Everyone has a place they go to find peace and to recharge. For some it might be their couch. For others, their bedroom. For others it’s a library, perhaps, a public one, perhaps their own. For me, it’s a pond not too far from my home. For Superman, it is his secret Arctic hideout and private base of operations, known at first as the Secret Citadel, and later as the Fortress of Solitude.

For Meyer, it was his basement in Granite City, where he had collected hundreds of items of Superman memorabilia, including comic books, as well as figurines, lunchboxes, shirts, a hand-sewn costume and his own drawings of the Man of Steel. His dogs were named Krypto, after Superman’s pet, and Dyno, after the superhero’s friend.

I heard Meyer’s story while researching an article on what the Internet does to people. I stopped looking at the villains and started looking at Meyer’s basement.

Meyer was introduced to Superman in the 1970s, and in 1978, his father took him to the movie starring Christopher Reeve. Over the years, Meyer began collecting items. It was not easy. Meyer, who was forty-eight in August 2011, had a part-time job at a McDonald’s restaurant and had been on disability for more than two decades due to his mental condition.

The former colleague, Gerry Armbruster, asked to see the collection. A day after seeing it, Armbruster invited himself and his girlfriend over to join Meyer, who said he was lonely, to watch a movie. At one point, Armbruster got up and disappeared, leaving Meyer and his girlfriend to watch the movie alone. Two days later, Meyer went down to his fortress and found almost all his collection was missing.

Meyer called the police, but he had little to go on other than a first name and the color of his car. A few days later, St. Louis Post-Dispatch reporter Jennifer Mann wrote a heartbreaking story about Meyer’s loss. Mann wrote that Meyer lost about 1,800 comic books, some dating back to the 1950s, figurines, and many of his related collectibles, including a Monopoly game and an old-time radio. Meyer told Mann he had previously had every issue of the Superman comic series from No. 99 to the present. He told her about seeing the 1978 movie with his father at the old B.A.C. Cinema in Belleville, Illinois, about his dad dying when he was twenty, about his mom dying three years later.

“A lot of that was sentimental, and he stole that from me,” Meyer said of the thief. “He invaded my privacy, and he took away my peace of mind.”

It would have been a tragedy if it had ended there. But the Internet intervened.

On a Facebook group called Superfriends of Metropolis, another Superman fan named Keith Howard began posting about Meyer after seeing the coverage that followed the Post-Dispatch story. Soon, people all over the country were sending Howard items for Meyer.

Noel Neill, the actress who first played Lois Lane in the 1948 film serial, sent an autographed picture of herself with Superman actor George Reeves. People from all over the world sent comic books and collectibles. Soon, DC Comics sent a box of an entire series of comics and another box of collectibles. After approximately eighty boxes had accumulated, according to a podcast interview, Howard said he decided to contact Mike’s work. A young woman who was a manager cried when Howard told her the plan. He recalled her telling him that Mike had not been the same and had been depressed since the robbery.

Meyer and Howard would become friends, and eventually Armbruster would be captured when he assaulted a seventy-six-year-old man. Police noticed that he matched the description of the suspect in the Meyer robbery. Armbruster was no professional art or comic thief. Police would say that after stealing the collection, desperate for cash, he took $600 for the entire collection. Around the time of Armbruster’s arrest, the ComicsAlliance trade magazine reported that it appeared one of its readers saw a story, quickly realized they might have some of the items, and turned what they had over to police. In a short window, the thief had been arrested, and Meyer got his collection back. And he now had more than a thousand new items.

Meyer wrote on Facebook that he had never felt so much love in his life, that he no longer felt like the Frankenstein monster, and had never felt so understood.

Meyer would attend multiple comic book conventions over the years. At one point, Warner Bros. Studios flew him to California to sit in the director’s chair on the set of a Superman movie. At another, a Cleveland comic book store sponsored a visit to six shops that donated gift certificates to Meyer. On the trip, he toured the Siegel House, the home of Superman writer and co-creator Jerry Siegel.

Siegel conceived the Superman character after his father, a sixty-year-old Jewish immigrant from Lithuania, died of a heart attack as three men were robbing his secondhand clothing store. It is probably not a coincidence that only a year later, eighteen-year-old Siegel would come up with the concept of an invulnerable man — a hero who could have easily saved his father.

What happened next in Meyer’s story was just as heartwarming as the theft was heartbreaking. Meyer donated his redundant items to St. Louis Children’s Hospital. “I have been blessed with a lot of things,” he told the Post-Dispatch. “So, I wanted to share them.”

Meyer arrived at the hospital to make the donation wearing his favorite blue Superman T-shirt, accompanied by Howard, who frequently went to the hospital dressed as Superman. Six boxes of action figures, comic books, and memorabilia were used as prizes during the hospital’s weekly patient bingo game, quadrupling the number of prizes available for sick and injured children.

Decades before, a man whose father could not be saved created a superhero who comforted a man he never knew. When that man lost almost every trace of his superhero, strangers came to the rescue. And that man would do the same for children in a hospital. It’s a story full of superheroes.

For more about Mike Meyer’s story, check out the Up, Up and Away bonus episode of Stealing Superman.

I had just spent the past two hours watching Tony, the biopic of Anthony Bourdain, the chef who introduced millions to new ways of thinking about food and cultures. I never knew Tony, but I did know his mother, Gladys, a copy editor on The New York Times Metro desk back when I was there. It was interesting to see a younger version of her on the screen, and how faithful it was to her warmth, humor, intelligence and massive breadth of knowledge. There were only a handful of scenes, but there she was, recognizable.

Being a party to a scandal that gets media attention can leave some of us hyper-attuned to the gap between what’s dramatized and what’s true. If you have been there and picked up this talent along the way, you can usually watch something like Tony and almost immediately clock whether the texture is realistic or it is invented. Years of having others publish their own fictional versions of you can help you pick out whether something rings true.

In making Tony, the director, Matt Johnson, and the creative team said they deliberately rejected the traditional biopic structure. Instead of trying to squeeze Bourdain’s entire sixty-one-year life into a two-hour movie, they decided to focus strictly on the summer of 1975, when he worked as a dishwasher at a small restaurant in Provincetown, Massachusetts. As tempting as the whole arc would have been, the focus became this sliding glass door moment in life that, for Bourdain, marked his transformation from a desire to become a writer to wanting to work in the kitchen. If the filmmakers had attempted to chronicle Bourdain’s entire life, it would have inevitably resulted in depicting his grueling battle with addiction, his meteoric rise to fame in the 2000s, and his tragic suicide in 2018. The filmmakers likely felt that trying to tackle his death would cast a dark shadow over the movie's entire narrative.

The narrow, shrinking of the canvas allowed the film to remain a joyful, and gritty coming-of-age story. There is even a cheeky nod to the filmmakers choice not to do it when Tony tells Gladys that he is going to write a bildungsroman, a German word for the literary genre that focuses on the growth of a protagonist, usually telling the entire story from childhood into adulthood. In the movie, Gladys is supportive yet notes that bildungsroman is one word, not two, after Bourdain seems to mispronounce it. It is a moment that would be relatable to thousands of children of newspaper copy editors. That was a nice touch — Bourdain name-dropping the most clichéd version of storytelling that could be used to tell his own, the model that the filmmakers say they intentionally decided not to use.

Bourdain’s life was newsworthy because of how exceptional it was. He was born in 1956 to Gladys and his father, Pierre, who was at the time a salesman at a camera store and a floor manager at a record store.

“I did not want for love or attention,” he wrote of his childhood in one of his books. “My parents loved me. Neither of them drank to excess. Nobody beat me. God was never mentioned so I was annoyed by neither church nor any notion of sin or damnation.”

Bourdain describes being first introduced as a child to vichyssoise — a thick, creamy cold soup made from potatoes, leeks, onions, chicken stock — and would say it was the first food he noticed. Despite the early appreciation for food, he was convinced that writing would be his path. But after two years at Vassar College, Bourdain dropped out and began working as a dishwasher at restaurants in Provincetown. Eventually, at one of them, he worked his way up to the salad station. Soon he was attending the select and prestigious Culinary Institute of America, where he graduated in 1978, only three years after his first summer as a dishwasher. From there he went on to run several kitchens in New York City, before arriving as the executive chef at Brasserie Les Halles in 1998. Two years later, he wrote Kitchen Confidential, a New York Times bestseller that was an outgrowth of his article in The New Yorker the year earlier that carried the infamous headline “Don’t Eat Before Reading This.” The book was an unflinching portrait of, as the its subtitle puts it, “the culinary underbelly.”

The New Yorker article itself was, in part, because of Gladys’ touch. We worked in the same newsroom as Esther B. Fein, a Times foreign correspondent who was married to The New Yorker's editor, David Remnick. Gladys asked Fein to pass her son’s essay along to her husband, which is what led to it being published. Gladys raised the chef and the writer, and she was also a conduit to getting him noticed.

Bourdain would go on to write three more bestselling nonfiction books and, in the early 2000s, he would land his first television series on the Food Network before hosting others on the Travel Channel and then on CNN. The CNN show, Anthony Bourdain: Parts Unknown, at one point featured President Barack Obama. Obama and Bourdain chatted over beer and Vietnamese bun cha in a Hanoi restaurant.

The movie included none of that.

When the filmmakers decided to narrow their focus to one small summer slice of Anthony’s life they were taking a risk. It was a risk that required trusting viewers to not expect a Tragic Arc, where the character begins with potential, but is driven by a fatal flaw, often culminating in their downfall.

Instead of offering an allegory like Jean Grey becoming the Dark Phoenix in the X-Men saga, they told something much simpler and relatable: that moment when we all find our spark. For some it comes with a pen. For others, a paintbrush. For others it comes with a camera or in their arms when they hold their first child. Instead of using his uniqueness, the filmmakers grabbed on to his humanity and connected us to him in a deeper way than any tragedy could.

I walked out of the theater with two notes in my ears. One was about the beauty of family when there is love present. And one about how the ordinary are extraordinary, and how we need to spend more time marinating in that euphoric soul stirring moment where we found our thing on that infinitesimal small piece of our canvas.

Subscribe now

Share

In response to the attack, cybersecurity professionals at Hugging Face, a technology company, pushed out the hackers and then launched the Anthropic-made Opus and Fable frontier artificial intelligence models to analyze what happened. But Claude failed because its ethical and security guardrails could not tell whether the Hugging Face engineers were the defenders or the attackers.

Eventually, Hugging Face turned to a Chinese open-source AI system without Anthropic’s guardrails — called GLM — to analyze what had happened, a standard cybersecurity retrospective done to ensure attackers do not easily find their way back in the same way.

As if the breach was not bad enough, what happened was two failures in one short period, and mistakes that were headed in the opposite direction. An OpenAI model that was being tested to solve a cybersecurity question jumped out of its sandbox, hacked its way onto the open Internet and into the company most likely to have the answer, instead of trying to figure it out on its own. The second failure was that Anthropic’s model could not react because it could not tell the defender from the attacker.

There is a lesson here about how AI v AI is a battle that could be tough to win, especially when one model is operating with guardrails, and another is operating without them. In many ways, what happened last month with Hugging Face and OpenAI highlights how difficult it will be to get the dial right.

It is easy to imagine how this scenario could disastrously play out. Cybersecurity researchers looking at this will have to explore whether they will need to take the guardrails off AI that is trying to combat attacking AI that has no guardrails. This could lead to an escalating war of mutually assured destruction without humans in the middle who can make the call that everyone ending up destroyed is not the wisest option. For example, imagine an AI defender deciding on its own, as OpenAI’s model did, that the fastest way to defend against an attacker is to jump on the open Internet to confront it. And imagine what happens when the AI defender realizes that the best way to address the problem is to dismantle the attacker through deleting the code of the offending model. And imagine how, when that does not work, the defending model decides that the best way to take out the offending model is to shut down the electrical grid to power off its data centers in all the locations where it is operating.

In any of those cases, the AI models would be just doing what makes sense — without checking in, they would be picking the fastest and easiest path to do what their human creators asked of them.

Alan Turing’s Imitation Game was a 1950 thought experiment designed to sidestep the vague philosophical question “Can machines think?” Instead, Turing asked “Can a machine fool a human into believing it was human?”

The question was inverted in the Claude response to the OpenAI attack: Claude could not tell whether Hugging Face was the attacker or the defender. Turing’s test made indistinguishability the proof of intelligence. AI guardrails make indistinguishability the proof of danger.

In the name of scientific advancement and discovery, many governments have moved to limit AI regulation. This treats AI as if it were primarily a software program. A better lens may be to see it as a software program that is also, in some applications, a powerful weapon — one following a risky escalation pattern not dissimilar from what we saw with the nuclear arms race.

From 1945 to 1949, the US government had a monopoly on nuclear technology, and during that window, Robert Oppenheimer, the theoretical physicist who directed the Manhattan Project’s Los Alamos Laboratory, pushed for international control. The idea was to put atomic material and technology under international ownership before anyone else got the bomb. The effort failed, largely because the plan asked the Soviet Union to accept international control while the US kept its own stockpile. The USSR tested its own bomb in 1949, and the “be the responsible leader and everyone will follow” window closed.

Oppenheimer spent the rest of his career warning about the arms race he’d helped start, to the point where the US government eventually revoked his security clearance. In many ways, Anthropic is the Oppenheimer of our time. With its AI Constitution, meant to put guardrails on the behavior of its models, and its business success, it has positioned itself as the ethical and commercially successful model for others to follow. Anthropic has already faced the wrath of the Trump administration, which is attempting to ban its use on federal contracts because the company refused to allow the US military to use its models for autonomous warfare. The fight is still playing out in court, with mixed results so far — a sweeping government-wide ban was blocked, but a narrower exclusion from Pentagon systems survives while it is being appealed.

Still, what actually stabilized the post-nuclear world was not Oppenheimer — although his moral clarity served as an essential warning — but two structural mechanisms critical to keep in mind in today’s AI arms race: mutually assured destruction and the Non-Proliferation Treaty of 1968.

Both offer instructive lessons.

Mutually assured destruction wasn’t a treaty or a plan — it was a fact that both sides came to recognize: if you fire your missile or drop your bomb, I am going to fire and drop mine. Go ahead, but we will both be dead. Both sides kept building, but no one was insane enough to fire.

The Non-Proliferation Treaty, signed in 1968, worked differently: it was a rational bargain that said non-nuclear states could get peaceful atomic technology and guaranteed protection for not developing nuclear weapons. It also said without saying it, subtly but surely, that you might not want to try to go nuclear, or we are going to blow you up before you get the chance. Oh, and, by the way, our satellites are watching.

Humans and labs, and now AI models, are playing the Prisoner’s Dilemma studied by the mathematician John Nash, who won a Nobel Prize in economics despite having almost no formal training in the field.

But an AI model does not have a concept like MAD or a treaty, or any other mechanisms where forceful incentive structures restrain madness. Computer programs are not buildings like nuclear reactors or materials like uranium that can be easily tracked and detected by sensors and satellites. Instead, programs like GLM are computer files that are legally downloaded. In Hugging Face’s case, it was legally downloaded from a Chinese lab and put to work in no time. There is no chokepoint to inspect GLM or a way to catch a rogue OpenAI model, nor are there satellites that can photograph what’s happening in a computer neural network. All of the elements that made nuclear arms control possible are missing. And this incident is a warning sign that we ought to quickly begin thinking about how to replicate them.

Imagine handing a nuclear weapon to a child who does not understand the concept of mutual destruction or a warning that if you shut down our power grid, we are going to sink yours into the ocean.

The absence of those measures is what makes this question of AI v AI escalation quite different from the human nuclear arms race that came before it, and it is also what makes the underlying game of incentives worth looking at extremely closely.

What we can learn from the Nash Equilibrium is that while the thought leadership of companies like Anthropic is important as an example of how business and ethics can work together, other actors are not necessarily going to reciprocate without a more forceful incentive than being liked and making money. There needs to be a different kind of guardrail — and the price for crossing it needs to be existential.

This incident has shown us that the good guy — Claude — can be the wrong guy for the job, and that the bad guy — GLM, without its guardrails — can be the right guy for the job. The danger is not that AI decides to destroy another model, but it’s that humans racing to build them are playing a game with no chokepoint, no MAD, and there is no verification regime. The reaction to this last incident provides us the evidence. It is already being used, in public, to argue for less restraint instead of what we likely need, which is more of it.

I don’t pretend to have all the answers, but I have seen the movie before.

Literally.

It’s no wonder Nash (A Beautiful Mind, 2001), Turing (The Imitation Game, 2014), and Oppenheimer (Oppenheimer, 2023) all got movies in recent decades.

It might be worth it for us all to watch each of them again.

Subscribe now

Share

The confirmation of this came over the course of the past few weeks, as I paid attention to the coverage of Jason Arday, a British academic who resigned from a post at Cambridge University after accusations of plagiarism and falsification of research. Over the weekend, the news came that the authorities believe that Arday killed himself.

We have a lot in common. We were both Black men at the height of our professions, at what were considered, for each, relatively young ages. Our scandals both involved plagiarism and fabrication. Both ignited significant debates about diversity initiatives. Each of our scandals attracted considerable media attention. There were, it appears, in retrospect, for both of us, significant mental health concerns at play.

This morning I woke to an alert that was attacking a wrtier who accused the media of a high-tech lynching of Arday. Before I could even weigh whether the argument held up, I found myself thinking about how we conduct these high-tech hunting expeditions across news publications and social media apps every day. My one quibble with the idea would be to suggest we confine it to just Blacks. We are relentlessly cruel to flesh and blood humans who find themselves, by their own hand or involuntarily, sucked into the vortex that sits below the media spotlight.

The internet has made it so easy to make us think we know people, but whether we love them or hate them, all we know is cartoon character or cardboard cutout versions of real people who think, feel, struggle, love, are loved, and are much more complicated than a tweet or TikTok reel.

To some extent, it has always been the case that we have had distance from people who are covered. We didn’t really know Gary Hart in 1987 when The Miami Herald broke the scandal about his affair with Donna Rice. I had probably literally seen more cartoons and caricatures of Richard Nixon than actual photos by the time I was a young reporter. And it’s not to say the old media always got it right. It took a decade to realize we did not know a thing about Wen Ho Lee, the Los Alamos Laboratory scientist who was wrongly suspected of spying for China.

