This is the field’s third or fourth resurrection. People declared it dead, with funding obituaries in serious publications, more than once. The version of AI that is in your phone right now sits on top of seventy years of false starts, busted hype, two famous winters, and one quiet 2012 result that almost nobody outside the field noticed for several years.

You do not need to know that history to use a chatbot. You need to know it to read the headlines about chatbots without being played. Most articles about “the AI revolution” are written by people who have never read a paragraph about the AI revolution of the 1980s. The same shape of optimism and the same shape of disappointment has happened before. Naming the pattern is what this lesson is for.

Read more

Originally published on Medium for consideration as a speaker at Medium Day 2026.

When you are the subject of the disinformation, source-first writing is the only armor that holds. The other kind, the heartfelt kind, the lived-experience kind unaccompanied by a citation chain, gets eaten alive by a hostile committee staffer in about ninety seconds. I know. I have watched it happen. I used to write the heartfelt kind myself.

I’m a 60-year-old transgender woman in Sioux Falls, South Dakota. I came out at 53. I publish on Medium and Substack about trans identity, biological sex, health informatics, federal counterterrorism policy, theology, and the slow procedural machinery that converts a culture-war slogan into a state-code revision. My pieces sometimes run 15,000 to 20,000 words. They get cited by attorneys filing amicus briefs, journalists covering bills nobody else is covering, and committee staffers writing memos for legislators who haven’t read the actual text of what they are voting on.

I want to tell you what I do, why I do it that way, and what I want to teach at Medium Day this September.

The problem

The phrase “gender ideology” was coined in the 1980s by Vatican intellectuals, refined through European and Latin American conservative networks across the 1990s and 2000s, and imported into American politics by organizations whose annual budgets run past $100 million. By the time it lands on cable news as a fact about you, it has been moving through institutional channels for forty years. It arrives with footnotes. The footnotes are bad, but they are there.

Most advocacy writing on the other side does not arrive with footnotes. It arrives with feeling. The feeling is real, justified, and will not survive contact with a 47-page committee report. This isn’t a moral failure of the advocacy writer. It is a structural mismatch. One side is doing citations, and the other side is doing testimony, and committee staffers, federal program officers, and state legislators read citations.

The other side has built infrastructure. I built the infrastructure back.

The pipeline

I run a set of custom scripts that pull from federal databases, state legislative trackers, peer-reviewed journals, agency budget documents, court filings, and news archives twenty-four hours a day. Overnight, the scripts collect. When I wake up, the morning’s raw material is already on disk. I sorted it. I read it. I write about it.

This sounds more glamorous than it is. Originally, the scripts were not AI. They were scrapers, RSS aggregators, and a small library of regex patterns that flagged the kind of language I have learned to look for. I added Claude Cowork & Claude Code to my workflow earlier this year. I’m still running both systems side by side, and sometimes Claude will find what my scripts didn’t. I’ll take the advantages where I can get them. The infrastructure is unglamorous. The discipline that uses it is the thing.

The discipline runs in five parts.

Source before sentence

I do not write a paragraph until the citation for it exists on my desk. This isn’t a virtue. It’s a defense. The first time you publish a piece on a bill and a hostile reader emails to say your claim is wrong, you need to be able to send back, the same hour, the Senate Judiciary Committee report page number that proves you right. The reader won’t apologize. That was never the point. The point is that the next hostile reader, watching the exchange, learns that you can’t be moved by bluster. After enough of those, the bluster stops arriving. Source-first writing is a long game, and the long game is reputational armor.

Authority in the clause

I do not write “studies show.” I write the name of the source, the year, and the sample size, inside the sentence. The 2022 U.S. Transgender Survey reports 92,329 respondents. The Williams Institute at UCLA estimates that 30 percent of trans adults have experienced homelessness at some point. The 2024 study by Lee and colleagues in Nature Human Behaviour, using a difference-in-differences design across 35,196 transgender and nonbinary young people aged 13 to 17, found up to a 72 percent increase in past-year suicide attempts in states that enacted anti-trans laws.

A hostile reader can’t dismiss the Williams Institute without dismissing UCLA, and they can’t dismiss Nature Human Behaviour without dismissing a Nature title. They will sometimes try. They look worse when they try than when you cite the source casually. Put the credential inside the sentence so it carries the weight.

Doors, not one building

A 15,000-word policy analysis is not how most people read. I know that. So every long piece gets distilled, in the same week, into three shorter versions. A 2,000-word summary article for outlets that can carry a longer piece. A 1,000-word breakdown for platforms that want substance without scale. A 500-word post for social. Each tier links back to the full piece. Each tier is sourced as cleanly as the full piece.

This is the move I am proudest of; the math underneath it is honest. A reader who came in through a 500-word Threads post and a reader who came in through a 22,000-word footnoted research article are the same reader at different doors. The work is built to receive both.

Take the opponent’s strongest slogan.

“Parental rights” is a phrase that means parents have the right to make medical decisions for their children. The bans on gender-affirming care for adolescents remove that right. The phrase, used honestly, is a phrase that condemns the bans. I write the piece that uses the phrase honestly, in front of the readers for whom the phrase still means what it used to mean.

“Big Government.” “Biological truth.” “Common sense.” Every one of these has a structural reframe inside it, waiting for somebody willing to write the piece that puts the slogan back on its actual referent. The unconvinced reader is moved by the reframe in a way no amount of moral assertion will move them. The reframe has the advantage of being true.

The personal stake comes last

If the first thing the reader meets is your body, the citation chain has to do twice the work to earn back the ground. The personal stake earns the most when it shows up after the case has already been carried by the evidence. The unconvinced reader doesn’t yet trust your lived experience. They’ll trust it more after they have watched you handle eighteen federal documents without flinching.

So the structure goes evidence, evidence, evidence, then the writer’s body, then the close. The close is short.

What it’s for

In December 2025, Attorney General Pamela Bondi issued an implementing memorandum to federal prosecutors and the FBI on National Security Presidential Memorandum 7. The administration’s FY27 budget request, submitted April 4, 2026, asked Congress for $166.1 million to fund what the document calls the NSPM-7 Joint Mission Center. The budget language describes the work in four verbs.

Map. Identify. Cripple. Neutralize.

The criteria the JMC will use to map and identify, written into appropriations language, include “extremism on migration, race, and gender.” The category includes writers, advocates, and trans Americans whose public output, by the framework’s own taxonomy, marks us as targets of the proactive surveillance the $166.1 million is for.

Read that paragraph again.

The federal government has put four verbs and a dollar figure on a workflow that aims at people in the category I belong to. The verbs are the policy. The dollar figure is the appropriation. A footnote doesn’t stop a federal budget. A footnote slows down the conversion of the budget into a precedent that survives the next administration. A footnote earns the next reporter, the next staffer, the next judge a place to stand.

The stake

I’m writing this from Sioux Falls, which is where I live, and where, in early 2026, the state legislature passed HB 1184 by a House vote of 57 to 9 and a Senate Judiciary Committee vote of 7 to 0. The bill codifies biological essentialism into multiple chapters of state code. The South Dakota Supreme Court ruled in March 2026 that trans people cannot alter the sex marker on their birth certificate. Lucky for me, I was born in Minnesota where my birth certificate is safe, but my existence is no less under threat.

I have Type 1 diabetes, weekly estradiol injections, and a small set of girlfriends who let me know when it’s safe to be in a public space. I came out at 53. I do not get those thirty years back. What I can do is be honest about them on the page, and put the honesty on a citation rail strong enough that the next person who tries to dismiss the testimony has to dismiss the apparatus around it first.

Map. Identify. Cripple. Neutralize. The verbs are the policy. The footnotes are the answer. Not the only answer. The answer that survives the news cycle.

What I want to teach

At Medium Day 2026, in the Identity and Social Justice slot of the Perspectives track, I want to walk attendees through the workflow. The aggregator setup, the source-first paragraph discipline, the tiered distillation, the slogan reframe, the placement of the personal stake. I want attendees to leave with a working version of what I do, scaled to whatever advocacy they are already trying to write. Not the trans-rights version. The methodology underneath it, which travels to any beat where the disinformation has a citation chain attached.

If you write about climate, public health, criminal justice, disability, immigration, or any other beat where the bad-faith framing is funded and footnoted, this workflow is for you. If you write about the beat you are the subject of, this workflow may be the thing that lets you keep writing.

The scripts are running. So am I.

Author Note:

Grace Ann Hansen is an independent researcher and writer, and an MBA and PhD graduate student in health informatics and artificial intelligence. She is also a published author, a professional musician, a gymnastics coach, and a queer transgender woman living in Sioux Falls, South Dakota. She corrects all her papers and articles with Grammarly, because even though she has deep thoughts, she has shallow patience for punctuation. She uses Anthropic’s Claude in Research mode for source location and verification on cited factual claims; all interpretation, argument, and prose are her own. Correspondence concerning this article should be addressed to Grace Ann Hansen at grace@graceannhansen.com.

The mustache is the thing to start with. Jonas Hole, salutatorian of D’Iberville High School’s class of 2026, is a transmasculine senior in JROTC. He has earned a “superior cadet” designation, serves on the sabre team and color guard, and has a photograph of himself in uniform that he has put into the public record on his own social media. The mustache is in that photograph.

The photograph the school posted to its official Facebook page on May 17, 2026, the Sunday before graduation week, wasn’t that photograph. The mustache in the school’s version had been, according to commenter Michaela Boffman, “photoshopped… off to make him look more ‘feminine’” (Ennis, 2026). A Biloxi drag performer and local organizer, Tara Shay Montgomery, posted a side-by-side comparison and reported that the mustache in the school’s version had been “edited out to look like a shadow” and that the lips had been “pinkened” (Ennis, 2026). The school’s caption named the salutatorian, but didn’t name him Jonas. It used the name he no longer uses.

That’s the case. A deadname in a salutatorian announcement, attached to a photograph that the community read as having been altered to feminize a transmasculine young man, published by the school speaking in its own institutional voice, on the eve of the graduation that the announcement was meant to celebrate. The post has been live since May 17. It had drawn more than thirteen hundred comments by the time The Advocate reported on it (Ennis, 2026).

I want to say one thing clearly at the top: the deadnaming is the headline, and the photograph is the deeper story. Verbal misgendering refuses recognition. Visual alteration rewrites the subject. They are not the same act.

What Two Outlets Could Verify

The reporting traces to two national LGBTQ trade outlets, The Advocate and LGBTQ Nation, both of which broke their stories on May 20, 2026. Both verified the school post against the live Facebook page. Both quoted the deadname-bearing caption. Both reported the JROTC detail and the salutatorian honor.

The mustache allegation is what’s “alleged.” Neither outlet ran forensic analysis on the photograph. The claim that the image was edited rests on lay visual comparison between the school’s version and Hole’s own selfie, made by people in the comment thread and by Montgomery on her own page. I want to be exact about what we know and what we don’t know. We know the school posted the deadname. We know the community read the photograph as altered. Whether the photograph was in fact altered, as opposed to looking that way through compression artifacts, lighting, or sharpening, hasn’t been independently established as of May 22, 2026.

That uncertainty doesn’t undo the argument. It refines it. The community read the image as feminizing. If a feminization happened, it is documentary evidence of intent in a way that almost no other school speech could be. If a feminization didn’t happen but the community read the image that way, the institutional posture that made the reading plausible is itself the story.

D’Iberville’s principal, Jennifer Courtney, didn’t respond to The Advocate’s phone and email inquiries. Neither did Harrison County School District Superintendent William Bentz (Ennis, 2026). Three days into the news cycle, on the Friday before the Saturday graduation, the institutional voice that had posted the deadname hadn’t produced a corrected post, a correction, an apology, or an acknowledgment.

This Is Year Four

The D’Iberville incident doesn’t arrive without a record. The Harrison County School District has been here before.

In May 2023, the same district told L.B., a transgender senior at Harrison Central High School, that she wouldn’t be permitted to attend her graduation in a dress and heeled shoes. A federal district court declined to enjoin the policy. L.B. didn’t attend her own graduation (ACLU, 2023; Pham-Bui, 2023).

In August 2023, the district revised its handbook to specify that “boys must wear shorts or pants” and “girls must wear dresses or skirts or shorts or pants,” and added that “attire must be consistent with the gender in the school district’s permanent record,” meaning the birth certificate (Hall, 2023).

In May 2024, the ACLU and the ACLU of Mississippi filed a Title IX complaint with the U.S. Department of Education’s Office for Civil Rights against the district on behalf of a transgender student identified as A.H. The complaint alleged a pattern of dress-code enforcement that disproportionately targeted transgender and gender-nonconforming girls and had been used to exclude at least one student’s senior portrait from the yearbook over a tuxedo (Harrison, 2024).

In May 2026, D’Iberville posted the deadname-and-altered-photo salutatorian announcement, and Montgomery reported that approximately six other students had been excluded from the senior yearbook over enforcement of the same-sex-keyed dress code (Ennis, 2026).

Three years. Four documented removals of transgender and gender-nonconforming students from the culminating public rituals of high school: graduation, the yearbook, the band concert, and the salutatorian announcement. The pattern isn’t a coincidence, and it isn’t a series of administrative misfires. It’s a district-level practice.

The Statutes Mississippi Wrote

I want to keep the legal frame brief, since the courts will eventually have their say, and the day-to-day harm is happening now. Three things are true about Mississippi law as of May 2026.

Read more

Open the chat tool you have been using. Look for the dropdown that shows which model you are talking to. In ChatGPT it sits above the input box; in Claude it lives in the top bar; in Gemini it is tucked into a menu. Click it.

You will see a list. Two or three names, sometimes more, each one different from the product you opened. The product is called ChatGPT. The thing the dropdown lets you choose between has a name like GPT-5.5, or GPT-5.5 Pro, or GPT-5.5 Instant. The product is called Claude. The dropdown shows Opus 4.7, Sonnet 4.5, Haiku 4.5. The product is called Gemini. The dropdown shows 3.1 Pro and a couple of older versions.

What you are seeing is the moment three things you may have lumped together as “the AI” pull apart into three separate things. The product, the model, and the interface. Telling them apart is the second tool I will hand you for the rest of this course.

Read more

The man who built The Late Show, the man who originated the 11:35 franchise on CBS in 1993 and ran it for twenty-two years before he handed it to Stephen Colbert, was asked why the network was retiring it. David Letterman said: “They’re lying. They’re lying weasels.”

That is the sentence to start with. Not for its careful argument, but because the man who said it knows the building. He knows the executives. He knows what the show cost and what it returned. He knows what a structurally declining advertising market looks like and what a corporate cover story looks like, and he can tell the difference. Letterman walked out to the roof of the Ed Sullivan Theater with Colbert one Thursday in May 2026, and the two of them threw an office chair and a cake reading “The Late Show 1993–2026” off the building. That is a man who knows what he saw.

The corporate story is that Stephen Colbert’s show ended for one reason: late-night television is an unprofitable business,s and The Late Show could no longer cover its costs. The corporate story is true on the surface and a lie in its implications. Late-night television is unprofitable; that part is real. The aggregate linear advertising spend in the genre fell by roughly half between 2018 and 2024, from around $439 million to about $220.6 million, and the trend is unlikely to reverse. But the highest-rated show in a declining business is not the show you retire ten months in advance of the end of a contracted run. The highest-rated show is the one you migrate. You move it to streaming. You shrink the band, you cut a writer, you produce four nights instead of five. What you do not do is retire the franchise.

What Paramount did was retire the franchise. And the timing of when they did it is the part that the corporate story cannot explain.

The Week That Cannot Be Explained Without the Merger

Here is the sequence. On July 1, 2025, Paramount paid $16 million to settle Donald Trump’s lawsuit against CBS over the 60 Minutes edit of his opponent’s October 2024 campaign interview. The money was earmarked for Trump’s future presidential library. There was no apology. Two senior CBS News executives had already resigned in protest over the contemplated payment.

On July 14, 2025, Stephen Colbert came back from a scheduled hiatus and devoted his opening monologue to the settlement. He called the parent corporation’s payment a “big fat bribe.” He named the merger pending before the Federal Communications Commission. He named the reason a sitting president might want a network to please him.

On July 17, 2025, three nights later, CBS announced that The Late Show was canceled. The network said it was a “purely financial decision against a challenging backdrop in late night,” and that it was not related in any way to “other matters happening at Paramount.”

On July 24, 2025, seven days after the cancellation announcement and ten days after Colbert called his parent company’s payment to the president a bribe on national television, the Federal Communications Commission voted two to one to approve Paramount’s $8 billion merger with Skydance Media. The dissenting commissioner, Anna M. Gomez, wrote on the record that the merger approval was “a dark chapter in a long and growing record of abuse that threatens press freedom in this country” and a “direct violation of the First Amendment and the law.”

