-Nuts. He was a millionaire by 1900.

W.K. Kellogg, watching this from the older brother's office, decided in 1906 that he too was going to start a cereal company, except in his version they were going to add sugar to the corn flakes, which his brother Dr. Kellogg considered an abomination against the dietary purity of the Sanitarium's mission. The brothers fought about it for the rest of their lives. They sued each other repeatedly. W.K. founded the Battle Creek Toasted Corn Flake Company on February 19, 1906, slapped his signature on the box ("Look for the signature W. K. Kellogg") to distinguish his product from the dozens of other "Kellogg's" cereal imitators that started popping up immediately, and within three years was producing 120,000 cases of Corn Flakes per day.

The cereal boom is the part most people don't know about. In the first decade of the twentieth century, Battle Creek was home to more than eighty separate cereal companies, all founded in the immediate wake of Post's Postum success and the Kellogg brothers' flake breakthrough. Korn-Kinks, Mapl-Flakes, Grape Sugar Flakes, Cero-Fruto, Force, Tryabita, Norka Oats. The whole town went mad for it.

The newspapers ran daily accounts of new cereal-company filings. A 1902 observer noted that everyone in Battle Creek "has gone daft over food cereal business." Most of these companies failed by 1910. The survivors, Kellogg, Post, Ralston (which acquired the failed Cero-Fruto in 1927), consolidated into the modern cereal oligopoly, and Battle Creek became "Kellogg's of Battle Creek" in the same way Carlsbad would later become "the place where Callaway is."

Why these places, and not others.

In both cases the cluster persists because of network effects. You manufacture cereal in Battle Creek because the rail logistics work, the grain mills are there, the wheat and corn come in cheap from the surrounding Michigan farmland, the boxing plants are there, the marketing departments are there, the test kitchens are there, and most importantly, the workers who know how to do this know how to do it in Battle Creek. You manufacture golf clubs in Carlsbad because the engineers who know how to design a clubhead are there, the shaft suppliers are there, the testing ranges are there, the PGA professionals you need to consult on prototype performance can fly into San Diego easily, and the workers who know how to do this work know how to do it in Carlsbad. The cluster generates the cluster. The first mover (Postum, Callaway) creates the conditions under which the rest of the industry has no choice but to set up next door, because the supplier ecosystem and the labor pool are already there.

And in both cases the product is, on inspection, slightly absurd.

Breakfast cereal is dried grain that you pour milk over and eat cold, which nobody in human history did until John Harvey Kellogg decided in the 1890s that hot oatmeal was too hard to digest. The product exists because a Seventh-day Adventist physician with a vegetarian fixation needed his patients to stop eating bacon and figured out how to make wheat palatable in a form that didn't require a stove.

The fact that this becomes the default American breakfast for the next 130 years, that it gets sold to children with cartoon mascots, that it gets fortified with synthetic vitamins, that it gets advertised on Saturday morning television, that the per-capita consumption of dry breakfast cereal in the United States peaks in the early 1990s at around ten pounds per person per year, is genuinely strange when you back up and look at it. We are eating the dyspepsia treatment of a Michigan health reformer because his brother and his ex-patient figured out how to mass-produce it and put it in a box with a tiger on the front.

Now flip coasts.

A modern titanium driver, with adjustable hosel and forged carbon-composite crown and tour-tuned aerodynamics, is a piece of precision engineering that costs $600 retail and gets used by middle-aged accountants to hit a 1.62-ounce ball off a tee at country clubs in Bloomfield, New Jersey. The R&D budget at Callaway alone runs into the high tens of millions per year.

The thing is engineered to within an inch of its life, face thickness measured in tenths of a millimeter, weight distribution optimized via finite-element analysis, the whole spectrum of materials science deployed in service of allowing the median user to hit the ball maybe seven yards farther than they would with last year's driver, which was itself engineered to within an inch of its life and produced exactly the same marginal improvement. The product is, considered economically, an annual upgrade cycle for an extremely expensive piece of leisure equipment that the average buyer uses fewer than thirty times a year.

Both products are luxury goods that present as commodities.

Both industries depend on the customer not asking too closely about what they're paying for.