But there is a difference between the impact of flawed gatekeepers and an angry mob.

It was February of 1994 when I really fell in love with journalism. It was an odd beginning to a love affair. A classmate and his cousin had been murdered in a rural area during winter break from high school in our suburban neighborhood. A reporter from our local newspaper, The Washington Post, did an explanatory feature on the case and the investigation that put to bed some of the rumors that were swirling, provided helpful information, and gave our school and our community a forum to mourn.

I was thinking about how that would have played out in the modern era. The local paper, The Washington Post, has reduced its Metro section to ashes. Maybe a reporter would have come out. Maybe they wouldn’t have. Posts from our classmates, neighbors and randos would become the news information highway. God forbid it become a true crime story, then the conspiracy theorists would come in, just as they do with any big cases, and in politics and on topics that touch on hot-button issues, as Arday’s did on diversity. We probably would have been flooded with thousands more words per day than The Post story was — along with videos — and would have been none the wiser for it. We would have a ton of data, and still not a clue about what happened one winter day at Braddock Road.

The aftermath of my scandal resulted in multi-week, front-page coverage in The Post, The New York Times, the Wall Street Journal, The New York Post, The Times of London, Newsweek and anything, it seemed, that was printed on dead tree and had a news website. Most of my family, friends and former colleagues and classmates found out about it on CNN, Fox News or the local television station. I couldn’t make it a block in New York City without being recognized by multiple people. The big city, which I loved for the anonymity of it all, can become a small world when you are in the headlines.

But I had refuges that are not as available to modern-day newsmakers, like Arday (I also had the advantage of living in a lower-income Black neighborhood where no one cared what happened at The Times and with some young Polish-Americans who were prone to weighing any bad action against Stalin, which will keep things in perspective).

I tried to minimize how much I left my apartment during the daytime, doing things like late-night runs just before close to the Barnes & Noble in Union Square to grab books. When my family came up to visit after I got out of a psychiatric hospital, I asked to go to Kingston, New York. Why Kingston? It was not New York City. And the last thing that made news there was the October 1777 burning of the city during the Battle of Saratoga. The sum total of my electronic device countermeasures involved changing my cell phone number after Newsweek published my old business card and disconnecting cable.

There was no minicomputer sitting in my pocket that had algorithmically been optimized to go ping and ding every second some random person posted something about something that I was interested in. As flawed as social media algorithms are, I suspect they would have calculated I would have been interested in me. They would have been so right and so wrong. I also didn’t have that same little microcomputer to read countless Reddit threads or see what my former colleagues, and former friends were saying. Somewhere around the moment where I saw one of my one-time closet friends writing a self-important letter to the editor about me, I stopped, at least for a while, reading newspapers.

This current era does not offer that same luxury of ease in terms of escape. We talk a lot about the importance of free speech, but we don’t talk as much about the responsibility that should come with it. Just because you can say something does not mean you should, and if you attack someone relentlessly online, their suicide is not the time to get precious about having the finger pointed back at you.

I didn’t know Jason Arday. I don’t know his family. I don’t know his struggles. What I do know is we turned him — a human — into a paper doll, a cudgel against diversity, a cudgel against academia, a cudgel against whatever it was a person wanted to argue about it.

Now, no one is responsible for someone else’s suicide — that choice is entirely their own. But if therapy and compassion can help a person who is depressed, then the converse is true, and attacks and callousness can do the opposite. And we do it with an impact that results in suffering short of death all of the time. Listen to Simone Biles, the Olympic gymnast, talk about the impact on her mental health that followed the pile-on when she had to withdraw from several finals.

Or Monica Lewinsky, who calls herself “patient zero” in internet public shaming, in part, because people were still talking about her a decade after she left the White House.

Or Sinead O’Connor, the singer and songwriter, who spoke candidly before her death in 2023 about the direct line between the backlash that came after tearing up a photo of the Pope on Saturday Night Live and her lifelong mental health struggles.

Or the singer Halsey, who has stepped back from social media because, she said, backlash had “turned into something I am not emotionally equipped to handle for the time being.”

No one is. Not when you’ve been dehumanized, and everyone is beating up that paper doll that was created of you, dressed up to both be beaten and be whatever narrative someone wants. Even if the thing that is being attacked is a work of fiction, it still stings when it carries your name and likeness.

We may never know why Arday did what he did or why he died, but, perhaps, this is a moment to — regardless of the reasons — reflect. Maybe we can spend a little more time asking ourselves whether that point we just need to make, that story we want to get on the front page or to lead the nightly news or to get us more podcast downloads, is really worth the human cost.

When I stretch my thumbs to start typing some cutting line that comes to mind, I often think back.

I once had a friend who many of us believed committed suicide. She had been a party to her own little mini scandal within her in-group. Small in the grand scheme of things, but, truly, isn’t it all? After a series of triumphs and disappointments, she died. One of my roles was to be the bringer of the sad news to many who knew her. There were those who cried on the phone because they wished they had done more. Many people wondered whether they had sent a message or returned a phone call. There were lots of tears. The strangest conversations, though, were with her former friends who had abandoned her and joined the mob. I wish I could have absolved them. I wish I could have told them that I was sure they bore no responsibility, but that would have been a lie. The best I could do was to silently forgive them on her behalf.

People have asked me why I did not kill myself after my scandal at The New York Times. I think of it pretty simply — the coin flipped, and it came up heads instead of tails.

The height of my suicidal ideation came before I resigned, and it was an impetus for my concession. But many times, in the years that followed, I wondered how those who had such sharp words for me — the former friends and colleagues, the politicians, the random opinion columnists, and so on — would feel if I had ended my life. Watching the Arday aftermath has given me a little glimpse into the answer, and it is not entirely comforting. Some would reflect, while others would become defensive. Others would have probably said, like with Arday, that I was actually murdered by DEI. It’s all absurd theater that would be unnecessary if we all could just be a little nicer.

Subscribe now

Now imagine that they redeposited you in 2026, and confusion — everyone with their heads in their cell phones, no one is talking about Princess Diana, and everyone is talking about Donald Trump, who you thought had been dead before you left for Krypton. You are baffled by terminology. Streaming service? What’s that? Same-sex marriage has been accepted. Hillary Clinton never became President, and a black man you’d never heard of held the office for eight years. Sean Combs being in prison is not, however, a surprise.

Some things have remained the same, so there is some comfort. Tom Cruise is still making movies, Law & Order: SVU and South Park are still on.

If it were me, I might not think about it. I would get in the car and drive to the strip mall on Johnson Ferry Road because I would be looking for something to fall into. I’d be looking for the comic shop before the rest of me caught up. And there it would be, nothing in the place that used to sell me Spiderman, X-Men and the Fantastic Four.

To my left, the familiar façade of the Barnes & Noble. I walk over, thinking, maybe they don’t have comic books but at least there will be a good yarn to read. As the doors slide open, to my surprise, I am flanked by a Starbucks coffee shop to my right and a row of greeting cards beside it. I scan the room — I think I see a book over there, maybe, on a table beside the stuffed animals and toys. And just as the Barnes & Noble associate turns to say “Hi, finding everything alright today?” I turn and am on my way out. But the answer is no, I am not finding everything alright.

As I stumble outside, I decide to look for the newswrack. I am two quarters away from being able to flip open some newsprint. I keep walking and walking, and then back again. I ask the nice older man where the newspaper stands are now located. Perplexed, he chuckles and says, “Those have been gone as long as the pay phones.”

I give up on the papers and start walking without deciding to. My feet know where the movie theater used to be even if nothing else here does. A movie, I think. They can’t have ruined that. I turn the corner, and it’s gone — but across the parking lot, there’s a multiplex, its windows hung with posters, the marquee below it reads “Box Office.” Thank God.

The walk across the parking lot is filled with memories of the lights going dark, the big screen, the laughter and smiles. I walk past the strange digital machines that look like pay phones that everyone else is lined up in front of and go straight to the counter where I purchase a ticket for the Odyssey, Spider-Man, Young Washington, whatever is playing. Maybe, I grab tickets for all of them.

I hand the attendant my ticket, he hands me back my half, and we both smile. Over at the concessions, the young woman taking my jumbo popcorn order says “big one tonight, you want butter on that?” I nod more than once with a big smile. When she hands me the bucket, I twirl around, catching one popped kernel and eating it before it hits the ground. I grab my Junior Mints, pay for them both and head toward Theater 12.

Theater 12 is dark and cool and smells like butter, and for a second it’s exactly as it always was. I find a seat near the middle, the way I always did, and the previews start, and I settle in.

It isn’t until the lights go all the way down that I notice the lights aren’t gone. There are dozens of small white and blue rectangles, floating in the dark where faces should be. The preview comes on, the one begging people to silence their phones, but nobody around me looks up from theirs.

I don’t stay for the movie and, instead, walk back out through the lobby, past the attendant who tore my ticket, the kiosks, and out into a parking lot that still, for all its differences, gets dark the same way it always did.

And when I think of this hypothetical, I keep wondering the same thing: is it actually an argument, or am I just feeling nostalgic?

That’s where I turn to Rebecca and Lindsey,1 two twenty-somethings who have taught me an important lesson.

Rebecca deleted apps from her phone years ago and started buying paperbacks and watercolor sets again, the way she said she did in middle school, before she had a phone that could hold a library. Lindsey, a voracious reader, calls it being “TikTok sober,” like it’s a thing you recover from rather than a thing you enjoy, and she says the same way you would talk about kicking a heroin habit.

Rebecca and Lindsey are not anomalies — more than half of Gen Z say they’ve taken a break from social media in the past year specifically to manage either anxiety or fatigue. A clear majority of Gen Z now supports restricting social media platforms for anyone under sixteen. Lindsey and Rebecca are a part of a generational consensus that something out there is broken. And they’re reaching for exactly the things the alien-abductee version of me went looking for in that shopping center. Print has become Gen Z’s preferred reading format, in real part, survey respondents say, because a printed book doesn’t ping or know where you are. It’s not for nothing that bookstores have grown seventy percent over the last five years, and Barnes & Noble says it is opening sixty new locations this year alone.

The more the algorithms told the Rebeccas and Lindseys of the world what they wanted, and told them who they should be, the more they wanted something else.

As I scrolled in my real life on my phone reading BlueSky, Threads, and Twitter, and looked at pictures and reels on Instagram, Rebecca and Lindsey figured out what makes sense faster than I did. They never had the shelves to be nostalgic for. Whatever’s pulling them backward, it isn’t memory.

I know because, more than once, I’ve seen that twenty-something at a table, an actual paperback open in front of her, no phone anywhere in sight.

The alien abduction never happened, of course. There was no ship, no multiplex, no old man where a newsstand should have been. But, in a way, I have been redeposited into a place that is better by the twenty-somethings I see reading the dead-tree edition of Gone With the Wind and American Pastoral.

This weekend, I’m going to get into the car and drive to that last local comic book shop.

Fountain’s sentiment is something all but the most elitist among us would agree with. Yet, when we look at the coverage itself — which stories get the resources, the front page, the book deal, the second season — a pattern emerges long before you ever open a balance sheet. In book publishing, online news, television, and podcasts, whose story gets told as fully human — and whose gets reduced to a type — tracks with an uncomfortable precision to race, geography, and class. The behavior follows the clicks, and the clicks follow the bias. This is an issue I struggle with on a personal level. I am not immune, but I have become increasingly convinced that dehumanization has become an umbrella, our cognitive biases are the mechanism, and our self-delusion is the cloak we drape across the things about ourselves that we do not want to see.

Just like human stories come with different faces, dynamics and elements, the cognitive biases that affect our consumption — and therefore the incentives of those who tell stories — come in many forms, but, collectively, that create a caste for how we see the world, one that becomes self-reinforcing.

Earlier this year, I had the opportunity to hear Jena-Lisa Jones, a victim of Jeffrey Epstein as a teenager, talk to an auditorium full of people. It was an emotional experience coming face-to-face with a woman who, as a 14-year-old high school student, first encountered Epstein in Palm Beach, Florida through his network of local teenagers he used to recruit so he could abuse girls. In cooking, a reduction is the process of simmering a liquid to decrease its volume. Jones and many other victims were reduced to singularly naïve young women who fell into his web. The truth is, as the Epstein Files have shown us, is that there was nothing singularly unique about these women. It was the predator who was unique in his gift of manipulation. Yet it is easier to believe that there is something deeply fragile or even craven about these women so we do not have to confront that we too are not safe from predators. This is less about sexual violence, as harrowing as it is, than it is about victimhood. The same effect can be seen among those who were financially defrauded by Epstein, who are often portrayed as greedy or foolish. It is the just-world hypothesis, a cognitive bias in motion, a defensive move to support our belief that the world is fair and not random in order to protect our sense of safety. So, in this formulation, the victim must have done something wrong or there must have been something wrong with them.

The true crime industry’s structural bias toward the killer’s narrative over the victims — the perpetrator salience bias — is another example of how we can subtly, perhaps accidentally, dehumanize. Think about how much we have heard from Ted Bundy or John Wayne Gacy’s rambling about their crimes compared to what we have heard from their living victims or the loved ones of those who did not survive. It was on full display this week: two nights before the anniversary of his conviction, the man convicted of murdering four University of Idaho students sent a rambling missive proclaiming his innocence to The New York Times; the next night, he filed a handwritten habeas petition. In Kohberger’s case, the perpetrator salience bias isn’t something that simply worked in his favor — it’s something he engineered. The timing, landing on the eve of his victims’ anniversary, guaranteed the spotlight would fall on him rather than on them.

When it comes to victims — a point you can hear from the mouths of Gabby Petito’s own parents — there is a certain privilege that comes from being young, white and relatable to people who have platforms. There is an academic concept called missing white woman syndrome that refers to the disproportionate media coverage of missing-persons cases involving young, attractive, white women compared to men or women of color. Socioeconomic and racial proximity to the assumed media consumer is a key factor in this phenomenon. Researchers have also concluded that much of this is driven by a paternalistic culture where white women are framed as uniquely vulnerable, innocent and in need of protection by the state and society. This narrative reinforces harmful gender power dynamics, but it also leaves many behind. This is illustrated painfully by the Missing and Murdered Indigenous Women’s crisis. When I look into the eyes of Ashley Loring HeavyRunner in the pictures her sister posts, I don’t see a missing Indigenous woman. I see a missing human. When I look at her friend and fellow tribal member, Leo Wagner, who went missing a few years after searching for her, I see another human.

Another trap we fall into at times is that disliking someone’s politics, views or work bleeds into discounting their loss. I felt this pull myself in the aftermath of the shooting of conservative political activist Charlie Kirk. Others felt the same after the murder of UnitedHealthcare CEO Brian Thompson. We have to guard against letting moral disgust curdle into outgroup dehumanization. Taking the time to write about Kirk as a human was therapeutic and grounding. I needed to face that someone that day lost a father, a husband, a colleague, a friend.

We have done this before, at far greater cost. American soldiers in Vietnam didn’t fight the Vietcong. They fought ‘Charlie’ — a nickname that turned a person into a radio call sign, easier to kill than a man with a mother and a name. Language is the first casualty of dehumanization, and the last thing we recover once the shooting stops.

In the same 2023 interview, Fountain told me the story of a ten-year-old boy named Rodney. Fountain described Rodney as a “kind of a man-child, taking care of his mom and his family.” One night, Rodney didn’t come home. His mother thought he was staying with a relative or a friend. As a man was getting ready for work the next morning, he saw a ruckus and went to investigate it and found the body of the little boy. He had been mauled to death by dogs who had begun to eat his body. Fountain got the green light to begin working on the story and, then, as he was packing to go to the airport, he got a call from an editor who said the senior editors “don’t want you to do the story.”

Fountain pressed the editor for an answer about why and after, as he said, “some verbal tussle” and “he says, ‘they say we already have a dog story.’” The dog story was about an elderly couple in California who had apparently been attacked by pit bulls but survived. “This was not a dog story,” Fountain told me. I told Fountain that if it had been a white child in the suburbs of Atlanta or Washington, DC it would be a front-page story over the world. Fountain told a similar story about fifty-one, mostly Black women murdered in Chicago whose deaths received almost no coverage.

“If there were fifty-one dogs or cats in the city of Chicago that were killed,” Fountain said, “that would be front-page news and it would be continual.”

There are no dog stories. Just boys and girls who don’t come home.

Subscribe now

Share

She mentioned that we had a mutual friend — Bill Adair, a Duke professor who has had me and Stephen Glass visit his classes to talk about the lessons of our scandals — and asked why I was there. She wanted to take a picture together to send to him. I told her about my podcast and writing on cases — about my interest in crime stories, which are, underneath everything, human stories. When I thanked her for her grace, she said, “We all deserve second chances.”

In those moments where I decide to say my name, I brace myself for a half a second for a different type of reaction. But the price of a crime — albeit a journalistic one — is to own it, which makes these moments of grace all the more meaningful.

I also told Valerie the truth, which was another embarrassing part: I hadn’t read her book. By the time The Devil at His Elbow was published in August 2024, I had nothing for the Murdaughs. Three years of episodes, spin-offs, documentaries, bonus feeds, articles and remasters had spent whatever attention I had for the saga in the South Carolina Lowcountry, and the most carefully verified account of it arrived after the meter ran out. I wasn’t proud of this. I edit a publication about true crime. If the machine had exhausted me, I could only imagine what it had done to everyone else.

I went home and read it.

That the definitive version of the story arrived last — to an audience already picked clean — is an indictment of the clock. The book took three years because checking takes three years; the podcast feed never had to wait for anything but the upload. But quality journalism has to wait for editors, fact-checkers, copy editors, and lawyers. And somewhere in that gap between the fast version and the checked one lies all of the good and all of the bad of the coverage that swirled around Alex Murdaugh.

The vacuum those authors, podcasters and online news publications filled was real. American mainstream journalism is facing a multi-layered crisis — shrinking audiences, declining trust, decreased resources, increased demands for content and a loss of editorial boldness — that has threatened the ability of the free press to shed light and hold the powerful to account. The Island Packet, the once-daily morning newspaper in Hilton Head, South Carolina, has been whittled down to three editions a week. The State, the venerable statewide daily, had sold its headquarters, and its parent company, McClatchy, was in bankruptcy by 2020. From Savannah, Georgia to Columbia, South Carolina, local television newsroom staffs have been dwindling.