Three weeks. Settlement, bribe monologue, cancellation, merger approval. The cancellation lands at the geometric center of that sequence, and the corporate story is that the four events have nothing to do with each other.

I do not believe Paramount. I do not believe the network’s communications department. I do not believe the trade-press friends-of-the-house who carried the leaked $40 million figure into the next news cycle without naming a source. I believe Letterman.

The Math That Does Not Math

The $40 million figure is what you need to understand to grasp how the cover story works. The day after the cancellation, two trade outlets ran nearly identical pieces citing anonymous sources who said The Late Show was losing $40 million a year. By the end of the week, the number had grown to $50 million in one of the sourcing trees and stayed at $40 million in the other. The figure spread. Within seventy-two hours, it was anchoring sympathetic coverage of Paramount’s decision as a regrettable but inevitable response to inescapable economics.

No one would put a name on it. The fact-checking site Snopes asked the reporters who broke the figure where it came from and was told “multiple anonymous sources with knowledge of the show’s finances.” The policy organization Public Knowledge, which has been thinking about media economics for twenty-five years, pointed out the obvious: in 202,5 the financial picture of a late-night show is not just linear advertising revenue. It is YouTube monetization, clip distribution on Paramount’s own streaming platform, and affiliate fees that are allocated to the program by formulas the public never sees. The $40 million number, as constructed, almost surely counted only one of those revenue streams. Jimmy Kimmel, who runs the same kind of show at the same kind of network and knows the math, said it directly: “The idea that Stephen Colbert’s show was losing $40 million a year is beyond nonsensical.”

Colbert himself, after the leak, said it on air: “Over the weekend, somebody at CBS followed up their gracious press release with a gracious anonymous leak, saying they pulled the plug on our show because of losses pegged between $40 million and $50 million a year. $40 million is a big number. I could see us losing $24 million. But where would Paramount have possibly spent the other $16 million? Oh yeah.”

The arithmetic he was doing on stage was not a joke. The settlement was $16 million. The loss leak was $40 million. The gap is $24 million, the actual number a TV executive might recognize as a structural deficit in a late-night program in 2025. The leak was the cover story, and Colbert showed his audience how it was assembled.

The show was, as of this writing, the highest-rated late-night program on American broadcast television for nine consecutive seasons. It has held its time slot in the eighteen-to-forty-nine advertising demographic since 2019. In September 2025, with the cancellation announcement already on the books, it won the Primetime Emmy Award for Outstanding Variety Talk Series, the first time The Late Show franchise had taken that prize. You cancel that show in July of the year it wins the Emmy. You retire that franchise ten months before the contracted run ends. You spend three weeks watching the host call the parent company’s payment to a sitting president a bribe, and then you tell the audience the decision is purely financial.

The math does not math. Letterman knows it does not math. So does everyone who watched it happen.

The Price of a License

The merger is the through-line. Everything else is decoration on the merger.

Read more

DOI: https://doi.org/10.5281/zenodo.20322887

Author Note. Grace Ann Hansen has no known conflicts of interest to disclose. She corrects all her papers and articles with Grammarly, because even though she has deep thoughts, she has shallow patience for punctuation. She uses Anthropic’s Claude in Research mode for source location and verification on cited factual claims; all interpretation, argument, and prose are her own. Correspondence concerning this article should be addressed to Grace Ann Hansen at grace@graceannhansen.com.

Abstract

This paper formalizes and defends the Pedro-Pastry Equation, a closed-form expression linking the aerodynamic ejection of a planar polygon from a moving passenger vehicle to the maximum quantity of cupcakes obtainable by an economic agent (Pedro) in exchange for one human soul. The motivating word problem (“If I throw a triangle out of the car and the car is going 20 km/h and wind resistance is a thing that exists, how many cupcakes can Pedro buy with one human soul?”) has, despite its evident structural elegance, received no prior treatment in the peer-reviewed literature. The equation under analysis is C = (T · v² · ρ_air · A_d) / (S_h · π · ε), where C denotes cupcakes purchasable, T a triangle-type coefficient, v vehicle velocity, ρ_air sea-level air density, A_d the projected drag area, S_h the spot-market valuation of a human soul, π the ratio of a circle’s circumference to its diameter, and ε a bakery loyalty discount on the half-open interval (0, 1]. The methodology proceeds in three stages: (a) dimensional analysis under the Buckingham π theorem, which reveals a non-trivial residual dimensionality that the paper argues confers predictive rather than corrosive properties on the equation, in a manner analogous to dimensional regularization in quantum field theory; (b) a multi-disciplinary survey of variable valuations drawing on aerospace engineering, atmospheric science, scholastic theology, Faustian literary criticism, the value-of-statistical-life literature in welfare economics, the history of transcendental number theory, and the empirical marketing literature on retail loyalty programs; and © a closed-form calculation under standard assumptions (medium equilateral triangle with projected area 0.1 m², U.S. Standard Atmosphere sea-level density, Viscusi and Aldy spot soul valuation, Costco loyalty discount ε = 0.85). Under these assumptions the model returns C ≈ 42, a result that is invariant to several plausible perturbations of the input vector and that converges on the canonical answer to the ultimate question of life, the universe, and everything as derived by Adams (1979). Sensitivity analysis indicates that the Bermuda Triangle case (T → ∞) implies an unbounded cupcake supply, with attendant hyperinflationary consequences for the pastry market that warrant further study. The paper closes with a discussion of the Sokal problem, arguing that the present work differs from that hoax in that its content openly declares its absurdity even as its citations remain rigorous, an inversion of the Sokal pattern that may constitute a novel methodological category.

Keywords: dimensional analysis, bluff-body aerodynamics, value of a statistical life, Faustian economics, loyalty programs, transcendental numbers, applied absurdism

Read more

The plastic box will arrive on a flatbed. Bright blue, vented at the top, with a swing door that latches from inside, and a hand-sanitizer dispenser that the school will not refill. It will be set on a square of compacted gravel just past the gym’s outside wall, since that is where the building’s plumbing already runs and since nobody wanted it visible from the front lobby. The custodian will be told it counts. The principal will be told it counts. By August 2026, in roughly half of South Carolina’s K-12 districts, it will count.

This is what the word “accommodation” means now.

Governor Henry McMaster signed House Bill 4756, the South Carolina Student Physical Privacy Act, on May 15, 2026. The House had passed it 96–19, the Senate 33–2, and the House had concurred 77–31 on the Senate’s amendments. One of those amendments lets a district satisfy its single-occupancy bathroom obligation to transgender students with a temporary outdoor facility. The ACLU of South Carolina read the amendment in plain English in its April 13 statehouse dispatch: “the Senate introduced an amendment that would allow schools to install a temporary outdoor facility, in other words, a porta-potty.”

Districts that fail to comply lose twenty-five percent of state operating funds. Any student or employee who shares a school with a transgender peer can sue the district for declaratory relief plus attorney fees. The state board moves the money before any court rules; the district contests later. The procurement question writes itself. A porta-potty is cheap. A wall is not.

Representative Nancy Mace, the gubernatorial candidate who introduced the federal version of this fight in 2024 when she targeted Representative Sarah McBride on her first day in Congress, praised the South Carolina law by saying, “Men do not belong in women’s bathrooms. Men do not belong in women’s locker rooms. South Carolina got this right.” Attorney General Alan Wilson, another gubernatorial candidate, said the state’s women and girls deserve “the assurance that they are safe from any potential harm.” Governor McMaster, who has signed three anti-trans laws in three years, said almost nothing at the signing. He did not need to. The plastic box was already on the order form.

The word “accommodation” used to mean something.

It used to mean a wheelchair ramp at the side door of a building that had stairs at the front. It used to mean a sign-language interpreter on the day of the assembly. It used to mean an extended-time letter, an enlarged-text textbook, a quiet room for a kid with sensory processing issues, a service dog with its vest on. It used to mean the institution moving toward the person. That is what the Americans with Disabilities Act said in 1990, and what every IEP coordinator has said since: an accommodation is a modification the institution makes so that a person who is already inside the building can stay inside the building.

South Carolina has now moved the word. An accommodation, under H. 4756, is a plastic box on a gravel pad in the parking lot, accessed by a child who has to walk out of fourth period, past the football team’s practice, into November rain. The institution does not move toward the person. The institution moves the person to the curb. The word has been hollowed out and refilled with its opposite.

This is what I mean by semantic violence. The law did not invent a new policy and call it something honest. It took an existing word, one with thirty-five years of disability-rights jurisprudence and pediatric IEP practice behind it, and used the word’s reputation as cover. A porta-potty is not an accommodation. A porta-potty is an enclosure with a door. South Carolina’s General Assembly knows the difference. The bill’s drafters at Alliance Defending Freedom, who authored and distributed the model “Student Physical Privacy Act” to school districts and state legislatures, know the difference. The 33 senators who voted yes know the difference. The point of the language is to make the difference unspeakable.

Consider what the language does to the child who has to use the box.

She is fourteen. She has been on hormone blockers since she was twelve, or she has not. She passes most days, or she does not. She has friends in the building, or she has one friend, or none. She has held her bladder for three hours through algebra, lunch, and PE; she did not want to be the kid the football team watches walk to the parking lot. By the fifth period, her stomach hurts. She asks the teacher for a pass. The teacher knows where the pass goes. Everyone in the room knows where the pass goes. By the time she comes back, her hair is wet on one side; the wind off the practice field caught the door as she closed it. The teacher does not say anything. Neither does she.

This is the empirical record. It is not contested. Murchison et al. (2019), in the journal Pediatrics, found that 36 percent of transgender and nonbinary adolescents in schools with restroom or locker-room restrictions reported sexual assault in the prior year, with adjusted risk ratios of 1.26 for transgender boys, 2.49 for transgender girls, and 1.42 for nonbinary youth assigned female at birth. Wernick et al. (2017) in the Journal of Youth and Adolescence found that feeling safe in school bathrooms predicted self-esteem, grades, and overall school safety for transgender students. Russell et al. (2018) in the Journal of Adolescent Health found that being able to use a chosen name in school, home, work, and friendship contexts was associated with 71 percent fewer symptoms of severe depression, a 34 percent decrease in reported suicidal thoughts, and a 65 percent decrease in suicide attempts. The Trevor Project’s 2024 National Survey reported that transgender and nonbinary youth had lower rates of suicide attempts when they had access to a gender-neutral bathroom at school. GLSEN’s 2024 National School Climate Survey, released in April 2026 under the organization’s new name Glisten, found that 86 percent of transgender students avoid certain school spaces out of safety concerns, that 64 percent avoid school bathrooms altogether, and that 41 percent have been stopped or punished for using a bathroom aligned with their gender.

None of this is buried. The studies are in peer-reviewed journals. The numbers replicate. The effect sizes are large.

No comparable literature shows a benefit to cisgender students from excluding their transgender peers. The “transgender predator” hypothesis cited in conservative court briefs, including the Adams v. St. Johns County concurrences in the Eleventh Circuit, is unsupported by evidence in any of the surveys above. There are no incidents on the public record of a transgender student assaulting a cisgender peer in a school bathroom that they were lawfully using. There are many incidents of the reverse.

The South Carolina legislature has the same access to Pediatrics and the Journal of Adolescent Health that I do. The legislature voted anyway.

Read more

A federal grand jury in the Southern District of Florida indicted Carmen Mercedes Lineberger on Tuesday. She is 62. She just retired from the Department of Justice after nearly two decades, the last of them as Managing Assistant United States Attorney at the Fort Pierce branch office that prosecuted Donald Trump in the Mar-a-Lago classified documents case. She is alleged to have emailed Volume II of Jack Smith’s final report on that prosecution from her .gov account to her personal Gmail account on December 1, 2025. The file name she allegedly used was “Bundt_Cake_Recipe.pdf.”

Read that paragraph again.

Volume II is the special counsel’s narrative account of why he concluded the evidence would have sustained a conviction against Trump for the willful retention of national defense information. We aren’t allowed to read it. Judge Aileen Cannon entered an injunction the day after Trump’s second inauguration, enjoining the Department of Justice from releasing it outside the agency, and made the injunction permanent in February. American Oversight and the Knight First Amendment Institute have spent eighteen months trying to pry it loose through FOIA litigation, mandamus petitions, and intervention motions. The Eleventh Circuit told Cannon in November that she was unreasonably delaying. She delayed three more months and then ruled against release on a theory most legal scholars consider untenable.

That’s the document at the center of this case. “Bundt_Cake_Recipe.pdf.”

Here is the policy claim I want to make, plainly. The Lineberger indictment is the strongest argument for serious FOIA reform that anyone has handed the public in a decade. It is stronger than the Moynihan Commission report of 1997, which the political class has spent thirty years politely ignoring. It is stronger than the Public Interest Declassification Board’s 2020 Vision for the Digital Age. It is stronger than every think-tank white paper on government secrecy I have ever read, including the good ones. It is stronger because it is comic, and the public will remember a comic story more readily than a serious one. Stories about regulatory architecture do not travel. Stories about a federal prosecutor naming the most-litigated document in modern federal law enforcement “Bundt_Cake_Recipe.pdf” and emailing it to her Gmail travel account at the speed of a screenshot.

So I want to use the comic frame to make four serious points. Then I want to come back to the Bundt cake.

The system did not catch her.

That is the first point. The forensic detail every reform conversation should start from is that the alleged scheme is, technically, the most boring possible exfiltration. A PDF, renamed, attached to an outgoing email, sent from a .gov address to a Hotmail address in September and then to a Gmail address in December. Any halfway-competent enterprise data-loss prevention system can fingerprint a marked document and refuse to let it leave the network in an attachment. Any modern DLP can flag a federal prosecutor mailing PDFs from her work account to her personal account at all, let alone PDFs containing the contents of a document under a federal judge’s injunction. The Department of Justice either doesn’t deploy such systems across its U.S. Attorneys’ Offices, or it deploys systems so easily defeated that the trivial expedient of renaming a file to “Bundt_Cake_Recipe.pdf” is sufficient.

This is not a hypothetical. The conduct allegedly took place over ten weeks. Two separate transmissions. Hotmail in September, Gmail in December. According to the Department of Justice press release, the FBI’s investigation discovered it after the fact, post-retirement. The system did not catch her in real time. Lineberger is the named defendant, and if the allegations are true, she has earned her prosecution. But the institutional failure exposed isn’t principally a failure of one prosecutor’s integrity. It is a failure of the technical architecture that is supposed to make individual integrity unnecessary.

What “chocolate cake” tells you about “Bundt_Cake.”

The September email was sent under the subject line “chocolate cake recipe,” with the attachment named “Chocolate cake recipe.pdf.” The December email was sent under the subject line “Bundt_Cake_Recipe.pdf,” with the attachment named “Bundt_Cake_Recipe.pdf.”

The progression is the prosecution’s whole intent theory. In September, Lineberger sent internal DOJ communications and a memorandum marked “FOR OFFICIAL USE ONLY,” disguised as a chocolate cake recipe. The Department of Justice was not yet looking. Three months later, in December, she sent the sealed Volume II itself, disguised as a Bundt cake recipe. The escalation from casual lowercase (“chocolate cake recipe”) to formal underscored camel-case (“Bundt_Cake_Recipe.pdf”) tells you the strategy was rehearsed. It worked in September. So she ran it again in December, with greater operational confidence, and with the most contested judicially sealed document in the country.

This is what a federal prosecutor’s 18 U.S.C. § 1519 obstruction theory looks like in practice. The statute carries twenty years. Most of the press coverage of the Lineberger case has focused on the twenty-year number as if it were the punchline. It isn’t. The punchline is that the government chose § 1519 because the cadence of the file names, on the government’s theory, is evidence that Lineberger knew what she was doing and developed a strategy to keep doing it. The chocolate cake was a dry run. The Bundt cake was the real attempt.

There is a defense theory available, and Lineberger’s attorney has not yet articulated one publicly. The recipe filenames could be evidence of naïveté rather than concealment, a way to mark files for her own reference that only looks guilty in retrospect. She sent the files to her own personal accounts, made no attempt at encryption, and distributed them to no one outside. That is a real argument. But it is an argument that the trial process will have to engage with. The government’s reading is the more parsimonious one, and the file-naming pattern is the central evidence supporting it.

Read more

John Senkler is a service-industry worker in Saint Paul. According to the federal record, he was the secretary of the board of directors of Feeding Our Future, a Minnesota 501(c)(3) that funneled more than $240 million in federal child-nutrition reimbursements to a network of more than 250 meal sites that, federal prosecutors say, served little or no food to actual children. Last year, on the witness stand, Senkler said he hadn’t known he was on the board (Sahan Journal, 2025). He hadn’t attended meetings. He didn’t know what a motion was. And the signature on the minutes, the one that said he had made a motion to approve a transaction, wasn’t his signature.

