Friday
On a dark, gray day in October 2025, the first dozen attendees of the Effective Altruism Global conference make their way to the Sheraton Hotel just north of Times Square in New York City. Cold and wet, I arrive at the Sheraton around 5 p.m., making my way through a lobby full of Brazilian businessmen holding squash racquets, and head down the stairs to a quiet hallway, where I find a table staffed by a young Englishman. After asking for my name, he prints it in lovely cursive on an index card and clips it onto a lanyard. Another young man gives me a pale gray shirt with delicate Victorian-era illustrations of birds, with captions in the style of amateur botanical sketches. They read:
1. the world is awful
2. the world is better than it once was
3. the world can still be much better
All are true.
●
While their personal motivations vary, most effective altruists trace their intellectual heritage back to the 1972 publication of an essay by the Australian philosopher Peter Singer entitled “Famine, Affluence, and Morality.” The essay is brief and repetitive, like biblical poetry, and urges readers to donate to relief efforts for the famine that was then afflicting Bangladesh. After a couple pages of such appeals, Singer proposes a thought experiment:
If I am walking past a shallow pond and see a child drowning in it, I ought to wade in and pull the child out. This will mean getting my clothes muddy, but this is insignificant, while the death of the child would presumably be a very bad thing.
Singer observes elsewhere that something like this really happened in the town of Wigan, in England: in 2007, two cops hesitated to dive into a pond after a ten-year-old boy who had tried to save his sister from drowning and then slipped underwater. Two elderly fishermen saved the sister but were unable to rescue the boy. Several minutes later, the boy’s stepfather arrived and pulled his dead body out of the water. “I don’t know why they didn’t go in,” the boy’s stepfather told the media. “I can’t understand it. If I had been walking along a canal and seen a child drowning I would have jumped in. You don’t have to be trained to jump in after a drowning child.”
Despite the assurance of Assistant Chief Constable David Thompson that the officers “acted correctly,” the Greater Manchester Police and most British media outlets refused to report the names of the officers on duty. Whether or not they had followed protocol, there was something obviously shameful about their willingness to stay on the banks. And yet, Singer pointed out, there is a disturbing amount of begging and cajoling that must go on before we feel compelled to provide even a minimal amount of aid to those suffering abroad. We do not want to give. What’s more, we are not quite expected to give.
“If it is in our power to prevent something very bad from happening, without thereby sacrificing anything morally significant, we ought, morally, to do it,” wrote Singer. “The uncontroversial appearance of the principle just stated is deceptive. If it were acted upon, even in its qualified form, our lives, our society, and our world would be fundamentally changed.”
Hearing the Parable of the Shallow Pond is a turning point for many young EAs. Put one way, it envisions ethics not as an academic football but as the science of helping others. Put another way, it instructs us to feel responsible for both what we do and what we do not do. Like many parables, it does not always find a receptive heart.
I first heard the Parable of the Shallow Pond from a prominent effective altruist named Zachary Robinson at a college near the farm where I worked along with several other young people. Robinson is a slight young man with friendly features, and his eyes darted cautiously across the small crowd as he ascended to a wooden lectern. I remember him explaining the parable and speaking briefly about his work. But frankly, my recollection of the afternoon is a little spotty, for shortly after he imparted his parable, a tide of objections rose up in my mind that effectively extinguished my ability to hear the rest of the talk. Sure—but who is this fictitious kid? You’re really asking me to give money to someone halfway across the world, with whom I have no connection and about whom I know nothing?
When Zachary finished, I remember rising with a queer feeling in my stomach. Back at the farm, my friends and I took a stroll through the communal garden, where we proceeded to stomp on effective altruism from all conceivable angles. I had just finished struggling through a class on The Phenomenology of Spirit, which made the simplicity of Zachary’s scheme seem especially insulting. I remember one of my friends snatching a ripe tomato off its vine and holding it before us as evidence: the prickly crown and the thin red fuzz, birthed from a skinny green plant that twisted its way around a trellis.
“The world is a mysterious place,” he declared.
We nodded. Wasn’t it? Just a year ago I had attended an urban high school, enjoying all the luxuries of a modern city, and now here I was out in the country, a good thirty miles from civilization, rising with the sun to harvest tomatoes. Life had borne me and my friends like leaves along like a mighty river, across the country and into the high mountains, and I found it difficult to imagine what the stakes would be in pretending that we had the right to administer pleasure and pain, life and death, to our fellow creatures. So I made a determined effort to forget the parable.