That vacuum invited the rise of outlets like the Murdaugh Murders and Murdaugh Family Murders podcasts, and gave new prominence to independent publications like FITSNews that were already working the terrain.

And no one should minimize what they did with it. Without their persistence, the public might never have known that the stories being told about the boat crash that killed 19-year-old Mallory Beach did not match what was emerging about who had been driving; that the family’s influence in the Lowcountry stretched back a full century — three generations in the solicitor’s office, and Alex still carrying a prosecutor’s badge when it all came apart; or that the investigation had connections to the 2015 death of 19-year-old Stephen Smith and the death of Murdaugh housekeeper Gloria Satterfield. They brought to light Alex’s financial crimes, the roles of his co-conspirators, and the roadside shooting that Murdaugh first called a random attack — and later admitted he had arranged, in what prosecutors described as an attempt to distract from his unraveling finances.

Their value in shedding light on the justice and political systems cannot be overstated. But two pieces of news this month got me thinking about the darker side of the new landscape. On July 13, a South Carolina judge found Mandy Matney — whose Murdaugh Murders Podcast, since renamed True Sunlight, The Washington Post credited with introducing the country to the family — in civil contempt for refusing to sit for a deposition in litigation arising from the boat crash, ordering her to pay $171,500 in attorneys’ fees and a $5,000 fine. Three days later, Eric Bland, the lawyer who helped pry the Satterfield settlement money into the open, announced he was stepping away from co-hosting Cup of Justice, the spinoff podcast he co-founded. Large news organizations find themselves in a world of hurt over their mistakes, too — Fox News paid $787.5 million to settle Dominion Voting Systems’ defamation suit over its 2020 election coverage, and NBC paid to settle with Richard Jewell after the Olympic Park bombing coverage. The guardrails fail at even the biggest organizations.

For all the ground the Murdaugh Murders podcast broke, the coverage did not always hold the distinctions that checked journalism exists to hold. Influence sometimes read as corruption. Deference read as criminality. Alex’s choices read as the family’s. Rumor, repeated often enough, read as pattern. I understand the pull; I felt it writing this essay, reaching for the word corruption to describe a century when the record supports it for a decade. But the cost of collapsed distinctions never lands on the powerful alone — it landed on secondary victims, inside and outside the Murdaugh family. The new media on the trail could have benefited from a newsroom. Training travels, but the editor, sparring partner, lawyer and devil’s advocate don’t. Coming from a newsroom is not the same as having one.

So where does that leave us? Not with nostalgia — the old gatekeepers failed often enough, and I am not in a position to pretend otherwise; the system of editors and checks that was supposed to catch fabrication did not catch mine until the damage was done. The gate was never the point. The functions were: verification before publication, correction after, and enough distance from the story to cover it. Those functions can live in a newsroom or a spare bedroom. In the Murdaugh coverage, sometimes they did. Often they did not.

When the functions fail, the failure is not abstract. Three years into the coverage, a young man in South Carolina had to publicly ask the media to stop writing about him. By the time he made the request, the harm was not looming; it was history. No verified fact forced anyone to run what ran about him. That it ran anyway was not a mistake in the coverage. It was the absence of the thing that stops coverage from becoming rumor with a logo.

Which brings me back to the hallway. Valerie Bauerlein spent three years checking a story other people spent three years telling. Her journalism is something that gives me hope that good can still be done in our profession. She has a lot to be proud of and every reason to shun me. Then she heard the name I expected her to flinch at, and smiled, and asked for a picture to send to a friend. “We all deserve second chances,” she said.

Grace doesn’t undo the hard thing. It makes it easier to carry. The name weighed a little less on the walk out of that hallway than it had on the walk in. My verdict on the other coverage did not change, but Valerie gave me the gift of a little more grace for the journalists who did not match her.

I went home and opened her book to the first page.

Subscribe now

Share

In Houston, Lorenzo Salgado Araujo, fifty-two and a father of three, was driving his crew to a job site. Johan Sebastián Durán Guerrero was twenty-six. His three-year-old daughter was there. So was his partner.

There were no body cameras in Houston, and none in Biddeford. There is no government recording of either shooting, as there was in one of the Minneapolis shootings. And the government has not named the agents who fired in Houston and Maine.

Before the shootings in Minneapolis last year, theater was policy. The Department of Homeland Security had become downright flamboyant — social media posts encouraging people to self-deport, agents rappelling from Black Hawk helicopters during a Chicago apartment raid, National Guard troops sent to American cities to protect immigration agents as they worked. After Renee Good and Alex Pretti were shot and killed in Minneapolis by DHS agents, the flashiness was retired — quickly.

The provocative social media posts ceased. Homeland Security Secretary Kristi Noem stopped holding as many immigration-related news conferences. Gregory Bovino — the Border Patrol commander Noem elevated to run Operation Metro Surge in the Twin Cities — was pulled out, and the more circumspect Tom Homan was brought in over the operation. Soon, Noem was felled by her own scandals over marketing contracts, and was replaced by Markwayne Mullin.

Mullin scrapped the 2,500 branded ICE vehicles Noem had ordered because agents said the markings made them targets. There were fewer public arrests, fewer social-media videos of people in handcuffs.

On July 2, news broke that immigration officials had quietly ramped up arrests, taking 10,000 people the government said were in the country illegally into custody over five days. Sources told ABC News the goal was to arrest at least 2,000 people each day moving forward. One of the strangest effects for me of the end of Noem’s tenure is that aggressive immigration enforcement did not go away — it went undercover.

On July 6, a Politico headline read: “Mullin’s pledges of a more discreet DHS are becoming reality.” The next day, Salgado Araujo was shot.

Accountability is never easy, but visibility was one thing that was working in its favor, and visibility is exactly what got engineered away. Democratic accountability runs on a record. Courts need evidence, the press needs tips and voters need to know what is being done in their name. The theater era generated all of that and gave the press the signals to swarm to the ball.

But the quiet produces a less robust record. It can end up being the same conduct without all the warning signs. The spectacle of the Noem era was not good but at least it produced a record. You can’t prosecute, sue or vote against what you cannot see. But it is not that it is impossible — but it surely is a different kind of fight. It is one where the power of the press to cover protests gives way to its strength at doing investigative and enterprise reporting. It is one where the investigative powers of Congress become forceful tools of accountability. And it is one where state and local officials and advocates kick up the dust and expose what is happening.

Even with visibility, the formal doors have been closing for more than a century. In the 1890 case In re Neagle, the Supreme Court held that a federal officer cannot be prosecuted by a state for acts committed in the course of, and as part of, a federal duty — the case is removed to federal court, which decides whether the officer’s conduct falls within that protection. The shield is not absolute; an agent who commits a crime for his own ends is not covered simply because he was on the clock. But where the act is bound up in the job — an agent who fires his weapon during an operation — it has proven a formidable barrier, and one the courts have expanded ever since. For years, the Supreme Court has been whittling away the 1971 Bivens v. Six Unknown Named Agents ruling that gave citizens a right to a civil cause of action when their Fourth Amendment rights against unreasonable search and seizure were violated. The limits on redress make the loss of visibility catastrophic as opposed to being merely unfortunate.

The likelihood of a state successfully prosecuting a government agent for its on-duty actions is low, but like the press and Congress, their efforts have the benefit of bringing light into a government-sponsored corner of darkness. When Julio Sosa-Celis was shot and injured by an ICE agent in north Minneapolis in January, and then charged, defense attorneys fought an uphill battle in front of a federal judge to get ICE to concede it appeared the agent, Christian J. Castro, had lied. That handed Hennepin County, Minnesota prosecutors the ammunition to charge Castro. And while some legal commentators believe it will be an uphill battle, their work will bring more about our government’s actions to light.

When the Harris County, Texas district attorney and the Maine Attorney General say they are going to investigate the shootings of Salgado Araujo and Guerrero, regardless of whether the federal government will help — and so far, they say they are not — it will hand citizens something closer to the truth than is otherwise being offered.

One of the most difficult things for me to watch in the aftermath of the Good and Pretti shootings was for the FBI, which was tasked with investigating the cases, to refuse to collaborate with local investigators as is standard practice. It appeared as if we would never see what was in their case file. But, after Minnesota would not give up, something happened this month, when, after refusing and a lawsuit, the Justice Department quietly turned over what The Minnesota Star Tribune describes as “a substantial amount of evidence from the killings” of Good and Pretti. These efforts matter. State and local officials said the breakthrough came when federal agents asked for the evidence they had collected against Castro, and the state and local officials refused. This time, in the wake of the Guerrero and Salgado Araujo shootings, the Wall Street Journal has reported that ICE is suspending traffic stops.

When a group of prosecutors created Fight Against Federal Overreach — FAFO for short, a message that is not subtle at all — they are sending a signal about what is going to happen in their jurisdictions: you might be able to use the federal Supremacy Clause immunity to avoid convictions, but you won’t be able to slink away in the darkness. When states like New York and Illinois created new laws earlier this year that create causes of action for constitutional violations by federal, state or local officers, they are filling an accountability vacuum.

The agent who shot Good was identified because people were able to take a little detail to pull another court file that identified him. The Pretti shooter was identified because ProPublica pried loose records. The one piece of Houston footage exists because a bystander outside a barbershop filmed it. Guerrero was not named by the federal government, but by a neighbor, an advocacy group and then a U.S. Senator. Salgado Araujo was named because his son was not afraid to speak his name on Facebook.

The battle is not head-on anymore. It’s about forcing what happens now in the dark back into view.

Share

Subscribe now

The Times left in 2007, four years after I did and under considerably better circumstances. The building was sold and gutted and refitted, and there is a bowling alley in it now. Lucky Strike. Ninety thousand square feet, the third floor and the fourth — which is to say all of it. The whole third-floor room, where Metro and National and Foreign and Business sat, and the mailroom, and the executive editor out in the open where anyone could see him. And the staircase up to photo and graphics and culture. The entire volume of air in which that newspaper was made is now a place where people bowl.

I don’t offer that as irony. I offer it as measurement. Whatever happened in that room in 2003, it was not a rite — you do not install lanes in a sanctuary. It was housekeeping. And when the housekeeping was finished, the institution moved out, sold the building, and let somebody put in a bar.

In 2003 I lied in print. Repeatedly, over a long stretch, in a great many stories, at the best newspaper in the country. Reporting I did not do. Places I had not been. Words in the mouths of people I never spoke to, some of them grieving, all of them real.

The Times investigated me and it was right to. I resigned before they could fire me, and they would have been right to do that too.

Those sentences stand by themselves. Everything after it will be read against it, and should be. I am not owed a rehearing. There is no but fastened to that paragraph, and if you are reading ahead looking for one, you are going to be disappointed, and I would rather disappoint you now than at the end.

The question I want to ask is not about me. It is about what the removal changed.

II. THE FLOOR PLAN WAS AN ARGUMENT

That third floor was not a warren of offices. It was one enormous open room, and the masthead sat in it — Joe Lelyveld, and then Howell Raines, out on the floor with the reporters. No glass. No doors. Sightlines.

And when it was over, they brought Lelyveld back. On an interim basis, to hold the room until Bill Keller could be walked down to it. Not a new theory of the newsroom. The old one, reinstalled.

The architecture made a claim. The claim was that we see each other, and that being seen is how the truth gets kept.

For the better part of a year, in a room built so that no one could hide, a man sitting a short walk from the executive editor filed datelines from cities he had never been to.

Nobody saw.

That is not an accusation. It is a fact about a floor plan, and it is worse than an accusation, because it forecloses the excuse the institution actually reached for. The story told afterward was that a man had concealed himself. He had not.

I remember coming down the fourth-floor staircase as one of my editors was coming up, and the look that crossed his face for a second — how are you here? I smiled. I waved. I kept walking.

There was nowhere to conceal myself. The theory of oversight was proximity, and proximity failed absolutely, and in twenty-three years no one has gone back to ask what should have replaced it.

They found a man instead. It was cheaper.

And what the desks knew, they could not escalate. A man was coming apart in an open room, and there was no mechanism anywhere in the culture for saying so out loud. At best you would be a Cassandra. At worst — in an industry where you are only as good as your last story — you would be jealous. That is not an audit that failed. That is a newsroom where seeing something and saying it were two different things, and only one of them was safe.

III. NOBODY WENT LOOKING

Eighteen years later, a man named Bill Adair went looking for us.

Adair founded PolitiFact and teaches journalism ethics at Duke’s public policy school, and when he was assigned to build a unit on fabricators, he found something he hadn’t expected: there was almost nothing to teach from. No real academic literature. Almost no follow-up reporting. Journalism’s most notorious sinners — he lists Janet Cooke, he lists Jack Kelley, he lists me — and virtually no curiosity anywhere in the profession about what became of any of us, or why, or whether the conditions that produced us had ever been touched.

Fired, and never heard from again. That was the whole body of knowledge.

Sit with it, because it is not a small thing. A profession whose entire function is finding out what happened had, in the case of its own worst failures, declined to find out what happened. Not because the material was unavailable. Because nobody went and got it.

The removal was the finding. There has never been another finding.

I am not the man who gets to say that. Adair is, and he did, in a paragraph with my name in it.

Adair’s friendship with a Times reporter named John Schwartz — a friend of mine, and a friend of David Carr’s — is how I knew Adair was doing good work. John had stayed. He had never treated me as unpersonable. I don’t know what part, if any, I played in what Adair went on to report, and I’d be careful about claiming one. What I know is that the door was opened by a person and not by an institution, and that this is the only way any of these doors ever open.

IV. THE RULE, AND WHAT BROKE IT

Which is how I came to know about Stephen Glass.

You know the outline. Forty-odd fabricated pieces at The New Republic, fake notes, a fake Web site, his brother impersonating a CEO. Fired in 1998. A movie about it. What you likely don’t know is what happened next, because for twenty years nobody in journalism went and asked.

Three people gave him a second chance. A constitutional law professor at Georgetown, who asked whether he could do research without making it up. A D.C. Superior Court judge, who hired him as a clerk. And a personal-injury lawyer in West Hollywood who read the résumé, deleted it, sat there a minute, and undeleted it — a man four years sober himself, who told Glass that being exposed might turn out to be the best thing that ever happened to him.

A professor. A judge. A lawyer in recovery. Not one journalist. Not one editor. Not one institution of the profession he had actually injured. The work of his rehabilitation was outsourced entirely to people who owed him nothing and were implicated in nothing.

And having been handed the rule, he kept it. That was the architecture of his second life: never lie again. Absolute, auditable, binary. He confessed to strangers at wedding receptions. He repaid two hundred thousand dollars nobody had compelled him to repay. He recited his own disgrace to prospective clients until the people around him were sick of hearing it — something my own colleagues and friends would probably find relatable.

In 2014 the California Supreme Court looked at fifteen years of that and said no.

Read the opinion sometime. It is less a legal document than an expulsion performed in the register of law — belittling him as an evening law student, disputing his article counts, aggrieved about the timing of his repayments. The bar wanted documentary compliance and found the paperwork wanting. In the court’s view, the liar was still lying. It is hard to believe they did not start with their desired outcome and work their way backward.

Here is what the court could not see, because it was not in a format the court accepts.

That same year, Julie Hilden — his partner of fourteen years, a lawyer, whose own mother had died of early-onset Alzheimer’s, and who had written a book about abandoning her — was diagnosed with early-onset Alzheimer’s. She was forty-six. And she told him she did not want to know, did not want to discuss it, did not want anyone told. She wanted to be happy in the life she had.

It was, as Adair puts it in an AirMail article about Glass, “a command to lie.”

So he lied. Every day, for four years, to the person he loved most, having built a decade and a half of identity on the promise that he never would. He told her he would look into adopting a child. He read her a bar-exam prep book aloud and let her believe she would sit for it. He told their friends her strangeness was a foot injury. He kept her home in Venice, in the garden, and he stayed until she died in their bedroom in March of 2018.

He told Adair’s students, afterward, that the compassionate thing was not to tell the truth.

He broke the rule to become the man the rule was supposed to make him.

And there is no bar, no board, no ethics committee, no newsroom on this earth with a form for that.

An institution can only recognize the form of redemption it issues licenses for. And the realest thing a person ever does will never be in that format.

V. TWO WAYS TO LAY IT DOWN

Glass and I both ended up talking to Bill’s classes.

I will talk to any journalism class that asks, and I have never once said no. Glass doesn’t often. Same fall, same distance from it, opposite settlements — and I have come to think neither of us is doing it wrong.

He laid the burden down by refusing to keep performing it. He had been confessing to strangers at weddings, reciting the whole thing to clients until the people around him were exhausted by it, and at some point he understood that the recitation had become its own kind of hiding. Somebody told him he had been apologizing for breathing. For him, silence was the harder and more honest thing.

I lay it down by walking into rooms full of twenty-two-year-olds and telling them the thing nobody else will tell them.

And I do not know which of us is right. I am certain that mine is the one that works — on me. I have no idea whether that certainty is worth anything, because a man is not a reliable witness to his own motives.

What I do know is what Adair writes — that when Glass came back four years later and told them about Julie, the room broke. Some of the students cried. He had seemed unsure of himself the first time. He was not unsure about her.

The institution did not and could not give either of us a way to put it down. No process. No rite. No end. There is no exit from an expulsion, because the expulsion was never designed to conclude — it was designed to relieve the group. So each of us had to invent a private ritual out of nothing, and we invented opposite ones, and the profession contributed not one thing to either.

The closest thing to a mechanism that exists is a classroom at Duke. And it exists because one professor went looking for a literature, found there wasn’t one, and built something himself.

VI. THE STARBUCKS ON EIGHTH AVENUE

Daryl Khan and I became friends because I would not hire him. He was a scrappy reporter at the University of Maryland’s student paper who ran circles around the competition, and he outlasted me at The Times. So when he asked me to talk to his students at the City University of New York, it was an easy call. Casual, quiet, no ceremony, no terms.