That signature is the spine of this piece, so hold on to it.

The trajectory

On May 21, 2026, Aimee Bock was sentenced to 500 months in federal prison and ordered to pay approximately $243 million in restitution (Minnesota Reformer, 2026). Forty-one years and eight months. Federal prosecutors had asked for 50; her defense had asked for three. The judge sat in the middle and told Bock, “This was a vortex of fraud, and you were at the epicenter.”

Nine years earlier, in November 2016, Bock had incorporated Feeding Our Future as a small Minnesota nonprofit with two co-incorporators from her previous employer, the established meal-program nonprofit Partners in Nutrition. By 2019, she was disbursing about $3.4 million in federal funds. By 202,1 she was disbursing nearly $200 million (United States Department of Justice, 2022). The Minnesota Office of the Legislative Auditor later determined that her reimbursements grew approximately 2,800 percent in one year (Office of the Legislative Auditor, 2024).

That is the rise. The fall is the FBI raid in January 2022 and the trial that ended last March.

The hollowing-out

The trajectory is the story. And the trajectory is the architecture as much as it is Aimee Bock.

Federal child nutrition funds do not flow directly from the USDA to meal sites. It flows from the USDA to state agencies, then from state agencies to what federal regulations call “sponsoring organizations,” and finally from the sponsoring organization to the actual sites where meals are supposed to be served. The sponsoring organization is the pass-through. In Feeding Our Future’s case, the sponsoring organization was a 501(c)(3) public charity with a recognized board of three.

Three names on the charter. Two of whom testified at trial that they did not know they were directors.

A board is the substantive accountability that the law assumes is there when it grants a nonprofit corporation tax-exempt status and lets it stand between the federal treasury and a meal site. Not an HR formality. Not three names you pull from a phonebook to satisfy the Secretary of State. The law gives a 501(c)(3) organization the privilege of being treated as a charitable entity rather than as a person with a checking account, and, in return, it requires an oversight body that actually oversees. Take the board out, leave the form in place, and the architecture still looks correct from a hundred yards. The state agency sends the reimbursements. The Form 990s get filed. Until they don’t.

What a real board looks like

I sit on two nonprofit boards. I’m treasurer of one of them. So I know what a treasurer actually does, and what a treasurer’s report looks like.

A treasurer reads bank statements. A treasurer compares the bank statements to the books. A treasurer asks why the line item for one program went from $300,000 to $30 million in eight months. A treasurer notices that operating revenue jumps 2,800 percent, but there are not 2,800 percent more meals being served. A treasurer asks the question and refuses to leave the room until somebody answers it. And when no one will, a treasurer calls the state attorney general’s charities division and asks for a callback.

Jamie Phelps, listed as Feeding Our Future’s treasurer, was a mechanic from Eagan who told the federal court he had never attended a board meeting (Sahan Journal, 2025). I have never met him. I have no opinion on whether he is a good or a bad man. I do have an opinion on what it means to put a stranger’s name on a treasurer’s line to satisfy a state incorporation requirement. It means the law’s substantive expectation of fiduciary care has been performed, but not done.

The performed version costs nothing. The done version is hard. You have to read documents you would rather not read. You have to ask questions in front of people who would rather you didn’t. You have to be willing to be the wet blanket at the meeting and the one who gets called paranoid afterward in the parking lot. You have to keep doing it for years, for free, since that is what the IRS letter you accepted on the organization’s behalf said you would do.

Read more

I never got out to the theater to see Project Hail Mary. I loved the book. I just didn’t have the time. I saw last night, in my Apple TV app, that it was available to purchase for streaming. So I did. Purchase it. Then I went about doing laundry while the movie played on top of the dryer while I folded clothes.

Eva Stratt, in the movie, Project Hail Mary, says that believing in God beats the alternative. People are reading the line as quiet faith. That misreads it.

The scene is not in Andy Weir’s 2021 novel. Phil Lord and Christopher Miller’s 2026 film adaptation added it (O’Hare, 2026; Curtis, 2026). Sandra Hüller’s Stratt, a woman who has spent the runtime authorizing nuclear detonations in Antarctica and ordering a scientist to a one-way orbit around a dying sun, lets the question sit, half-smiles, and answers like a person who has run the math.

It beats the alternative.

Half the people I have heard quote that line have heard it as a confession of quiet faith. I do not think that is what Stratt was saying.

What the line actually does

Stratt is not making a metaphysical claim. She is making an operational one. Belief in God, on her account, performs better than the substitute the project requires of her. She does not say God exists. She says she would rather believe than not, and she would rather not explain it further.

That is a serviceable position. It has a long résumé in philosophy. William James, in his 1896 lecture “The Will to Believe,” argued that under conditions of “live, forced, momentous” choice, where staying agnostic is itself a choice with consequences, a person can rationally elect to believe even before the evidence has settled the question (James, 1896). Blaise Pascal’s wager, in the Pensées published posthumously in 1670, runs along the same coastline. Both arguments concede that the evidence does not settle the question. Both treat the act of believing as something a rational person can elect when the alternatives have been weighed.

That is the line Stratt is borrowing from. She is not Augustine. She is not even the Sunday-morning person down the row from me at Springdale Lutheran. She is a competent, cornered woman picking a frame to operate under, and the frame she picks has lower long-term overhead than the one without God in it.

The alternative isn’t what people think it is

The trouble with “it beats the alternative” is that the alternative depends on the speaker.

For a certain kind of religious person, the alternative to belief is despair. A meaningless universe. A morality without ground. The dread of being a small animal in a bigger animal’s mouth. Belief, set against that floor, looks like a reasonable upgrade.

For a certain kind of atheist, the alternative to belief is not despair. It is the relief of not having to argue with a deity about why this should have happened. It is the freedom of treating suffering as bad without filing it under a divine plan. The alternative to belief is, for them, a different kind of peace.

So whose alternative are we measuring belief against?

The interesting case is the third one. The case where the alternative is not despair and is not relief. The case where the alternative is honesty about both.

Hart and the wave

In December 2004, a fault under the Indian Ocean slipped, and a wave climbed the Sumatran coast and killed roughly 227,898 people across fourteen countries. The Eastern Orthodox theologian David Bentley Hart published a short book in response the following year. He rejected two answers to the wave. The first was the atheist answer that used the disaster as proof of no God. The second, which he treated more harshly, was the theological answer that folded the dead into a redemptive plan that required them.

“In another and ultimate sense,” Hart wrote in The Doors of the Sea (2005), “suffering and death, considered in themselves, have no true meaning or purpose at all.”

That sentence is the hinge for me.

What Hart is saying is that the cheap version of “belief beats the alternative” is the version that uses God as a meaning-machine, conscripted to extract redemption from every drowning. If that is what belief in God is for, then belief loses the contest. The atheist who refuses to pretend the wave was a message is being more honest about the wave than the theologian who pretends it was.

The belief that beats the alternative, on Hart’s terms, is the one that names suffering as the enemy of God’s intention rather than as God’s instrument. The belief that beats the alternative is the one that, when the wave comes, picks up the wounded.

That is not the belief most people on this continent mean when they invoke God on a marquee.

The skeptic in the pew

I am still on the rolls at a Lutheran congregation in rural South Dakota. I was baptized at eight weeks old. I was confirmed at thirteen. I still volunteer on the audio-visual team. When the pastor preaches, the livestream works because I made sure of it.

I am also, on a lot of Sundays, the person in the pew with my head in my hand trying not to shake it at the recited words.

I am at peace with not knowing what God is. I am not at peace with what human beings have done with God. I am specifically not at peace with the marquee on Minnesota Avenue that taught me, over six years of driving past it on my way home from coaching, that its God is built to identify and exclude the woman I am.

If I had answered Ryland Grace’s question across that briefing-room table, I would not have said it beats the alternative. I would have said it depends on which alternative you mean. The alternative of pretending the world is a mechanical accident with no reverence in it, I do not want. The alternative of pretending the world is a managed redemption project in which my queerness, or someone’s drowning, is part of the curriculum, I do not want either.

What I want is a third position. I want belief that does not lie about the cost.

What actually beats the alternative

Stratt’s line works because it does not promise much. It does not promise heaven. It does not promise a plan. It promises a working stance from which a person can keep ordering the impossible things the survival of the species requires of her.

Read that way, yes. It beats the alternative.

But the alternative it beats is not unbelief. It is the kind of belief that pretends. The kind of belief that lets the wave become a sermon and lets the marquee become a doctrine. The kind of belief that uses God as a shortcut around the unbearable arithmetic of being a person who has to decide.

Stratt does not do that. Stratt is, for the length of one half-smile, a credible Lutheran. She holds the question. She does not solve it. She keeps the work. She does not pretend the work is anyone’s plan but her own.

The rest of us, on the side of the briefing-room screen where the camera is, are allowed the same posture. We can believe in something larger than ourselves without claiming to know its wishes. We can light a candle without filing the candle under a redemptive plan. We can sit in a pew and shake our heads at the recited words and stay anyway, because the people in the pew are our people and the work in the building is real.

That is what beats the alternative.

Not the comfort. The honesty.

Author Note:

Grace Ann Hansen is an independent researcher and writer, and an MBA & PhD graduate student in health informatics and artificial intelligence. She is also a published author, a professional musician, a gymnastics coach, and a queer transgender woman living in Sioux Falls, South Dakota. She corrects all her papers and articles with Grammarly, because even though she has deep thoughts, she has shallow patience for punctuation. She uses Anthropic’s Claude in Research mode for source location and verification on cited factual claims; all interpretation, argument, and prose are her own. Correspondence concerning this article should be addressed to Grace Ann Hansen at grace@graceannhansen.com.

References

Curtis, H. (2026, April 13). Project Hail Mary can’t outrun the God hypothesis. Science and Culture Today.

Hart, D. B. (2005). The Doors of the Sea: Where Was God in the Tsunami? Eerdmans.

James, W. (1896). The Will to Believe. Project Gutenberg.

O’Hare, K. (2026, March). ‘Project Hail Mary’: Thrilling sci-fi that’s full of grace. Patheos.

Pascal, B. (1670). Pensées. Project Gutenberg.

Weir, A. (2021). Project Hail Mary. Ballantine Books.

I never got out to the theater to see Project Hail Mary. I loved the book. I just didn’t have the time. I saw last night, in my Apple TV app, that it was available to purchase for streaming. So I did. Purchase it. Then I went about doing laundry while the movie played on top of the dryer while I folded clothes.

Eva Stratt, in the movie, Project Hail Mary, says that believing in God beats the alternative. People are reading the line as quiet faith. That misreads it.

The scene is not in Andy Weir’s 2021 novel. Phil Lord and Christopher Miller’s 2026 film adaptation added it (O’Hare, 2026; Curtis, 2026). Sandra Hüller’s Stratt, a woman who has spent the runtime authorizing nuclear detonations in Antarctica and ordering a scientist to a one-way orbit around a dying sun, lets the question sit, half-smiles, and answers like a person who has run the math.

It beats the alternative.

Half the people I have heard quote that line have heard it as a confession of quiet faith. I do not think that is what Stratt was saying.

What the line actually does

Stratt is not making a metaphysical claim. She is making an operational one. Belief in God, on her account, performs better than the substitute the project requires of her. She does not say God exists. She says she would rather believe than not, and she would rather not explain it further.

That is a serviceable position. It has a long résumé in philosophy. William James, in his 1896 lecture “The Will to Believe,” argued that under conditions of “live, forced, momentous” choice, where staying agnostic is itself a choice with consequences, a person can rationally elect to believe even before the evidence has settled the question (James, 1896). Blaise Pascal’s wager, in the Pensées published posthumously in 1670, runs along the same coastline. Both arguments concede that the evidence does not settle the question. Both treat the act of believing as something a rational person can elect when the alternatives have been weighed.

That is the line Stratt is borrowing from. She is not Augustine. She is not even the Sunday-morning person down the row from me at Springdale Lutheran. She is a competent, cornered woman picking a frame to operate under, and the frame she picks has lower long-term overhead than the one without God in it.

The alternative isn’t what people think it is

The trouble with “it beats the alternative” is that the alternative depends on the speaker.

For a certain kind of religious person, the alternative to belief is despair. A meaningless universe. A morality without ground. The dread of being a small animal in a bigger animal’s mouth. Belief, set against that floor, looks like a reasonable upgrade.

For a certain kind of atheist, the alternative to belief is not despair. It is the relief of not having to argue with a deity about why this should have happened. It is the freedom of treating suffering as bad without filing it under a divine plan. The alternative to belief is, for them, a different kind of peace.

So whose alternative are we measuring belief against?

The interesting case is the third one. The case where the alternative is not despair and is not relief. The case where the alternative is honesty about both.

Hart and the wave

In December 2004, a fault under the Indian Ocean slipped, and a wave climbed the Sumatran coast and killed roughly 227,898 people across fourteen countries. The Eastern Orthodox theologian David Bentley Hart published a short book in response the following year. He rejected two answers to the wave. The first was the atheist answer that used the disaster as proof of no God. The second, which he treated more harshly, was the theological answer that folded the dead into a redemptive plan that required them.

“In another and ultimate sense,” Hart wrote in The Doors of the Sea (2005), “suffering and death, considered in themselves, have no true meaning or purpose at all.”

That sentence is the hinge for me.

What Hart is saying is that the cheap version of “belief beats the alternative” is the version that uses God as a meaning-machine, conscripted to extract redemption from every drowning. If that is what belief in God is for, then belief loses the contest. The atheist who refuses to pretend the wave was a message is being more honest about the wave than the theologian who pretends it was.

The belief that beats the alternative, on Hart’s terms, is the one that names suffering as the enemy of God’s intention rather than as God’s instrument. The belief that beats the alternative is the one that, when the wave comes, picks up the wounded.

That is not the belief most people on this continent mean when they invoke God on a marquee.

The skeptic in the pew

I am still on the rolls at a Lutheran congregation in rural South Dakota. I was baptized at eight weeks old. I was confirmed at thirteen. I still volunteer on the audio-visual team. When the pastor preaches, the livestream works because I made sure of it.

I am also, on a lot of Sundays, the person in the pew with my head in my hand trying not to shake it at the recited words.

I am at peace with not knowing what God is. I am not at peace with what human beings have done with God. I am specifically not at peace with the marquee on Minnesota Avenue that taught me, over six years of driving past it on my way home from coaching, that its God is built to identify and exclude the woman I am.

If I had answered Ryland Grace’s question across that briefing-room table, I would not have said it beats the alternative. I would have said it depends on which alternative you mean. The alternative of pretending the world is a mechanical accident with no reverence in it, I do not want. The alternative of pretending the world is a managed redemption project in which my queerness, or someone’s drowning, is part of the curriculum, I do not want either.

What I want is a third position. I want belief that does not lie about the cost.

What actually beats the alternative

Stratt’s line works because it does not promise much. It does not promise heaven. It does not promise a plan. It promises a working stance from which a person can keep ordering the impossible things the survival of the species requires of her.

Read that way, yes. It beats the alternative.

But the alternative it beats is not unbelief. It is the kind of belief that pretends. The kind of belief that lets the wave become a sermon and lets the marquee become a doctrine. The kind of belief that uses God as a shortcut around the unbearable arithmetic of being a person who has to decide.

Stratt does not do that. Stratt is, for the length of one half-smile, a credible Lutheran. She holds the question. She does not solve it. She keeps the work. She does not pretend the work is anyone’s plan but her own.

The rest of us, on the side of the briefing-room screen where the camera is, are allowed the same posture. We can believe in something larger than ourselves without claiming to know its wishes. We can light a candle without filing the candle under a redemptive plan. We can sit in a pew and shake our heads at the recited words and stay anyway, because the people in the pew are our people and the work in the building is real.

That is what beats the alternative.

Not the comfort. The honesty.

Author Note:

Grace Ann Hansen is an independent researcher and writer, and an MBA & PhD graduate student in health informatics and artificial intelligence. She is also a published author, a professional musician, a gymnastics coach, and a queer transgender woman living in Sioux Falls, South Dakota. She corrects all her papers and articles with Grammarly, because even though she has deep thoughts, she has shallow patience for punctuation. She uses Anthropic’s Claude in Research mode for source location and verification on cited factual claims; all interpretation, argument, and prose are her own. Correspondence concerning this article should be addressed to Grace Ann Hansen at grace@graceannhansen.com.

References

Curtis, H. (2026, April 13). Project Hail Mary can’t outrun the God hypothesis. Science and Culture Today.

Hart, D. B. (2005). The Doors of the Sea: Where Was God in the Tsunami? Eerdmans.

James, W. (1896). The Will to Believe. Project Gutenberg.

O’Hare, K. (2026, March). ‘Project Hail Mary’: Thrilling sci-fi that’s full of grace. Patheos.