(The cereal aisle in your supermarket is a study in this. A box of Corn Flakes is, in raw-material terms, maybe twenty cents of corn and sugar. It retails for five dollars. The packaging, the brand, the shelf placement, the cartoon mascot, the cross-promotion with whichever Disney movie is out this summer, the entire markup is the marketing infrastructure built up over 120 years of selling people something they didn't know they wanted until the Kelloggs and the Posts taught them.)

(A golf club is, in raw-material terms, a few dollars of titanium, a few dollars of carbon-fiber composite, and a graphite shaft that might cost the manufacturer thirty bucks. The retail price is $600. The markup pays for the R&D, the tour sponsorships, the marketing, and the brand premium that lets Callaway charge $600 while a comparable no-name club from a Chinese OEM that probably came off the same Vietnamese assembly line costs $150.)

The thing that really makes the two cities the same city, though, is the relationship between the founding wellness ideology and the eventual industrial output.

In both cases, the original wellness project was making a specific moral claim: that the white American middle class was sick because of how they were living, that the cure involved discipline and the right routines and the right consumption habits, and that the city in question was the place where you could come to be taught how to live correctly. Battle Creek taught you how to eat. Carlsbad taught you how to bathe and how to take the waters. The wellness was a service, delivered by professionals, in a controlled environment, to paying customers who came from somewhere else.

The eventual industrial product is the exported, mass-produced version of that wellness claim. Corn Flakes is the Sanitarium diet in a box, sold to people who will never go to a Sanitarium but who can afford ten cents a week to eat the way the rich people at the spa were eating. The titanium driver is the equipment necessary to play a sport that originally functioned, in turn-of-the-century America, as a wellness practice for the same demographic that went to Battle Creek and Carlsbad, fresh air, exercise, the right kind of leisure, the right kind of company. Both products are what's left over after a moral project got industrialized.

You can't actually go to the Sanitarium anymore. (Dr. Kellogg died in 1943, the Sanitarium went into receivership during the Depression, the building got bought by the federal government and is now a federal office complex.)

You can't really go to the original Carlsbad spa either, the well still exists, you can still fill water jugs at a self-serve dispenser on Carlsbad Boulevard for 70 cents a gallon, there's a small bathing facility, but the actual spa hotel was demolished decades ago and what's left is mostly a roadside attraction. The wellness project is dead in both places. The industrial cluster that grew out of the wellness project, however, is alive and producing at scale and exporting the leftover idea of the wellness project to the entire country.

You eat Frosted Flakes because John Harvey Kellogg believed your colon needed flushing.

You buy a new driver every two years because Ely Callaway believed Carlsbad was a healthier place to engineer titanium than Temecula was.

The chain of causation in both cases runs from a slightly cracked nineteenth-century theory of bodily improvement, through the accident of a particular founder picking a particular town, through the network effects that consolidated an entire industry into a few square miles, to the end product on your kitchen table or in your garage. The product no longer references the theory. The theory is the soil the product grew out of, and it's still in there if you know how to look, but the company itself has long since forgotten and has moved on to figuring out what color to make the next box.

The towns are also, structurally, the same town in another way: they're both monocultures.

If Kellogg's leaves Battle Creek, and the Ferrero acquisition of WK Kellogg in 2025 has people there genuinely scared about whether Ferrero will keep the headquarters there beyond the contractually mandated period, the city loses its identity, its tax base, and somewhere between a quarter and a third of its remaining industrial employment in one shot. There is no Plan B. There never was.

The city merged with Battle Creek Township in 1983 specifically because Kellogg threatened to leave if voters didn't approve the merger, and the voters approved it because they knew what would happen otherwise. The cereal industry has been shrinking in Battle Creek for thirty years already, 1,700 layoffs in 1997-99 when Kellogg started moving manufacturing to Mexico, more cuts through the 2010s, the 2023 split of Kellogg into Kellanova (snacks, Chicago) and WK Kellogg (cereal, Battle Creek) which left only about 850 cereal workers in the city. The Ferrero deal might be the end. Or it might be a reprieve. Nobody knows.

The west coast is in better shape, for now.

