Register now!

Alex Bronzini-Vender is a 20-year-old writer from New York, NY.

On a cool June evening in Berkeley, I watched a man named Victor try to shoot a beam of light out of his penis. Minutes earlier, he had raised his hand because he found the volunteer onstage before him attractive. And now here he was, facing her in front of a crowd of perhaps 60 people, having been instructed to do precisely this by a woman named Aella, a pseudonymous camgirl and self-described sex researcher. She had promised the men in the room that the exercise, if undertaken in earnest, might actually result in hookups, real ones. Now she told her first subject to imagine himself gathering all of the energy in his body into his pelvis and then sending that energy outward as a beam of light, straight through his trousers and into the room.

Victor pointed out that this seemed a little absurd, which it manifestly was. But he did it anyway, standing there and trying — and you could see him trying — to imagine himself emitting light from the tip of his penis. Aella reported that Victor wasn’t shooting enough light, and he readjusted. She and the volunteer watched. The rest of us did, too. And then the volunteer delivered her verdict, which was that Victor had come up short. She had, in fact, “felt more” from him a moment earlier, before he’d begun to imagine light streaming out of his phallus.

This was Manifest, and Victor’s ordeal belonged to its official programming. The conference — this was its fourth iteration — convenes each summer at Lighthaven, a verdant complex of gardens and halls in a converted Berkeley hotel that rationalists keep as a gathering place. At Manifest, rationalism has little to do with Descartes; it’s a Bay Area subculture built around the site LessWrong, a “community blog” preoccupied with mathematics, philosophy, genetics, and the apocalyptic risks (and opportunities) presented by artificial intelligence.

Around 800 people attend Manifest each year. It takes its name, and its nominal subject, from Manifold, a prediction market. At the opening assembly, Austin Chen, co-founder of Manifold, asked everyone who had ever bet on it to stand, and very nearly the whole audience got to its feet. Officially, then, this is a conference about forecasting. But “a conference about prediction markets” describes Manifest about as well as “a conference about effigies” would describe Burning Man. Prediction markets are what the attendees have in common, but they are not, mostly, what they came to discuss.

Any attendee could propose a session on any topic throughout the weekend, and several hundred sessions took place over the three days, a dozen or so underway in any given hour. At one session, attendees proposed political systems that might replace representative democracy (among them “futarchy,” a governance framework in which voters set overarching goals and prediction markets determine the specific policies used to achieve them). At another, a married couple gave a practical guide to IQ-screening one’s unborn children, drawn from their own experience doing the same. At the very moment that I watched Victor try and fail to emit photons from his manhood, a few doors down, a mathematician was explaining his PhD thesis on the topology of three-dimensional shapes.

Aella’s session was titled Get Hotter With Aella (Reinforcement Learning: Hot Girl Feedback). Reinforcement learning with human feedback is a technique for teaching a machine learning model to do what a human wants it to do. The model produces an output, a human rewards or penalizes it, and the cycle repeats, each round nudging the machine’s behavior closer to what the human prefers. Here, the model was Victor, and Aella, the human in the loop, was teaching him to be desired.

It didn’t work. Victor left the stage no closer to seducing his woman of choice. Round after round, the men failed. Now another man, Alex, raised his hand to face a new woman, Sam, while the room looked on. Alex’s trouble, Aella determined, was his stiffness; he was holding himself like he was bracing for a flu shot.

So she set about loosening him. “Tense your abdomen,” she told him. “I want large circles.” Alex made the circles with his arms, but Sam reported that he still felt “very stiff” to her. Everything about him, Sam added, was “feeling very robotic.”

Then Aella told Alex to touch Sam. “Reach out,” she said, “and touch Sam in a way that feels okay.” Alex smiled, a wide and sudden smile, and swung his arm up and over Sam’s shoulders in a single great swoop, planting his hand at the base of her neck, his face fixed in an awkward rictus.

“Let’s put that hand down,” Aella said. “Now let’s try it again. Step closer.” Alex stepped closer and reached for Sam a second time. Sam reported that everything about it still felt “very robotic.”