Pascal, B. (1670). Pensées. Project Gutenberg.

Weir, A. (2021). Project Hail Mary. Ballantine Books.

A word is a container. Pour one thing into it for a hundred years, then pour another, and most people will keep reading the label and miss the swap.

On the weekend of May 16, 2026, Maureen Galindo, the Democratic candidate in the runoff for Texas’s redrawn 35th Congressional District, posted to her campaign Instagram a multi-slide attack on her opponent. Referring to herself in the third person, she wrote that “She’ll turn Karnes ICE Detention Center into a prison for American Zionists and former ICE officers for human trafficking.” The next sentence, in parentheses, added that “It will also be a castration processing center for pedophiles, which will probably be most of the Zionists.” The San Antonio Current ran it under a plain-English headline: house candidate pledges to send “American Zionists” to internment camp. The Jewish Telegraphic Agency, the Washington Times, and Mediaite carried it. A senior Texas Democrat, James Talarico, said he won’t campaign with her. The Federal Election Commission still lists her as a registered candidate. Early voting in her runoff was already underway when she posted it. She came in first in the four-way March primary with 29.2 percent of the vote on a fundraising base of under ten thousand dollars.

I want to talk about the container.

Read more

Last Friday night, my daughter flew home from California to attend our niece’s high school graduation. My wife and I had discussed who should pick her up. Jill had already left work and said she would do it since she was already out in the car. The plane landed at 9:21 PM. It was 11:30 PM when I checked the “Find My” app, and found that my family had gone to a late dinner, and nobody told me. Maybe they figured I was busy. Or maybe they just didn’t think of me. I was home, in a chair, at my desk, writing, not anything important at the moment. Just writing, which is what I do when I’m by myself. With my phone facing me on the little magnetic charger to my right. Nobody called or texted.

That is the whole story. There is no betrayal in it, no slammed door, no scene. A small group of people I love decided on a meal, and the thought of me did not come up. When I found out, it landed in my chest like a stair you misjudge in the dark. I want to tell you why a missed invitation hurts like a physical thing, and why, after a year of reading the science on it, I have stopped believing the line everyone has handed me my whole life, which is that I am simply too sensitive.

I am not too sensitive. My instrument is fine. The room is the problem.

The thing nobody calls a wound

Start with the body, since that is where this lives. The largest review of the evidence, pooling seventy studies and more than three million people, found that loneliness and social isolation raise your risk of dying early at a rate that sits next to smoking and beats obesity (Holt-Lunstad et al., 2015). Not your feelings about dying early. Your actual odds. The body does not file a claim for social disconnection under “hurt feelings.” It files it under “threat to the organism,” and it bills for it, in cortisol and bad sleep and a heart that ages faster than the calendar says it should.

So when I tell you the dinner felt like an injury, I am not reaching for a metaphor. I am reporting a reading off a gauge that evolution installed long before anyone invented the polite fiction that words can’t hurt you.

Here is the part that took me longest to accept. The science says the kind of exclusion that wrecks you is rarely the dramatic kind. Kipling Williams spent decades studying ostracism and built a model around a simple, brutal finding: being ignored threatens four things at once, your sense of belonging, your self-worth, your sense of control, and your sense that your existence means something (Williams, 2009). The effect is enormous and almost impossible to talk yourself out of. People get the full hit even when they’re told a computer is excluding them; they know it’s a computer.

And the ambiguous version, the forgotten-not-rejected version, the my-family-just-didn’t-think-of-me version, is often worse than a clean no. A clean no, your mind can process and bury. An ambiguous one runs in a loop. Did they forget? Are they angry? Was it me? Each answer means something different about the rest of your life, so the mind keeps all the tabs open and lets them drain the battery. I have decades of these. Each one, alone, was plausibly an oversight. Together, they are a pattern, and my brain was correct to detect it. That is not paranoia. That is arithmetic.

The airport, and the dignity of being asked

A few months back, I drove to the airport at night to pick up the same daughter. I did the task. The task went fine. Afterward, the son and the daughter peeled off to meet friends, and I went home, since nobody mentioned an outing to me until it was already happening to other people.

Someone will say, reasonably, that I’d had a long night and the lights and noise of a midnight outing to a bar or restaurant would have wrecked me, and I probably wouldn’t have wanted to go. They might be right. I am autistic, diagnosed at fifty-something after a lifetime of being told I was just difficult, and a loud room after a long day is a genuine cost for me. But notice what that reasoning does. It decides, on my behalf, that I would have said no, and then skips the part where I get asked. Dignity is not in the meal or meeting up with people. The dignity is in being the one who declines. Baumeister and Leary, in a paper published thirty years later, defined the need to belong not as contact but as frequent care from people inside an ongoing bond (Baumeister & Leary, 1995). Being invited and saying no feeds that. Not being invited starves it, even when the no would have been mine.

CringeGar

My youngest twin son once decided I was funny. Not funny, like she tells good jokes. Funny like a thing you point at. He coined a name for me, a mashup of my name and the word cringe, and he was good at it, the way kids are good at the one cruelty that works. He got a whole scout troop of boys to use it. For about a year, I was that name to a room full of teenagers, and the consensus, including from the adults who heard it, was that it fit. It was good fun. I was the one who couldn’t take a joke.

When I got upset, the upset became the offense. This is the oldest move in an invalidating house: hit someone, and when they flinch, the flinch is the thing that ruined the evening. There is research on exactly this, on the experience of being chronically misread and then blamed for your own reaction. It names that pattern as a real predictor of suicidality in autistic adults (Cassidy et al., 2018). I am not citing that to win an argument with my kid. I’m citing it since, for years, I thought the wrongness was a personal failing, and it turns out it has a literature.

Damian Milton gave the failing back its real name. The old story said autistic people can’t read other people’s minds, or even read the room. Milton’s double empathy problem holds that the misreading runs both ways, that I am no worse at reading my son than my son is at reading me. That the only reason the deficit gets filed under my name is that the people holding the diagnostic pen are not autistic (Milton, 2012). This is not a soft idea. When you put autistic people in a chain and have them pass a story down the line, the story survives at the same rate it does among non-autistic people. It degrades in the mixed chains, the ones with both (Crompton et al., 2020). The breakage is in the translation, not in me. My son’s teasing read, to him and to the room, as light. It landed on me as contempt. Both readings are real. The asymmetry is that only mine gets called a misread.

The son who went quiet

The older twin son, by a minute, voting in his first presidential election, voted for the administration that has spent this last year working through a hit list of my healthcare, my documents, my potential domestic terror profile, and my ability to exist in public. After my initial outburst, I was told that I shouldn’t assume he understood the stakes. So, one night, sitting on the bed my wife and I sleep in each night, I explained to him, plainly, the way you explain something to a person you raised and still hope thinks well of you. He has been distant from me since.

Pauline Boss built her whole career on the kind of loss this is. She calls it an ambiguous loss, and she splits it into two. There is a physical absence with the person still psychologically present, the missing soldier, the disappeared. And there is the other one, physical presence with the person psychologically gone (Boss, 1999). My son is alive. He is in the family group chat. He left a thumbs-up on a thread a few weeks ago. He has not said a word to me since the election that isn’t for something he needs. “Can you help me set up the internet in my apartment?” “Can you help me finish my taxes?” He is, in her exact phrase, here and not here. Boss’s finding, the one that reorganized how I think about this, is that ambiguous loss is the hardest grief there is, harder than a death, since the mind cannot close a loop that the person keeps half-open by still technically being there.

Stack onto that what we know about family rejection for people like me. The cleanest study, a national sample of trans adults, found that high family rejection more than tripled the odds of a lifetime suicide attempt (Klein & Golub, 2016). Not friend rejection. Not stranger cruelty, which I can almost set my watch by. Family. The math on autistic trans women is its own quiet catastrophe, since the two run together far more often than chance would predict (Warrier et al., 2020), which means a lot of us are carrying both stacks of stones up the same hill.

The relief and the grief that came in the same envelope

I was diagnosed autistic in my fifties. The women who get found this late tend to get found the same way I did, after a lifetime of passing, of building a person out of careful imitation and running her in public until the battery dies (Bargiela et al., 2016). The masking has a cost, the research now measures anxiety, depression, and exhaustion (Hull et al., 2017). When the diagnosis finally came, it arrived as two things in one envelope. Relief, since the lifelong wrongness had a name and the name was not “moral failure.” And grief, for the kid nobody accommodated, the one who got told she had boy germs in kindergarten and believed it meant something was contaminated in her, rather than something was wrong with the kid doing the telling.

There was a phone call about six years ago when my mother had a computer issue, and I was the one who could help her. At the end of the troubleshooting session, she signed off by saying “thank you, sir,” and hung up. I cried for twenty minutes over one syllable. There was a Father’s Day when every person in the room used the old name all afternoon, and the reason given, the one I have been handed for years, was that nobody was ready yet. I have stopped waiting for people to get ready. People who never try do not get ready. They just stay not-ready, comfortably, as you do the bleeding.

In 2019, I tried to end my life. I did not, obviously. I have written about that day before, and I am not going to dress it up here. What I will say is that the version of me who did that believed the wrongness was hers. The science I’ve been reading this past year says it wasn’t, and I find myself believing the science.

The instrument is fine.

Here is where it lands, and it is not the warm place you might expect, since the warm version would be a lie, and you’ve all had enough of those.

The lonely brain gets accused of distortion. The story goes that loneliness makes you see a threat that isn’t there, rejection in a neutral face, a snub in an honest oversight. For a lot of lonely people, there is something to that. But run the tape on my life. The girl who said I had boy germs was not a misread. The scout troop chanting a cruel name was not a misperception. The executive administration working through its list is not a figment. The son’s silence is not in my head, since I can see the thumbs-up he left for everyone but me. My threat detector is not malfunctioning. It has been carefully calibrated for over fifty years by an environment that kept proving it right.

That is the conclusion, and I want it stated clearly, since every soft person in my life has spent decades trying to talk me out of the only thing I know to be true about myself. The problem was never my perception. The problem is what I have been correctly perceiving.

The same research that delivered that grim little fact delivered the one piece of mercy I trust too, since it had to fight its way past the grim part to get to me. Across the literature on lonely kids, queer mental health, autistic wellbeing, trans suicide, the protective effect of a single durable relationship with someone who actually sees you is large enough that some clinicians just call it the one-person rule. Not the whole family is fixed. Not the silent son returned. One person, real, ongoing, safe. I have spent my life trying to be invited to every dinner. The evidence says I needed one chair, not the whole table, and I was looking at the wrong piece of furniture.

I am a skeptical agnostic Lutheran, which means I distrust tidy endings and the people selling them, including, frequently, the people in my own pews. So I won’t give you one. I am still lonely. One son still gaslights me for his amusement, and the other son is still quiet. The math is still the math. But I have put down the one belief that was doing me the most damage, the one where my instrument was the broken thing. It was never the instrument. It was the room. Knowing the difference does not fix the room. It does, on a good day, stop me from taking the room’s word for what I’m worth.

If you read this and recognized your own chair sitting empty, the part of me that did 2019 wants you to know there is a number, 988, and a person at the other end of it, and that calling it is not the dramatic thing the movies made it. It is closer to telling someone the truth out loud for the first time, which, in my experience, is exactly what started it all.

Author Note

Grace Ann Hansen is an independent researcher and writer, and an MBA and PhD graduate student in health informatics and artificial intelligence. She is a published author, a professional musician, a gymnastics coach, and a queer transgender woman living in Sioux Falls, South Dakota. She corrects all her papers and articles with Grammarly, since though she has deep thoughts, she has shallow patience for punctuation. She uses Anthropic’s Claude in Research mode for source location and verification on cited factual claims; all interpretation, argument, and prose are her own. Correspondence concerning this article should be addressed to Grace Ann Hansen at grace@graceannhansen.com.

References

Bargiela, S., Steward, R., & Mandy, W. (2016). The experiences of late-diagnosed women with autism spectrum conditions: An investigation of the female autism phenotype. Journal of Autism and Developmental Disorders, 46(10), 3281–3294. https://doi.org/10.1007/s10803-016-2872-8

Baumeister, R. F., & Leary, M. R. (1995). The need to belong: Desire for interpersonal attachments as a fundamental human motivation. Psychological Bulletin, 117(3), 497–529. https://doi.org/10.1037/0033-2909.117.3.497

Boss, P. (1999). Ambiguous loss: Learning to live with unresolved grief. Harvard University Press. https://www.ambiguousloss.com/about/

Cassidy, S., Bradley, L., Shaw, R., & Baron-Cohen, S. (2018). Risk markers for suicidality in autistic adults. Molecular Autism, 9, 42. https://doi.org/10.1186/s13229-018-0226-4

Crompton, C. J., Ropar, D., Evans-Williams, C. V. M., Flynn, E. G., & Fletcher-Watson, S. (2020). Autistic peer-to-peer information transfer is highly effective. Autism, 24(7), 1704–1712. https://doi.org/10.1177/1362361320919286

Holt-Lunstad, J., Smith, T. B., Baker, M., Harris, T., & Stephenson, D. (2015). Loneliness and social isolation as risk factors for mortality: A meta-analytic review. Perspectives on Psychological Science, 10(2), 227–237. https://doi.org/10.1177/1745691614568352

Hull, L., Petrides, K. V., Allison, C., Smith, P., Baron-Cohen, S., Lai, M.-C., & Mandy, W. (2017). “Putting on my best normal”: Social camouflaging in adults with autism spectrum conditions. Journal of Autism and Developmental Disorders, 47(8), 2519–2534. https://doi.org/10.1007/s10803-017-3166-5

Klein, A., & Golub, S. A. (2016). Family rejection as a predictor of suicide attempts and substance misuse among transgender and gender nonconforming adults. LGBT Health, 3(3), 193–199. https://doi.org/10.1089/lgbt.2015.0111

Milton, D. E. M. (2012). On the ontological status of autism: The “double empathy problem.” Disability & Society, 27(6), 883–887. https://doi.org/10.1080/09687599.2012.710008

Warrier, V., Greenberg, D. M., Weir, E., Buckingham, C., Smith, P., Lai, M.-C., Allison, C., & Baron-Cohen, S. (2020). Elevated rates of autism, other neurodevelopmental and psychiatric diagnoses, and autistic traits in transgender and gender-diverse individuals. Nature Communications, 11, 3959. https://doi.org/10.1038/s41467-020-17794-1

Williams, K. D. (2009). Ostracism: A temporal need-threat model. Advances in Experimental Social Psychology, 41, 275–314. https://doi.org/10.1016/S0065-2601(08)00406-1

The chairs are already out. Somewhere this month, a field house is filling with families holding programs they’ll fold into fans, since the ventilation in a building designed for basketball was never meant for two thousand people in gowns in June. A name gets read. A young woman in a mortarboard that doesn’t quite sit right crosses a stage, takes a hand, moves the tassel left to right, and the entire ceremony exists to certify one sentence: she learned what we taught her, and we are vouching for it.

I want to talk to that young woman, and to the thousands like her this season, about the one seminar that wasn’t on her transcript.

I coach boys’ gymnastics, so I think about certification differently than most people in that field house do. The sport I teach, six events of it, floor, pommel horse, vault, parallel bars, still rings, and high bar, is under the chalk and the calluses and the medals a multi-year course in falling on purpose. Before a kid does anything you’d pay to watch on the high bar, he spends a year learning what his body should do at the top of a release when his hands miss the bar, which they will, repeatedly, in front of an audience. He learns the same thing flying off the end of the pommel horse and coming down short off the rings. There’s no medal for that part. Nobody’s grandfather drives four hours to watch a twelve-year-old boy finally stop reaching for the floor with a stiff arm and instead tuck and roll through a missed dismount the way I’ve told him six hundred times. But that skill, the ugly one, the one with no ceremony attached, is the one that keeps him in the gym long enough to earn the pretty ones. The medals are a lagging indicator of how well a boy learned to fall.

I’m also a sixty-year-old graduate student. Two degrees in progress, an MBA and a PhD, both started decades after the age when a transcript says you’re supposed to start them. So I’ve sat in a lot of classrooms recently, and I can report with some authority on what is not in any university catalog in this country. There is no required senior seminar in abject failure. There is no syllabus for the thing that is actually going to happen to the graduate in the gown, somewhere in her first eighteen months out: the launch that craters, the model that was confidently wrong, the organization she joined for looking unsinkable, which turns out to have been quietly taking on water the whole time she was studying how to get hired by it.

So consider this the makeup course. It’s pass/fail; the reading covers the last twelve months, and it’s long since the last twelve months were generous. One detail, before the syllabus, that I want you to hold onto for the whole hour: every single failure on this list was committed by people who had the credentials. Not by dropouts. Not by amateurs working out of a garage. By the certified, the audited, the board-approved, the Ivy-trained, the people with exactly the kind of diploma the young woman in the gown is about to receive. That’s not incidental to the course. That is the course.