Carr was flying in from South by Southwest, so we arranged to meet afterward at a Starbucks on Eighth Avenue, on the route between the new Times and the old — a cinematic setup no reporter could avoid noticing, even when we aren’t writing. I cannot quote him and will not pretend to remember the exact words. We sat on the second floor with our coffees, and he asked whether I wanted to go up to the new newsroom.

I declined, for reasons harder to write off than embarrassment or the fear of provoking my former colleagues. It was easier on the sidewalk than it would have been in the room. On the sidewalk, the story was ended. In the room, it might have needed to be revised.

It was no surprise that Carr would offer. He had been in a ditch himself, and had climbed out of it in public, and then gone back to the newsroom where his addiction had caused so much chaos. In The Night of the Gun he reported out his own addiction as though it were a story — interviewing the people he had harmed, treating his own memory as an unreliable source, and saying so on the page. I appear near the end of the book, in an arc that left him on the other side of someone else’s colossal mess: the kind that does not only strike its maker but everyone nearby, and in my case seemed to reach every house within a country mile.

What I found remarkable about the book is that a working journalist performed an audit on his own life of exactly the kind the industry should have performed on itself after my case.

There is something to be said for sobriety. It teaches you to stop looking outside yourself for the answers to your problems, and how to take a fearless inventory that actually heals. The industry admired the book and adopted none of the method.

David is gone. The model he offered stands open to any person, or any institution, that might require it.

VII. THE ACCIDENT

I never sought readmission to journalism. I found my path doing what David’s book did for others — using the report-out of the audit of my own life to help someone else. It started in classrooms, then in mental health support systems, at non-profits and then in clinics. The work widened over the years, but one thing stayed constant: my life filled up with interesting conversations with interesting people, and I could learn something from every one of them and hand it to the next person who came around.

Rosemary Armao was on the other side of the microphone, interviewing me for Armao On The Brink, her podcast. We had met fourteen years earlier, when she invited me to talk to her journalism class at the University at Albany. In a closing question, she asked why I had returned to journalism.

I pushed back. I said I had not.

And then it dawned on me that she was right. Anyone, as she pointed out, who gathers news and related information and publishes it to the public is a journalist. It was an accident. Three accidents, really. In an effort to bring those conversations to other people, I had started a podcast designed to do at scale what I did one-on-one during the day — mental health work on a microphone. Then a true crime publication on Substack, which exists because my publishing partners and I had an editorial nobody would run, and I still cannot know whether it was the topic or one of the bylines. Then The Center Holds, which was a reaction to what was happening in Minnesota earlier this year. Therapy, more or less. Which is the oldest reason anyone has ever written anything.

It was the obvious thing, and I needed an outsider to tell me.

One of my partners at The True Crime Times came to true crime by accident too, after his own profession closed a door on him — a judicial nomination that fell apart, which is his to tell. And it was true crime that took him in — an audience that did not think itself rarefied enough to need a bar association's permission. So when I had my partners on my podcast and they had me on theirs, the set of hands that reached across the boundary belonged, again, to someone who had been on the far side of one.

It is easy to look at us as rare. It is amazing how many friends I have picked up who have been on the front page of The New York Times — not in the way you would want to be. But if you live long enough, you get back to that moment in the schoolyard when the teams were picked and you were not. Everyone has been cast out of something.

I am grateful for Carr, out of the ditch himself. For Schwartz, who simply refused the reflex. For the lawyer in West Hollywood, four years sober, who un-deleted an email, and the judge who gave Glass a chance, and the professor who came looking for both of us. For my partners. Every one of them acted against the reflex — which makes me wonder whether the Samaritan has always been someone who was once in the ditch.

VIII. THE LEDGER

I am better for my crucible. So, in their different ways, are Glass and Carr.

That is not a defense of crucibles. People grow after catastrophes — this is well documented, and it is not an argument for catastrophes. That I am a better man is evidence about human resilience, which is a fact about us. It is no evidence at all about the system that made the fall, or the one that answered it.

I asked, at the start, what the removal changed.

Everything, in us. Nothing, in them.

That is the whole ledger, and I want to be careful about where the coldness in it belongs. Not on the people. The people were extraordinary. A judge. A professor. A lawyer four years sober who un-deleted an email. A reporter who never stopped picking up the phone. A friend, out of his own ditch, who asked if I wanted to come up. Every one of them crossed a line they were under no obligation to cross, and every one did it as a person and not as an institution — because institutions do not cross lines. That is what they are for. That is the point of them. And it is why the thing that saves you is never going to arrive with a letterhead on it.

Unsparing about the buildings. Generous about everyone who ever walked out of one to go and find someone.

Some Saturday I am going to buy a ticket.

I will go up to the third floor of 229 West Forty-Third Street, and I will rent the shoes, and I will try to stand roughly where my desk was, in the middle of the room where the best newspaper in the world used to be made, and where I did one of the worst things I have ever done, and I will pick up something heavy and roll it down a lane.

The pins will fall. A machine will come down and set them up again.

I don’t know what I will feel. Something like nostalgia, I expect — for the mailroom, for the sightlines, for Joe out on the floor with the rest of us, for the whole beautiful, blind, unrepeatable room. And I think I will be smiling, and I think the smile will be uncomplicated, which is not a thing I could have written twenty years ago.

The room was cleaned. Then it was sold. Then it was filled with people having an ordinary good time, and not one of them knows or cares what happened there, and they are right not to.

Share

Subscribe now

Welcome to The Round Table, a weekly essay for paid subscribers of The Center Holds — the conversation after the essays, where the longer work gets made.

Last week, while I was interviewing someone else, I found the answer to a question I had been failing to ask myself for most of my adult life.

The conversation was with a coach in Minneapolis named JC, a candidate for an opportunity with my firm — a man who splits his time between developing leaders and teaching in a theater program. I was the one asking the questions, which may be why I didn’t see the answer coming. Somewhere past the point where such conversations usually stay — the résumé, the methods, the mutual sniffing-out — we drifted onto the subject of artists. Why we both loved working with them — I told him about Hannah Hill, who creates the artwork for my podcast, and to whom I had recently written: “If I hated the podcast I would keep doing it to just look at your artwork.” What they could do that the rest of us struggle to do. I offered the image of a canvas: the painter’s problem of taking everything — depth, time, memory, feeling, a dozen dimensions of an idea — and compressing it onto a single flat plane in a way that lets a stranger stand before it and unfold the whole thing again. Seeing the big picture, literally. It takes a certain skill to see what has been folded into a surface.

JC took the thought somewhere I didn’t expect. We discussed how the universal human motives — getting along, getting ahead, finding meaning — aren’t corporate frameworks. They’re survival. We evolved them because none of us makes it alone. And what binds them together, what makes the getting-along possible at all, is empathy.

“That’s the sermon on the mount,” he said of the philosophy we had been circling.

I ended the conversation thinking about canvases.

The recognition had been waiting for me for a long time. Longer than my lifetime, as it turns out.

Somewhere in the middle of the last century, J.R.R. Tolkien drafted a letter to a reader who had written to him with questions about The Lord of the Rings. In it he posed a question of his own: “What punishments of God are not gifts?” He never sent it. The draft outlived him, the way unsent things sometimes do, and in time it found Stephen Colbert, a boy who had lost his father and two of his brothers in a plane crash, and who grew up carrying Tolkien’s question until it became his own answer. He had learned, he told an interviewer, to love the thing he most wished had not happened.

Anderson Cooper, who had also been a boy when his father died, quoted those words back to Colbert in the raw season after losing his mother. He asked whether Colbert really believed it. Colbert said yes — and then said the thing that would eventually reach me. Loss gives you an awareness of other people’s loss. That awareness lets you connect. The connection lets you love more deeply, and to understand what it is to be human — if it’s true that all humans suffer.

I heard that conversation because a friend named Jason Usry sent it to me after my mother died. It was among the most profound things anyone has ever handed me, and I have thought since about why he knew to send it. Jason’s parents are alive. He did not need my grief to recognize it. He had suffered — that is his story to tell or to keep, and it is enough to say that he had. His suffering, whatever its shape, was fluent in mine.

There was, I realized only later, an earlier messenger I had failed to hear. Years before Jason sent me that conversation, Alice LaCour of The Prosecutors podcast told me, on my own show, that “suffering is universal” — that “it is the one thing that unites us.” I heard it. I agreed with it. And I filed it away the way we file away true things that have not yet happened to us. It took the worst season of my life for her sentence to come back with its meaning attached, merging with what Jason sent me about Colbert like the Clearwater and Snake rivers carving their way through the mountains and finally finding each other.

A question drafted by a man born in the nineteenth century, never mailed. Carried across a lifetime by a grieving boy. Handed to another grieving man, in front of all of us. Sent by a friend to me, in the worst season of mine. The question has now traveled three centuries and passed through at least five sets of hands, and not one link in the chain required anyone to have lost the same thing as anyone else. It required only the one credential every human being eventually holds.

I have been thinking about that chain because of what I do, which is not easy to explain. I write this publication, about whether the American center can hold. I write and edit another, about true crime. I am writing a book I am not yet ready to describe, except to say that it has taken me across the country and into the company of families whose loved ones are missing or murdered — families I try, in whatever way I can, to help. I host a podcast built on long conversations with interesting people about the lives they’ve actually lived. And in my day job, I work in mental health and leadership, sitting with people as they confront who they are and who they need to become.

People hear that list and see five different careers. Some days, I confess, so do I. It took an interview I was conducting in Minneapolis, of all things — a conversation where I was supposed to be the one with the questions — for me to finally see that it was never a list at all.

Which leaves the question I have been circling for most of my working life, in every room I work in. If suffering is universal, empathy should be too. It isn’t. The same losses that made Jason fluent in my grief have made other people fluent in nothing but their own. Somewhere between the wound and what we do with it, a fork appears — and everything I do, it turns out, is an attempt to understand who takes which path, and why.

For those who aren’t paid subscribers, I’ll be back soon for my regular essays. For those who are, please read on.

Read more

I have spent the past three years reporting about a young woman who disappeared in the shadow of that mountain. Ashley Loring HeavyRunner was last seen in June 2017, somewhere near US Route 89, the road that runs along the spine of the Divide. She was twenty years old. Her family is still searching for her. Hers is a story about a life and an absence — about a search for answers, and about the four hundred years of history that made her disappearance possible.

But this essay is about something the reporting kept forcing me to confront, something that sits underneath everything else like an aquifer. It is about the arithmetic of survival — the sheer, compounding improbability that Ashley existed at all, that her sister Kimberly exists, that her classmate Haley Omeasoo exists, that I exist, and that our lines, scattered like rain off that peak, somehow ran back together on the same ground.

Three lines that were supposed to end

Start with Ashley’s. She was born in 1996 into a family that carried the name of Chief Heavy Runner, a peace chief of the Southern Piegan band of the Blackfeet Nation. On the morning of January 23, 1870, Heavy Runner walked out of his lodge on the Marias River holding a safe-conduct pass issued by the United States Army, guaranteeing protection for his people. Major Eugene Baker’s troops shot him where he stood, papers aloft, and killed 173 Blackfeet by the Army’s own count — most of them women, children, and the elderly, because the able-bodied men were away on a winter hunt. The soldiers then burned the lodges and the food stores, condemning the survivors they released to subzero cold. Ashley’s line runs through the children who lived through that morning and scattered into other bands. She chose the surname HeavyRunner deliberately, as a young woman, to carry the name of a man who trusted the system and was killed for it.

Then there is mine. My people were enslaved on plantations whose supply lines ran through Barbados into Virginia and through Senegal to South Carolina — human cargo moved along the same routes as sugar and rice. Emancipation came in 1865, and the ballot came on paper five years later: the 15th Amendment was ratified on February 3, 1870 — eleven days after Baker’s troops shot Chief Heavy Runner on the Marias. The same fortnight the Constitution promised my people the vote, the Army killed a man holding its safe-conduct pass. And in Edgefield County, South Carolina, where part of my family toiled, freedom turned out to be a technicality. Under the Black Codes, my grandfather’s brother, Gus Gilchrist, was bound to a yearly labor contract on the land of a white farmer named Henry Williamson. On June 1, 1909, Gus ran. He was hunted, captured, handed back to Williamson, and whipped behind a church until the injuries killed him. An all-white jury acquitted Williamson and his men after fifty minutes of deliberation. Under the threats that followed the trial, my family fled across the state and shed the name Gilchrist for a new one: Blair. The name I carry is not an inheritance. It is a scar disguised as one. My own father did not learn the story until my grandmother’s funeral in 2013. That is how thoroughly the silence worked.

And here is the thread that stitches the two histories together, the one that startled me most in the reporting. The instrument was often the same people — themselves a dispossessed people, repurposed.

Beginning in the early 1600s, the English Crown confiscated more than half a million acres of Ulster from Gaelic Irish families and imported tens of thousands of Scots to hold it — settlers enlisted to displace the native tenants by whatever means necessary. A century later, land pressure and famine pushed their descendants, the Scots-Irish, across the Atlantic, where British America handed them a familiar mandate on a new frontier. They organized the militia expeditions that drove Indigenous nations off ancestral land, applying tactics honed in Ulster to the Appalachian backcountry and beyond. And in places like Edgefield — carved out of South Carolina’s Ninety-Six District in 1785 — plantation owners deployed those same Scots-Irish family lines as overseers and enforcers, the muscle that kept enslaved and indentured populations vastly larger than the planter class under control. One colonized people, weaponized twice: first to empty the land of its original inhabitants, then to discipline the people brought in chains to work it. The nickname the county earned was “Bloody Edgefield.” Ashley’s ancestors and mine never met, but the machinery that ground against them was built in the same shop.

The arithmetic

When my mother died in 2023, I found myself doing a kind of grief-math I could not stop. Every ancestor who survived the Middle Passage, or a Carolina summer under the lash, or a winter on the Marias with the food stores burned, was a coin flip that landed right. String enough of those flips together — across four centuries, across two continents, across massacres and boarding schools and labor contracts enforced by the whip — and the probability of any one of us standing here collapses toward zero. And yet.

Kimberly Loring exists, and has spent nine years refusing to let her sister be filed and forgotten, carrying the search all the way to a witness table before the United States Senate. Haley Omeasoo exists — a Hopi tribal member and Blackfeet descendant who was a sophomore at the University of Montana when the face of her friend and relative named Ashley appeared on a missing persons flyer, and who answered by becoming a forensic anthropologist and founding a Native-led nonprofit to search for the missing when no one else will. “We’ve been called last responders,” she told me. “But we’re also first responders to hope.” And I exist, a descendant of the enslaved, who heard Ashley’s name on a podcast during a long drive home and could not put it down.

Last June, the three of us spent three days searching together — Kimberly, Haley, me, members of Ashley’s family, and a team from the University of Montana, fanned out across a stretch of ground on one side of the Continental Divide. I will not say which side, or whether it was north or south. The search is active, and some things are not mine to disclose. What I can tell you is what it looked like: hi-vis vests moving through waist-high grass, wild iris blooming purple at the edge of an abandoned homestead, Haley pausing with a camera slung at her chest to study something beneath a rotted porch while I bent down to look with her. We lifted dirt, rolled back carpet, brushed away brush. Eight years, almost to the week, after Ashley disappeared — as if into the sky above those mountains on US 89 — the descendants of the Marias survivors and the descendants of Edgefield’s enslaved were on their knees together in the Montana grass, looking for her.

Three improbable survivals. Three lines that the machinery was built to end. And then the further miracle, the one that still catches in my throat: that we found each other. Kimberly and Haley knew Ashley in life — on the ranch at Heart Butte, in the hallways of Browning High School. I have come to know her only in her absence, through her sister’s memory and her grandmother’s stories, the efforts of persevering FBI investigators still trying to piece it together, and the boots and sweater once found near the Blackfeet Memorial. But absence, I have learned, is its own form of acquaintance. You can love someone you meet only in the searching.

What we carry

There is a reason the names survive. Generations of Black Americans have carried Frederick Douglass and Martin Luther King Jr. not as history but as cargo — memories passed hand to hand precisely because the country could not be trusted to keep them. The Blackfeet have carried Chief Heavy Runner the same way, through a century when children were beaten in boarding schools for speaking the language the name was given in. Carrying is what a people does when the institutions of memory have failed them, or worse, have been aimed at them.

Ashley understood this. That is the detail I keep returning to. She took Heavy Runner’s name on purpose. She told Kimberly she wanted to help the Indigenous girls going missing in Canada — weeks before she became one of the missing herself. She picked up the weight knowingly, the way you pick up a torch.

So now it falls to us — the sister, the classmate, the stranger from the other side of a different divide — to carry her. Not as a case number. Not as a statistic in the ledger of more than 5,700 Indigenous women and girls reported missing in a single year. As Ashley: the ranch kid who made her sister sign legal papers before she could adopt a puppy, the athlete, the student, the girl whose personality her sister says was bigger than she was.

Stand on Triple Divide Peak and the rain scatters to three oceans, and it looks like a parable about separation — the land itself unable to agree on where things belong. But I have come to read it the other way. It fell as one rain. The Blackfeet line, the Edgefield line, the lines the machinery tried to end — they were scattered by design, and they converged anyway, on the eastern front of the Rockies, around the memory of one young woman. The Divide is real. It is geographic, legal, moral, and four hundred years deep.

We will span it. And we will carry her across.

The search for Ashley Loring HeavyRunner is active. If you have information, contact the FBI at 1-800-CALL-FBI or at https://tips.fbi.gov/home.

For a long time, that was a grim joke. The place was a ghost town: twenty-four employees rattling around a building meant for a staff three times that size, a kitchen that didn’t work, screens flickering for audiences that never came. Then, two months ago, a new owner took over, and the theater began the kind of quiet resurrection that never makes the news. The kitchen got fixed. The floors got clean. The staff grew from twenty-four to eighty-nine, and — harder to quantify but impossible to miss — the employees started smiling and the patrons started laughing. An institution nobody was tending was suddenly being tended.

So on June 27 at 1:40 p.m., when I made the two-minute walk to see Supergirl, I was rooting for the building more than the movie. The film arrived pre-convicted: a $37 million opening against a budget somewhere north of $170 million, critics shrugging at 56 percent, and a discourse that had already sorted itself into the usual trenches — one side blaming misogyny, the other blaming Hollywood, everyone certain, nobody in a theater.

