Forgotten Pioneers

On the sixth of January, in Buffalo, New York, a chemist named Edith Flanigen died at ninety-six. She held one hundred and nine United States patents. The molecular sieves she invented sit inside the refinery that made the gasoline in your tank, inside the air conditioning in your dashboard, inside the oxygen concentrator at your grandmother's bedside, and inside the filtration used to strip radioactive contamination from the water at Fukushima Daiichi.

Eleven days later, in Fredericksburg, Virginia, a mathematician named Gladys West died at ninety-five. For forty-two years she worked at a naval weapons laboratory, programming a room-sized computer to describe, with steadily improving precision, the exact irregular shape of the Earth. That description is the reason the blue dot on your phone knows which lane you are in.

Neither death was much noticed by the automotive press. The reason to begin here is not sentiment. It is that both women spent their working lives on a particular kind of problem — the patient, unglamorous, probability-driven work of preventing a bad outcome that has not happened yet — and this series has spent two parts arriving at a question their careers help answer.

The question Part Two left open

The second part of this series opened with a puzzle. We are a species that engineered safety into hospitals, into buildings, into ships and into aviation long before we were willing to engineer it into the car. Our most primal instinct is to stay alive and out of harm's way. And yet for roughly the first sixty years of the automobile, the industry's settled position was that the crash could not be beaten and was not its problem.

That part offered one answer: institutional inertia hardened into doctrine, and doctrine is expensive to dislodge. It also told the story of the bombers — the returning aircraft riddled with holes, the instinct to armor the places that had been hit, and the statistician who understood that the armor belonged where there were no holes at all, because the aircraft hit there were the ones that never came back. Survivorship bias. The most important evidence is the evidence you cannot see.

That story carries a second lesson, and this part is about that one. Somebody had to be capable of reasoning about the aircraft that were not in front of him. That is a specific cognitive act: assigning real weight to an outcome nobody has observed, and paying a cost today against a probability you cannot point to. It is exactly the act the automobile industry could not perform for sixty years.

And it turns out we know a good deal about who performs it, and who does not.

That sentence is a tightrope walked in the dark, blindfolded, during a hurricane. So let us be exact about what is being claimed, because the precise version is both more defensible and considerably more interesting than the version that will be assumed.

It is not a claim that women are better engineers. It is not a claim about courage, or care, or moral character. It is not a claim about biology — the research it rests on measures perception and belief, not chromosomes, and finds differences that vary by domain in ways no innate account explains. And it is not a claim that men cannot do safety work: the second part of this series is largely a monument to men who did it superbly, several of them at ruinous personal cost.

The claim is narrower, and it has three parts. Safety engineering demands an unusual and specific shape of judgment. There is a large, replicated body of evidence about how that shape of judgment is distributed. And there is a documentary record — the case file this article is mostly made of — showing what actually gets built when people who possess it are allowed anywhere near the drawings.

I. The evidence

What the research actually found

Start with the outcome, which is not in dispute. Men are more often the victims of accidents than women, across essentially every category anyone has counted. For every hundred thousand American drivers, men are roughly three times as likely to be involved in a fatal crash. Male pedestrians in the United Kingdom are struck around eighty percent more often than female pedestrians. Men drown more, and die of accidental poisoning more, throughout the Western world. American women report wearing seat belts substantially more often than men. Men run yellow lights more.

Some of that gap is exposure — men have historically driven more miles. But not all of it, and not close to all of it. The residue is behavior, and behavior has been studied for fifty years. A meta-analysis by Byrnes, Miller and Schafer in 1999 assembled more than a hundred and fifty papers and concluded that the literature clearly showed men more likely than women to take risks. Weber, Blais and Betz found the same in 2002 across five domains of decision — financial, health and safety, recreational, ethical and social — with men perceiving less risk in four of the five. A large German sample replicated it.

What none of that work explained was why. Knowing that one group takes fewer risks tells you nothing about the machinery producing the difference, and the machinery is where this argument lives.

A 2006 study in the journal Judgment and Decision Making set out to take that machinery apart. Six hundred and fifty-seven participants rated their likelihood of engaging in risky activities across four domains and — this is the part that matters — separately reported three distinct judgments about each: how probable a bad outcome was, how bad it would be if it happened, and how much they expected to enjoy the activity if nothing went wrong.

Decomposing it that way produced a finding far more specific, and far more useful, than the headline everyone quotes.