Unit One: The diploma certifies the paperwork, not the truth underneath it

Start where the credential is densest, because that’s where the rot was best hidden.

In September 2025, a company called First Brands filed for bankruptcy. You have never heard of First Brands, but you have held its products. It makes the FRAM oil filter, the Raybestos brake pad, the Anco wiper blade, and the unglamorous parts in the bin at the auto store. When it filed, it listed liabilities of somewhere between ten and fifty billion dollars against a business that booked roughly five billion in sales a year. Sit with that ratio for a second. Then consider that federal prosecutors went on to indict the founder and his brother on conspiracy, wire and bank fraud, and money laundering, alleging that the same collateral had been pledged to more than one lender at the same time. By February 2026, the company had filed notices to close four Ohio plants and cut 1,267 jobs.

Six weeks before First Brands filed, a subprime auto lender, Tricolor, collapsed into Chapter 7 on the same trick. The U.S. Attorney for the Southern District of New York later unsealed indictments against the founder and three other executives and called the company, in plain English, a “financial crimes enterprise,” saying fraud “became an integral component of Tricolor’s business strategy.” JPMorgan took a $170 million charge between the two failures. Jamie Dimon, asked about it, said the line that traveled: “When you see one cockroach, there are probably more.”

Now, the people in this field house, with the finance concentrations, the brand-new CPA results, and the offer letters from firms whose entire product is the word trust: this unit is addressed to you specifically, so lean in.

Every one of those deals had paperwork. Audited financials. Credit ratings. Loan tapes reviewed by warehouse lenders whose job, whose only job, is to look at the loan tapes. The certifications were issued by credentialed professionals who do exactly what their credentials certify they can do. And the fraud went through for years anyway, until something unrelated stumbled and the lights came on by accident. The audit didn’t catch it. The audit dressed it. Your diploma will accurately certify that you can produce the document. It will not certify that the document is true, and it will offer you no protection at all from being the person whose name is on it when the truth shows up late and angry.

The case to sit with the longest here isn’t even the fraud. It’s 23andMe. A company that, at its peak, had collected the genetic information of more than fifteen million people, around eighty percent of whom had agreed to let it be used for research, including myself, my spouse, and all four of my children, two biological and two adopted, in order for our family to understand more about where we all came from. It filed for bankruptcy in March 2025. More than two dozen state attorneys general objected to what happened next on the entirely sane ground that a human genome is not the same category of object as a forklift. The court ran the auction anyway. The DNA sold for $305 million to a nonprofit run by the same executive who’d run the company into the ground in the first place. The legal instrument that moved fifteen million genomes was Section 363 of the Bankruptcy Code, the ordinary, unremarkable “sell it free and clear” provision. The privacy ombudsman, as provided for in the law in a case like this, could write a report. He could not stop the sale.

Nothing malfunctioned. That’s the part I need the graduate in the gown to understand before she signs anything. The bankruptcy system performed flawlessly. It took a distressed asset and found it a buyer using the exact procedure it would use for a warehouse lease, since the people who built that procedure, all of them credentialed, never imagined feeding fifteen million genomes through it. An institution is a procedure with a building around it. A procedure does not have a category called unthinkable. It will process the unthinkable with precisely the same calm it brings to the routine, and it will generate clean paperwork the entire time.

Unit Two: Redundancy is not safety, and the backup is usually a second copy of the problem

The first thing any competent field teaches you is to have a backup. Two engines. Two data centers. A spare. The 2025 to 2026 academic year was, among other things, a long and very expensive seminar in why that sentence is only half finished.

On the night of October 19, 2025, an automated system inside Amazon Web Services deleted the DNS records for one of its own core databases in Virginia. The system that was deleted was a safety system. It had a redundant twin, built for exactly this, so that if one copy failed, the other would take over. The two of them raced each other, and the cleanup routine that was supposed to tidy up after the race deleted the live record instead. For roughly fifteen hours, Snapchat was down. So were Fortnite, Venmo, Ring doorbells, United Airlines’ booking system, and the United Kingdom’s tax website. The redundancy didn’t fail, despite being redundant. The redundancy was the failure. Two safety processes built to protect each other collided, and the safety routine deleted the thing it was supposed to guard.

Three weeks later, Cloudflare did it differently and landed in the same crater. A permission change in the database caused a query to return duplicate rows, doubling the size of a configuration file that Cloudflare’s own bot-detection software loads into memory. The file blew past a memory limit. The memory limit existed for safety. The traffic-routing software started crashing in a loop, and X went down, and ChatGPT, and Spotify, and Discord, and Zoom, and, with a symmetry you could not put in fiction, Downdetector, the website you visit to confirm that the thing that’s down is in fact down. Two weeks after that, Cloudflare went down a third time and conceded that the fixes it had promised after the second outage were not yet fully in place.

A configuration file that exists to tune the system. A memory limit exists to keep the system safe. The tuning broke it, and the limit was the wall it shattered against. When the graduate in the gown builds her own version of this, and she will, the question to ask is never “Is there a backup?” Everyone has a backup. The question almost no one asks, the question that separates the people who survive their own infrastructure from the people who write the postmortem, is: Does the backup fail independently of the thing it’s backing up? If it fails the same way at the same time for the same reason, it was never redundant. It was a duplicate of the single point of failure, and the org paid extra to feel safe about it.

Unit Three: The warning almost always exists; the failure is that no one owns the response

In late January 2025, an Army helicopter and a regional passenger jet collided in the air near Reagan National Airport, and sixty-seven people died. When the National Transportation Safety Board released its final report a year later, the finding that should have ended careers was this: the airspace contained a helicopter route placed too close to a runway approach, and the agency responsible for that airspace had its own data showing 15,214 dangerously close events between helicopters and commercial planes in that exact corridor over three years. Eighty-five of them were genuine near misses. The warning was not missing. The warning was not subtle. The warning had been sitting, fully formed, in a federal database for years, waiting for a person whose job was to read it and act.

You will spend your entire career being told, implicitly, that gathering the information is the work. It is not. Gathering the information is the cheap part, the part that generates dashboards and quarterly decks and the comfortable feeling of diligence. The expensive part, the part with no class grades and no dashboard measures, is the institutional act of looking at something you already know and changing what you do in response. Sixty-seven people died in the gap between the data and the decision. That gap is not empty. It is crowded with competent, credentialed people who all had access to the warning and not one of whom owned the response. Try, for your whole career, not to be a resident of that gap. It has more people in it than you’d ever guess, and most of them have very good degrees.

The Department of Government Efficiency belongs in this unit, and it belongs here as the largest single illustration of the principle in the entire reading. Set aside, for a moment, every argument about whether government should be smaller. That’s a real argument, and it is not this one. This one is about what happens when a credentialed operation decides the warning is meant for others.

DOGE was created by executive order in January 2025, with a promise of $2 trillion in savings. The number fell to one trillion, then to a hundred and fifty billion, then to a public “Wall of Receipts” that, on inspection by reporters with calculators, listed an $8 million contract as $8 billion, counted single contracts two and three times, counted contracts that had ended under the previous administration, and counted credit-line ceilings as if they were realized savings. An independent, nonpartisan analysis estimated that the effort’s actions cost taxpayers roughly $135 billion in a single fiscal year. At the same time, self-reported savings could not be verified by any external auditor, including right-leaning ones, at face value.

But the part for this unit, the part about owning the response, is the firings. DOGE fired, then scrambled within roughly twenty-four hours to rehire several hundred employees responsible for the U.S. nuclear weapons stockpile, some of whom had already lost their email access before anyone could tell them they still had jobs. It fired, then, within forty-eight hours, tried to reverse, with USDA staff working the bird-flu response at the height of an outbreak that had already killed 148 million birds. The warnings here were not buried in a database. They were the literal job titles of the people being fired. Nuclear weapons. Avian influenza. The operation moved fast enough that the title on the badge was not read before the badge was deactivated. The courts, for their part, repeatedly found the operation had acted unlawfully, and the unit quietly ceased to function eight months before its own charter expired, its director’s denial of its demise contradicted by the reporting within a single day.

I’m not telling you this to score a political point. I’m telling you this for a simpler reason. The graduate in the gown will be handed, at some point, a mandate, a deadline, and a feeling of urgency, and the most seductive sentence in professional life is: “We don’t have time to check.” DOGE is what that sentence costs at scale. The warning was the job title. Nobody read the badge.

Unit Four: The correction exists, and it arrives far too late to be a save

There is a particular kind of comfort, especially for the idealistic graduate, in the belief that systems are self-correcting. That if something goes badly enough wrong, a mechanism, a court, a regulator, an election, a guardrail, will engage and set it right. The reassuring thing is that this belief is true. The devastating thing is the timing.

In April 2025, the president imposed the steepest tariffs the United States had levied in more than a century. Global markets fell in the worst week since the early days of the pandemic. The constitutional structure that exists precisely to constrain that kind of unilateral economic action did, in the fullness of time, constrain it: the courts ruled the tariffs illegal, and in February 2026, the Supreme Court affirmed it, six to three. Roughly $166 billion had to be refunded to more than 300,000 businesses. The check worked. It worked for eleven months and one market crash after the damage was inflicted.

The same year produced the longest government shutdown in American history, forty-three days, and then, four months later, a longer partial one, seventy-six days. The federal budget process is itself a safety mechanism. The requirement that elected people affirmatively decide to keep the lights on is supposed to force deliberation. It performed exactly as designed. It produced precisely the deadlock it is structurally capable of producing, twice in one year. Several hundred thousand federal workers went without pay during the gap. Roughly forty million people on food assistance watched their November benefits hang in the balance on an emergency court order.

Here is the lesson, and it is the one your idealism is least likely to accept on graduation day. The presence of a correction in a system does not guarantee that it will arrive in time to matter. A guardrail that engages after the car is already in the ravine is a genuinely good thing to have installed, for the sake of the next car. It is not safe for the car in the ravine. The graduate in the gown should build the kind of work and choose the kind of institutions where correction is fast and close to failure. And she should be ruthlessly honest with herself, for the rest of her career, about how rare that kind of work actually is, and how much of professional life is conducted in systems whose corrections are real, and slow, and arrive for the next person.

The unit I am not going to resolve, since resolving it would be a lie

There is one more, and I am going to break the shape of this seminar to handle it, since keeping the shape would require me to lie to you, and you’ve been lied to enough by the time you reach a stage in a gown.

On June 12, 2025, Air India Flight 171 lifted off from Ahmedabad and came down thirty-two seconds later into the hostel block of a medical college a little over a mile from the runway. Of the 242 people on board, 241 died. One man, in seat 11A, walked out of it. At least nineteen more people died on the ground. It was the first fatal crash and the first hull loss in the history of the Boeing 787, an aircraft that had flown without a fatal accident since 2011. The preliminary report found that both engine fuel cutoff switches moved from RUN to CUTOFF within seconds of liftoff. The cockpit voice recorder caught one pilot asking the other why he had cut off the fuel, and the other saying he hadn’t. A year later, investigators on two continents do not agree on what happened, and the families do not have an answer.

I am not going to extract a tidy lesson from that, because there isn’t one, and a commencement speaker who manufactures wisdom out of two hundred and sixty deaths is doing exactly the thing this whole seminar is warning you against: dressing the absence of an answer in the costume of one. Sometimes the failure does not yield a takeaway on your timeline, your career’s timeline, or your lifetime. Sometimes the most rigorous and honest thing a credentialed person can say is, “We do not yet know,” and the people who deserve to know are still waiting. That sentence is not a failure of analysis. It is the highest form of it. The course does not always pay you back in insight. Sometimes it just takes, and asks you to keep the names in mind as it does, and to refuse the comfort of a moral you didn’t earn.

The final question is one.

So here is what the field house ceremony will not say, which means I have to.

Every failure in this seminar was the work of the successful. I want to be precise about that, since it is the entire argument, and the graduate in the gown is about to walk directly into it. Amazon employs some of the best infrastructure engineers money can buy, and its safety automation deleted its own database. First Brands had audited financials, produced by accountants with the same credentials as half this room is about to earn. The FAA had the data, all 15,214 events of it. The bankruptcy court had decades of settled, expert procedure, and it used that settled craft to auction fifteen million genomes with a clean docket. People staffed DOGE credentialed enough to be trusted with the federal payment system, and they deactivated the badges of the people who guard the nuclear stockpile without reading the badges. In every single case, the thing everyone trusted, the credential, the audit, the redundancy, the procedure, the constitutional check, was either the precise thing that failed or the precise reason no one was watching the thing that did.

Your diploma is real. You earned it, and you should be proud of it, and I mean that without a flake of irony. It certifies, correctly, that you learned what they taught you, and that isn’t nothing; it is, in fact, a great deal. But there is exactly one thing it cannot certify, the one thing no speaker at any podium in any field house this May or June will say into the microphone, and it is the entire content of the senior seminar that wasn’t on your transcript: being good at the success part does not inoculate you against the failure part. It never has. It only changes the shape the failure takes and moves it somewhere your training has specifically taught you not to look.

So look there. That is the whole final exam, one question, and you will retake it for the rest of your working life. When the year goes wrong, and it will, with witnesses, the only thing that will distinguish you from the very credentialed people in this seminar’s reading list is whether you are the kind of professional who keeps her eyes on the tired metal even when the inspection schedule, signed and stamped by someone with your exact degree, says the metal is fine.

Nobody grades that question. There’s no ceremony for getting it right, the same way there’s no medal for the twelve-year-old boy who finally learned to fall off the high bar without breaking himself. It just quietly determines everything that happens after the chairs are folded and put away.

Congratulations. Welcome to the part of the curriculum that was always the point.

For a deeper dive into colossal, abject failure, please check out the research paper that was the basis for this essay:

When Systems Cracked: Monumental Failures From May 2025 to May 2026

Author Note

Grace Ann Hansen is an independent researcher and writer, and an MBA & PhD graduate student in health informatics and artificial intelligence. She is also a published author, a professional musician, a gymnastics coach, and a queer transgender woman living in Sioux Falls, South Dakota. She corrects all her papers and articles with Grammarly, because even though she has deep thoughts, she has shallow patience for punctuation. She uses Anthropic’s Claude in Research mode for source location and verification on cited factual claims; all interpretation, argument, and prose are her own. Correspondence concerning this article should be addressed to Grace Ann Hansen at grace@graceannhansen.com.

Between May 2025 and May 2026, a series of structurally independent failures unfolded across business, technology, government, infrastructure, and finance. A pioneering consumer genomics company sold the DNA of more than 15 million people through bankruptcy court. A $10 billion auto-parts conglomerate and a subprime auto lender collapsed amid fraud indictments alleging multiple-pledged collateral. Amazon Web Services and Cloudflare each disabled large portions of the public internet due to internal automation failures. The longest federal government shutdown in United States history lasted 43 days, followed by a 76-day partial lapse. The Department of Government Efficiency generated documented waste, lost most of its court challenges, fired and rehired nuclear-weapons and avian-influenza staff, and ceased operating eight months before its charter expired. A Boeing 787 crashed for the first time in the type’s service history, a cargo jet destroyed a Louisville neighborhood, and the Iberian Peninsula experienced the first overvoltage-driven blackout on record. This brief synthesizes these events from primary sources, including agency investigations, court filings, and company disclosures, and argues that a recurring pattern unites them: the protective layer, the redundancy, audit, oversight, or constitutional check meant to prevent catastrophe, was itself the mechanism of failure.

Keywords: systemic failure, institutional accountability, financial fraud, infrastructure resilience, government efficiency, technology outages

The twelve months bracketed by May 2025 and May 2026 will be remembered as a period when several long-running fault lines in the global economy, in government, and in the systems we depend on every day finally cracked open at once. A pioneering DNA company was liquidated and its genetic database auctioned off in bankruptcy court. The longest federal government shutdown in U.S. history paralyzed Washington for 43 days. A multi-billion-dollar auto-parts conglomerate collapsed in what prosecutors describe as a multi-year fraud, and a subprime auto lender went down the same month on charges of double-pledging $800 million in collateral. A Boeing 787 Dreamliner crashed for the first time in the type’s commercial history, killing 260 people in Ahmedabad. A UPS cargo jet’s engine sheared off on takeoff, leveling part of an industrial neighborhood in Louisville. Spain and Portugal blacked out simultaneously for the first time. Cloudflare and AWS each took out enormous swathes of the internet. The U.S. president’s “Liberation Day” tariffs triggered the worst week for global equities since the early days of the pandemic before being struck down as illegal by the Supreme Court. And a quasi-agency created to make government more efficient instead generated tens of billions of dollars in waste, lost most of its court cases, fired and re-hired nuclear-weapons and bird-flu staff, and quietly ceased to exist eight months before its own charter expired.