If Callaway, TaylorMade, and Cobra all left Carlsbad, and there's some early evidence of erosion, sector employment was down 16% between 2018 and 2020 even though golf participation was up, the city would lose its identity and a sizable chunk of its commercial-real-estate base, though Carlsbad is a more diversified economy than Battle Creek and would survive the loss in absolute terms. But "Carlsbad without golf" would be exactly as meaningful as "Battle Creek without cereal." The thing that makes the place that place is the cluster. The cluster is the city.

Both towns are, in the end, the same kind of place. They are mid-sized American cities that staked their entire civic identity on a single oddball product category that they happened to invent the modern version of, that the rest of the world now consumes without ever thinking about where it comes from, and that grew, in both cases, out of the wreckage of a nineteenth-century wellness project that promised to fix the white middle class and didn't quite deliver but left behind, as a kind of consolation prize, an industry that's lasted four to five generations and counting.

Same as it ever was, basically.

The wellness project failed. The byproduct of the failed wellness project became the city. The city forgot what the wellness project was about. The product on the supermarket shelf and the product in the pro-shop case are the descendants of the same idea, which is that there's a right way to live and that the right way involves buying something the city is selling.

You can drink the Carlsbad mineral water at the dispenser on Carlsbad Boulevard for 70 cents a gallon. You can buy a box of Frosted Flakes for $5.99 at the Meijer in Battle Creek. The water has the same mineral content as Karlovy Vary's. The flakes contain the same corn-based carbohydrate that Dr. Kellogg fed his patients to clear their bowels. Neither product does anything for you that you couldn't accomplish with a kitchen tap and a bowl of plain oatmeal. People buy them anyway, because the brand is the wellness project compressed into a unit of consumption, and the unit of consumption is what the city has left to sell.

OK, let's try another angle.

Google is a search engine. It's a device by which you ask natural-language queries of the internet. This is extremely useful for obvious reasons, but in case they aren't: outside the social media walled gardens, whose content is basically frivolous, nothing on the internet is discoverable. The internet, as an architecture, contains no provisions for finding things. Search engines are how you turn a title of a paper into a PDF, a title of an obscure book into a list of stores that carry it (or a PDF), a half-remembered news story from ten years ago into the article you read, and more other use cases than any one person can know or enumerate. They are also, more often than not, how you navigate large and complicated websites.

It's possible to imagine a world without general-purpose search engines. In that world, every use case of Google's core product is a separate skill: the tools for turning a title of a paper into a PDF are completely different from the tools for turning the title of a book into a link to buy it, or for finding the 10-K filings of a publicly traded company, or for finding porn.

You are no doubt aware that there are many different websites for finding porn. Imagine a world where everything is like that but worse. Porn is one of the largest market segments in the world, and has one of the largest hobbyist communities in the world – so it's able to direct more resources toward usability and accessibility than almost anything else. You are no doubt also aware of the number of posts on this very website sharing usability tips for AO3, which is a porn website. Imagine a world where you need a completely separate set of usability tips for any given task.

A very common use case for search engines is working around the design problems of DNS. Top-level domains make sense as a technical protocol, but are terrible UX. Gmail is called Gmail, not gmail.com. "Dot-com businesses" solved this name/address mismatch by making their address their name, which worked when it was 1998 and widespread internet access was new, but this was always unsustainable: when everything has .com in its name, nobody sees the .com anymore. (Standard advice for businesses: own the .com of your name. And the .net, and the .org, and every other TLD worth registering, because otherwise, if you come to the attention of the scammers and spammers – and it doesn't take much – they'll register yourco.horse and use it to steal your customers' life savings, which is also terrible UX.)

AOL keywords solved this problem. AOL keywords were also a manually curated directory provided by one single ISP. This was always unsustainable. Instead, we have search engines. The tool you use to turn the title of a paper into a PDF is the same tool your grandma uses to turn Gmail into gmail.com: Google, the flagship product of a company whose flagship product is the general-purpose interface of the biggest advancement in communications technology since the printing press.

Imagine if there was a single company that owned finding information in books, and if the same company also owned TV. That's not Google. They don't own any of that. What they own is a product that you and your grandma both use.