Read the last two rows again, because they are what rescue this argument from the crude version of itself. The difference is not that women flinch at catastrophe — severity was the weakest of the three mediators. And in the social domain the difference disappears entirely; in some studies it reverses outright. This is not a general disposition toward caution. It is something much more particular: a difference in how likely a bad outcome is judged to be.

A developmental study by Hillier and Morrongiello had already found the same split in children, and stated it more starkly than the adult literature ever has. Shown pictures of physically risky situations, boys' overall risk judgments were predicted by how severe they thought an injury would be. Girls' judgments were predicted by how vulnerable they were to injury of any kind. The authors drew the obvious conclusion: girls avoid situations carrying any real likelihood of getting hurt; boys avoid them only when the injury looks bad enough.

Now hold that against the first sixty years of the automobile. Only when the injury looks bad enough. That is not a caricature of the industry's position. It is a precise description of it.

The fifth domain, which nobody quotes

The 2006 study did one more thing, and it is the single most useful result in this entire article.

Alongside the four risk domains, the researchers tested a fifth category of choice they called the positive domain: decisions offering a small, certain cost in exchange for a small probability of a large reward. Buying the ticket. Making the call. Paying a little, now, against long odds later.

In every other domain tested, women were less likely to engage. In this one the result reversed. Women reported being more likely to take the bet — and the study found this was partly explained by their more optimistic judgments about how probable good outcomes were, and how good they would be.

This is where the argument stops being about temperament and becomes a claim about competence at a particular task. Safety engineering is not brave work and it is not timid work. It is probabilistic work: reasoning carefully about rare events, assigning real weight to outcomes nobody has yet observed, and accepting a guaranteed small loss to buy a chance at a large invisible gain. The evidence indicates that this specific combination of judgments is not evenly distributed across the population. It also indicates that the people the industry hired, promoted and listened to for sixty years were drawn almost entirely from the half that scores lower on precisely the judgment the work requires.

The industry did not lack engineers. It lacked people willing to take the bet.

There is a grim confirmation of all this sitting in Part One of this series. When the industry finally did build a safety apparatus, it built it around a proxy human being that was, for fifty years, a man of average height and average weight — and women in comparable crashes turned out markedly more likely to be seriously injured. The blind spot was not only in what got built. It was in who the builders were able to imagine being hurt.

II. The test

If the claim is true, the record should look a certain way

An argument like this is worth very little standing on its own. Fortunately it makes a prediction that can be checked.

If safety work is disproportionately the work of people who weight the probability of harm heavily, and if women were largely kept out of the rooms where cars were designed, then the handful of contributions women did manage to make should be strikingly lopsided. They should cluster rather than scatter. And they should cluster on the preventive side — the first collision, the crash that does not happen — rather than on the survival side that occupied the industry's official effort.

So assemble the list: every woman with a documented, checkable claim on some part of the automobile. Then sort by which collision the contribution addresses.

Seven to one. Communication, visibility, lane discipline, sobriety, position — every one a device for preventing an impact rather than surviving one. The prediction holds. What follows is the case file behind it.

III. The case file

The woman who made the cabin warm

Margaret Wilcox was born in Chicago in 1838, into a country where a married woman could not reliably hold a patent in her own name. Her earliest inventions were filed under her husband's. That changed with the married women's property acts — Illinois passed its statute in 1861 — and by 1890 she was filing as Margaret A. Wilcox, which is the only reason we know about her at all.

Between 1890 and 1909 she took out at least seven United States patents, and the range is the tell. This was not a woman with one idea.

The 1893 patent was written for railway carriages, and it did not solve a comfort problem. It solved an exposure problem. A Chicago winter inside an uninsulated rail car was genuinely dangerous, and the only heat on board came from portable stoves and gas lamps with an unfortunate habit of setting the carriage on fire. Wilcox's answer was to stop carrying a second source of heat and start using the enormous quantity the engine was already throwing away.

Her design put a combustion chamber beneath the car inside a water-jacketed casing divided by vertical partitions and horizontal diaphragms. Hot water rose through pipes running along the floor at both sides of the carriage; cooled water returned to the bottom of the casing to be reheated. Circulation was driven by a spiral conveyor geared to the axles, so the system pumped harder the faster the train ran and needed no power source of its own. It had one flaw she never solved: no way to regulate it. You set it when you boarded and lived with the consequences three hours later.