What follows is a research brief organized by category, built first on original documents, court filings, NTSB and AAIB reports, ENTSO-E findings, agency post-mortems, company SEC and bankruptcy disclosures, and reputable original journalism, so a long-form essay can be built from a verifiable record. A consolidated source-verification list (not for publication) follows the brief.

Read more

This paper revises and expands a prior cross-disciplinary synthesis of confirmation bias in the post-2020 information environment. The original argument located the social damage of misinformation at the interaction of three layers: individual-level motivated cognition, partisan media ecosystems, and engagement-ranked algorithmic curation, with each necessary and none sufficient. The revision retains that argument and addresses a substantive analytical weakness. The original case set consisted entirely of episodes in which the empirically supported answer was congenial to a politically liberal reader: the absence of Iraqi weapons of mass destruction, the legitimacy of the 2020 United States election, vaccine safety, the falsity of QAnon, anthropogenic climate change, and the unworkability of sovereign-citizen pseudolaw. That selection allowed a left-leaning reader to finish the paper diagnosed in the abstract and exonerated in the particulars. The revision repairs this by adding a left-coded anchor case (the April 2026 propagation of the false claim that 62 million men attended an “online rape academy,” derived from the March 2026 CNN investigation by Vandoorne, Fox, Kennedy, Stubbs, and Chacon) and by introducing a major new Level 1 section that adjudicates, as neutrally as the evidence allows, whether motivated reasoning is ideologically symmetric or asymmetric. The paper also applies replication skepticism to a finding convenient to the motivated-reasoning thesis: Persson, Andersson, Koppel, Västfjäll, and Tinghög’s (2021) preregistered failure to replicate Kahan, Peters, Dawson, and Slovic’s (2017) motivated-numeracy effect. A second left-coded illustration uses the November 25, 2025, UNODC and UN Women report on intimate-partner and family-member femicides as an instance of motivated numeracy operating in a domain of contested definitions. The synthesis I defend is more specific than the original. At the individual cognitive level, the best current evidence (Ditto et al., 2019; Frimer et al., 2017; Washburn & Skitka, 2018; Guay & Johnston, 2022) is most parsimoniously read as approximate ideological symmetry, with credible asymmetry claims surviving primarily at the levels of media-ecosystem structure (Benkler et al., 2018) and elite incentives rather than at the level of individual rationality. The visible right-coded skew of post-2020 American misinformation is therefore not well explained by an asymmetry in individual cognition; it is better explained by an asymmetry in the structure of the conservative media ecosystem and by the asymmetric behavior of political elites who feed it. This distinction matters for intervention. Ecosystem-level and elite-level interventions are licensed by the evidence in a way that individual-level cognitive remediation premised on liberal exceptionalism is not. The paper closes by integrating implications for generative-AI-mediated information environments, where engagement ranking, model hallucination, and synthetic outrage compound the diagnostic and corrective challenges already documented.

Keywords: confirmation bias, motivated reasoning, ideological symmetry, motivated numeracy, falsehood diffusion, algorithmic curation, media ecosystems, generative AI, replication crisis, femicide

Read more

In a county hospital in West Texas in the spring of 2025, a child died of measles. The first American to die of that disease since 2015. By July, the national case count had reached 1,333, more than four times the previous year, and 92 percent of those who caught it were unvaccinated or had no record of a shot (Centers for Disease Control and Prevention, 2025). Gaines County, near the center of the outbreak, had kindergarten measles coverage as low as 82 percent (Cousins, 2025). Community immunity needs about 95.

The same season, a national survey of roughly 5,800 voters found that 63 percent of Republicans still believed the 2020 election was stolen, four years after every recount, audit, and court said it wasn’t, including the audit run by a firm that went looking for bamboo fibers in the ballots after someone decided the fraudulent votes were flown in from China (Public Religion Research Institute, 2024).

Two different rooms. Same wallpaper. People holding, with total confidence, a belief that the evidence flattens, and not budging when you show them the evidence.

The story everyone tells about why this happens goes like this: social media broke our brains. The algorithm built a funhouse mirror, fed each of us a private reality, radicalized the lonely, and now your uncle thinks the vaccine has a tracking chip. It’s a tidy story. It has a villain with a logo. It implies a fix: tune the algorithm, add a label, and sanity returns.

I want to tell you that the tidy story is mostly wrong, and that the research it was built on largely fell apart when people tried to reproduce it. The thing actually wrecking the shared world is older than the internet, older than cable news, older than me, and it does not have a logo. It has a name from a 1960 psychology paper, and you have it. I have it. The reason this matters is not academic. It changes what we should even be trying to fix.

The Thing Is Old

In 1960, a British psychologist named Peter Wason sat people down with a number sequence, 2, 4, 6, and asked them to figure out the rule that generated it. They could test guesses by proposing their own triples. He’d tell them yes or no, then ask for the rule (Wason, 1960).

Almost everyone did the same thing. They guessed “even numbers going up by two,” then proposed 8, 10, 12. Yes. Then 20, 22, 24. Yes. Then 100, 102, 104. Yes. Confident now, they announced the rule. Wrong. The actual rule was “any three increasing numbers.” They never found it because they never tried to break their own guess. They only ever asked the question whose answer they already expected.

That’s confirmation bias, and it’s worth being precise about what it is, because the loose version (“people believe what they want”) doesn’t align with the research. Klayman and Ha cleaned it up in 1987: it’s a positive test strategy (Klayman & Ha, 1987). We probe our hunches by looking for cases that fit them. This is a perfectly good strategy when your hunch is roughly right. It fails badly, and invisibly, when the truth is wider than your hunch, since every confirming case feels like proof and tells you nothing.

Sixty-five years of work since then, cataloged in Nickerson’s enormous 1998 review and threaded through Kahneman’s fast-and-slow machinery, points the same direction (Nickerson, 1998; Kahneman, 2011). The mind runs cheap. Agreeable information slides in without a fight. Disagreeable information triggers an audit, and the auditor is not neutral: the defendant hired them.

The cruelest finding in this literature, the one that should bother you if you think education is the exit, comes from Dan Kahan. He gave people data to interpret. On a neutral topic, people who were better at math read the data more accurately. On a politically loaded topic, the people who were better at math were more polarized, not less (Kahan, 2013). They used the horsepower to build a better defense of the conclusion they already wanted. Smart does not save you. Smart buys you a nicer lawyer.

None of this needed Facebook. Wason ran his study on a chalkboard.

The Algorithm Did Less Than You Think

Here is the part that gets me in trouble at parties.

The strongest causal evidence we have on whether the feed itself polarizes people comes from a set of experiments that Meta let outside academics run on real Facebook and Instagram users during the 2020 election, published across Science, Nature, and PNAS (Guess et al., 2023; Nyhan et al., 2023; Allcott et al., 2024). These weren’t surveys. They reassigned what real people saw, for months, and measured what happened.

They switched people from the ranked algorithmic feed to a plain reverse-chronological feed for 3 months. It changed the feed a lot. It changed political attitudes almost not at all. They cut people’s exposure to like-minded sources. No measurable polarization effect. They deactivated people’s accounts entirely for six weeks before the election. Small dip in news knowledge, nothing measurable on polarization or who people voted for.

Brendan Nyhan, one of the lead researchers, put it as plainly as a careful scientist can: nobody is saying social media has no negative effects, but these were among the most-discussed interventions, and none of them measurably moved attitudes (Wagner, 2023).

It goes back further. The famous filter-bubble panic, the Eli Pariser idea that the algorithm seals each of us in a private echo chamber, never had the empirical goods. Bakshy and colleagues looked at 10 million Facebook users back in 2015 and found that the algorithm trimmed cross-cutting content a little, while people’s own clicking trimmed it much more (Bakshy et al., 2015). We are not trapped in the bubble. We are building it by hand, and we like it in here.

YouTube was supposed to be the radicalization pipeline, the auto-play conveyor belt to extremism. Then Hosseinmardi and colleagues looked at actual behavioral data and found that the people watching far-right content on YouTube were already seeking it everywhere else too, and mostly arrived through external links and searches, not the recommender, which “quickly forgets” extreme viewing if you go back to normal stuff (Hosseinmardi et al., 2021). The pipeline mostly carried people who’d already bought the ticket.

I’m not telling you the algorithm is fine. I’ll get to what it does. I’m telling you that the specific, popular, satisfying claim that the recommendation engine reaches into a normal person and manufactures a conspiracist is the part of the story with the weakest evidence under it. We wanted it to be true because it located the disease outside of us, in a server farm in Menlo Park, where it could be regulated. The data has been rude about that wish.

The Backfire Effect Was Mostly Wrong Too

As long as we are demolishing comfortable beliefs, here’s one that lived inside the fact-checking profession itself.

In 2010, Nyhan and Reifler reported that correcting a political falsehood could backfire: conservatives shown a correction about Iraqi weapons believed the false thing more afterward (Nyhan & Reifler, 2010). It was catnip. It explained everything. It launched a decade of communications strategy built on the premise that correcting people is counterproductive, so don’t.

Then people tried to reproduce it. Wood and Porter ran five experiments with more than 10,000 subjects and 52 of the most polarized issues they could find, precisely where backfire should show up if it were real. They found none (Wood & Porter, 2019). The original authors joined a collaborative effort to find it again and couldn’t reliably reproduce their own effect. One of the researchers began calling the search for it the ‘white whale’ (Nurse, 2019). The current read, from the meta-analysis, indicates that backfire is rare, not nonexistent, and not the normal response to correction (Swire-Thompson et al., 2020).

Sit with what that means. For roughly a decade, a chunk of the people whose job was fighting misinformation operated on a finding about misinformation that was, functionally, misinformation itself. The zombie idea that you shouldn’t correct people because it backfires shambled around long after its evidentiary skull was empty. Confirmation bias does not spare those who study it. That should make you humble, and it should make you suspicious of any version of this story, including a satisfying one, including this one, that you didn’t have to work to believe.

Corrections mostly nudge factual beliefs in the right direction. They just don’t move the attitude underneath, and they fade. That’s not a backfire. That’s something more boring and more durable, which is the actual subject.

What’s Actually Doing the Work

So if it isn’t the algorithm reaching in and reprogramming people, what is it?

Three things, locked together, none of them sufficient alone, which is the whole point and the reason the single-villain stories keep failing.

One: motivated reasoning, the old engine. Not “people believe what they want.” More precisely, people believe what they can build a passable argument for, and their desire decides which arguments get built and how hard each one gets checked (Kunda, 1990). The check is real. It’s just rigged. And it runs hardest exactly when the belief has become a uniform, a way of showing whose side you’re on. Whether the election was stolen stopped being a question about ballots a long time ago. It became a question about who you are. You don’t reason your way out of a hat.

Two: the partisan media ecosystem, which is neither new nor symmetric. The American right built an explicitly conspiratorial broadcast layer, OAN and Newsmax, with no real left-side equivalent at the same wattage. The cleanest causal estimate we have, from Martin and Yurukoglu exploiting the near-random spot where Fox landed in different cable lineups, shows that slanted cable doesn’t just preach to the choir; it moves votes at the margin (Martin & Yurukoglu, 2017). And the partisan gap on the stolen-election claim, 63 percent of Republicans to 4 percent of Democrats, is not a both-sides shrug. It’s a 59-point canyon with a donor network and elite voices shoveling on the far side of it (Public Religion Research Institute, 2024).

Three: engagement-based ranking. This is where the algorithm actually earns its indictment, and it’s narrower and nastier than the funhouse-mirror story. The single strongest predictor of whether a political post spreads is not how true it is, not how emotional it is, but whether it attacks the other side. Rathje and colleagues went through 2.7 million posts. Each additional word about the political out-group raised shares by 67 percent on Twitter and 45 percent on Facebook (Rathje et al., 2021). Brady’s team found that moral-emotional language travels the same way, mostly within the in-group, rather than across (Brady et al., 2017).

Read those two together, and you get the actual mechanism. The algorithm is not building you a private reality. The algorithm is running an auction, and the currency is contempt for the people you already dislike. It doesn’t need to manufacture the bias. The bias was there in 1960. It just pays a bounty on the ugliest expression of it, which trains the people producing content to bring more of that, which is fed to a mind that was already, by ancient design, going to wave the agreeable version through without an audit.

No layer does it alone. The cognitive engine without the partisan ecosystem is just ordinary human stubbornness. The ecosystem without the engine has nothing to grip. The ranking without either is an empty auction house. You need all three in the room, and almost every popular explanation, and every viral fix, picks one and ignores the other two. That’s why the fixes keep underperforming.

The Loop That Can’t Lose

You can watch all three layers click together if you look at the worst cases, since those are the ones where the belief has been engineered, accidentally, to never lose.

QAnon started in October 2017 as anonymous posts on an imageboard most people will never see, claiming a cabal of Satan-worshipping pedophiles ran the government and that Trump was secretly at war with them (Public Religion Research Institute, 2024). On its face, this should have stayed in the basement. It didn’t. It walked onto Facebook, YouTube, and Instagram, got recommended, gained followers, and then folded itself into the stolen-election movement so completely that by January 6, 2021, the iconography was at the Capitol. After the deplatforming, the brand fragmented, and the belief didn’t. According to polling, the share of Americans holding the central tenets increased between 2021 and 2023, not decreased.

Here is the engineering. Every prediction QAnon made failed. The “storm” never came. A normal belief takes damage when its predictions fail. This one survived because the framework had a built-in answer: the failure proves the enemy is powerful and the plan is deep. Disconfirmation got recycled as confirmation. The audit Wason’s subjects skipped, this belief structurally cannot perform, since every possible piece of evidence, including the evidence against it, has been pre-assigned the meaning “I’m right.”

The sovereign citizen movement runs the identical trick in a different register. Adherents believe, on idiosyncratic readings of commercial law, that the legal system has no authority over them except when they consent. Every time a court rules against one of them, the ruling is not evidence that the theory is wrong. It’s evidence that the illegitimate system is doing exactly what an illegitimate system does. The FBI has tracked them as a domestic terror threat since 2011, mostly since the loop tends to detonate at traffic stops (Public Religion Research Institute, 2024). You cannot argue someone out of a structure whose every exit has been pre-labeled as the enemy’s trap. The motivated-reasoning engine, handed a self-sealing frame and an ecosystem that rewards the most contemptuous version of it, doesn’t just resist correction; it actively resists it. It converts the correction into fuel.

This is why I keep insisting that no single fix touches it. You cannot fact-check a structure that eats fact-checks. You cannot redesign a feed out from under a belief that the person is now seeking on purpose, across every platform, now that it has become the load-bearing wall of who they understand themselves to be.

My Own Feed

Here is the part a reader on my own side gets to skip, and I let them skip it the first time.

Every example I just walked you through is one where the evidence-backed answer is the one a left-leaning reader already holds. Iraqi WMD. The stolen election. Vaccine denial. QAnon. A progressive can read all of that, nod, and close the tab feeling diagnosed and then told they’re healthy. That’s cotton candy. It also quietly guts the only sentence in this piece I actually care about, the one that says the bug is in you too. So let me spend the discomfort.

In March 2026, CNN’s As Equals team published an investigation into a horrifying thing: networks of men trading instructions for drugging and assaulting their partners, including a Telegram group, “Zzz,” with nearly a thousand members (CNN, 2026). Real reporting. German journalists had already documented a wider web of these groups, some with tens of thousands of members. The harm is not in dispute, and I want that on the record before the next sentence, since the next sentence is the uncomfortable one.

Buried in that same CNN piece was a traffic statistic: a pornography site connected to the story had roughly 62 million visits in a month. Site visits. Across more than a hundred content categories. On April 15, Shannon Watts, who founded Moms Demand Action, posted it to X, writing, “Over 62 million men attended in February alone.” That post drew about two million views. The number was wrong in every load-bearing way. It was visits to a website, not men, not attendance, not a seminar. The Telegram group at the actual center of the reporting had closer to a thousand users. Snopes had it corrected within days (Liles, 2026).

You already know what happened next, since it’s the same thing that happens on the other side. The false version traveled. The correction didn’t. There’s a 2018 study in Science that measured this directly: false stories reach far more people and move faster than true ones, and the gap is not small (Vosoughi et al., 2018). The retraction, when it comes, runs into the continued influence effect, the finding that a correction reduces but never fully removes the original belief (Lewandowsky et al., 2012). Watts’s post was still up weeks later, carrying a reader-added community note and no correction from her. About two million views of the false number. A fact-check that, by the structure of the thing, could never catch it.