So Aella set about making Alex less robotic. She had him touch Sam “as robotically as you possibly can,” and then with as little robotness as he could manage, and then as “a robot pretending to be human,” and then as “an actual human.”

“It’s interpretive inspo,” she explained to him. “Just go on vibes.” It did not help.

“I feel like I’m not connecting with you at all,” Sam told Alex. “We’re looking into each other’s eyes but not seeing anything.”

At this point, Aella had gone quiet. She said that looking at Alex gave her a feeling of “grief.” She asked him what he wished to do to Sam. He thought for a moment and then replied that he wished to hug her. Aella gently coached him toward telling Sam that he wanted to hug her. Finally, Alex asked for a hug. Sam declined.

That the reinforcement learning didn’t take with this particular group of men was, in the end, beside the point. All ideologies rest on a theory of what humans are, or ought to be — there’s John Stuart Mill’s homo economicus, Hannah Arendt’s homo faber. Aella’s theory of man, and Manifest’s, holds that humans, with sufficient feedback and rigor, can and should train their brains to function like computers, that we can set aside the superstitions, the insecurities, the neediness and pride, the whole tangle of stuff that makes seducing a woman by means of imagined scrotal light feel humiliating — that, with enough work, we can become homo manifesticus.

“Davos man,” Samuel Huntington’s term for the world-bestriding elite of our era, takes its name from the Swiss town that plays host to the World Economic Forum’s annual summit. Davos man is staid, professional, and credentialed to the teeth. He would sooner die than be seen before a roomful of strangers imagining himself shooting light from his scrotum.

Perhaps Davos man needs no instruction in seduction, or perhaps he’s just too vain to ask for it. Either way, Manifest man labors under no such reservations. He seeks to optimize every facet of his life. At Manifest, the optimizing began early: Lighthaven’s first session, on Friday afternoon, recast the basics of social interaction in financial terms.

It was called “how 2 socialMax Manifest 😎,” and it was led by a 23-year-old quantitative researcher at a trading firm — let’s call him Josh. He proposed to teach the room what he called “portfolio theory applied to socializing.” As it went, friendship was a kind of trading book. Like in any trading book, one should always maximize expected value when socializing. One should not, Josh advised, “give a fuck about fairness” — that is, whether an individual reciprocates the effort you give them — because fairness is a drag on returns. According to Josh, asking someone a great many questions as you strike up a conversation is “OP” (“overpowered,” a high-return move at almost no cost).

The single most overpowered move of all, though, is vulnerability. The trick, Josh explained, is to disclose something that makes you look bad, precisely when almost no one else is willing to do the same. You take the hit to your personal vanity, and the friendship deepens faster than it otherwise could. By way of example, Josh told the room about his porn habit. He was, he explained, a “gooner.” He’d confessed as much to the last stragglers in a thinning room at a previous conference; one of them stuck around to talk to Josh, and he later gave Josh a free ticket to Manifest. Josh’s vulnerability had paid off! He recommended the strategy to the rest of us.

If Davos man ever had a porn addiction, it’s unlikely that he’d ever admit it. But then, Davos man probably wouldn’t need Josh’s instruction in the first place. Socialization comes naturally to him. The relevant skills were installed early, at the right schools, among the right people, so thoroughly that he has long since forgotten they are skills at all. At Manifest, though, there is a certain existential loneliness, a specifically sexual strain of angst, that did not lift once in the three days of the conference. Multiple sessions every day centered sex, love, and romance — that is, how to master them, or merely obtain them. These neuroses practically announced themselves at the door: the first thing I saw at Lighthaven, more or less, was a whiteboard by the entrance on which someone had written, “Single? Check out FEMBOYS.”

In previous years, Manifest attendees reportedly participated in orgies, the occurrence of which were later revealed to the public through prediction markets, with people betting on whether an orgy had indeed taken place during the conference. At the time of publication, though, the odds that “there is an orgy” at Manifest this year stood at five percent on Manifold. And on the conference’s final day, one Manifold user created a market on whether he would manage to make out with someone before it ended. By evening, it had resolved to “NO.”