Yet it began to seriously nag at me a few months later, after I had gotten to know a cow on the ranch named Boo. Boo’s mother had died unexpectedly after his birth, and the student body had brought him onto the main campus and fed him milk out of a baby bottle. Now, fully grown, he had only good feelings toward humans. When I went out to the fields in the morning, he would gallop toward me, hoping to get scratched between his eyes. Sometimes he would help me move the metal irrigation lines by nudging them with his big, black head. Then, one day, I went out into the fields to find that all the cows, including Boo, had been trucked to slaughter at a big meatpacking operation, the kind that stretches endlessly along the side of Western highways.
I knew it was coming, of course—I was living on a ranch. I had butchered hundreds of chickens, and I wasn’t squeamish about the mechanics of slaughter. By growing alfalfa, I had been fattening Boo myself. But he was there, and then he was gone. It was the first time anyone I knew had died. Alone in the field, I wept.
Over the following weeks, I began to feel that there was something misguided about our disembodied food system, and in my musings I returned to the parable. Boo would be eaten by someone who’d never known him. In return, the ranch received a trivial amount of money, and the student farmers trudged through cold mornings without a friend. We had gained something insignificant and lost something tremendous. We had miscalculated.
Saturday
The first morning of the conference, several hundred young EAs tramp into the Sheraton’s lobby through a light rain. During the introductory ceremony, I take my seat beside a very elegantly dressed Indian man with an Enron sticker on his laptop.
“Where did you get that sticker?” I ask him.
“My boss,” he replies calmly.
I look around the table—his buddies keep a straight face. Finally, they burst out laughing.
“It’s just a joke,” they exclaim. “Of course he doesn’t work for Enron! He runs the effective altruist club at the University of Wisconsin, Madison.”
“I’m focused on Global Health and Wellbeing,” he tells me proudly.
I give him an approving nod. Global Health and Wellbeing is the most ancient of the EA departments, and confers a certain noble modesty, not unlike being a Hufflepuff. It is premised entirely on the contrast drawn by “Famine, Affluence, and Morality”: that there is something repugnant about the coexistence of extreme wealth and extreme poverty. This fact is strikingly illustrated by a bar graph that Giving What We Can has erected outside of the conference hall. The chart displays the wealth of the world’s upper percentile as a skyscraper at the far right, teetering evilly over nine unimaginably squat little flats—the global poor. It seems both easy and right to knock it down, like a tower of blocks.
The public imagination can generate no shortage of explanations for global poverty. Perhaps it can be ascribed to American imperialism, or to the sinister machinations of global capital, or maybe somehow to the starving themselves. But EAs tend to shrug off these explanations. As the EAs see it, the reason a child is dying, right now, of dehydration, is because you have refused to buy him a bottle of water. By donating seventeen dollars to a charity called Evidence Action, you can provide ten Kenyan children with clean water for an entire year.
To free up cash for projects like these, effective altruists have built out an impressive toolkit of money-saving devices. For instance, most of the EAs at my table are not eating breakfast. Instead they are drinking Soylent, a nutritional shake which is perplexingly named after the nutritional wafer made from human remains in the 1973 dystopian thriller Soylent Green. Soylent is a nutritionally complete meal replacement, meaning that it contains all 28 vitamins and minerals required to maintain and regenerate the human body. It tastes sweet and slightly stale, and is almost redundantly smooth, like liquefied pudding, clotting unpleasantly in your esophagus before sliding into the pit of your stomach, where it sticks like cement. Scientifically, you can drink only Soylent for the rest of your life and remain in perfect health.
“It can replace any meal,” announces one of the EAs, sniffing his Soylent nostalgically and evidently recalling some nightmarish phase of his working life. “But not every meal.”
Soylent was first cooked up in San Francisco in 2013 by an electrical engineer named Rob Rhinehart, and received an enthusiastic blast of venture capital from Y Combinator and Andreessen Horowitz. Initially marketed as a futuristic lifehack, it is now primarily sold in suburban Walmarts, where its gentle, sludgy texture is apparently indispensable to people with throat cancer and irritable bowel syndrome. It is packaged like a geriatric protein shake, such as you might find in a pharmacy next to the laxatives.
I suspect that neither the digestive benefits nor the micronutrient content, however, are what recommend Soylent to EAs, who are not as health- and longevity-obsessed as their second cousins in the tech sector. Rather, the many bottles of Soylent, tossed with happy abandon across the tables and floors of the conference rooms, serve as a persistent reminder of the solvability of problems. One has only to read the many descriptions of poisonous mushrooms in a mycological field guide to feel a gruesome awe at the millennia of trial and error involved in determining what is healthy for us to eat and what is not. Well, here it is.
As such, Soylent is a good introduction to effective altruism: both to its pugnacious quest to solve problems (like spending money on food) and its blithe disinterest in marketing those solutions to the general public.