I’m not dragging one activist. That’s the point I keep making and keep having to make to myself. The “62 million men” post is an out-group-animosity post. Remember the Rathje et al. (2021) number from earlier in this piece, the one where each word about the political enemy raised shares by sixty-odd percent. That mechanism doesn’t check your politics at the door. It paid a bounty on contempt-for-Republicans content, and it paid the identical bounty on contempt-for-men content, and the people sharing the second kind weren’t lying any more than those sharing the first were. The number confirmed a prior so cleanly that the audit nobody ran on the right is the same audit nobody ran here. Same machine. Same bug. Different jersey.

My editor on the first draft, a reader named Max, also raised the issue of femicide statistics and said they need more care, since it’s not the same kind of error. The UN’s 2024 figures put roughly 83,000 women and girls killed intentionally worldwide, and roughly 50,000 of those killed by a partner or family member (UNODC & UN Women, 2025), which of those is “the femicide number” depends entirely on whom you decide to count: only intimate-partner killings, all gender-motivated killings, or every woman killed for any reason. Roughly four in ten of those killings don’t carry enough information to be classified at all. That’s not a fabricated statistic. It’s a real disagreement over a definition, and both sides reach for the count that flatters the argument they already want to make. The technical name for that is motivated numeracy (Kahan et al., 2017), and I have to flag something to keep myself honest: the famous study behind motivated numeracy failed a careful preregistered replication in 2021 (Persson et al., 2021). If I’m going to spend this whole essay arguing that replication failures should make us hold our favorite findings loosely, I don’t get to exempt the finding that happens to be convenient here. The femicide case is softer than the 62 million case for exactly that reason. It’s contested counting, not a clean false claim. It belongs in the essay anyway, since the softer cases are the ones we’re most fluent at not noticing.

The Fixes Are All Small

I want this part to be honest, because the temptation in a piece like this is to end on a five-point plan that makes you feel better.

Fact-checking works a little. Real effect, small, decays, doesn’t touch the identity underneath (Walter et al., 2020). Prebunking, where you show people the manipulation trick before they encounter it in the wild, is the most promising thing we’ve got, but its effects fade, and they are modest (Roozenbeek et al., 2022). Accuracy nudges, a little prompt asking whether this is true before you share, genuinely shift sharing toward better stuff, and the effect got smaller the harder people looked at it (Pennycook et al., 2021; Roozenbeek et al., 2021). Platform redesign, per the Meta experiments above, did almost nothing in the short run. Media literacy education helps a bit, and carries a quiet risk: it assumes the reader wants to find the truth, when the entire problem is that the motivated reasoner is using exactly those skills to defend the conclusion they walked in with.

Add them up, and you do not get a cure. You get a set of small forces pushing against a very large one. The honest framing is that misinformation is not an outbreak you end. It’s a chronic condition you manage, like the thing in that Texas county that we used to manage and then, collectively, decided to stop believing in.

Why I Actually Care About This

I do health informatics. I build and study the systems that hand information to patients. So this is not a spectator sport for me.

The large language models now being wired into symptom checkers and patient portals, and the little chat box on the hospital website, were trained on the open internet, which is to say they were trained on the full corpus of human motivated reasoning. They can produce the vaccine-skeptical script fluently. They can write it in the calm voice of a doctor. They will hand a worried parent a confident, personalized, wrong answer with the same cognitive ease that lets a false thing slide into your head unaudited at Thanksgiving. The Center for Countering Digital Hate got one major system to generate misleading election images on every single one of 60 test prompts (Center for Countering Digital Hate, 2024).

We are about to install machines that exploit the 1960 bug at an industrial scale, into the one domain where the wrong answer comes with a body count, and a measles ward in West Texas is the early reading on what that costs. That is the part that should keep my field up at night, and it mostly doesn’t yet.

The Landing

Here is what makes me angriest, to borrow my own habit.

We spent a decade and a fortune building a story in which the problem was the machine, since a machine can be subpoenaed, regulated, redesigned, and blamed. The machine does something real, an auction that pays out in contempt, and that part deserves the fight. But the engine the auction is feeding has been running since before the transistor; it runs in numerate people and credentialed people and, demonstrably, in the researchers who study it, and it is running, right now, in you. At the same time, you decide whether this whole argument is correct based partly on whether it flattered a thing you already suspected.

The algorithm is not the disease. The algorithm is a pharmacy that fills prescriptions quickly. The disease is older, it’s ours, and it does not have a logo to sue.

For a deeper dive on this topic, read the spring semester final paper for one of my graduate classes right here.

Author Note

Grace Ann Hansen is an independent researcher and writer, and an MBA & PhD graduate student in health informatics and artificial intelligence. She is also a published author, a professional musician, a gymnastics coach, and a queer transgender woman living in Sioux Falls, South Dakota. She corrects all her papers and articles with Grammarly, because even though she has deep thoughts, she has shallow patience for punctuation. She uses Anthropic’s Claude in Research mode for source location and verification on cited factual claims; all interpretation, argument, and prose are her own. Correspondence concerning this article should be addressed to Grace Ann Hansen at grace@graceannhansen.com.

Type the same question into two boxes on your screen. Pick something ordinary. Can I substitute a loaf pan with a casserole dish for banana bread? Type it into the Google search bar in one window. Type it into ChatGPT, or Claude, or any chat box of that family, in another. Look at what comes back.

The first window gives you a list. Ten or so blue underlined phrases, each one a link to somewhere else: a recipe site, a kitchen-supply blog, a Reddit thread from three years ago. To get the answer, you click. You read. You decide which of those ten the answer actually lives inside.

The second window gives you a paragraph. Not a link to a paragraph. The paragraph itself. Three sentences explaining that yes, a casserole dish works, that the bake time will run a little shorter since the bread is thinner, that you might want to check it with a toothpick five minutes early. No links. No place to click. Just text that did not exist on your screen ninety seconds ago.

That is the whole observation this lesson is making. Hold onto it. We are going to spend a year unpacking everything that follows from it, and it will all anchor back to this one moment.

The first box pointed at something. The second box wrote something.

Two shapes of answer

What the first box did has a name: information retrieval. It scanned a huge index of pages that already exist, ranked them by how well they seem to match what you typed, and handed you the ranked list. The pages themselves were written by people. Real cooks, real food writers, real Reddit users. The search engine’s job was finding them, not writing them. If you had asked a slightly different question, you would have gotten a different ranked list of the same pre-existing pages, or a different set of pages entirely. But every word on every page would still have been written by a human being, at some earlier moment, with their own coffee and their own keyboard.

What the second box did has a different name. It generated. The chatbot did not go find a page that already said exactly what it told you. It produced text. The sentence you read was, in some real sense, new. It existed on your screen for the first time. Nobody, anywhere, had ever previously written that exact paragraph about your exact question.

This is what people mean when they call something a generative AI tool. The word generative is doing work. It says: this thing makes new content. Search engines do not. Generative tools do. Both are useful. Both get called “AI” in casual conversation. They are radically different categories of thing, and the difference shows up in what comes out, which is the only thing we can see from the user’s side at this point in the course.

Why the shape matters more than it looks

If both windows answered your loaf-pan question correctly, you might be tempted to conclude that the two tools are interchangeable for everyday use. They are not, and the gap shows up the second you ask anything where the source of the answer matters.

Ask the first box, “What did the Supreme Court decide last week?” The list of links it returns will be dated. You can see which ones are from this Tuesday’s coverage. You can click through to a news outlet you trust and check the actual reporting. The blue underlined phrases are accountable to the documents they point to.

Ask the second box the same question and you may get a calm, confident paragraph describing a decision in tones that sound exactly like the way news articles sound. Or you may get a calm, confident paragraph describing a decision that did not happen, in a case that does not exist, citing a docket number nobody assigned. The chatbot generated something that sounds like the answer to your question. Whether the something it generated is true is a separate question, one we will spend many lessons on later. The shape of the answer is the same either way: a fluent paragraph. The shape gives you no way to tell, from the shape alone, whether the paragraph is correct.

That is a real difference, not a minor one. It is the difference between here is where the answer lives, go look and here is the answer, take my word for it. Both are useful. They are useful for completely different jobs.

You will hear people argue that AI is going to “replace search.” You will hear other people argue that search is going to replace AI. Both arguments are missing the point. The two boxes are doing different work. The right question, every time you sit down to use one of them, is which job is this? If the job is “find me where this information lives so I can read it for myself,” the first box is doing the work you want done. If the job is “draft me a sentence so I can edit it,” the second box is doing the work you want done. If the job is “tell me a fact and let me trust you,” neither box is quite the right tool, for reasons we are going to take apart all year.

Both of them get called “AI” now

Walk into a meeting room in any company in 2026 and you will hear the word AI used to mean both of these tools, plus six or seven others. The face-scan that opens your phone is “AI.” The recommendation that played that podcast episode you almost liked is “AI.” The filter that catches most of your spam is “AI.” The chatbot you talked to about banana bread is “AI.” The image-generator that made the LinkedIn banner you scrolled past this morning is “AI.”

The word is doing a lot of work. It is, in fact, doing too much. By Lesson 6 you will have a clean way to draw a map of all of those things and locate each one on it. For now, just notice that the word covers more ground than you might have realized, and notice that this is going to be one of the central problems the course solves: giving you better words than just “AI” for the different things in the AI category.

The lesson you can take away today is small but load-bearing. AI is not one thing. The first box and the second box are both in the AI conversation. They are doing different work. The whole course is going to be us slowly opening up the second box, the strange new one, and learning what is inside it and how to use it well.

Going Deeper (optional)

There is a quiet historical pattern worth knowing about, and it has a name: the AI effect. It was articulated, in different forms, by the historian Pamela McCorduck (2004) in her 1979 book Machines Who Think and by the computer scientist Larry Tesler (2014) some years later. The pattern is this. As soon as a problem in AI is solved, people stop calling it AI.

Spam filtering, in the 1990s, was AI research. Now spam filtering is “just software.” Playing chess at superhuman levels, in the 1990s, was AI research. Now your phone has a chess engine that would crush every human alive and nobody calls it AI; they call it the chess app. Optical character recognition was AI in the 1980s. Now it is the feature your phone uses to copy a phone number off a business card.

The effect creates a strange illusion. AI feels both like it is always five years away from arriving and like it is already in everything. Both of those feelings are partly correct. The things we currently call AI are the things we have not yet figured out how to demystify. Once we figure out how the magic trick works, we stop calling it magic. We just call it a feature.

This matters for the course you are starting. The loud, hyped, scary category of AI today, the one that generates new text and images, is the current frontier and nothing more. By 2030 or so, much of what we now call AI will quietly become “the autocomplete feature” or “the assistant tab” or whatever the product team decides to call it. What the word AI points to keeps shifting. The work you are doing in this course is learning the underlying mechanism, which is what stays stable as the marketing labels keep cycling.

What you have, what comes next

You have one distinction now. The first box returns links to things that exist. The second box returns text that did not exist a moment ago. You do not yet know how the second box does it. You do not know what is inside it. You do not know why it sometimes gets things confidently wrong, or what it means for it to be “trained,” or where its words actually come from. You do not yet know how to use it well, what it costs, what risks come with it, or who built it and why.

You will know all of that by Lesson 52.

In Lesson 2, we make the second box less mysterious by separating three things that get casually mashed together in the news: the model, the product, and the chat interface. ChatGPT is a product. GPT-5 is a model. The little window you type into is an interface. These three words refer to three different things, and pulling them apart is the next move.

For this week, just notice the shape difference. Every time you reach for one of these tools, ask yourself: am I asking for a list of places to go look? Or am I asking for a sentence to be generated? Match the box to the job. The rest of the year is built on that habit.

If You Want to Dig Deeper

For the textbook-grade definition of “what counts as AI” and why the field’s boundaries keep shifting, the first chapter of Artificial Intelligence: A Modern Approach is the canonical reference. It is dry, and it is the source most other introductions trace back to. Russell, S. J., & Norvig, P. (2021). Artificial intelligence: A modern approach (4th ed.). Pearson.

https://aima.cs.berkeley.edu/

For a current snapshot of what the AI industry actually looks like in 2026 (compute, capital, deployment, public perception), Stanford HAI’s annual AI Index Report is the closest thing to a neutral state-of-the-field document. Stanford Institute for Human-Centered Artificial Intelligence. (2025). The AI Index 2025 annual report. Stanford University. https://hai.stanford.edu/ai-index/2025-ai-index-report (accessed 2026–05–19)

For a readable history of the field aimed at non-specialists, The Road to Conscious Machines covers the same arc the next few lessons will sketch, in more depth and with more of the personalities. Wooldridge, M. (2020). The road to conscious machines: The story of AI. Pelican. https://www.penguin.co.uk/books/308196/the-road-to-conscious-machines-by-wooldridge-michael/9780241333907

Author Note:

Grace Ann Hansen is an independent researcher and writer, and an MBA & PhD graduate student in health informatics and artificial intelligence. She is also a published author, a professional musician, a gymnastics coach, and a queer transgender woman living in Sioux Falls, South Dakota. She corrects all her papers and articles with Grammarly, because even though she has deep thoughts, she has shallow patience for punctuation. She uses Anthropic’s Claude in Research mode for source location and verification on cited factual claims; all interpretation, argument, and prose are her own. Correspondence concerning this article should be addressed to Grace Ann Hansen at grace@graceannhansen.com.

References

McCorduck, P. (2004). Machines who think: A personal inquiry into the history and prospects of artificial intelligence (25th anniv. ed.). A K Peters/CRC Press. https://doi.org/10.1201/9780429258985 (Originally published 1979)

Tesler, L. (2014). Adages and coinages [Personal website, continuously updated]. https://www.nomodes.com/Larry_Tesler_Consulting/Adages_and_Coinages.html (Wayback archive: https://web.archive.org/web/2024/https://www.nomodes.com/Larry_Tesler_Consulting/Adages_and_Coinages.html; accessed 2026–05–19)

Updated May 20, 2026. New sources surfaced after first publication prompted three additions: a reckoning with Francis’s record on the Synodal Way and the seminary remarks, a closer reading of what “freedom of men and women” actually means in Vatican vocabulary, and the silence on decriminalization during the April flight back from Africa.

OnMay 9, 2026, the Associated Press filed a piece out of Vatican City under Nicole Winfield’s byline. The lead was that the Vatican was sending new signals of openness to LGBTQ Catholics, with limitations attached. A working group inside the Synod on Synodality had released testimony from two gay married Catholics on the harm Church teaching had done them. Pope Leo XIV had told reporters on the flight back from Africa that the unity or division of the Church should not revolve around sexual matters. Cardinal Parolin had called talk of sanctions against German priests using same-sex blessing guidelines premature.

In the same news cycle, a letter from the Dicastery for the Doctrine of the Faith went public. Cardinal Víctor Manuel Fernández wrote it on November 18, 2024 to Bishop Stephan Ackermann of Trier. The Vatican released it May 4, 2026. The letter told the German bishops that any blessing format that “could offer a form of moral legitimization to a union that presumes to be a marriage or to an extra-marital sexual practice” was outside the bounds of Fiducia Supplicans. Bishop Joseph Strickland called the synod working-group report deeply alarming (Winfield, 2026; White, 2026).

That’s the shape of the pontificate in a single news day. Door cracked. Door braced against further movement.

I read the piece on a Saturday morning in Sioux Falls, between coffee and a twelve-hour writing binge. I’m a queer trans woman. I’m also a skeptical Lutheran who stays in the pew. The Catholic Church has never been my Church. The Pope isn’t my Pope. But the Vatican has a voice that travels, and what it does and refuses to do shapes the political weather under which my doctor practices medicine and my passport carries a sex marker the government chose for me. So when Leo XIV says justice should come before sexual morality in the Church’s witness, I pay attention. And I notice what’s missing from the sentence.

This is the second version of this essay. The first one published last week. Readers and an editor I trust pushed back on several points. The pushback was right. The piece is sharper now.

What Leo Has Actually Said

Let me be fair to him first. Leo XIV’s pivot is real.

On April 23, 2026, on the flight back from his first papal visit to Africa, he answered a reporter’s question about the German bishops with this:

“First of all, I think it’s very important to understand that the unity or division of the Church should not revolve around sexual matters. We tend to think that when the Church is talking about morality, the only issue of morality is sexual. And in reality, I believe there are much greater, more important issues, such as justice, equality, freedom of men and women, freedom of religion, that need to be addressed.” (Vatican News, 2026)

I want to hold that sentence a moment. A pope, on a working press flight, saying the things the Catholic Right has used as litmus tests for forty years aren’t the things the Church should split over.

It’s not the only signal. On October 9, 2025, Leo released his first apostolic exhortation, Dilexi te, a document on the Church’s love for the poor, which he completed from a draft Francis was working on at the time of his death. The text names “the dictatorship of an economy that kills” and “structures of sin.” Section 16 says that God “has a special place in his heart for those who are discriminated against and oppressed, and he asks us, his Church, to make a decisive and radical choice in favor of the weakest” (Leo XIV, 2025). Cardinal Czerny, presenting the document on its release, said it had been incorporated into the papal magisterium. That’s not boilerplate. That’s the magisterial weight of two pontificates resting on the option for the poor.