“You’re all fucking animals,” Aella offered, by way of welcome, just after noon on the second day of Manifest, “Obviously.” We’d each been assigned an animal and asked to make its sound, all 50 of us at once, as part of a group icebreaker. I was a lion.

Then, under the California sun, Aella ran an experiment, the main event, in what she called “live polls.” She would name a spectrum, and we were to physically arrange ourselves along invisible axes, each of us walking to the spot that accurately represented our place on the distribution. Whereas Josh had encouraged us to make a single confession, Aella proposed to extract confessions from all of us at once, continuously, for the better part of an hour.

We started gently enough, ordering ourselves by height to get the hang of it. Then, with no particular change in tone, Aella had us sort ourselves along two axes: who had the dirtiest butt, and whether or not we picked our noses. Then the spectrums turned to sex.

“Stand over here,” she said, “if there is anyone in this room you would say yes to a date with.” We sorted.

“Now, anyone in this room you would say yes to casual sex with?” We sorted again. She proposed that those who had clustered on the “yes” side consider having sex with one another.

“Now, who is horny, right now, at this moment, and who is not?” The crowd rearranged itself into a live map of the room’s desires.

Then the axes began to multiply. She had us plot ourselves in two dimensions at once, wealth against relationship structure, so that she could call out “my richest poly people,” then “my richest monogamous people.” Then, it was the degree of mental illness against the kind of person we were drawn to — the more mentally ill toward one end, the less so toward the other, the attracted-to-the-mentally-ill on one axis and the attracted-to-the-mentally-stable on the reverse — so that the room resolved, briefly, into a scatterplot of its own psychiatric tastes. Once more, Aella proposed that the people most attracted to mental illness and the people most afflicted by it seek out one another; a relationship, perhaps, might follow. Somewhere in the middle of all this, Aella asked whether we liked pineapple on pizza, crossed it with whether we’d had a happy childhood, and announced she could detect a correlation.

We lined up by body count. (Mine was on the lower end of the room’s.) We lined up by IQ — first by our guesses, and then, before the entire room, by our measured scores, for those of us who had taken an official test. We lined up by how much childhood trauma we had experienced. We counted off the number of times, past the age of 18, we had soiled our own pants.

Finally, Aella wanted us to order ourselves by whether we experienced above- or below-average desire to engage in sexual intercourse with our parents. Already we’d bared our sexual histories, bowel control, appetite for casual sex, present state of arousal, and more. But this was a bridge too far, even for the Manifest crowd. The participants fled toward the extreme right-end of the strip of astroturf we’d assembled on.

Aella patiently reminded us why she had framed the question in terms of averages; the laws of statistics required most of us to cluster near the middle of the range. We could not all be outliers.

I was not, to put it mildly, enamored of the exercise. But I was there to understand homo manifesticus, goddamnit. Could a taboo, restated as a normal statistical distribution, become normal itself?

While most of us, myself included, planted ourselves firmly on the below-average side of the turf, a modest contingent had ventured across to the other side. My parents are not dead, nor are they elderly. And so I inched toward the center.

Four hours had passed since Aella’s live poll, and I now sat in a small back room, a converted lodging of the old inn. The evening light came in sideways and died against the far wall. Those of us who had arrived too late for chairs folded ourselves down onto the floorboards. I was there to watch a married couple, Rachel and Jonathan Wallace, offer their “Practical Guide to Superbabies.”

Rachel’s “autistic special interest,” she explained, introducing herself, was “repro tech and super babies.” These methods, she stressed, were available to all of us in the audience. You undergo ordinary IVF to produce a cohort of embryos; you biopsy a few cells — a sample of the trophectoderm — from each and read off their DNA; then you predict the future health, and intelligence, and cancer risk of the child each embryo might become; and you implant the best one.

The Wallaces had lately done this themselves. They had screened their embryos for IQ. The winning one, Jonathan said, was the far outlier at the right edge of the resulting distribution. Rachel was six weeks pregnant with the result, their fifth child. Their experiment had recently gone viral online, Jonathan added, and now it was “in the room with you.”