She died in Los Angeles on the thirtieth of March, 1912, before the automobile had any use for the idea. Closed passenger cabs had only appeared around 1907; the first crude car heaters around 1917. It took until the 1929 Model A for Ford to route engine heat into a passenger compartment in a way that actually worked. Every car built since runs her loop — hot coolant out of the engine, through a heater core, a fan pushing air across it, warm air into the cabin. Cars, trucks, buses, trains, aircraft. The principle has not changed in a hundred and thirty years because there is nothing left to improve. The heat is free, and it is already there.

The signal she never patented

Florence Lawrence was born Florence Annie Bridgwood in Hamilton, Ontario, on the second of January, 1886. Her mother, Charlotte Bridgwood, was a touring stage actress who performed as Lotta Lawrence and eventually handed the surname to her daughter. The child toured with the company as Baby Flo, the Child Wonder Whistler. She took her first film part in 1907, appeared in something close to three hundred pictures, and in 1910 became the first screen actor in America to have her name printed above the title — which is why she is remembered, when she is remembered at all, as the first movie star.

She was also a serious, hands-on car enthusiast at a time when almost no women and not many men were. She bought her own car, worked on it herself, and described the relationship in a line worth quoting exactly: "A car to me is something that is almost human."

In 1914 she built what she called an auto signaling arm — a mechanical flag on the rear fender, raised or lowered by a button beside the driver's seat, to tell the vehicle behind which way she was about to turn. Then she built its companion: a sign at the back of the car that flipped up to read STOP whenever the brake pedal went down. A turn signal and a brake light, in the same year, from the same person, in an era when a car announced its intentions by means of the driver's arm and the goodwill of everyone nearby.

She never patented either one. Her mother later tried to file on her daughter's behalf and failed. By 1918 a Boston outfit called the Naillik Motor Signal Company was selling motorized versions; by 1939 Buick had made electric turn signals standard equipment; by the 1950s they were on everything. Lawrence received nothing, and no credit. She died in Beverly Hills on the twenty-eighth of December, 1938, by her own hand, at fifty-two.

Two clarifications the record demands. She was not provably the first person on earth to conceive of a directional signal — trafficators were being worked on in several countries at once — but the evidence indicates she arrived at hers independently, and she was unquestionably early. And the windshield wiper in this family belongs to her mother, not to her: Charlotte Bridgwood patented an automatic, roller-driven windshield cleaner in 1917, and Florence went on to serve as president of Bridgwood Manufacturing. That story, alongside Mary Anderson's hand-cranked 1903 original, is told in full elsewhere in these pages.

What belongs here is the trade she made: a flag and a button, costing almost nothing, bought against the probability that the driver behind her would guess wrong.

A device conceived in 1914 by an actress who never earned a cent from it is, by this estimate, still preventing twice as many crashes as we lose to the phone in everyone's hand. It works perfectly. It costs nothing. It requires a finger.

The mother who changed the law

On the third of May, 1980, in Fair Oaks, California, thirteen-year-old Cari Lightner was walking to a church carnival with a friend when a car struck her from behind and killed her. The driver's blood alcohol content was around 0.20, roughly twice the legal limit. He did not stop. He had a string of prior drunk-driving arrests and had been out of jail two days. He was sentenced to two years.

Her mother, Candace Lightner, was told by police what everyone in 1980 already understood: that drunk driving was rarely prosecuted seriously, and that this outcome was normal. Four months later, on the fifth of September, she and Cindi Lamb — another mother, another child — incorporated Mothers Against Drunk Drivers in Sacramento.

What followed is the most effective road-safety campaign in American history, and it did not involve inventing anything at all. Lightner lobbied Governor Jerry Brown daily until he convened a state commission and made her its first member. California passed minimum fines and mandatory jail for repeat offenders in 1981. Reagan appointed a national commission in 1982. Legislatures passed more than four hundred drunk-driving laws across the country. On the seventeenth of July, 1984, the national minimum drinking age went to twenty-one. By 2004 all fifty states had adopted the 0.08 threshold.

Look at what Lightner actually did, in the terms this article has been using. She took an event the entire legal and cultural apparatus had classified as an accident — unfortunate, unpredictable, nobody's fault — and reclassified it as a probability that could be driven down by design. MADD went so far as to campaign to replace the word accident with the word crash, a linguistic intervention aimed squarely at how likely people believe a bad outcome to be.

That is the 2006 study's central finding, executed at national scale by a woman with no engineering training and no institutional power of any kind. She did not think the injury was more severe than anyone else thought it was. She thought it was more preventable.