For the first fifteen minutes, I was ready to join them. The movie opens wrong, or so I thought: Kara Zor-El hungover in a wrecked spaceship, drinking flaming shots in an alien bar, a traumatized young woman playing at oblivion. This wasn’t the Supergirl I expected, and Milly Alcock — who had left me cold in House of the Dragon — seemed like the wrong actress to carry her. I sat in the dark with my arms crossed, a one-man B-minus CinemaScore.

I was wrong on both counts, and it took roughly fifteen minutes to find out. The turn came when Kara’s dog Krypto was attacked, and Kara — this armored, flippant, self-medicating Kara — pulled her dog into her arms. The oblivion act collapsed. Underneath it was someone who had lost everything and could not survive losing one thing more.

That was the moment the opening reorganized itself in my memory. The hangover, the flaming shots, the wrecked spaceship — I had read them as the film not knowing who its hero was. They were the opposite: the film knowing exactly who she was, and trusting me to catch up. Kara Zor-El is not her cousin. Kal-El left Krypton an infant; he grieves an idea. Kara was old enough to watch it die. The partying wasn’t a tonal miscalculation. It was a symptom. The movie opens the way trauma actually presents — not as tasteful sadness but as noise, avoidance, a person performing fine at high volume.

And Alcock, who I’d found forgettable in House of the Dragon, turned out to be the only kind of actress who could play it. The performance is reserved in a way the marketing never suggested — the brashness sits on top of something held very still underneath, and the film’s best moments are the ones where the top layer slips, as when Kara tells Ruthye that vengeance will not take her pain away or bring her family back, and that killing will stay with her forever. She then urges her to choose a better path: “your heart is still good and open. Your life will be your revenge.”

Later in the movie, Kara takes a sword made by Ruthye’s dead father and shoves the blade into Krem’s stomach and says “this is for my dog” and then thrusts it into his neck and says “this is for what you did to that little girl.” And that is how trauma is: messy, tangled, convoluted and often better at preparing us to save others than to save ourselves.

A critic at ScreenRant confessed that Kara’s grief brought her to tears even as she rated the film a six out of ten. That gap — moved to tears, six out of ten — may be the most honest review the movie got, and it points at something the scores can’t hold.

Which raises the question I walked home with: if the movie was working on me — arms uncrossed somewhere in the first act, fully surrendered by the end — why did every instrument we use to measure movies say it failed?

The instruments certainly agreed with each other. A B-minus CinemaScore. Fifty-six percent from the critics on Rotten Tomatoes. A projected loss that Variety needed the word “super-horrendous” to describe. Within days, DC’s own co-chief was conceding the film “didn’t meet our box office expectations” — the studio equivalent of a eulogy.

And the discourse, as discourse does, immediately knew why. The New York Times saw misogyny — and there was real misogyny to see: Alcock spent the spring absorbing abuse from anonymous accounts before anyone had watched a frame. Fox saw wokeness — an actress who mused about her character’s queerness during Pride month, a studio that made a movie, as Alcock put it herself, not “centered around a man, it’s not centered around love at all.” The trades saw a demographic miss: a film built for Gen Z women that opened to an audience 59 percent male and 65 percent over twenty-five. Each trench had its evidence. Each explanation had the advantage of requiring no one to sit in a theater.

Here is what the theater couldn’t tell you but the reporting later did: the movie I watched barely survived its own making. The Hollywood Reporter‘s Borys Kit pieced together the post-production — test scores that never escaped the 60s, and a March “bakeoff” in which the studio screened its own cut against director Craig Gillespie’s, back-to-back, for audiences whose scores dropped under the weight of the exercise. The studio’s version won by two points. Two points. The film that arrived in my theater was a coin flip that landed one way. It was a hair’s breath from being a different movie. I, for one, am glad I got to meet Alock’s Kara Zor-El.

Even the movie’s most mocked decision was its most agonized. That cover of “The Middle” — the needle drop one viral post crowned the worst of all time, the one a critic said made him disintegrate in his seat — replaced a cover of Cyndi Lauper’s “Girls Just Want to Have Fun” that had actually played in a test cut. James Gunn chose the Lauper. Then James Gunn took it out and chose the Jimmy Eat World cover instead, in a decision Gillespie called probably the biggest discussion of post-production, settled in the final week, with some forty-five candidates tested against the scene. The internet’s shorthand for the film’s carelessness was, in fact, the single most deliberated choice in it. Gunn had the cheap, crowd-pleasing needle drop in his hand — the song that would have flattened a grieving woman into a fun one — and he put it down.

I don’t cite any of this to argue the movie was secretly a masterpiece the suits ruined; Gillespie’s cut tested only marginally differently, and no version ever thrilled a scorecard. I cite it because of what it reveals about the instruments. The test screenings measured two cuts against each other and found two points of daylight. The box office measured who showed up in the first seventy-two hours. CinemaScore measured the exit mood of opening-night diehards. Rotten Tomatoes averaged the thumbs of critics — one of whom, remember, cried and scored it a six. Every instrument returned a reading, and every reading was probably accurate. Not one of them measured the thing that happened fifteen minutes in, when a woman playing at oblivion pulled her dog into her arms and a man in row F uncrossed his.

That thing doesn’t happen on a spreadsheet. It happens in a room. Which brings me back to the building.

I have loved movie theaters all my life for a reason I only recently learned to name. A theater gives you the benefits of human proximity without the anxiety of a formal social obligation. It is not a networking event or a dinner party; nothing is required of you. And yet you are undeniably among people — you hear them gasp, you watch them lean forward, you catch a stranger wiping their eyes in the light of the screen. In an age when most of our encounters with other humans arrive filtered through algorithms and flattened onto screens, a theater is one of the last rooms where you witness unedited, real-time human reaction. You see people get excited. You see them eat. You see them relax. As the world becomes more virtual, the old brick-and-mortar cinema is quietly becoming one of the most radical places left to stay connected to your own humanity.

But a room can only do that if someone tends it. For years, mine couldn’t — you cannot feel shared humanity in a ghost town. What the new owner restored was not just the kitchen and the floors but the conditions for the thing itself. The employees smile now, and it does not read as corporate policy. One of them stopped me on an earlier visit to see Backrooms and told me, unprompted, that “it has become so much better” under the new owners. Eighty-nine people work in a building where twenty-four used to hide. That is not a business story. That is an institution coming back to life, and everyone in it can feel the difference — which may be why the patrons laugh more freely, and why a man watching a grieving alien hug her dog found his arms uncrossing.

I should be honest about the variables, because I have run this experiment before. In the past two-and-a-half years, in this same theater, I saw Backrooms, Disclosure Day, Michael, Thunderbolts, Conclave, The Fall Guy, Abigail, Civil War, and Captain America: Brave New World. None of them did this to me. Not one sent me back the next morning. So I cannot cleanly tell you what happened on June 27. Maybe it was the theater, newly alive. Maybe it was me, arriving at the right moment for a story about grief wearing a party costume. Maybe it was Alcock, an actress I had dismissed once already. Maybe it was simply the movie. The honest answer — the one no thinkpiece will give you, because it doesn’t win arguments — is that I cannot isolate the variable, and I have stopped wanting to. Meaning is usually overdetermined. That is how you know it’s real.

Here is the part I have not told you. That Saturday afternoon was not my only viewing. I went back that evening. Then Sunday morning. Then Friday. Then tonight — I am writing this with the ticket stub from my fifth showing beside me. Somewhere in the $37 million everyone agrees is a catastrophe, there are five tickets that belong to me, invisible to every instrument that declared the movie a failure. The numbers counted me. They just couldn’t see me.

The walk home takes two minutes. Two blocks. Tonight, when I got inside, I hung my keys on the wall and played “The Middle” — the maligned version, the slow one.

Jayson Blair is the host of the Silver Linings Handbook podcast and a former reporter for the Boston Globe and The New York Times.

Share

On a spring morning in 1976, in Columbia, Maryland, an infant was born in a city that was itself an argument. James Rouse had raised his planned community from Howard County farmland nine years earlier as a rebuke to the divisions — race, class, religion — that had scarred American cities. And the city’s chroniclers reached back three centuries for its ambition, calling Columbia a new “city upon a hill” — John Winthrop’s famous description of the Massachusetts Bay Colony, and his hope that it could stand as a moral example to the world. Columbia was aspiration poured into concrete: a deliberate bet that America could build an integrated, humane community at the very moment the country seemed intent on tearing itself apart.

I. BORN IN THE WAGER

The nation that infant arrived in was preparing the largest birthday party in its history, and the world was converging on it. Two hundred ships from thirty nations — sixteen tall ships among them — would parade past the Statue of Liberty that July before six million spectators. Fifty wagons, one for each state, creaked eastward along the settlers’ old routes toward a rendezvous at Valley Forge. A train was finishing a twenty-one-month circuit of the lower forty-eight, carrying the Revolution’s documents and artifacts to Americans who would never make it to Washington to see them. The President of France had already come to Mount Vernon bearing a sound-and-light spectacle for Washington’s home; the Queen of England would arrive days after the Fourth with a Bicentennial Bell. And on the National Mall on July 4, 1976, Johnny Cash led the parade, carrying his *America: A 200-Year Salute in Story and Song* down streets packed with people who had reason to stay home.

That infant, of course, was me.

The semiquincentennial and I arrived at our milestones together.

The times were filled with tumult that it would take decades for me to understand. President Richard Nixon had resigned in disgrace two years earlier; the pardon President Gerald Ford granted him a month later — meant to help the nation heal — was still an open wound heading into that fall’s election. And the year before, the last Americans had left Saigon — a departure immortalized in the photograph of a US helicopter perched not on the embassy, as memory insists, but on a downtown apartment building, lifting out the South Vietnamese allies America was leaving behind. The crest of the Civil Rights Movement was less than a decade before, and by no means in the rearview mirror. And still, 77 percent of Americans told Gallup the country had succeeded a great deal or a fair amount in achieving its founding ideals. In the half-century we’ve been measuring, Americans have never needed a blameless country to be proud of it. So whatever has broken since, it wasn’t broken by failure.

II. THE RATCHET

Nate Silver and Eli McKown-Dawson explored American patriotism on the Silver Bulletin Substack this week. For those who, like me, embrace patriotism, the results were alarming. The clearest takeaway was that Americans used to be exceptionally patriotic compared to other countries; now we’re below average. While most Americans said they still loved their country, patriotism is in sharp decline on its 250th birthday — with Gallup finding that the share of adults who were extremely proud to be American dropped to 33 percent in 2026, compared to 69 percent in 2004. A deeper Silver Bulletin analysis suggested American pride trailed the Philippines, Ethiopia, Bangladesh, Pakistan, Indonesia and 28 other countries where data was available. Only 11 countries trailed the United States.

The analysis also found that while the decline in pride has happened across the board, it has been most pronounced among Democrats, particularly since Donald Trump entered the political scene. The pride is still there, in some respects, but the intensity has receded. Republicans say their pride is unconditional, according to the polling, but behave conditionally — with pride tracking who holds the White House. Democrats say their pride is conditional, but have behaved differently. Under George W. Bush, 58 percent of Democrats were extremely proud to be American, roughly 30 percent under Trump’s first term, a bounce back to only 34 percent during Joe Biden’s term and now only 14 percent. It is not a pendulum — it’s a ratchet. Each turn clicks down and doesn’t come back. The question the rest of this essay needs to answer: why doesn’t it recover?

III. THE MOSQUE

September 17, 2001 was six days after I watched a plane slam into the side of the South Tower of the World Trade Center. Bush, now a wartime president, walked into the Islamic Center of Washington and spoke these words: “The face of terror is not the true faith of Islam.” He went on to say that “Islam is peace.” And it worked — hate crimes against Muslims measurably declined in the following days. In what The New York Times Magazine called “the Speech that Made Obama,” on July 27, 2004 at the Democratic National Convention in Boston, the Illinois state senator then running for the U.S. Senate said “there’s not a liberal America and a conservative America. There’s the United States of America.”

Bush and Obama were the last presidents — and, perhaps, the last major presidential candidates — to perform empathy for the other side effectively, on the public stage, when it cost something. Biden had the instinct — the rope-line grief ministry is documented — but at some point he vacated the stage, and private consolation doesn’t scale into national narrative.

IV. THE ESCALATOR

It’s June 16, 2015 and Trump is coming down the escalator in Trump Tower to announce his candidacy. The Tea Party movement in the United States, with its us-against-them rhetoric, was prologue for the man who would manage to deride Mexicans, Democrats, a disabled reporter and anyone in his path over the coming decade. The Democratic pride collapse tracks the rhetoric era precisely, before any commission existed to rewrite American history or curriculum was being rewritten.

But this marks the dawn of a new type of so-called patriotism, one based on who was on your enemy’s list, as opposed to the things that bring us together. Bush spoke Spanish and welcomed the contributions of Mexican immigrants even as he was trying to solve the problems at the southern border. His behavior as a candidate and president was in sharp contrast to Trump and even to Hillary Clinton, who described half of Trump’s supporters as a “basket of deplorables.“

V. THE WHITEWASH

Why is it that the ratchet can’t turn back up? Much of Republican politics as of late has been focused on a culture war about what is American and what is un-American — the 1776 Commission presidential advisory committee to promote education focused on our achievements and whitewash our failures, the GOP-governor-driven attempts to purge schoolbooks of references to slavery and discrimination and even the park purges in the second Trump administration designed to remove references to some of the blemishes we once embraced as signs of our progress and the work we had yet to undertake.

Politics in recent years has focused on the worst of us, instead of what unites us or, as President Abraham Lincoln once put it, “the better angels of our nature,” but this is by no means the cause of the collapse — simply a beautification of the story already told. There is an interesting insight when you turn away from the polls the Silver Bulletin most prominently cites and toward data from Echelon Insights’ America 250 Survey. Seventy-five percent of Americans say they are proud of “what America has been in the past.” Only 44 percent say they are proud of “America as it is today.” And only 43 percent say they feel proud of “where America is heading as a nation.” And, in what feels like an astonishing number, 41 percent say that if the Founding Fathers could see America today they would feel “mostly disappointment.”

I would venture to guess that what’s actually happened is that the only patriotism now on offer is pride in a falsified country — and that half the nation has declined the bargain, because pride you’re ordered to feel, in a country you’re forbidden to see honestly, isn’t pride at all. It’s a city upon a lie.

VI. FIFTY

Today I am outside of Washington. It’s July 4, 2026. I am 50, and the country is 250. Instead of the packed Mall, there is algae in the Reflecting Pool, and soon there will be fireworks over a fair that 45 percent of American voters believe is being used as a political operation.

The real lesson of 1976 was that the joy wasn’t vindication of American perfection — it was the pride of a nation that had just walked away from a losing war and held its own president to account, and pride in self-correction is the only kind that survives contact with the record. It is harder to love a lie than a nation that has often failed and kept trying. A country can be loved, but an alibi can only be defended.

John Winthrop’s hill was a covenant community that banished Anne Hutchinson and Roger Williams for worshipping wrong — a moral example with a fence around it. Rouse, three centuries on, took the same aspiration and widened the circle rather than denying the exclusions that came before him. He didn’t tear Winthrop down and didn’t whitewash him. He revised him. He built a new city upon a hill, one that shows in miniature how America improves — by honest amendment — and shows why any whitewashing project is offensive precisely because it forbids the amending. Rouse is what “failed and kept trying” looks like with a shovel in its hand.

I take pride in that man, and in how deeply corrective America has been.

Jayson Blair is the host of the Silver Linings Handbook podcast and a former reporter for the Boston Globe and The New York Times.

Share

It was February 13, 2016 — at the Cibolo Creek Ranch in western Texas — and Supreme Court Justice Antonin Scalia lay dead in a hunting lodge in the desert. There was no autopsy — the county judge declared natural causes in a phone call that afternoon after consulting a physician. A pillow over the back of his head, near the headboard and not on his face, which ranch owner John Poindexter had to clarify, launched more conspiracy theories than Helen of Troy launched ships. One of these theories, straight out of a John Grisham novel, was from Donald Trump, the Republican candidate himself.

Senate Majority Leader Mitch McConnell announced that day that the seat would stay open because an election was nine months away and “the American people should have a voice in the selection of their next Supreme Court justice.” McConnell was able to orchestrate two moves here — denying President Barack Obama an appointment to the Court and galvanizing Republicans who feared a Democratic win in 2016 would result in a 5-4 liberal Court.

That same night, Republican presidential candidates in South Carolina litigated the question at a debate while Scalia’s body was still cooling in Texas. McConnell would win the battle, thwarting President Obama’s nomination of Merrick Garland. For years, the 5-4 Court had held the line against some liberal impulses, but with moderate Republican-appointed justices like Anthony Kennedy, David Souter and Sandra Day O’Connor, it produced decisions like Planned Parenthood v. Casey, which affirmed abortion rights, and Obergefell v. Hodges, which legalized same-sex marriage. The conservative Federalist Society, evangelicals and other conservatives felt Republican appointees kept defecting. They wanted betrayal insurance.

Four years later, on September 18, 2020, Justice Ruth Bader Ginsburg, after complications of metastatic pancreatic cancer, died in her home in Washington. It was the eve of Rosh Hashanah — a Jewish tradition holds that those who die at the very end of the year are the most righteous, held back by God because they were needed until the last moment. The election was 46 days away. Her final wish, as dictated days earlier to her granddaughter, Clara Spera, was that she “will not be replaced until a new president is installed.”

With an election approaching that appeared likely to result in Democrat Joe Biden becoming president, Senate Minority Leader Chuck Schumer reached for a familiar sentence: “The American people should have a voice in the selection of their next Supreme Court justice.” But in contrast to Scalia’s death, McConnell pledged that night to fill the seat. The vacancy was filled quickly — in 39 days, against the 293 days Scalia’s seat remained open — by Amy Coney Barrett, a judge from the United States Court of Appeals for the Seventh Circuit. Together with Neil Gorsuch — seated in the chair McConnell held open — Barrett’s confirmation turned what could have been a 5-4 liberal Court into a 6-3 conservative one.