“Any of you,” Rachel said, “could do this tomorrow,” though, she said, you should resist the impulse to make your child as tall as possible. The ideal height for a man, setting aside the social advantages of being tall, is, where longevity is concerned, about five foot nine.

Actually, genetically engineering your offspring, in the Wallaces’ view, is your moral obligation. The couple shares the rationalist conviction that superintelligent AI is coming for us all; one of the better ways to keep pace with the machines, the Wallaces said, was to build better humans. We’d best begin now, then. So they provided helpful bits of advice, too, like which clinic was cheapest (CNY Fertility) and which drugs took the worst edge off the inflammation from the egg retrieval (GLP-1s).

There was one small issue, though. A predictor is only ever as good as the genomes and the life outcomes you train it on, and the world’s genetic data sits locked away inside national biobanks, out of reach of the Wallaces. Prying it loose would be “a political project, unless somebody wants to fund genome heists.”

“Legal genome heists, of course,” Jonathan added.

The couple asked for a show of hands. Who here, Jonathan wanted to know, would sit for an IQ test and let their genome be sequenced, for science? The Berkeley Genomics Project would be glad to pay.

“Spit in the tube,” Rachel said, “to defeat the artificial general intelligence.”

Hands went up across the room. Manifest man, it seemed, eagerly enlisted his own genes in service of the common good.

Many of Manifest’s attendees hold a deep, low-grade distrust of the unaided human being. If you believe the species is an unfinished draft soon to be beaten in its intelligence by computers — that is, if we don’t revise it gene by gene — you will not place much faith in its current hunches, its decisions, votes, and elections. You will want such decisions moved somewhere safer. You will want them moved into the prediction market.

Robin Hanson is an economist at George Mason University and, more than any living person, the author of the normative argument for prediction markets. In Hanson’s ideal world, voters would settle the values a society wishes to pursue, and markets would determine which policies are likeliest to implement them. (“Vote on values, bet on beliefs,” his mantra goes.)

Hanson built one of the first corporate prediction markets in 1990, inventing the scoring rule that runs many of the markets traded on today. The rationalist movement’s intellectual genealogy runs through Hanson. It was through his blog, Overcoming Bias, that Eliezer Yudkowsky first became known online, and it was there that the two staged their foundational 2008 debate over whether AI would rapidly “FOOM” into superintelligence. Yudkowsky went on to write the scene’s defining texts, chief among them Harry Potter and the Methods of Rationality. (At 661,000 words, Yudkowsky’s work of Harry Potter fanfiction is longer than Infinite Jest and Gravity’s Rainbow put together.) And in 2003, Hanson talked the Pentagon into running the “Policy Analysis Market,” in which traders would bet on the future of the Middle East — on coups, and assassinations, and terror attacks, as its critics put it — reasoning that a market in such horrors would predict outcomes better than the CIA.

Two senators discovered the Pentagon program, denounced it as a federal betting parlor on terrorism, and quickly killed it. Fortune, picking through the rubble, called Hanson someone who “would have us bet on terrorism — not to mention discard democracy and cryogenically freeze our heads.” All three were true. Hanson favors futarchy; he has arranged to have his head frozen; and he would, in fact, like us to bet on very nearly everything.

When the Commodity Futures Trading Commission moved in 2024 to banish Kalshi from the United States, the platform’s lawyers argued that a ban’s “only effect would be to confine all election trading activity to unregulated exchanges.” Hanson harbors no such modesty. He has spent 40 years insisting that almost any question, however political, however intimate, is best answered by getting people to bet on it. Whereas Josh had merely proposed that we run friendship as a trading book, Hanson would run much of human experience that way.

If the rationalist movement has a founding father, then, it was this ebullient man now presenting before us. Here, though, Hanson was preaching to the converted. “As you may have heard, prediction markets are great,” he told us. “The main reason they’re great is that we have this data that says they consistently beat alternatives in head-to-head comparisons for accuracy.”