The chemist in the gasoline

Edith Flanigen joined Union Carbide in 1952, out of Buffalo, at a moment when there were almost no women in industrial chemistry. Four years later she began working on molecular sieves — zeolites, porous crystals whose pores are sized to sort molecules by shape, letting some through and trapping the rest. In 1968 she was made head of the company's molecular sieve research team, which was an extraordinary position for a woman to hold in 1968. In 1973 she became Union Carbide's first female Corporate Research Fellow, and in 1982 its first Senior Corporate Research Fellow.

Over a career spanning six decades she invented or co-invented more than two hundred novel synthetic materials and accumulated one hundred and nine United States patents. The one that matters most to a car magazine is zeolite Y — a catalyst that substantially increases the quantity of gasoline recoverable from a barrel of crude, and makes the refining process cleaner and safer while doing it.

She also synthesized gem-quality emeralds, which Union Carbide sold both as jewelry and as components in early laser work. She was the first woman awarded the Perkin Medal, American applied chemistry's highest honor, and she received the National Medal of Technology and Innovation from President Obama.

Flanigen is the hardest case in this article to fit to the argument, and it is worth saying so plainly rather than forcing her. Zeolite Y is not a safety device. Nobody survived a crash because of it.

But look at what the work is. A molecular sieve does one thing: it separates what should pass from what should not, at a scale nobody can see, by anticipating the shape of a molecule that has not arrived yet. Her sieves are now used to pull water out of automotive air conditioning lines, to concentrate oxygen for people who cannot breathe unaided, to replace the phosphates that were poisoning rivers through laundry detergent, and to strip radioactive contamination from water at a ruined nuclear plant. That last application is the tell. When something goes catastrophically wrong and the problem is separating the lethal thing from the harmless thing, this is the technology you reach for.

She spent sixty years building filters for disasters. Some of the disasters had not happened yet.

The shape of the Earth

Gladys Mae Brown was born in 1930 in Sutherland, Virginia, on a farm, in the Jim Crow South. She walked three miles to a segregated one-room schoolhouse with a leaking roof and one underpaid teacher. She worked out early that the valedictorian and salutatorian of her high school would win scholarships to Virginia State College, and she made sure to be one of them.

In 1956 she was hired as a mathematician at the naval proving ground at Dahlgren, Virginia — the second Black woman ever hired there, and one of only four Black employees. She stayed forty-two years.

Her early work was range and bombing tables. In the 1960s she was part of an award-winning study establishing the regularity of Pluto's motion relative to Neptune. Then she moved to satellite geodesy, which is the unglamorous science of describing the actual shape of the Earth — not the sphere of a schoolroom globe, and not even the smooth ellipsoid of a textbook, but the lumpy, irregular geoid produced by an uneven distribution of mass, tugged on by gravity and tides and the planet's own rotation.

She programmed an IBM 7030 to grind through satellite altimeter data and produce models of that shape at steadily increasing resolution. She became project manager for Seasat, the first satellite capable of remotely sensing the oceans, and cut her team's processing time in half. In 1986 she published the data-processing specification for the Geosat radar altimeter, a technical document about extracting more accurate geoid heights from orbit.

None of that is a car story until you consider what it is for. A satellite navigation system works by trilateration against orbiting clocks, and its accuracy is bounded by how well you know the shape of the thing you are standing on. Get the geoid wrong and every position derived from it is wrong by the same amount. West's models are part of why the error budget of modern satellite navigation is measured in centimeters instead of city blocks — which is the difference between a device that tells you roughly where you are and one that can tell a car which lane it is in.

Lane-keeping. Automatic emergency braking that knows a curve is coming. Collision warnings tied to map geometry. Every one of those is a first-collision technology, and every one of them rests on knowing the shape of the Earth to a precision that a woman at a naval weapons lab spent four decades improving, on a machine with less memory than a modern key fob.

She received essentially no public recognition until she was in her eighties. In 2018 the Air Force inducted her into its hall of fame. She finished a doctorate at seventy.

The rooms, and the panel they could not draw

In 1955 General Motors hired Suzanne Vanderbilt into its styling department. She was part of a group Harley Earl assembled, promoted, and — with a tin ear that has echoed for seventy years — allowed the public relations department to name the Damsels of Design. Vanderbilt loathed the name.

Earl's reasoning was commercial and, for 1955, unusually clear-eyed: women were making an increasing share of household purchasing decisions, including about cars, and GM had nobody on staff who could think about the product from that side. So he hired ten women. Six went into automotive interior design; the rest went to Frigidaire.