For a more familiar example, take domestic crime, which costs the United States at least $300 billion annually through policing, prisons and courts, not counting the social costs of community violence and instability. This despite the fact that crime has received more formal attention over the last several thousand years than perhaps any other problem in human history. Substantial chunks of most major religious texts are devoted to crime and punishment; it is a critical dividing line between the worldviews of the major parties.
Alright, but what if I told you that the famous decline in violent crime that began in the 1970s and 1980s occurred simultaneously with the intensive regulation and eventual ban of leaded paint and gasoline? Or that violent crime rates in America since the 1930s track extraordinarily closely with levels of lead in blood samples from preschoolers taken 23 years earlier? Or that protecting a population from lead exposure has been shown not only to raise cognitive ability and improve impulse control but also simply to reduce violent crime by as much as 56 percent?
So instead of making murals about urban decay, or donating to the New York Public Library, or architecting even bigger prisons, a good effective altruist will book a flight to Malawi to meet with government officials who are interested in mandating alternative lead-free paint formulas in line with contemporary medical recommendations—a program that requires a small handful of people to draw up a single solid meta-analysis and hold a few conversations with a few officials who already more or less agree with the project. In 2024, following just such an effort by the Lead Exposure Elimination Project and its partners, the country of Nigeria agreed to implement enforceable regulations on the sale of lead paint. That is what the effective altruists call a sound intervention.
●
Lunch is served around noon in the buffet hall, which remains at the same thunderous decibel level that it attained early that morning. Not one effective altruist stops talking long enough to chew. Every 25 minutes, everyone in the room shakes their partner’s hand and heads to another meeting. At any given moment, at each banner and at all the ballroom tables in between, there are several hundred conversations going on simultaneously, all echoing off the Sheraton’s enormous ceilings.
Three long tables are set up with coffee, sandwiches, salads and soups. The Soylents are arrayed in a phalanx on the back wall, and little bags of sweet potato chips are always available. What’s more, everything is vegan, even (rather ambitiously) the turkey, provolone and honey-mustard sandwiches.
At first glance, the alliance between vegans and effective altruists might seem obvious. A consciousness that has already been expanded to encompass all the world’s poor is surely big enough to fit the world’s animals. A declaration about reducing harm surely comes with a suggestion about nonviolence. And any believer in a moral framework that attempts to quantify suffering should be immediately alarmed by the numbers involved: 85 billion land animals and trillions of fish are raised and slaughtered annually in terrifying conditions.
Indeed, effective altruism has historically had a very strong connection to animals. In 1975, a couple years after publishing the Parable of the Shallow Pond, Peter Singer wrote the provocative Animal Liberation, a book he would later describe as “one of the best strategic decisions that I made” and a work that had the most “positive, practical impact” of any in his career. In turning his attention to animals he was following a well-trodden path: many of the same nineteenth-century English reformers who preached utilitarianism and abolition also agitated for animal welfare. Jeremy Bentham, the founding father of utilitarianism, had explained the connection in this way:
The French have already discovered that the blackness of skin is no reason why a human being should be abandoned without redress to the caprice of a tormentor. It may come one day to be recognized, that the number of legs… [is] equally insufficient for abandoning a sensitive being to the same fate….The question is not, Can they reason? nor, Can they talk? but, Can they suffer?
“Why,” Bentham asked, “should the law refuse its protection to any sensitive being?”
Peter Singer makes a similar argument in Animal Liberation, providing a usefully detailed description of modern animal farming and drawing the same parallels between animal and human rights. Yet his account is somewhat less earnest than Bentham’s, and reads more like a thought experiment—a long appendix to “Famine, Affluence, and Morality.” The lesson is that if you have truly understood the Parable of the Shallow Pond—that you must judge what you do by all of its consequences—then you must reconsider many seemingly innocent or almost praiseworthy habits, including eating fish and drinking milk.
The logical appeal of Singer’s consequentialism, however, contrasts sharply with the movement that spawned directly from Animal Liberation. There are a great number of animal-rights activists at the conference, hoping to squeeze a few extra cents out of the EA grant makers, who are notoriously fickle and exacting. These activists see animal liberation as a real, glorious future, and they intend to bring this future about by infiltrating, protesting and disrupting, just like twentieth-century leftist revolutionaries—many of whom, incidentally, were vegan or vegetarian themselves. There are vegan graffiti artists and vegan punk bands1 and vegan terrorists. Most of them are women.