Today, though, Hanson wouldn’t be the main authority on his own project’s progress. He soon ceded the floor to Proph3t, the pseudonymous founder of a funding platform called MetaDAO.

On MetaDAO, Proph3t explained, startups raise money not from venture capital but from prediction markets. Then the markets, rather than any founder or board, decide what the company ought to do with the cash. Companies governed by decision markets would make better decisions, and so compound their capital faster, and so out-compete every company that let mere human individuals decide things.

“You can imagine a world,” Proph3t said, “in which [futarchy] is just the way that you have to raise money, because otherwise investors think that you’re trying to steal their money.”

He had raised 12 million dollars, he told us, and was hiring people who believed futarchy would be “the future of human civilization.” Ordinary venture capitalists, he said, coddle their founders, but MetaDAO, by way of decision markets, would be different — “the Navy SEALs for founders.” The young Proph3t spoke with evangelical passion.

One member of the audience asked whether MetaDAO was trafficking in unregulated securities. Proph3t cheerfully allowed that it might be. But Trump is in office, he pointed out, and MetaDAO’s policy advisors considered the odds of “an enforcement action on this administration” to be low.

“Be brave,” Hanson called from his seat. “Come on, create the future.”

“Break the law,” someone in the audience added, and the crowd laughed.

Manifest man is anti-institutional to his core — no conviction is aired more gleefully in a nighttime open-ended Q&A session with two heretic bloggers: population geneticist Razib Khan and theoretical physicist Steve Hsu. Both speakers, in their way, have been cast out of the mainstream. The New York Times hired Khan as a columnist in 2015 and dropped him within a day after his writings on race and genetics drew public outcry; Hsu resigned his administrative post at Michigan State University in 2020 after the graduate students’ union mounted a campaign to remove him over views they said were sexist and racist. Khan and Hsu had every reason to resent the institutions that had expelled them, and over the next hour they obliged in airing them.

“Academics,” Khan offered, “are pussies,” a class of individuals selected, over years on the tenure track, for conformity and timidity. Khan produced the abandonment of the GRE as Exhibit A; under ideological pressure, graduate programs by the hundreds had dropped the test since 2019. In doing so, Khan argued, they had allowed “retarded people into graduate school,” the committees too cowardly to defend a measure they privately knew to be effective. “One rabid dog activist,” he said, “can herd a whole group of sheep.”

In some sense, Manifest man is a hedonist. He k-holes in bathrooms. He crushes White Claws on Lighthaven’s rooftops. And on the first night of the conference, I came upon several bottles of prescription amphetamines plainly visible by the entrance to the complex. Manifest has a vaguely Woodstock-like vibe, too, minus the public nudity and the public sex. (There was, however, a great deal of public cuddling, often in group form, called “cuddle puddles.” One attendee had gone so far as to try to get a cuddle puddle onto the official schedule, though it had, in the end, been cancelled.)

But Manifest’s anti-institutionalism is not the anti-institutionalism of Woodstock. The flower children wanted out of the draft, out of the whole apparatus of American power. Manifest man is not resisting control as such; he is resisting a specific form of control, that of the credentialed expert class. He will tell you so himself, if you ask. At a panel, I put a version of this to the organizers. The AI labs are accumulating wealth and political power at extraordinary speed. Did it trouble them that Manifest might, within 50 years, curdle into just another Davos — another self-satisfied gathering of the people who run the world? It did not trouble them in the least. But merely becoming Davos as it presently is — stodgy and unspirited — is not enough. Instead, Manifest Man aims to become a better Davos: younger, stranger, and with considerably more group cuddling.

These people have had their hands on real power before, and recently. On my final morning of the conference, I rode the shuttle from the hotel to Lighthaven, and a few rows ahead of me sat Caroline Ellison. Not long ago, Ellison ran Alameda Research, the trading firm at the dark heart of Sam Bankman-Fried’s FTX — the fraud whose billions had, for one brief and giddy stretch, made this very community a genuine force in American politics. The statistician Nate Silver held court at Manifest two summers ago, and David Shor — “the most influential data scientist in the Democratic Party,” according to Vox — was currently at Lighthaven to speak on “The Politics of AI.” Also, as one well-connected rationalist told me, lowering his voice as he said it, figures of even greater influence appeared in the blocked-off “private events” that peppered the schedule — those parts of the weekend that were not for me, or for most of the people there, ever to see.