What they produced is a list of things now so ordinary that nobody wonders where they came from. Child-proof door locks. The glove compartment as we understand it. Storage consoles. Illuminated vanity mirrors. And, in Ruth Glennie's 1958 "Fancy Free" Corvette shown at the Feminine Auto Show, General Motors' first retractable seat belt.

Now the detail that makes this section belong in this article. The Damsels designed every element of the interior — seats, doors, trim, fabrics, detailing, color — with one exception. They were not permitted to design the instrument panel. That was considered a masculine responsibility and it was closed to them.

Read that against Part Two of this series, which ended on the observation that the people who take the fewest risks were largely absent from the rooms where safety was designed. Here they are actually in the room. They are in the building, on the payroll, at the drawing boards. And the one surface in the entire car where the driver's attention lives, where every control and every warning is placed, where the question of what a person can see and reach and operate without looking away from the road is settled — that surface is off limits to them by policy.

When Earl retired after the 1958 show, his successor Bill Mitchell had little interest in the arrangement, and the group dispersed. Vanderbilt stayed twenty-three years anyway.

The woman who finally got the panel

Mimi Vandermolen was born in the Netherlands in 1946 and moved to Toronto as a child. In 1965 she became the first woman to study industrial design at the Ontario College of Art. She joined Ford in 1970, worked on the Mustang II and the Granada, and by 1979 was a design specialist. Then she was pulled onto Team Taurus.

The 1986 Ford Taurus is remembered for its exterior — the aerodynamic shape that broke American sedan styling out of a decade of right angles. Vandermolen led the interior, and the interior is the more consequential half.

She curved the dashboard so that the controls came to the driver instead of requiring the driver to lean forward, which is what virtually every previous American car had demanded. She replaced push-pull climate switches with rotary dials. She gave buttons raised tactile bumps so they could be identified by feel. She rebuilt the seats from scratch, a process that consumed two and a half years and a hundred thousand miles of prototype testing — the most extensive seat evaluation Ford had ever conducted.

To force her own designers to understand the problem, she made the men on her team wear false fingernails while operating the controls, and reportedly threatened to put them in skirts so they would find out what getting into the car was actually like.

Every one of those decisions is a safety decision wearing the clothes of a convenience feature. A control identifiable by touch is a control operated without looking. A dial that can be turned blind is a second of eyes on the road that a push-pull switch would have cost. A dashboard curved toward the driver is a driver who is not leaning out of position at the moment something happens. Nothing about this shows up in a crash test rating. All of it shows up in whether the crash occurs.

In 1987 Vandermolen was promoted to design executive for small cars across North America — the first woman in the industry to hold a job of that scope — and went on to lead the 1993 Probe inside and out. The Taurus became the best-selling car in the United States. Ford called the whole program the rounded-edge revolution and largely credited the shape of the body.

Twenty-nine years after General Motors told a room full of women designers that the instrument panel was not theirs to touch, one of them finally got it. She used it to make a car you could operate without taking your eyes off the road.

IV. The rest of the record

The names that do not carry a section

A handful of others belong in the file, either because their contribution is smaller, or because it is documented less securely, or because we have already told the story elsewhere.

Bertha Benz financed her husband's company and then, in August 1888, took the Patent-Motorwagen and two of her sons on an unannounced sixty-six-mile drive from Mannheim to Pforzheim — the first long-distance automobile journey anyone had attempted. She was proving the machine worked, and she was also debugging it. When the wooden brake blocks wore down she stopped at a cobbler and had him nail leather over them, which is the first brake lining in history. She also cleared a fuel line with a hatpin and insulated a wire with a garter. Every part of that trip is a person calculating what could go wrong and solving it.

Mary Anderson patented the hand-operated windshield wiper in 1903; Charlotte Bridgwood patented the automatic version in 1917. Neither made money from it. Their story is told at length in Blame It on the Rain.

June McCarroll, a physician in the California desert, was run off the road into a sand ditch by a truck in 1917 and concluded that the road needed a line painted down the middle of it. The idea is so completely absorbed into modern life that it is genuinely difficult to see as an invention. Her claim is contested — Michigan credits a man with an earlier line — and we have given the whole dispute its own article.