By contrast, Singer’s band of effective altruists are overwhelmingly male, and were made by their Creator to write code, not to eat lentil soup and use ridiculous phrases like “animal rights.” They do not rail against hunters and butchers and Big Ag. They prefer to consider “win-wins,” another term of art borrowed from Wall Street, which refers to situations in which none of the parties involved are forced to make sacrifices.
An infamous example of hardcore Singerian marginal thinking as applied to animals is a nonprofit called the Shrimp Welfare Project. Small and unloved, shrimp are not provided by most Western countries with even minimal standards of humane handling. Their treatment is gruesome, to say the least; they spend their entire 120-day existence in a pool of their own excrement, as a result of which they overwhelmingly become depressed, lethargic, anorexic, cannibalistic or enter hypoxic shock. A small number of female shrimp are removed in order to perpetuate the stock, and upon removal, their eyestalks are “ablated”—severed or crushed—which depresses reproduction-regulating hormones and accelerates ovulation. After several months, all the shrimp are taken out of the water and left to asphyxiate for as long as an hour in a bucket of ice.
To the average vegan activist, this is a perfect example of the perverse excesses allowed by a society that does not recognize animal rights, either culturally or in law. Protests, billboards, consumer outreach and sabotage with explosives might all be in order.
But Andrés Jiménez Zorrilla and Aaron Boddy, two entrepreneurial effective altruists, saw things differently. Fish and crustaceans are the only mass-farmed animals without stunning regulations. So they partnered with the Norwegian company Optimar, which manufactures heavy-duty electrical stunners capable of stunning over ten metric tons of shrimp per hour. With money raised from a host of EA grant makers, they began purchasing stunners wholesale at a price of $55,000 per unit and giving them—for free—to shrimp farmers.
Millions of dollars’ worth of expensive machines, delivered to them personally by tens of thousands of smiling vegans, must have caught all those Vietnamese, Honduran, Sri Lankan, Indian, Indonesian, Ecuadorian, Thai, Venezuelan and Nigerian shrimp farmers by surprise. But they were happy to route their shrimp through a stunner before slaughter and add a “Humanely Stunned” label on the bags. For their part, the Shrimp Welfare Project was happy to become the single most effective animal charity on the planet, with 1,400 shrimp stunned per dollar donated.
Lateral thinking or sleight of hand? Many traditional animal advocates were outraged. They believe that the Shrimp Welfare Project is the latest in a string of “welfarist” compromises, with the most famous being the introduction of “cage-free” eggs in the 2000s—that is, eggs laid by chickens who have been removed from cages but otherwise remain in exactly the same dank barns as before.
Just a few blocks down the street from our hotel, for instance, local activists are protesting Marriott International for serving eggs laid by caged hens. They’re participating in something called RAGE (or, Resistance Against Greed and Exploitation): a global pressure campaign to hold companies to animal-welfare promises which, in Marriott’s case, were extracted nearly a decade ago. Isn’t it cruel to make these activists stand out in the cold, pleading with Marriott for another empty deadline? Still, effective altruists smile on these cage-free campaigns, which, despite some corporate flip-flopping, have over the past couple decades freed several billion hens worldwide from cages whose minimal space requirements induced gruesome self-mutilation, and whose wire struts irreparably mangled the sensitive bottoms of their hens’ feet. The activists, they say, prefer the romance of a hopelessly distant revolution to interventions that improve the lives of animals in the present—billions of chickens and trillions of shrimp.
●
By late afternoon, my head is swimming from a full day in the Sheraton’s grand ballroom. I stagger into a quieter, gray-toned lecture hall where a young researcher named Mal Graham is addressing a small crowd. Graham holds a B.A. in physics and another B.A. in philosophy from Oxford and a Ph.D. in Engineering Mechanics at Virginia Tech, where their research was on “dynamic gap-crossing movements in jumping and flying snakes.” (By the standards of the conference, this is a fairly normal academic background.) The resume implies that Graham is interested in another possible theater for continued EA-vegan collaboration—the young, very trippy field of Wild Animal Welfare.
“The literature estimates suggest that up to a billion birds are killed annually in North America from ‘bird-window collisions,’” Graham is informing the room. “But… we have to remember that protecting them from a single collision does not make them immortal. So under some philosophical views that are predominant within EA, the question becomes, ‘Is the expected value of the time we gain them by preventing them from having that collision better than the expected value from the difference in suffering between their new death and their old death?’”