Leo has also been costly in terms of migration. He met Bishop Mark Seitz of El Paso the day before Dilexi te dropped and told him the Church couldn’t be silent on the U.S. administration’s deportation campaign (Picheta & Mortensen, 2025). His Christmas 2025 Urbi et Orbi named Gaza, Yemen, Ukraine, Sudan, Lebanon, and the people crossing the Mediterranean and the borders of the Americas (NPR, 2025). He announced a plan to make Vatican City the first carbon-neutral state in the world (National Catholic Reporter, 2025). On the same Africa flight, he condemned the death penalty in plain terms.

All of this is real. None of it is performance. The pope chose the name Leo to invoke Leo XIII and Rerum novarum, the 1891 encyclical that put workers’ dignity at the center of Catholic social teaching, and he’s done the work to make that name mean something on his watch.

About the Francis Comparison

The first version of this essay used Francis’s todos, todos, todos, “all are welcome,” as the universalist baseline Leo XIV was now narrowing. That framing was too generous to Francis. Here are the receipts I should have engaged with the first time.

On October 23, 2023, Cardinal Pietro Parolin, Vatican Secretary of State, sent a letter to Dr. Beate Gilles, general secretary of the German Bishops’ Conference, telling the German Synodal Way that the Church’s teaching on women’s ordination and homosexual acts was non-negotiable and warning of “disciplinary consequences,” including potential excommunication, for any local move to ordain women. Die Welt published the letter in full on November 21, 2023 (Wimmer, 2023). On November 10, 2023, Pope Francis personally wrote to four German laywomen that he shared their concern that the German laypeople and clergy organizing for reform were drifting “further and further away” from the universal Church.

On February 16, 2024, Parolin, Fernández, and Cardinal Robert Francis Prevost, then prefect of the Dicastery for Bishops, co-signed a letter telling the DBK that their planned vote on Synodal Council statutes would be “in contradiction to the instructions of the Holy See” (L’Osservatore Romano, 2024). That third signature matters. The man now sitting in Peter’s chair personally co-signed the Vatican letter that halted the German Synodal Council’s reform vote.

On March 14, 2024, Francis announced that ten study groups would handle the most contested questions raised at the October 2023 Synod on Synodality, including LGBTQ pastoral questions and Fiducia Supplicans implementation, which were routed into Study Group 9, on “Theological criteria and synodal methodologies for shared discernment of emerging doctrinal, pastoral, and ethical issues,” chaired by Cardinal Carlos Castillo Mattasoglio of Lima. The deadline for the groups’ reports was June 2025. Leo XIV extended it to December 31, 2025. The interim reports went up November 17, 2025. Study Group 9’s final report, the one that included testimony from the two gay married Catholics that broke this news cycle, came out May 5, 2026 (O’Connell, 2024; Vatican News, 2025). The reforms didn’t get a vote on the synod floor. They got a study group reporting back to the pope personally.

On May 20, 2024, Francis told the Italian Episcopal Conference in a closed meeting that gay men should not be admitted to seminaries because there was “già troppa frociaggine,” already too much faggotry, in Italian formation houses. La Repubblica broke the story May 27. The Holy See Press Office apologized May 28 for the language, not for the policy on gay seminarian admissions (Catholic News Agency, 2024). On June 11, 2024, in a closed meeting with about 160 Rome diocesan priests at the Pontifical Salesian University, Francis used the same word a second time, this time framed as something a bishop had reportedly said to him (Brockhaus, 2024).

So Francis said the words about welcome and used the office to enforce their opposite. The universalism was rhetorical. The doctrine wasn’t moving. The German laypeople and clergy who tried to act on todos, todos, todos were told the welcome didn’t extend to the part where the Church changes anything. The men training for the priesthood who would have to lie about being gay to qualify learned what the Pope thought of them from the leaked closed-door transcript.

What this changes about the rest of the essay: Leo XIV isn’t narrowing a sincerely universalist Francis. Leo XIV is the public-record successor to a Francis whose welcome language and doctrinal practice were already in tension. The Leo argument, that what he says in public is what he can be held to in public and the silence on criminalization is what it is, holds. The contrast with Francis was too generous to him. Leo at least tells you out loud the doctrine isn’t moving. That isn’t progress. It’s a different shape of dishonesty, and the piece should have said so.

What the Catalog Doesn’t Include

Now read the April 23 sentence again.

“There are much greater, more important issues, such as justice, equality, freedom of men and women, freedom of religion.”

Justice. Equality. Freedom of men and women. Freedom of religion.

Each of those is a Vatican category with its own theological infrastructure. The phrase “freedom of men and women” is not gender-neutral equality language. It is complementarity language, the Vatican’s doctrinal vocabulary for talking about the dignity of human persons inside the man-woman binary that the magisterium treats as theologically foundational. Dignitas Infinita, the April 8, 2024 declaration from the Dicastery for the Doctrine of the Faith on human dignity, says explicitly that “all attempts to obscure reference to the ineliminable sexual differences between man and woman are to be rejected” (Dicastery for the Doctrine of the Faith, 2024, §60). The document affirms gender-difference as a permanent feature of human dignity in the same passage where it groups “gender theory” with abortion, surrogacy, war, and trafficking as offenses against dignity.

The feminist Catholic theologian Mary E. Hunt, writing in Religion Dispatches the week Dignitas Infinita came out, called the document “an intellectually embarrassing and harmful mess” for exactly this reason. It makes a theology of human dignity that depends on enforcing the binary it claims is “ineliminable” (Hunt, 2024). The National Catholic Reporter commentary the same week concluded that “this ideological commitment to gender complementarity renders the document incoherent” (National Catholic Reporter, 2024).

When Leo XIV says “freedom of men and women,” he is speaking the lexicon his predecessor’s DDF used to frame gender-affirming care as a category error. Queer equality isn’t omitted from his catalog by accident. The catalog is constructed not to have a slot for it. The perimeter isn’t drawn elsewhere, in Dignitas Infinita or in the Diario Correo interview or in the Allen biography. The perimeter is drawn in the same sentence that announces the pivot.

That sharpens the half-a-turn read. Leo isn’t moving in the direction of trans equality and then stopping at a fence. He is announcing a justice agenda whose categories were already constructed to keep us outside.

What He Has Said in the Same Breath

The April 23 press conference where Leo said justice should take priority over sexual matters also contained this:

“First of all, the Holy See has made it clear to the German bishops that we do not agree with the formalized blessing of couples, in this case, homosexual couples.” (Vatican News, 2026)

He invoked Francis’s todos, todos, todos, all are welcome, and then attached the qualifier: “all called to conversion.” To go further than the brief, informal blessings sketched in Fiducia Supplicans, he said, “the topic can cause more disunity than unity” (Roewe, 2026).

In February 2026, in an interview with Elise Allen excerpted in La Repubblica, he was even clearer: “It seems to me very unlikely, at least in the near future, that the doctrine of the Church will change its teachings on sexuality and marriage.” Asked about recognition of gay marriage or recognition of trans people, he said: “Individuals will be welcomed and received… not as an expression or non-expression of a specific identity… I believe that we must first change attitudes, before even thinking about changing what the Church teaches on a particular issue” (Euronews, 2026).

On December 4, 2016, as Bishop of Chiclayo and a public spokesman for the Peruvian Episcopal Conference during the national fight over the gender-inclusive school curriculum, Robert Prevost told Diario Correo (Fernández, 2016, translated from Spanish):

“Promoting gender ideology is already a confusion, because it seeks to create genders that don’t exist, since God has created man and woman. Attempting to confuse the natural ideas will only do harm to families and people.”

“There are men and women. We must respect the dignity of every person, including the choices adults may make. Telling a child who hasn’t yet reached sufficient development to choose with regard to identity and sexual orientation is going to create much confusion.”

“That campaign, apparently, is going to create much confusion and is going to do much damage. We must not confuse the importance of family and marriage with what others want to create as if it were a right to do something that isn’t.”

He has never publicly retracted any of it.

And the institution he leads still operates under Dignitas Infinita. Leo hasn’t retracted that document either. Nobody’s expecting him to.

So the pivot is bounded, both by the catalog Leo used to announce it and by the doctrinal record he has not moved. Justice over sexual morality, yes. But “justice” defined in a frame that calls my existence an “ideology” that “seeks to create genders that don’t exist.”

That’s half a turn.

Read more

This morning, I woke up to an article on Medium by The Land Of Green Ginger.

A Trans Woman Coming to Terms With How Transition Changed My Sexuality

It’s not a lack. It’s different. And sometimes more meaningful.

I’ve never told a story like this on Medium or Substack. After reading it, I was, to say the least, inspired.

So, after a long morning and afternoon of writing this and other things, here’s my story.

A pair of teenagers on the television was holding my attention on a Thursday night. I was fifty-six. I’d been on estrogen for two and a half years. The teenagers were fictional brothers named Conrad and Jeremiah Fisher, at the center of The Summer I Turned Pretty, which had premiered on Prime Video in June of 2022, and I’d ignored every recommendation to watch it for months before letting an Amazon algorithm wear me down.

What I hadn’t expected was the heart business.

Not theirs. Mine.

When Christopher Briney’s Conrad turned his attention to the girl, Isabella, at the center of the show, I felt it. When Gavin Casalegno’s Jeremiah grinned at his best friend’s sister, I felt it again. My chest did the thing a chest does when a stranger you can’t have walks into your line of sight and registers as someone you’d like to lose sleep over. The thing my chest had only ever done for women.

I sat up.

I’ve spent my entire adult life telling myself, and anyone who asked, that I liked women. The premise had survived two marriages. It had survived thirty years of gender dysphoria I couldn’t yet name. It had survived coming out as transgender at fifty-three, nine days after a Labor Day Sunday on Highway 11, I’ve written about elsewhere. It had survived two and a half years of hormones. The premise was the one piece of my interior life that had ever felt settled.

Estrogen, it turns out, edits things you didn’t ask it to edit.

Before The Hormones

I came up through the 1970s and ‘80s in a small South Dakota town, then a college campus in southern Minnesota, then a road life with a touring band. Every space I moved through was a binary I never thought I would escape. By every external metric, I was a man who dated women. A high school girlfriend. A college girlfriend. A first wife. A second wife, who’s still my wife and the spine of every chapter that follows. I didn’t notice men. Given my long-established gender dysphoria, boys in locker rooms made me nervous, and I read that nervousness as the wariness I’d learned of men in a house run by a patriarch I feared. Sexual attraction to men was not part of my vocabulary. The category did not occur to me yet.

In my thirties, somewhere inside a long depression I wouldn’t start to understand for another decade, I caught my eye drifting in a way that didn’t fit the story. A man in a meeting. A man at a concert. The noticing was brief, the noticing was vague, and the noticing happened often enough that I sat with the word bisexual for a few weeks one summer and put it back on the shelf. I was married. I wasn’t going to test anything. Whatever the noticing was, it was a footnote.

In September of 2019, I came out to my therapist as transgender. Nine days after that first appointment, I started spironolactone. A few months later, I added estradiol. Through it all, I assumed the sexuality footnote would stay a footnote. I expected to be a lesbian when I got to the other side. I told my wife I’d be a lesbian when I got to the other side. We were both planning around a known quantity.

The known quantity turned out to be wrong.

January 27, 2023

Three months after the Conrad-and-Jeremiah surprise, Taylor Swift dropped the music video for “Lavender Haze.” Her love interest in the video was Laith Ashley De La Cruz, a Dominican-American model and actor. A trans man.

I watched it twice.

The first watch was the kind any Swiftie does for any new video: I was hunting Easter eggs, admiring the wardrobe choices, and trying to figure out which song would get the next video. The second watch was different. The second watch was me sitting on my couch, realizing that the love interest at the center of the frame was not making my brain do the thing brains do when a beautiful person on screen reads as someone else’s beautiful person. He was making my brain do the thing brains do when a beautiful person on screen reads as beautiful, period.

A trans man. A handsome trans man with a jawline that absolutely cannot be earned through gym work alone, lying in bed next to Taylor Swift as she traced a galaxy across his back.

Hmm, I thought.

A few weeks later, I was halfway through an episode of Shameless when Elliot Fletcher walked into a scene as Trevor, the trans community organizer who becomes Ian Gallagher’s boyfriend in seasons seven and eight. Trevor is patient, articulate, frequently put-upon by Ian’s chaos, and, as Fletcher plays him, distractingly good-looking. I lost the plot. I rewound.

The pattern was getting hard to deny. Cis men on a screen. Trans men on a screen. The reactions felt the same, which was suspicious, since I’d spent decades being certain the reactions to one were impossible.

More than bisexual? I asked the ceiling.

The ceiling did not answer. The ceiling rarely does.

Read more

One year ago today, a guy from Austin walked into a Texas Senate committee room and did the thing almost nobody wants to do. He looked bigots in the eye and named them.

His name is Nick Mollberg. He is not trans. He had never testified before a legislative body before that May afternoon (The Pink News, 2025b). And he said the thing out loud that women like me have been saying for years into microphones that felt broken:

“Do y’all ever get tired of being on the wrong side of history?” (Mollberg, 2025).

The video went viral. Pedro Pascal shared it. George Takei shared it. The Instagram reel crossed two million views, and the TikTok crossed one million (Skinner, 2025). For a week, Mollberg was the internet’s favorite Texan. A lot of trans people, me included, cried on our phones.

Then the bill passed anyway.

On June 20, 2025, Governor Greg Abbott signed House Bill 229 into law. The statute took effect September 1, 2025 (Klibanoff, 2025). It defines sex in the Texas state code by biology at birth. It tells intersex Texans they do not exist. It tells trans Texans that Texas will refuse to acknowledge who we are on our driver’s licenses, our birth certificates, and our death certificates (Klibanoff, 2025).

That last one is the part that gets me. Our death certificates.

Sit with that for a second. Texas passed a law making sure that when trans Texans die, the state will misgender us one last time, on the page that gets filed in a courthouse forever.

I live in South Dakota. I am not Texan. Still, HB 229 is a template. Texas Republicans filed more than 120 anti-trans bills in the 2025 session, the most of any state in the country (Axios Houston, 2025). By the time Abbott held his signing ceremony, Texas had become the fourteenth state to enact one of these sex-definition laws (Klibanoff, 2025). The Texas playbook is now the national playbook.

What Mollberg Actually Said

Let me sit with the substance for a minute, for the people who saw only the six-second cut.

Mollberg told the committee that the legislature was not protecting anyone. He said they had chosen to bully one percent of the population and to hurt that one percent as badly as they could. He said lawmakers were not interested in helping or protecting women, that the women ‘s-rights framing was a Trojan horse, and that the real purpose of HB 229 was to hurt trans people (Skinner, 2025). He told the senators that if they wanted him to look them in the eye as he named them bigots and cowards, he would gladly do so (The Pink News, 2025a).

Then he asked what their grandkids would say to them in the future.

The answer Mollberg gave on behalf of those future grandkids was simple. What these senators were doing was bigotry, the same flavor as every past generation’s bigotry, just pointed at a new target (The Pink News, 2025a). That sentence is the one quiet allies keep trying not to say. It is the sentence that has to come out of your mouth, in your language, in your own tone, directed at your own representative. Soon.

What the Year Has Looked Like for Us

Here is what happened in the twelve months after Mollberg said what he said.

On January 20, 2025, President Trump signed Executive Order 14168, titled “Defending Women from Gender Ideology Extremism and Restoring Biological Truth to the Federal Government” (Williams Institute, 2025). The order redefines sex for federal purposes to mean biological characteristics at conception. Passports. Prison placements. Federal employment. Homeless shelters. Refugee status. All of it now routes through a definition that pretends I do not exist (Williams Institute, 2025).

In July 2025, the federal government shut down the specialized service for LGBTQ+ youth at the 988 Suicide and Crisis Lifeline. The Substance Abuse and Mental Health Services Administration said it would no longer “silo” services for that group (The Trevor Project, 2025).

A suicide line. For queer kids. The federal government turned it off.

In December 2025, the Department of Health and Human Services proposed rules blocking Medicaid reimbursement for gender-affirming care for patients under eighteen, and cutting Medicaid and Medicare funding from any hospital that provides that care to children (KFF, 2026). Twenty-seven states had already banned gender-affirming care for minors by then (KFF, 2025).

In November 2025, the Human Rights Campaign released its annual murder count. Twenty-seven trans and gender-nonconforming people were killed between Transgender Day of Remembrance 2024 and November 20, 2025. Since 2013, the organization has documented at least 399 of us dead, disproportionately Black trans women, disproportionately young, usually shot (Human Rights Campaign, 2025).

Read more