Helen Blair Bartlett earned a doctorate in mineralogy from Ohio State in 1931 on an AC Spark Plug fellowship, joined the company's ceramic research department, and stayed until she retired in 1966. Spark plug insulators had been made of stacked mica, which leaded gasoline fouled so quickly that drivers cleaned their plugs constantly. Bartlett's work on sintered alumina — a patent application filed with colleagues in 1935 — produced an insulator with the mechanical strength, high-temperature electrical resistance and thermal shock resistance the job actually required. She accumulated seven patents and briefly left General Motors to work on the Manhattan Project before returning. This one is honestly a reliability story more than a safety story, and it is included here for exactly that reason: the record is not uniformly safety-shaped, and pretending it is would be dishonest.

Hedy Lamarr requires care, because the popular version of her story does not survive scrutiny. With the composer George Antheil she filed in June 1941 and received, in August 1942, United States patent 2,292,387 for a secret communication system: a frequency-hopping scheme, synchronized like two player pianos, meant to keep a radio-guided torpedo from being jammed. That patent is real and it was hers. What is not true is that she invented frequency hopping — prior art predates the filing by at least a year — or that the patent was classified as vital to national security, a claim disproved by its own routine processing. The Navy declined it. She was inducted into the National Inventors Hall of Fame in 2014. The straight version is impressive enough without the embroidery.

Helene Rother fled the German advance across Europe, arrived in Detroit, and in 1943 was hired by General Motors to design automotive interiors — the first woman known to have done that job at an American automaker. She had previously designed hat pins, jewelry, furniture, and comic books.

Alice Ramsey, in 1909, became the first woman to drive across the United States: some three thousand eight hundred miles, of which roughly a hundred and fifty were paved. She founded the Women's Motoring Club and in 2000 became the first woman inducted into the Automotive Hall of Fame. Denise McCluggage was a New York Herald Tribune reporter who raced professionally, co-founded AutoWeek, and was still filing copy for it when she died in 2015. Janet Guthrie, an engineer before she was a driver, became the first woman to race in the Indianapolis 500 and the Daytona 500 in 1977.

Those last three did not invent anything. They are here because the argument requires them: they are the control group. Given access to the automobile on any terms at all, women did all the same things men did with it, including the reckless ones. Guthrie drove flat out at Indianapolis. Whatever the pattern in this article is, it is not a pattern of people who did not want to go fast.

V. The case against

Every serious objection, and what survives them

An argument that only sounds good to people who already agree with it is not an argument. Here are the strongest objections, stated as forcefully as the people who hold them would state them.

VI. Where do we go from here

What the claim actually asks for

If the argument is right, it does not ask for a quota. A quota would be a way of appearing to have acted on it without acting on it.

What it asks for is that the automobile industry take seriously a proposition it has never once tested: that the judgment safety work requires is a specific and unevenly distributed capability, that hiring and promotion for safety roles ought to select for it, and that sex is a real but imperfect proxy for it — useful precisely because the industry has spent a century selecting against it by accident.

It also asks a narrower question that can be answered immediately. Sixty-two percent of new cars in the United States are bought by women. The regulatory apparatus that certifies those cars as safe was built around a male proxy body and, as Part One of this series laid out, still has not been fully rebuilt. A dummy modeled on an actual female body has existed since 2023. American regulators published design details for their own advanced small-female dummy in 2025. It is still not required.

The objection raised against requiring it is cost. A fully instrumented female dummy runs to about a million dollars, roughly twice its predecessor.

Sit with the shape of that objection for a moment, because it is the argument of this entire article compressed into a single procurement decision.

A certain, immediate, quantifiable cost — one million dollars, on a line item, this fiscal year. Against an uncertain, distant, statistically distributed benefit: some number of women, whose names nobody will ever know, who will walk away from crashes that would otherwise have injured them, in accidents that have not happened yet and mostly never will.

That is the positive domain. That is the identical bet the industry declined in 1893, and 1914, and 1959, and 1980. And it is being declined again, right now, in the same terms, for the same stated reason, by the same kind of room.

Margaret Wilcox took that bet with a water jacket and a spiral conveyor. Florence Lawrence took it with a flag and a button and never saw a dollar for it. Candy Lightner took it with a kitchen table and four hundred laws. Edith Flanigen and Gladys West took it for forty-two years each and died eleven days apart in January, largely unremarked, having between them made the fuel more efficient and the map more accurate for most of the human race.

The industry has spent a hundred and thirty years being handed the winning side of that wager by people it declined to hire. At some point the pattern stops being a curiosity and starts being a recruitment strategy.