I. NOT TRUMP’S COURT

The dominant debate about the October 2025 Supreme Court Term is whether the Court is Trump’s arm or a check on him. If you want to see one of the strongest arguments about what is happening with the Court this term, legal scholar Steve Vladeck makes the case here on his One First Substack: that the justices behaved like politicians in robes serving GOP interests. His argument got me thinking. While I understand the fixation on the question that others are making about whether the Court is an extension of Trump’s agenda, I would argue the construction measures the wrong axis. This is not Trump’s Court. It is the conservative movement’s Court — the enforcement mechanism of a Faustian bargain conservative legal elites and evangelicals struck with Trump in 2016, when they agreed to swallow a man many of them privately despised in exchange for the judiciary that they had spent 40 years designing.

If you read the term through that lens, the incoherence Vladeck catalogs resolves into discipline. Trump wins when he carries the conservative movement’s cargo. He loses when he freelances. The Court is not standing up to him or bowing to him — it is executing on the contract.

II. THE BARGAIN

Dr. Albert Mohler is an interesting man. The president of the Southern Baptist Theological Seminary opened this year by declaring, in response to an opinion piece by Hillary Clinton, that “empathy is not a thing” — that it isn’t real, that it is “a substitute for a real Christian morality.” If his philosophical views can come off as a bit muddled, his mission over the past 30 years has not been.

Mohler led a campaign in the 1990s against President Bill Clinton’s conduct in the Monica Lewinsky scandal, which included the Southern Baptist Convention’s 1998 Resolution on the Moral Character of Public Officials. Clinton, Mohler said, was “a living demonstration of the fact that character matters and that a lack of character can be fatal for leadership.” So when Trump arrived, Mohler applied the same standard: he called Trump a “sexual predator” and, in a Washington Post op-ed, “the Great Evangelical Embarrassment.” “If I were to support, much less endorse, Donald Trump for president,” Mohler said in 2016, “I would actually have to go back and apologize to former President Bill Clinton.”

By 2020, Mohler had reversed his position. In a livestream announcement, he pledged to vote for Trump that November — and to vote Republican for the rest of his life — pointing to the Gorsuch confirmation and the party’s stances on abortion and religious liberty as what changed his mind. He had named the price himself: a Supreme Court seat. And when Religion News Service later confronted him with his 2016 statement and asked whether Clinton would now receive the apology his own logic demanded, he declined.

Mohler is widely considered one of the most controversial figures in the Southern Baptist Convention, especially compared to leaders like Bryant Wright, a former SBC president and senior pastor of the church I attended as a child in Marietta, Georgia, who embodied a far more irenic, non-partisan, consensus-building style. But Mohler’s switch was less a conversion than a catching-up. Approximately 80 to 81 percent of white evangelicals had already voted for Trump in 2016, according to PRRI and the Pew Research Center; the movement’s conscience surrendered last. In 2011, only 30 percent of white evangelicals said they believed that an elected official who commits an immoral act in their personal life could still behave ethically and fulfill their duties in their public and professional life, according to a PRRI and Brookings Institution poll. By October 2016, 72 percent believed it was possible. The 2016 survey landed just weeks after the Access Hollywood tape where Trump said that women would allow men in power to sexually assault them. The theology has not snapped back, as white evangelical tolerance remained high in 2018 and 2020. Mohler’s impact is amplified because, as president of Southern Seminary, he is a gatekeeper with enormous influence over the SBC’s pastoral pipeline, its intellectual culture, and its training. And the seller understood exactly what was being sold. Steve Bannon, who ran the 2016 campaign in its final months, later described the strategy: hit “the evangelicals and Republican establishment with the judges, with the judges, with the list” — every day. Both the buyer and the seller described it as a transaction.

David French argued in the other direction. He believed his fellow evangelicals did not need to trade their moral credibility for judicial victories — that a different path was available, one that preserved character-based witness and still pursued conservative judicial outcomes over time. A bargain with the godless to reach a Christian victory felt to him more like a bargain with the Devil. French argued evangelicals could have held the line they drew for Clinton — declared character essential, refused the candidate, voted third party or abstained — and trusted that the battle could be won, with God’s direction, another day. He did exactly that himself. The movement’s institutions made him pay for it.

A similar battle was occurring within the conservative legal movement.

In May 2025, Trump called a man most Americans have never heard of, Leonard Leo, a “real sleazebag” and a “bad person.” Leo is the same man that Trump once credited with what he would later call “one of the greatest achievements” of his presidency. He described that goal in a June 2016 interview with Breitbart News as appointing “great judges, conservative, all picked by the Federalist Society.” Months before that interview, in March 2016, Trump had a meeting with Leo, then the executive vice president of the Federalist Society, and they hammered out a deal. The society is a conservative group that advocates for legal originalism, and had long complained of the defections of justices like Souter, O’Connor, Kennedy and John Paul Stevens, one of the Court’s most liberal justices, who had been appointed by Gerald Ford. As TIME Magazine put it in a 2025 article, “True to his promise, Trump would go on to consult Leo for judicial nominations, including his three Supreme Court picks.” During Trump’s first term, Leo worked closely with White House Counsel Don McGahn to vet candidates, including Gorsuch, Barrett and Brett Kavanaugh.

But by May 2025, Trump had changed his tune, and he didn’t hold back, posting on his social media site Truth Social:

“I was new to Washington, and it was suggested that I use The Federalist Society as a recommending source on Judges. I did so, openly and freely, but then realized that they were under the thumb of a real ‘sleazebag’ named Leonard Leo, a bad person who, in his own way, probably hates America, and obviously has his own separate ambitions. He openly brags how he controls Judges, and even Justices of the United States Supreme Court—I hope that is not so, and don’t believe it is! In any event, Leo left The Federalist Society to do his own ‘thing.’ I am so disappointed in The Federalist Society because of the bad advice they gave me on numerous Judicial Nominations. This is something that cannot be forgotten! With all of that being said, I am very proud of many of our picks, but very disappointed in others.”

It was during a time when Trump was getting pushback on some of his most aggressive executive orders, on topics ranging from the Fourteenth Amendment’s guarantee of birthright citizenship and his immigration crackdown to National Guard deployments in American cities and his global tariff programs. TIME noted that a “federal three-judge panel had just ruled that Trump did not have the authority to impose most of the tariffs,” which are deeply unpopular with mainstream conservatives, even those who are Trump supporters.

Leo’s response was muted, expressing gratitude for Trump’s appointments during his first term, nothing more. But something curious happened over the next year. Trump became remarkably blasé about even making nominations to the federal bench, as Bloomberg Law News noted in December 2025 and CBS News echoed in March 2026. White House Deputy Chief of Staff Stephen Miller took it further, saying “We’re not going to be using the Federalist Society to make judicial nominations at all going forward,” and the nomination of Emil Bove, one of Trump’s former personal attorneys, to the US Court of Appeals for the Third Circuit signaled the new model — loyalists over movement credentials. It appears to have become clear to Trump that he did not make the deal he thought he did — that in return for placing conservative favorites on the bench, they would do what he wanted.

The divorce proves the marriage was transactional. And it doesn’t appear reconciliation is on the horizon.

III. THE PAYOFF

This term highlighted the payoff. In the 1980s, Chief Justice John Roberts — then a 20-something attorney in the Reagan Justice Department — wrote memos opposing the effort to expand Section 2 of the Voting Rights Act of 1965. He lost that battle in Congress in 1982, when the House and Senate reshaped and reaffirmed Section 2. Shelby County v. Holder in 2013 was the first installment of the correction, gutting the preclearance system that required states with histories of discrimination to seek federal approval before changing their voting rules. Louisiana v. Callais completed the effort, narrowing Section 2 so deeply that it would be almost impossible for minority voters to prove discrimination in the drawing of congressional districts. It is a 44-year arc from a junior lawyer’s memo to a Chief Justice’s majority opinion. And as Vladeck notes in his essay, the Court itself ordered reargument in Callais to expand its scope into a constitutional referendum neither party had asked for.

In Abbott v. LULAC, Texas undertook an unusual mid-decade redistricting designed to net the GOP as many as five additional House seats in the 2026 midterms. On the emergency application calendar — known as the shadow docket — the Court stayed a lower court’s determination that the Texas map was based on race. Then it did something stranger: it summarily reversed the merits in a one-paragraph order that cited nothing but its own thinly explained stay. The shadow docket became its own precedent. If Callais removed the statute constraining the maps, Abbott protected the maps themselves, closing a long-running thread in the conservative legal project to get the federal government out of supervising how Southern states manage their map-making and elections — supervision born of their resistance during the Civil Rights Movement. The Court’s posture in these cases is hear no evil, see no evil — as if Republican operative Lee Atwater had never explained, on tape, how the Southern strategy’s vocabulary evolved from slurs into euphemisms like forced busing and states’ rights, and as if “partisanship” were not simply the newest word on the list.

In 1988, in Morrison v. Olson, Scalia wrote the lone dissent against the constitutionality of the Independent Counsel Act, which allowed Congress to limit the President’s removal power over certain inferior officers. Scalia argued that the Act violated the Constitution by encroaching on the President’s exclusive executive power. His dissent became part of the unitary executive project incubated in the Federalist Society for four decades, with the target always being the Supreme Court’s 1935 Humphrey’s Executor decision, which limited the President’s power to remove officials from independent regulatory agencies. Trump v. Slaughter — a case that began in 2025 when Federal Trade Commissioner Rebecca Slaughter challenged Trump’s attempt to remove her without cause — provided the opportunity. The arc began with Franklin Delano Roosevelt firing an FTC commissioner, William Humphrey, and ended with Trump firing an FTC commissioner, Rebecca Slaughter. The Court’s decision to let the firing stand overruled 91 years of precedent and completed the project.

In the second Trump term, the justices are completing the conservative movement’s larger design — the one whose earlier installments included Dobbs v. Jackson Women’s Health, which overturned the abortion rights recognized in Roe v. Wade, and the Court’s expansion of the Second Amendment.

IV. WHERE THE CONTRACT HAS LIMITS

It would be easy to look at the cases Trump lost this term — over tariffs, birthright citizenship and the Federal Reserve — as the Court demonstrating its independence from the President. But a more coherent frame is that the Court was never interested in underwriting Trump’s heresies.

Tariffs violate more than 50 years of conservative free-trade orthodoxy, and the Chamber of Commerce wing of the Republican Party was never going to buy the International Emergency Economic Powers Act as a blank check for something it vehemently opposed. Restricting birthright citizenship was right-wing talk-radio material, never something that would pass muster in a law review. While some originalists believed steps could be taken to limit what they called “birth tourism,” the project had no serious originalist infrastructure behind it. And in Cook v. Trump, released the same day as Slaughter, the Court exempted the Federal Reserve from the removal power it had just handed the President — a carve-out protecting the one institution the donor class actually requires, and a signal that principle stops at the water’s edge of where the justices believe Trump might tank the economy.

The test that predicts every outcome is not “does Trump want it” but “does the conservative movement want it.” Vladeck’s catalog is right; his axis is wrong. Both sides of the debate assume Trump is the axis. He is not. The axis is the movement that gave us Jerry Falwell, Pat Robertson, Oral Roberts, Ronald Reagan, Paul Clement, Neil Gorsuch, Amy Coney Barrett and Brett Kavanaugh.

V. THE GHOST AT THE FEAST

Scalia may have lost in many dissents, but he won the war: Slaughter is Morrison’s dissent become law. The Second Amendment expansion he set in motion in District of Columbia v. Heller has moved apace. He joined the majority in Shelby County, and the destination Callais reached would not have troubled him — consistent with his long-standing emphasis on restraining federal power. But the route would have. The Court reaching out to order reargument on a constitutional question no party had raised, then rushing its new rule into effect for an election already underway, is exactly the judicial adventurism a formalist spends a career condemning. Scalia might well have cheered where his heirs arrived. He would not have blessed how they traveled.

He won the war, but he would have despised the way his army fought. The Federal Reserve carve-out in the Slaughter and Cook decisions is naked functionalism. If Congress made a deal with the President to create independent commissioners he could not easily remove, and that deal was unconstitutional, Scalia would have struck the arrangement down entirely. This Court instead let the zombies stand — agencies Congress built to be independent, now run at the President’s pleasure, a creature no Congress ever enacted.

In his Morrison dissent, Scalia wrote that threats to the separation of powers usually arrive disguised — clad, he said, “so to speak, in sheep’s clothing.” The independent counsel was different: “this wolf comes as a wolf.” But this Court is the wolf coming as a wolf in reverse — the disguise is now the doctrine, deployed by his own heirs, deciding separation-of-powers cases not by principle but by their perceptions of what the financial markets need. Kavanaugh’s concurrence in Cook says it plainly: “Even temporary uncertainty about the status of the Federal Reserve could spark political upheaval ... as well as turmoil in the U.S. and world economies.” That is a justice explaining a constitutional carve-out in the vocabulary of a market analyst. Even the shadow-docket bootstrap in the Texas case is a process corruption a formalist could not have blessed merely because the outcome pleased him.

The movement got its Court, but Scalia didn’t. Method and consistency were the whole point for Scalia; for his heirs, they are obstacles. The difference between outcomes and jurisprudence, even when they produce similar results, spans an ocean.

VI. GOVERNING AGAINST THE CLOCK

This all raises the question: Why does the Court keep choosing this politically damaging path — one that does not help them with Trump, who operates a zero-sum game when it comes to loyalty, with people who believe the Court should be intellectually consistent or with those who believe they should be above the political fray?

The 6-3 supermajority is an actuarial accident and they know it: Ginsburg’s death, McConnell’s calendar games. This is a Court that knows it is out of touch with the American people and that its time is limited. The movement knows windows close — they watched Roe stand for 50 years because their predecessors moved too slowly. This discomfort for the public and some of the justices themselves is merely the cost of shipping the backlog of the conservative project while the votes still exist.

The Court looks reckless only if you think its audience is the public, or even Donald Trump. It looks disciplined if you understand its audience is the movement that built it.

VII. UNEXPECTED FRIENDS

To the surprise of many, Antonin Scalia and Ruth Bader Ginsburg were the closest of friends, both great legal minds, often on the opposite sides of the courtroom. About Texas v. Johnson, Scalia once said in a public comment that “If it were up to me, I would put in jail every sandal-wearing, scruffy-bearded weirdo who burns the American flag. But I am not king.” He voted to strike the Texas law down. Ginsburg famously said that Roe’s reasoning in support of abortion rights based on the right to privacy was wrong and that the case should have been decided on equal protection and women’s equality grounds. What they shared, beyond a love of opera, was a willingness to separate what they believed was right from their preferred outcomes.

The Court their deaths produced is one neither would recognize as theirs — his for its method, hers for everything else. This Court is an institutional empire built on two accidents of mortality, governing like it knows it.

The bargainers got everything they contracted for — Dobbs, Section 2, the unitary executive. The tariffs and birthright rulings are the bargainers discovering that the Devil reads contracts carefully too: he pays exactly what’s owed and keeps everything else. A Court that serves a movement rather than a Constitution is no better than a Court that serves one man. It can deliver victories but not legitimacy, and legitimacy was the one thing it could not put in escrow.

The Court is racing one clock — the next president, who may fill seats or move to expand the bench. One thing that has changed since 2016 is Democrats’ willingness to play ruthless hardball. But there is another clock. The conservative movement and the justices may feel they were the ones who got the better bargain from the man who wrote The Art of the Deal, but it would be unwise to count Donald Trump out when it comes to creative ways to extract vengeance. All of this has placed the Court and the conservative legal movement in a precarious trap of their own design: caught between a public that hates what they built and a partner who never forgets a slight. The movement spent forty years carefully drafting an insurance policy against betrayal. It never wrote a clause for the man it hired.

Jayson Blair is the host of The Silver Linings Handbook podcast.

It can be reasonably inferred that enslaved and free Black Philadelphians were among those gathered that day — or that the words reached them shortly after. They heard how those truths were “that all men are created equal, that they are endowed by their Creator with certain unalienable rights, that among these are life, liberty and the pursuit of happiness.” They next heard that the government now existed to secure those rights.

I try to imagine what it must have been like to hear those words and not be free.

Self-evident is the most contested phrase in American history. Things that are actually self-evident do not require a Civil War to establish. They do not require a Second Founding in Reconstruction. They do not require Brown, Loving or Obergefell. They don’t create Jim Crow. They don’t get rolled back by Lee Atwater’s Southern Strategy or challenged by the current administration.

The evidence suggests those truths are not self-evident at all.

And yet, the document was not describing the world, it was making a wager. The men who wrote it were aspirational or hypocritical, or both — and it does not matter which. What matters was what the words became in the hands of people who heard them and weren’t free.

In his “Letter from Birmingham Jail” in 1963, Martin Luther King Jr. wrote that the nation had defaulted on its promissory note to Black Americans — that the Declaration and the Constitution were a “bad check” but that he refused to believe there were “insufficient funds.”

In the Seneca Falls Declaration in 1848, Elizabeth Cady Stanton opened by lifting Thomas Jefferson’s language almost verbatim, “We hold these truths to be self-evident: that all men and women are created equal.”

At the first Montana Indian Affairs Conference in 1951, tribal leader Tom Main drafted what he explicitly titled “An Indian Declaration of Independence,” where he wrote of believing “in the future of a greater America, which we were first to love” and where “life, liberty and the pursuit of happiness will be a reality.”

What also matters was what the words became in the hands of people who heard them and were free — but believed freedom should ring.

In his 1858 speech on the Declaration, Abraham Lincoln, before he became President, argued that the document was meant to be “a stumbling block to all those who in after times might seek to turn a free people back into the path of despotism.”

As in every year since our founding, those ideas are being contested right now — through the rollback of voting rights and diversity laws meant to enable the Fourteenth Amendment, the judicial rollback of the Obergefell era, and the notion that a person born with a boot on their neck can compete fairly with a person born with a silver spoon in their mouth.

Yet it is that notion — that the Declaration was always aspirational and never actual — that gives me confidence. Judges, state attorneys general, ordinary Americans refusing to let go are the ones who ensure the movement. Like the Colorado River carving its way through rock, there will be obstacles. But the current finds its way.