Ellison, for the time being, was riding the public bus to Lighthaven with the rest of us, but only for now. The fortunes that bankroll this world have a way of regenerating, and the vanished FTX billions are giving way to the freshly vested equity of the AI labs. (Had Bankman-Fried’s fraud stayed hidden, one rationalist told me, Alameda Research’s stake in Anthropic alone would eventually have made customers whole.) Davos man may not yet have noticed what is underway in Berkeley. But Manifest man is auditioning to replace him all the same.

Everything that looks like freedom at Manifest is actually its opposite. Distrust of experts passes for populism; it’s the enthronement of a new and impersonal expert, the market. Their hedonism seems Dionysian; Manifest attendees’ vice of choice is ketamine, because it’s supposedly better for your health than alcohol. Josh’s vulnerability might look like intimacy; it was, to him, a strategic concession.

But Manifest man shares exactly one thing with Davos man: the Baconian conviction that knowledge is power — except Manifest man believes it, really believes it, with a sincerity that puts Davos man to shame. To Manifest man, knowledge is not quite what it is to the rest of us. For Bacon, knowledge was extracted from nature by a human investigator, and for Davos man, knowledge is a positional asset, certified by the right institutions. But for Manifest man, knowledge belongs to nobody at all. Only markets can truly “know,” and humans are merely their sensors.

Last summer, the commentator Gil Duran announced to the subscribers of his newsletter, The Nerd Reich — soon to be a book of the same name billed as an exposé of Silicon Valley “technofascism” — that a “tech cult compound” was rising in Berkeley. But Lighthaven is hardly a “walled, surveilled compound,” as The Guardian wrote in 2024 — there’s a shoulder-height wooden fence around the place. Rationalists are often cast as secretive and cultlike, but if there’s anything to fear about them, it’s their belief that there should be no secrecy.

There is nothing Manifest man is unwilling to know about himself, by way of the markets, and there is exceptionally little that Manifest man is willing to hide from his fellow rationalists. On Manifold, throughout the duration of the conference, you could bet on anything, down to the likelihood that attendees would injure themselves. Manifest man believes, on some level, that if we know everything about one another, all the time, we will become better versions of ourselves.

“You don’t hate journalists enough,” mumbled one attendee at the opening circle when I introduced myself as one. But for a movement often considered hostile to the fourth estate (its adherents spent the summer of 2020 at war with The New York Times, which had prepared to publish the real name of Scott Alexander, against his will), I otherwise encountered little hostility among the rationalists. Manifest man was kind to me, and kind in ways I frankly found perplexing, the chief of which was that he offered me ketamine even as I wore a bright green press badge around my neck, a badge whose whole semiotic function was to announce that anything said or done in my vicinity might end up in print. Maybe this was Josh’s “overpowered” play, a strategic offering meant to maximize expected value, or maybe they don’t believe in withholding damaging information at all.

On the second night of Manifest, a couple, whom I’ll call Patrick and Dana, cuddled together on a beanbag chair. I’d been intrigued by the session’s title, “Relationship Manifold market — Ask us anything.” Sure enough, ten or so Manifold traders sat around them, betting on whether the relationship would last six months.

Patrick was British, a lawyer for six years before he quit to trade energy. Dana was American, a theater critic. The couple had been together since New Year’s, and at some point in the months that followed, Patrick had done a very rationalist thing: he had opened a market on the two of them. Would they still be together on the first of July?

Before their session began, the odds of their six-month companionship stood at 75 percent. A trader in London stood to win 1,300 mana if the couple stayed together; a man named Andrew stood to win 1,300 if they didn’t.

“Why did Andrew fucking hate you?” someone asked. “He doesn’t,” Patrick replied.