This product makes Google $300 billion of revenue in a year (source: Google, cross-checked against quarterly findings found using Google), or twice as much as the global smart-home product market, and about as much as the most charitable estimates for the global adult entertainment market. It is structurally impossible for Google, as an institution, to care about smart-home product markets. Google would very much like to find other markets to care about, but it can only care about markets that are bigger than porn.

Cloud computing is bigger than porn. (Proof: porn needs servers, but servers don't need porn.) Google has Google Cloud Platform. GCP. It's the third of the Big Three, with about 13% of the market, it comes with an anti-recommendation in the name, it's Google's largest non-advertising segment, and most people have never heard of it. Google's second-largest non-advertising segment is consumer subscription services, like selling YouTube subscriptions to people who don't know how to set up an ad blocker. Expecting Google to care about smart-home products is like expecting a lawyer to fish a dropped penny out of a porta-potty, DeBoer revealed himself to be an out-of-touch boomer the second he bought Google Nest, and the effect of one single company continually launching consumer-grade product lines that it cannot possibly care about on public sentiment toward the tech industry in general (i.e. industry in general) is the kind of negative externality that may lead one to wonder whether or not it would be government overreach for a more credible administration to politely but firmly suggest to Alphabet, Inc., ticker symbol GOOG, that it's time to start thinking about dividends.

Above-average Classical Greek monumental building (Parthenon):

That said, it's not certain that 𐀅𐁆𐀪𐀵𐀍 da-pu₂-ri-to-jo does mean 'of the labyrinth' (it's a genitive) or what it's actually referring to if it does.

The fact that the first sign is da rather than ra (Linear B doesn't distinguish between r and l) is notable but not without precedent in Pre-Greek words: cf. δάφνη ~ λάφνη 'sweet bay', Ὀδυσσεύς ~ Ὀλυσσεύς ~ &c. 'Odysseus/Ulysses'. The second sign, pu₂, usually specifically stands for /pʰu[...]/ (as in e.g. pu₂-te-re 'planters', cf. φυτέω 'plant') and we don't have any other clear examples of it standing for /bu/, but then /b/ is very rare in Mycenaean Greek anyway (in later Greek, β usually results from earlier /gʷ/, but that sound change hadn't happened yet).

The word shows up in up to two documents, both from Knossos:

KN Gg (1) 702:

pa-si-te-o-i me-ri *209 1 da-pu₂-ri-to-jo , po-ti-ni-ja me-ri *209 1

pansi thehoihi: meli 𐃨 1 daburinthojo (?) potnijāi: meli 𐃨 1

to all the gods: honey, 1 amphora to the Lady of the Labyrinth (?): honey, 1 amphora

KN M-(-) 745:

a-ka-[ ]-jo-jo , me-ṇọ[ da-pu₂-ṛị[-to-jo ]po-ti-ni-j̣ạ ri *166+WE 2̣2̣[

This one is damaged, obviously, and harder to parse, but it's the same formula.

The Classical equivalent of da-pu₂-ri-to-jo po-ti-ni-ja would be λαβυρίνθου ποτνίᾳ, and the Potnia in question is (all but) certainly a goddess. It would make sense for the da-pu₂-ri-to to be a temple or a place, but we really can't say much more about it. -νθ- is a common morpheme in Pre-Greek place names, but even if it's present in da-pu₂-ri-to-jo (which is not a given!), it's more widespread than the Minoans ever were:

This map marks a Λαβύρινθος in Crete with a question mark based on exactly this discussion; there is no good reason for it IMO.

As for 'labyrinth' originally meaning 'palace', in post-Mycenaean Greek λαβύρινθος just means 'labyrinth, maze', both literally and metaphorically, and not even particularly the one Daedalus constructed for the Minotaur (that actually shows up in literature surprisingly late; the story as we have it is from the late Hellenistic and even Roman period, and while it is older we don't really know what it looked like earlier on). The association with Minoan palaces comes entirely from Arthur Evans, who excavated Knossos and was so impressed with its complexity that he suggested it could be Daedalus' labyrinth. Popular imagination kind of ran with it, but it should be kept in mind that that is literally the entirety of the argument.