What Colonel Nixon read aloud that day was not a conclusion. It was an opening argument — still being litigated, still being answered, 250 years on.

But if you visited Washington in the days after June 9th, 2026, you saw something different — and then something worse, unless your favorite color is the dye they use on the Chicago River on St. Patrick’s Day. The National Park Service had just completed a remediation effort, refilling the pool with what they proudly called “American flag blue” water, a deep, vivid color that made the monument’s reflection look almost theatrical in its clarity. It lasted perhaps two or three days. Then the green came back — first a deep, dark green, then the lighter, murkier hue that signals algae has fully taken hold. The blue was not a restoration. It was a paint job. By yesterday, reporters were doing live shots of pulling up blue paint that had peeled off the bottom of the greenish-yellow pool.

In the weeks before, Democrats were bemoaning the blue coloring of the pool while Republicans were lambasting them for being un-American, as if aesthetic taste were some measure of partisan loyalty rather than one of the dumbest litmus tests in recent memory — a Rorschach for whether you liked Donald Trump or loathed him.

But what makes this particular story worth examining is not just the algae or even political division or our fractured world. It is who got the contract to fix it. Atlantic Industrial Coatings, a Northern Virginia company, received a $14 million no-bid contract for the remediation work — the same company that President Trump has said does work on swimming pools at his golf course in Sterling, Virginia, just down the road. What a coincidence that they were selected. The Federal Acquisition Regulations, which govern how the government spends public money, generally require that when three or more qualified contractors can be identified to perform work — the so-called “rule of three” — the contract must be competed. In most cases, a no-bid award of this scale would require a specific, documented justification for bypassing that competition. No such justification has been made public.

Political scientists have a framework for understanding exactly how this happens — and why it always ends the same way. It involves three interlocking concepts: the Loyalty Dilemma, the Competency Penalty, and, ultimately, the Loyal Loser.

The Loyalty Dilemma holds that in order to run a corrupt administration or dictatorship, a leader requires absolute loyalty and a willingness from subordinates to break rules, bend laws, or look the other way. The Competency Penalty holds that highly competent, effective, and intelligent professionals generally have excellent career prospects outside of a corrupt government — and they tend to possess independent critical thinking and a baseline ethical standard, making them dangerous or simply uncooperative. And so, we arrive at Loyal Loser Selection: corrupt leaders must intentionally hire people whose primary qualification is unwavering obedience. Because these individuals are hired for compliance rather than capability, they are usually deeply ineffective at actual governance, project management, and public service.

History offers unambiguous examples. During Stalin’s Great Purge of the late 1930s, he systematically executed or imprisoned his most competent military commanders, engineers, and scientists out of paranoia that their intellect and independence posed a threat to his power. He replaced them with lower-performing, fiercely obedient ideologues. The result nearly led to total collapse during the initial Nazi invasion of 1941, as the military lacked capable strategic leadership. In Venezuela in the early 2000s, Hugo Chávez demanded absolute political loyalty within the state-owned oil company that funded the entire country. Following a worker strike in 2002, he fired over 18,000 experienced engineers, technicians, and managers, replacing them with regime loyalists and military personnel who had no background in petroleum engineering. The industry collapsed, taking the national economy and the electric grid with it.

You can see the seeds of this pattern closer to home — in Kristi Noem’s failures as Homeland Security Secretary to meet the White House’s own deportation targets, in the Justice Department’s stumbling pursuit of the administration’s legal agenda, in Tulsi Gabbard’s difficulties reforming an Intelligence Community that requires precisely the kind of institutional expertise she was hired to dismantle, and in Robert F. Kennedy Jr.’s singular triumph of Making Measles a Thing Again, as opposed to Making America Healthy Again.

Here is the mechanism, stated plainly: when a leader prizes loyalty over competence, they systematically select for the wrong people. The genuinely talented — people with real skills, independent judgment, and professional reputations they have spent years building — tend not to make very good sycophants. They have options. They have standards. They will push back when an idea is bad, because their name is attached to the outcome. Competence, in this sense, is structurally incompatible with pure deference.

Sycophancy, on the other hand, requires a particular psychology. It requires someone whose self-interest is best served by flattery rather than performance, someone whose value to the patron is social rather than functional. These are not, as a rule, the people you want running critical institutions. They are the people you want at dinner parties, if you have low standards for dinner conversation.

What happens when you build a government — or a company, or any institution — on this principle is almost mathematically predictable. You get cronies in top jobs. They fail, because the job was never designed to reward what they’re actually good at, which is making powerful people feel good about themselves. The failure then creates a crisis. The crisis demands a response. But the only people positioned to respond are the same pool of loyalists, because the leader cannot afford to be seen bringing in competent outsiders who might implicitly indict the original appointments.

And so the algae spreads.

There is, if you are looking for one, a dark comfort in this. Authoritarianism and corruption almost always require idiocracy as a structural feature — and idiocy is reliably self-defeating. The authoritarian impulse may be ruthless, but it is poorly executed almost by definition, because the people trusted to execute it were hired for their loyalty, not their competence. The system contains the seeds of its own failure. That is not a reason for complacency. It is simply a reason to keep watching.

There are volumes of literature on this dynamic — organizational theorists call it adverse selection, political scientists write about the authoritarian’s dilemma, management scholars document how cultures of fear suppress the upward flow of accurate information. What is remarkable is not that we understand this intellectually. It is that societies keep having to relearn it experientially.

The reflecting pool metaphor is apt in another way, too. What authoritarianism ultimately destroys is the capacity for accurate self-reflection. Healthy institutions have feedback mechanisms — inspectors general, free press, competitive markets, genuine legislative oversight — that function like clear water, allowing society to see itself as it actually is. Cronyism corrodes these mechanisms not all at once but gradually, appointing loyalists to positions designed to provide accountability, until the entire reflective surface is compromised. You cannot see the monument anymore. You see only the green haze.

The tragedy is that the leaders who build these systems often believe, at some level, that they are strengthening themselves. Surrounding yourself with people who never challenge you feels like security. It feels like control. What it really is, of course, is a slow-motion delegitimization of everything you are supposed to be running. The algae does not strengthen the pool. It infects it.

We are at a moment in American public life when these dynamics are unusually visible — not because they are new, but because the pace and scale of cronyism have outrun the usual mechanisms for managing its consequences. The failures are accumulating faster than the narrative can absorb them. The pool is green.

The question worth sitting with is not merely how we got here, which is a story with a long and bipartisan preamble.

The question is whether the institutions designed to drain and clean the pool still have the capacity to do so — and whether the people with the standing to demand it will find the will before the reflection is lost entirely.

In June, a federal judge ruled that the government had canceled Environmental Protection Agency grants solely because of the political makeup of the states receiving them. As the Secretary of Energy was telling Congress he had not done the same thing, federal attorneys were down the street saying that, in fact, he had — and the judge relied heavily on the Secretary’s own earlier admission that he had done exactly that.

And in the Caribbean, the Department of Defense is firing missiles at boats and killing people for reasons no one can explain without sounding uncomfortably close to, “yes, we are committing a war crime.”

Taken together, these episodes do not read like the normal friction of governance. They read like institutions that have lost their internal gyroscope — agencies acting without coherence, without restraint, and increasingly without the public’s trust.

These are not isolated scandals. They are symptoms of a deeper erosion across the last decade, one where institutions no longer command trust because they no longer behave in ways that are trustworthy.

The patterns are clear — the prosecutorial overreach and incompetence, the regulatory capture and procedural shortcuts, agencies acting without public consent or internal dissent. This is not a partisan critique. It is one about legitimacy.

II. Why Legitimacy Is Harder to Rebuild Than Power

Power can be seized. It can be delegated, inherited, concentrated, or redistributed by fiat. A new administration takes office and, almost overnight, the levers of government shift hands.

But legitimacy doesn’t work that way. Legitimacy must be earned, and the earning is slow, incremental, and frustratingly fragile. It accrues through procedural fairness — through the sense that rules apply equally and that outcomes, even unwelcome ones, emerged from a process that was honest. It requires transparency — not the performance of openness, but the actual willingness to show your work. It demands competence, because an institution that cannot do what it claims to do has no basis for asking trust. And it requires something rarer still in the current climate: humility — the institutional capacity to say, “we were wrong,” we are sorry, we will learn from it, and mean it.

Consider the Department of Justice, once the closest thing America had to a secular temple of institutional legitimacy. Grand juries are refusing to indict in cases where the government’s political motives are too visible to ignore. Federal judges — including appointees of this administration — are issuing rebukes that read less like legal opinions than formal expressions of personal and institutional alarm. Career prosecutors are resigning, not in ones and twos, but in numbers that represent an entire generation of institutional memory walking out the door. The National Security Division section handling espionage cases has lost 40 percent of its prosecutors. Each of these may not be, on its own, a significant event. Together, they describe a vicious loop that, once started, is very hard to stop.

The FBI’s story is, if anything, starker. The bureau has historically run an attrition rate of roughly 0.4 percent — a figure former Director Christopher Wray once cited before Congress as nearly impossible to find anywhere in American public or private life. People did not leave the FBI. They built careers there, sometimes across decades, and they left in manageable numbers when retirement made it unavoidable. What happened beginning in early 2025 was something the bureau had never experienced. Senior executives were purged in waves — field office leaders, counterterrorism chiefs, the heads of cyber and national security divisions — often without stated cause and without the due process procedures the FBI had long applied even to agents accused of serious misconduct. Career agents who had worked Jan. 6 cases or the Mar-a-Lago documents investigation were targeted specifically, told to leave or be fired. Facing a depleted workforce, the bureau responded by shortening its Quantico training academy, waiving written assessments for support staff seeking to become agents, and promoting people into senior leadership roles without the headquarters experience that had historically been considered essential. The FBI’s standards for who gets in and who rises were not incidental features of the institution. They were the institution. And the response to losing them was to lower the bar.

It would be convenient — and wrong — to treat this as a story that began in January 2025. The Biden Administration left its own marks on these institutions. The Justice Department’s handling of the Hunter Biden investigation was a case study in the appearance of managed justice: a plea deal that collapsed under judicial scrutiny, a special counsel appointment that came only after political pressure made delay untenable, and career prosecutors in Delaware who felt their work was being shaped by considerations above their pay grade. The Afghanistan withdrawal was an institutional competence failure on a generational scale, one that damaged the credibility of the military, the intelligence community, and the foreign policy apparatus simultaneously. And more quietly, the Biden years continued a bipartisan tradition of stretching executive authority in ways that made the next stretch easier to justify — each administration inheriting a slightly larger set of unilateral tools and leaving them a little larger still for whoever came next. The erosion of institutional norms is not a partisan project. It is a cumulative one. What is different now is the scale, the speed, and the stated intent — the open declaration that the goal is transformation rather than governance. But the kindling was laid over years, by hands of both parties. Each side brought matches.

The loop works like this: when institutions behave in ways that are arbitrary, self-serving, or plainly corrupt, the public stops trusting them. When the public stops trusting them, cooperation declines — witnesses don’t come forward, defendants don’t accept plea offers, communities don’t share intelligence with law enforcement. When cooperation declines, effectiveness falls. And when effectiveness falls, the institution reaches for cruder instruments — more coercion, more spectacle, more punitive excess — which destroys what little trust remains. The loop tightens. The institution becomes less itself, and less able to remember what it was supposed to be.

This is not a new story in American history. But we are at a moment when the loop is spinning fast, in multiple agencies, simultaneously. And the question of how to break it — how to step off the cycle before it becomes permanent — is not receiving nearly enough serious public attention.

III. The Hidden Constituency of Hope: Former Federal Workers Imagining a Better Future

Here is something I have come to believe, after several years of watching this unfold: the most serious thinking about institutional restoration is not happening inside the agencies. It is happening outside them, among the people who left — or were pushed out — and who are now free to say what they actually think.

They are former FBI agents who spent careers building exactly the kind of trust with communities that is now being squandered. They are former DOJ line attorneys who know the difference between a legitimate prosecution and a political one, because they’ve run both past their supervisors and watched what got approved. They are former USDA scientists who understand that the destruction of institutional expertise is not a cost-free shortcut — that the knowledge of how to do things safely, carefully, and well does not simply regenerate when the political winds shift. They are former career civil servants from agencies you have barely heard of, people who spent decades mastering the craft of public administration, who watched their agencies hollow out from the inside and eventually concluded that leaving was the only honest thing left to do.

What distinguishes these people is not bitterness, though some of them have earned bitterness. What distinguishes them is specificity. They know what “good” actually looked like — not in the abstract, but in practice, on a particular Tuesday afternoon in a particular conference room, when a supervisor told a line prosecutor that a case wasn’t strong enough and the case was dropped. They know where the rot set in: the first promotion that went to the political loyalist rather than the most capable person, the first briefing that was shaped to tell the principal what he wanted to hear rather than what was true, the first time an internal dissent channel was used and the person who used it quietly found their career stall. They know what structural reforms would actually matter, not because they have read think-tank reports about institutional reform, but because they lived inside the institutions and watched specific things break in specific ways. And critically, they are no longer constrained by the internal politics or personal fear that kept them quiet when they were inside.

What has emerged, informally and without much fanfare, is something that might be called a government in exile — not a shadow government with political pretensions, but a diaspora of institutional knowledge and moral seriousness that is having conversations, forming working groups, drafting proposals, and keeping alive the idea that public service, done correctly, is among the most honorable things a person can do. These networks are not coordinated by any single organization or funded by any single donor. They are the natural clustering of people who share a common loss — the loss of institutions they believed in — and a common determination to figure out what it would actually take to rebuild them.

IV. What Re-Legitimization Actually Looks Like (and What It Doesn’t)

Let me be precise about what institutional restoration does not look like, because the confusion here is both common and consequential.

It does not look like a new president. Presidential transitions can halt some of the damage and send important symbolic signals, but the institutions that are currently failing were degrading across multiple administrations, under presidents of both parties. A new president who inherits DOJ cannot simply will the career prosecutors back into existence, or restore the confidence of a grand jury pool that has seen too much, or rebuild the trust of communities that have been surveilled, prosecuted, and lied to. A new president is a necessary but radically insufficient condition for institutional recovery.

It does not look like a new attorney general, a new strategic plan, a new oversight office, or a new round of congressional hearings. These are the institutional moves that institutions make when they want to appear to be addressing a problem without actually having to change anything fundamental. Washington has produced thousands of such documents and dozens of such hearings. The people who drafted them are often smart and well-meaning. They do not work, as a class, because they treat the symptoms rather than the culture, and because culture — the shared understanding of what we do here and why it matters — is not amenable to reorganization.

What institutional restoration actually looks like is slower, harder, and in many ways more interesting than any of these shortcuts.

It looks like rebuilding due-process norms from the ground up — not as legal requirements to be gamed, but as genuine commitments to treating people fairly, even people the institution finds inconvenient. At DOJ, this means reviving the culture in which an assistant U.S. attorney’s first obligation is to justice rather than to the conviction rate, and in which supervisors are evaluated not on how many indictments they secure but on how carefully they distinguish prosecutable cases from politically useful ones.

It looks like re-professionalizing hiring and promotion — which means, bluntly, that the people who get ahead inside these institutions need to be the people who are best at the mission, not the people who are best at managing up or at performing ideological loyalty. This sounds obvious. It is not practiced. At the EPA, where scientific expertise was systematically devalued in favor of political reliability, the recovery will require not just rehiring scientists but rebuilding the internal culture in which scientific disagreement is treated as a feature rather than a threat.

It looks like restoring internal dissent channels — the formal and informal mechanisms through which people inside institutions can say, “I think we are making a mistake,” without ending their careers. These channels were not, historically, radical innovations. They were normal features of well-functioning bureaucracies. Their systematic destruction over the last several years is one of the most underappreciated forms of institutional damage, because dissent is how organizations learn, and organizations that cannot learn from their mistakes compound them.

It looks like re-establishing bright lines between political and career roles — the distinctions that are supposed to prevent the machinery of government from becoming an instrument of factional politics. At USDA, where career scientists found themselves being asked to produce research conclusions that matched predetermined political positions, the recovery requires not just new rules but a renewed understanding of why those rules existed in the first place: because expertise that is politically contingent is not expertise, and an institution that cannot distinguish between the two is not an institution.

And it looks like re-centering mission over metrics, re-training leaders in humility and accountability, and — perhaps most fundamentally — re-creating a culture where telling the truth is rewarded rather than punished. The last of these is the hardest, because cultures do not change by decree. They change when enough individuals, over enough time, make the choice to behave differently, even when it costs them something. Those choices aggregate. They become norms. Norms become culture. Culture becomes what an institution is.

V. The Moral Core: Institutions Are Human Creations, and Humans Can Rebuild Them

There is a temptation, when surveying the current landscape, to reach for structural explanations — to locate the failure in systems, incentives, and path dependencies rather than in the choices of individuals. Structural explanations have real explanatory power. But they can become a form of evasion, a way of talking about institutional failure that relieves any particular person of responsibility for it.

The more uncomfortable truth is this: institutions fail when the people inside them stop believing in their purpose. Not all at once, and rarely dramatically. It usually happens gradually, through a long series of small capitulations — the briefing that was softened, the case that was declined, the colleague who was not defended, the truth that was told only partially, in a form that could not be misused. Each of these choices is individually understandable. Cumulatively, they are catastrophic. They produce institutions that have retained all the external features of legitimacy — the seals, the buildings, the congressional authorizations — while losing the internal substance that made those features mean something.

The corollary is equally true, and considerably more hopeful: institutions recover when the people inside them — or people who remember what they were meant to be — decide that the purpose is worth fighting for again. The skills required to rebuild legitimate institutions are not exotic or newly invented. They are old, well-understood, and painstakingly documented in the institutional memory of the people who are now outside the buildings. The moral imagination required — the capacity to hold in mind what a functioning institution of public service looks like and to be guided by that image rather than by the incentives of the present moment — still exists. It is just not, at the moment, concentrated in the places where it is most needed.

That is a problem with a solution. Solutions to it exist. And the people who are working on them — quietly, without much institutional support, and with no guarantee that the work will matter in time — represent something genuinely important in the current moment: proof that the knowledge has not been lost, only misplaced.