I made a Manifold account on the spot and began doing my due diligence. Were they monogamous? “Yes.” Did they want kids? “Yes,” but they differed on how many.

“If you did break up,” someone asked, “what would be the top five causes, in rank order?” Patrick volunteered that he had cheated in three of his four previous relationships.

I asked them to close their eyes and raise a hand if they were in love with the other person. Dana objected to the format: “I think we should do a spectrum thing.” So instead they held up their fingers to rate the amount of love they felt for one another, and Patrick was more in love with Dana than vice-versa.

“Have you taken anything that reasonably correlates with IQ?” one woman who worked as a security engineer, asked the couple. “One of the top predictors of relational compatibility is IQ being within a certain band of each other.”

How many people had they each met that weekend and felt “very horny” about, she asked? “None,” both said.

Had either person, she continued, ever been in a relationship where “the heavens parted, the trumpets blared,” where they could feel the rest of their life becoming possible together? “Including this time,” she added. No, neither had.

She had been watching their bodies, too. “Why have you not touched him much?” she asked Dana. On Manifold, the likelihood that they’d stay together tanked.

At some point, I sensed the security engineer’s strategy. She was asking questions likely to make the couple angry at one another. I asked Dana and Patrick whether their decision to open the market gave people an incentive to get between them, to ask questions that would turn them against one another. “That would be hilarious,” Patrick replied.

I tracked down the engineer afterward. Could she really justify destroying a budding relationship for the sake of a few Manifold points?

What she had been doing, she told me, was trying to “anneal” the relationship. Annealing is a metallurgist’s term: “When you take a metal and you hammer it, the crystalline structure gets repeatedly fewer defects, and it becomes stronger,” she said. Whatever survived the evening, if it did, would be harder than what had walked in. She considered this “a valuable public service.”

And if the relationship shattered under the hammering, instead? That, too, was a service. “That which should break, or that which will break, should break sooner,” she told me. Why sooner? “Because someone’s got to be the villain.” It was better to shatter the relationship, cheaply, in a Berkeley loft than in two years, expensively, somewhere else.

I offered the obvious objection, that a market on a couple’s survival gives strangers a reason to pry them apart. Without missing a beat, she agreed. “Information discovery is the fundamental purpose of a prediction market,” she told me. If her questioning hastened the outcome, that would be “desired and expected.”

In the week before American forces seized Venezuelan dictator Nicolás Maduro, a Special Forces Master Sergeant who had helped plan the operation placed about 13 bets on Polymarket — that Maduro would soon be out of power, that American troops would soon be on Venezuelan soil, and so on. He turned $33,000 into roughly $400,000. This spring, he was arrested for it. And in April, dozens of freshly created Polymarket accounts bet on an imminent US-Iran ceasefire shortly before the president announced it on Truth Social.

The White House has told staffers to avoid trading on non-public information — a sensible enough precaution unless you happen to believe, as the truest prediction market devotees do, that insider trading is a public good. To them, more information in the market means more accurate prices, and more accurate prices, the argument goes, means a clearer picture of the future.

But the line between observing reality and shaping it is not always clear. Knowledge is power, yes, but power also can construct knowledge. Having taken financial stakes in particular outcomes, those in positions of power might very well steer our politics toward those outcomes. That’s the problem with financializing every aspect of our politics and our lives: you create constituencies with pecuniary stakes in bending reality toward certain outcomes.

And there may be reasons, beyond mere weakness, why a couple would resist “annealing” themselves before a crowd of traders. Facing the couple on the beanbag before her, the engineer had a deadline. The market was set to resolve in two-and-a-half weeks, and she had taken a stake in the breakup. If the couple was going to fail, she needed them to fail quickly. How to accelerate the collapse, to kill the uncertainty?

She thought for a moment. “Could you turn to each other, one at a time, and say, ‘I’ve enjoyed our time together. It’s been wonderful, but it’s time for something else’?”

The pair considered it for a moment.

“No,” Dana said. “I couldn’t do that.”

A month later, the market had resolved to “YES.”