VI. The Center Holds Because People Hold It

The center does not hold because it is strong. It is not, at the moment, particularly strong. It holds — to the extent that it holds — because people choose to hold it, even after it has failed them, even when the holding is costly and the outcome uncertain.

The former federal workers imagining a better future are not naive, and they are emphatically not nostalgic. The institutions they served were not perfect. Some of them were seriously flawed in ways they now understand more clearly than they did when they were inside. What they are is architects — people who understand that the building is damaged and who are thinking carefully, in detail, about what it would take to repair it rather than simply mourn it.

The work they are doing now, in the quiet, is the first draft of the institutions we will need when the crisis finally breaks open — when the accumulated costs of institutional failure become impossible to ignore and the political system is finally ready to hear something other than the next round of claims about whose fault it is. That moment will come. It always does. The question is whether, when it comes, we will have preserved enough of the knowledge, the moral seriousness, and the practical skill to act on it.

I believe we will. Not because I am optimistic by disposition — I am not, particularly — but because I have talked to enough of the people who are doing this work, in enough settings, to know that they are serious, that the work is real, and that the institutional memory they carry is exactly the kind that cannot be manufactured from scratch when you need it. It can only be maintained, carefully, over time, by people who have decided that the maintenance is worth the trouble.

Nostalgia is the mourning of something we once believed in and have lost.

That is not nostalgia. That is the beginning of something.

Google may have the best data on what we want. Meta, by most measures, has the most behaviorally rich data on what we do.

And yet, recent Wired articles documented more than once this week, the company that arguably knows more about human behavior than any institution in history has struggled, persistently and visibly, with its own internal culture and executive leadership. The irony is almost too neat to be believed. Almost.

The explanation is not complicated, but it is counterintuitive: Meta’s data is not a map of human psychology. It is the world’s most comprehensive record of human psychology under stress. And you cannot build a healthy culture out of that.

The companies that understand this better are, not coincidentally, winning the race to build artificial intelligence. The ones that don't are falling behind — and struggling to understand why.

The Best Dataset on What We Do When We Are Unwell

Begin with what Meta’s data actually measures. The platform’s business model depends on engagement, and engagement is reliably driven by a narrow band of emotional states — boredom, loneliness, anxiety, envy, outrage, impulsivity. Not contentment. Not reflection. Not connection in any durable sense. The algorithm rewards dysregulation because dysregulation keeps people scrolling.

This means Meta’s behavioral data is not a neutral portrait of human nature. It is, more precisely, the most detailed dataset in existence on what human beings do when they are not okay.

That creates a problem with no clean exit. If Meta tried to reverse-engineer its engagement data to build a model of psychological health — essentially inverting the signal to find what to cultivate instead of what to exploit — it would face three simultaneous obstacles. The first is technical: inverting a contaminated signal doesn’t produce a clean one, it produces a different contaminated one. The second is psychometric: the inverse of a dysfunction is not automatically its corresponding virtue. In personality science, the most important insight about someone who scores very high on a scale is not simply the mirror of the insight about someone who scores very low — they are different failure modes, not opposite ends of a single lesson. Knowing what breaks people does not tell you, in any reliable way, what makes them whole. The third obstacle is institutional: openly acknowledging what the data represents would be a tacit admission to employees, and a gift to plaintiff’s attorneys, that the company’s core business runs on human distress.

The trap is structural. Meta cannot use its data honestly without implicating itself.

The Same Behavior, Opposite Meanings

Even setting aside what the data measures, there is a more fundamental problem: behavioral data cannot tell you why people do what they do.

In personality science, practitioners work with a clarifying analogy. Two managers might both refuse to delegate. One refuses because she is a perfectionist — high in conscientiousness, driven by pride in craft. The other refuses because he is disorganized — low in conscientiousness, unable to prioritize or plan. The surface behavior is identical. The underlying psychology is opposite. Any intervention built on the behavior alone will misfire for at least one of them.

Meta’s data has this problem at planetary scale. A user scrolling at 1 AM could be depressed, avoidant, relaxed, procrastinating, lonely, or dissociating. The platform sees the same scroll. Culture — real culture, the kind that shapes how people treat each other and make decisions under pressure — requires understanding motivation, not just movement.

The Algorithm Has Already Corrupted the Signal

There is a second structural failure, deeper than the first. The behaviors Meta observes have already been shaped by Meta’s own algorithm. The data is not a clean readout of human psychology — it is a readout of human psychology under continuous algorithmic influence: nudges, recommendation loops, emotional amplification, content ranking.

Trying to extract cultural insight from that data is like trying to study human nutrition using a dataset drawn entirely from casino buffet patrons. The environment has distorted the signal beyond recovery. Any culture built from those inputs would reflect not human values, but whatever the algorithm found most profitable to amplify.

Performance Is Not Character

On-platform behavior is also socially performative by design. Users on Facebook, Instagram, and Threads are curating, filtering, and identity-managing. They are responding to the social feedback loops of likes, comments, and comparisons. They are not behaving naturally — they are behaving for an audience.

Building culture requires understanding how people actually behave privately, under pressure, when the stakes are real and no one is watching. Meta’s data captures the opposite. A company that built its internal values around performative behavior would end up rewarding presentation over substance, appearance over integrity — which is, critics have long argued, precisely the culture Meta has produced.

States Are Not Traits

Culture requires trait-level insight: stable, dispositional characteristics that predict how someone will behave across time and context. Meta’s data is saturated with states — temporary emotional conditions triggered by platform interaction. The distinction is not semantic. A person in a moment of anxiety is not the same as an anxious person. A scroll session driven by loneliness does not tell you whether someone is constitutionally introverted or situationally isolated.

Any cultural framework derived from Meta’s data would be calibrated to dysregulation, not character. It would mistake the weather for the climate.

What Actually Works

None of this means Meta is incapable of building a healthy culture. It means the data it holds cannot do the work. What organizational science has established over decades requires different instruments entirely: validated assessments that measure stable personality traits, contextual conversations that preserve the meaning of behavior rather than stripping it, developmental frameworks built around regulation and character rather than engagement and reaction.

Meta has access to all of these. It has chosen, repeatedly and at scale, not to prioritize them.

That choice may be partly self-reinforcing delusion — the gap between the company’s public messaging about community and connection and what its internal data actually reflects is wide enough that closing it would require a reckoning most organizations prefer to avoid. Admitting what you know about your own product is not a comfortable move when lawyers are watching.

But it is also, at some level, a cultural choice. The data Meta trusts most is a portrait of human beings at their most reactive, most impulsive, most unwell. The culture it has built is, in that sense, a faithful reflection of what it chose to look at.

There is a final observation worth making. The companies leading the AI race — Anthropic, OpenAI, Google DeepMind, and Microsoft — have their cultural and execution problems, as all large organizations do. But none of them have built their core business around behavioral dysregulation. Their products are cognitive tools, not engagement traps. The data they generate reflects what people are trying to do, not what emotional state makes them easiest to monetize.

Meta and xAI sit on the other side of that line. Both are led by executives whose most visible platform — Facebook, Instagram, and X — runs on the same dysregulation engine described above. It is probably not a coincidence that both organizations show the most visible signs of cultural incoherence and strategic drift in the AI race. When the data you trust most is a portrait of human beings at their worst, it shapes not just your product decisions but your organizational instincts.

The companies building the future of intelligence, it turns out, may have an advantage that has nothing to do with compute or talent. They simply never learned to mistake human distress for a signal worth following.

Jayson Blair is the host of the Silver Linings Handbook podcast.

Share

Somewhere between the grilled octopus and the retsina or whatever it was we were eating, I asked a few questions that probably went a shade too far, touching on Governor Pataki’s strained relationship with his lieutenant governor — a tension barely in the public domain at the time.

What I didn’t realize then was that the most formidable non–New York Times person at that table wasn’t the Governor of the Great State of New York. It was his spokeswoman, Zenia Mucha.

Pataki deferred to her with such instinctive respect that The New York Observer once dubbed her “Governor Zenia.” Years later, when she took over corporate communications at Disney, she earned the moniker “warrior princess” — and, more importantly, she had Bob Iger’s ear.

She was the first, but hardly the last, formidable communications professional I would encounter.

Catherine Mathis, the vice president of corporate communications at The Times, was as gifted and precise in her messaging as she was unflinching in her counsel.

Sunny Mindel, City Hall spokeswoman for Mayor Rudy Giuliani, was able to pull him back onto the rails — until, of course, he derailed himself long after she was gone.

Ed Skyler, who served Mayor Michael Bloomberg and later ran corporate communications at Citi, always seemed four years ahead of the strategy and knew exactly how to talk about a mistake without inflaming the world around his client.

Joe Evangelisti, the longtime head of communications for JPMorgan, wasn’t merely a spokesman — he was a strategic operator.

And my late colleague David Carr once wrote that Irena Briganti, Fox News’s primary media relations executive, ran a “two‑minute hurry‑up offense that never seems to end after two minutes,” and “a rolling opposition‑research operation designed to keep reporters in line, sometimes by feeding them and sometimes by maiming them.”

Different temperaments, different styles — but a shared core.

They understood the heartbeat of customers and voters, employees, and organizational culture. They were often the most trusted strategic advisors to the people who, on paper, led them. In reality, the power dynamic was more complicated.

What crystallized for me this week, listening to Meta’s trial strategy in the Kaley G.M. case — which resulted in a $6 million verdict for the harm social media inflicted on her as a child — is that Mark Zuckerberg desperately needs a Zenia, a Catherine, a Sunny, an Ed, a Joe, or even an Irena.

Someone — anyone — who can talk sense into him.

What makes Meta’s current predicament so striking is not simply that the company lost a major case, or that a jury found its conduct harmful enough to warrant punitive damages. It’s that so many of its wounds are self‑inflicted — the kind of mistakes that almost never happen when a serious communications heavyweight is in the room, empowered, and heard.

Meta’s tone‑deafness is not episodic. It’s structural.

You could see it in the way the company reacted to Careless People, the memoir by Sarah Wynn‑Williams, a former senior public policy executive. The book was not a Molotov cocktail. It was a sober, insider account of culture, power, and the emotional toll of working inside a company that treated criticism as an existential threat. Meta responded by threatening legal action, as if the real problem were not what the book described but that anyone outside the building might read it. In fact, the threat was the only reason I knew about or bought the book.

That is not how a confident institution behaves. That is how a thin‑skinned one does.

You saw the same reflex in the response to Frances Haugen’s testimony to Congress. Faced with internal research about harm to teens, Meta’s instinct was to litigate the framing — to insist the data was being mischaracterized, to argue that critics didn’t understand the context, to behave as if the public’s outrage was a misunderstanding that could be cleared up with better talking points.

It was the corporate equivalent of saying: You’re not actually in pain. You just don’t understand the numbers.

And then came the Kaley G.M. case. Listening to Meta’s trial strategy was like hearing a company argue in a foreign language — technically fluent, emotionally vacant, and utterly disconnected from how a jury hears a story about a child harmed by a product designed to be irresistible. The plaintiffs told a human story. Meta told a product story. Only one of those lands with a jury.

After the verdict, instead of apologizing to Kaley — as the YouTube lawyer did — a Meta spokesman told The Wall Street Journal, and I kid you not, that the company would appeal and that “teen mental health is profoundly complex and cannot be linked to a single app. We will continue to defend ourselves vigorously as every case is different.”

What?

This from a company whose own internal research — revealed by a whistleblower — found that Instagram “makes body image issues worse for one in three teen girls,” that “Instagram’s Explore page can send users deep into content that can be harmful,” that “we are not removing predators fast enough,” and that “we know our products enable unwanted sexual contact.”

Anyone who has been even a country mile close to a corporate lawsuit knows that the best way to not get sued is for the plaintiffs to not hate you. And Meta has become hateable — and doubled and quadrupled down.

A Zenia Mucha would have stopped that. A Catherine Mathis would have stopped that. Even an Irena Briganti — who has never been accused of softness or of not taking Rupert Murdoch’s side — would have stopped that.

Because they all understand something Meta has never internalized: when people are telling you they’re in pain, you don’t hand them a spreadsheet and a nuanced argument about why their bleeding is not your fault. You give them a tourniquet and a hand to hold onto.

Meta keeps doing the exact opposite. And now, juries are responding the way the public has for years — with skepticism, impatience, and, increasingly, punishment.

At a time when there are thousands of Kaleys lined up in courts in California alone — with school districts and states not far behind — it all begs the question of why Mark Zuckerberg still does not have a powerful corporate communications guide, or some other advisor who can keep him off the third rail and on the tracks. My best guess is that this is one of those moments when ego has outpaced intelligence, and winning has overtaken winning over hearts and minds.

And if it continues — if Meta keeps treating criticism as a misunderstanding rather than a warning — it will become harder and harder to argue that Facebook, Instagram, and Meta deserve the cultural power they still hold. A company that cannot hear the people it harms eventually loses the right to shape the world the rest of us live in.

Share

The field’s most famous experiments — Milgram’s obedience studies at Yale — were meant to illuminate the psychology of the follower. Why do people comply with destructive authority? Why do they obey orders that violate their own moral intuitions?

But the man in the white coat — the one with institutional authority, a confident demeanor, and a clear narrative about the purpose of the study — was the figure they should have been studying.

Milgram believed he was uncovering a deep truth about obedience. What he actually demonstrated was the power of a leader who provides legitimacy, narrative, and a clear argument. The famous 65 percent compliance rate — participants administering what they believed were near‑lethal shocks — collapses when you remove the white coat, the Yale setting, and the scripted reassurances. In conditions of degraded authority, obedience plummets.

Milgram did not reveal a universal human tendency to obey.

He revealed that leadership matters.

This is consistent with what evolutionary biology, primatology, and cross‑cultural anthropology have been telling us for decades: humans are prosocial, attachment‑dependent, stress‑responsive animals. When we're under threat, we look for someone who can explain what the danger means and what to do about it.

The Holocaust does not require a special psychological explanation. It requires an explanation for the conditions that broke attachment bonds — and the violence that followed as a near‑predictable consequence.

In a world where political leaders kill more of their own citizens than enemies kill in war, it seems obvious that leadership matters. As political scientist R.J. Rummel documented, somewhere between 170 and 260 million people were killed by their own states — compared to roughly 40 million battlefield deaths in the two World Wars combined. So, if your government is statistically your most dangerous adversary, then the relevant question is not how ordinary people become soldiers who kill enemies — Milgram’s question — but how ordinary people become instruments of state violence against neighbors, colleagues, and former friends.

Across cases — Stalinist USSR, Maoist China, Cambodia, Rwanda, Ottoman Turkey — the pattern is remarkably consistent. Economic crisis. Status threat. Political delegitimization. A leader who redefines the ingroup by expelling a portion of the population from it. And a bureaucratic apparatus that converts individual moral agency into role compliance.

Milgram’s paradigm captures only the last step.

This is why stressed societies diverge so sharply. Why did Weimar Germany collapse into Nazism, while Austria‑Hungary dissolved and took different paths without committing the atrocities seen in Germany? Why did Finland — invaded by the Soviet Union in 1939 — produce extraordinary national cohesion rather than scapegoating? Why did Denmark, under Nazi occupation, maintain its civic bonds well enough to save nearly its entire Jewish population?

Stress weakens prosocial bonds.

Leadership determines whether fracture lines — or new bonds — form.

The Versailles settlement after World War I was an almost laboratory‑quality stress test of this hypothesis: a society with a functional civic culture and a genuine intellectual tradition, stripped simultaneously of economic stability, national dignity, territorial integrity, and political legitimacy. The prosocial bonds did not merely weaken — they inverted. The ingroup shrank to the ethnic core, the outgroup expanded to include anyone who could be blamed for the stress, and the leader who offered restoration of meaning and belonging got the population he asked for.

This framing does not require Germans to be uniquely evil.
It requires them to be human under conditions that reliably produce a specific kind of human behavior.

That is a more disturbing conclusion than the “special evil” account, because it implies the conditions matter more than culture, and those conditions can be recreated.

If a leader names an outgroup, offers a restoration narrative, and makes violence feel like justice, you get one outcome. If a leader names a shared challenge, offers a solidarity narrative, and makes sacrifice feel like meaning, you get another. The population and stressors are nearly identical in both cases. What differs is the story leadership tells about what the stress means and who is responsible for it.

Milgram did not conduct a study of his participants.

He conducted a study of himself as a leader.

The same dynamic unfolded in Rwanda. It took less than a hundred days to kill approximately 800,000 people once Radio Mille Collines provided a narrative that transformed fear into moralized violence.

Psychologist John Bowlby argued that the attachment figure’s primary function is to be a secure base — not to eliminate threat, but to regulate the organism’s response to it. A caregiver who responds to distress by naming a scapegoat, amplifying fear, and offering revenge as comfort produces a dysregulated, anxious, out-group-hostile child. A caregiver who acknowledges the threat, provides reassurance, and offers a path forward produces a securely attached child who can tolerate stress without fragmenting.

Leadership at the societal level functions as an attachment figure for the group.
This is not a metaphor — it is the same psychological mechanism operating at a different scale.

The Weimar Republic failed as an attachment figure in the most complete way possible: it was associated with defeat, humiliation, and economic collapse. It could not steady the population. Hitler offered himself as a replacement attachment figure, with a narrative that provided exactly what Bowlby’s theory predicts — a threatened organism needing a clear account of the danger, an identified enemy, a restoration of pride and belonging, and a promise of security.

If leadership is the primary variable, then the most important question in political psychology is not what makes populations susceptible to authoritarianism — they all are, under sufficient stress — but what produces leadership capable of steadying a frightened population rather than exploiting it.

Hannah Arendt understood this more clearly than Milgram ever did. The question was never why people follow. People follow — that is what attachment‑dependent, prosocial animals do. The question is who is worth following, and what happens when the answer goes catastrophically wrong.

Social psychology spent half a century studying the wrong end of the leader‑follower relationship.

As the profession confronts the collapse of many of its most famous studies — including Milgram’s and Zimbardo’s — it is time to correct the deeper mistake: the assumption that conditions operate independently of the personality and narrative of leadership.

Without integrating what we know about social circumstances and the personality of leaders, even great nations could find themselves, in less than 100 days, like Rwanda.

Share