If each creature on the earth was represented by a dot, and the human dots could fit on a quarter, then the wild animal population would fill a FIFA regulation-size soccer field. The overwhelming majority of these animals may well inhabit a living nightmare. In New England, for example, a warming climate has encouraged the population growth of a tick called Dermacentor albipictus; tens of thousands of these ticks attach to a single moose, until large patches of its skin are invisible underneath great carpets of bloated insects. All sucking at once, the ticks can kill a moose—usually a calf—over the course of several months through anemia, blood loss and exhaustion. Each biological order has its own demons. Langur monkey populations in India will crush the infants of defeated tribes in order to trigger estrus in their mothers. In tropical waters, the fat-lipped cookiecutter shark will suction onto live prey, twisting out plugs of fresh meat. And the less said about parasitic wasps, the better.
There is only one major organization devoted to alleviating the suffering of wild animals. It is called the Wild Animal Initiative, and it occasionally funds researchers who have bright ideas about how to interrupt the endless cycle of suffering in nature. But WAI’s research remains primarily speculative; meanwhile, the yearly budget for both farmed and wild animal welfare is under three hundred million dollars, which is several times less than what Americans spend every year on Halloween costumes for their pets.
Feeling vaguely nauseous at the magnitude of evil in the world, I wander out of the talk. “Trauma,” a school counselor once told me kindly, “is getting hurt and not being able to do anything about it.”
Sunday
On the final day of the conference, I hop out of bed bright and early, determined to pack my Sunday with a respectable number of chats. However, half-asleep, I grab the wrong coat on my way out the door. I usually wear a blue puffer jacket, but today I accidentally snatch a very chic brown canvas windbreaker instead.
Forty-five minutes later, progressing rapidly through the stages of frostbite that precede amputation, I stagger through the doors of the Sheraton. I make my way to the buffet tables and plant myself beside the tanks of coffee, which have miniature burners beneath them. I draw up a cup for myself and absentmindedly nibble on some vegan cookies.
Truthfully, I imagine that part of the reason I did not immediately turn back for the winter coat was that I expected my willingness to self-sacrifice would win me some points with the altruists. Wasn’t sustained exposure to the New York City winter proof of my indifference to the pleasures and privations of the material world? Were Peter Singer drowning in a very cold pond, I clearly would not hesitate to pull him out and engage him in a chat.
My first chat of the day is with a rangy Dutch fellow named Sjir Hoeijmakers,2 who bears a beguiling resemblance to the infamously hot English actor Tom Hiddleston. Sjir and I are scheduled to walk around the perimeter of the hotel. He looks me up and down.
“If you are going to be chilly,” he says with a soft accent, “then I’m happy to stay inside.”
I assure him that my windbreaker is warmer than it looks, and follow him out into the blistering cold.
Sjir is affectionate and eminently calm—the very best of the effective-altruist personality types. He does not seem indifferent to pleasures or privations. In fact, he is very tall and healthy-looking. Sjir is also the CEO of Giving What We Can, which is the sanest and most publicly legible of the EA charities.
“We encourage people to pledge 10 percent of their income to charity,” he says. “And we make some recommendations as to where we think their donation could have the largest impact.”
Like veganism, the 10 percent figure is an old-school Singerism. It was inspired by the observation that two major religious traditions—Jewish tzedakah and the Christian tithe—recommend donating 10 percent of one’s income to charity. According to the Giving What We Can website, over eleven thousand people have answered the call to pledge, directing over sixty million dollars toward good causes every year.
At this point in the essay, a pledge of six thousand dollars per person may actually strike you as a little small. One young acquaintance of mine, upon discovering effective altruism, promptly underwent surgery and donated his kidney to a random stranger. I explained that, upon seeing pictures of him in the hospital, weak and frail and grinning broadly, I thought: that is an altruist.
“Oh, we’re not against that at all,” Sjir says affably.
“Can I ask what you donate?”
“Sure,” he says. “I donate 50 percent of my income.”
“How do you feel about that?”
“It gives me joy, and it leaves me more than enough to cover my needs. Many of the things that provide the most meaning to my life are free, such as my relationships with my family and loved ones.”
“Don’t you think you should be giving more?”
Sjir explains, “Some people work themselves very hard, and donate a lot of money, but they put so much pressure on themselves that they become unhappy, and may even burn out. And then they don’t have much of an impact at all.”
●
Most of Sunday afternoon is taken up by a job fair, which occupies half of the Sheraton’s main ballroom. I perambulate idly through the stalls, rubbing some elbows and half-hoping to land myself a sweet summer internship.
But I am not in luck. Of fifty stalls, a whopping 27 are devoted to AI research, AI journalism, AI alignment, AI governance, AI solutions…
What is this? Have I stumbled into a trade show, an expo of third- and fourth-tier artificial-intelligence startups, looking to skim some spoonfuls for themselves off the top of that rampaging gravy train; that utterly obvious, omnipresent, unmissable, all-consuming pop-culture phenomenon? Who is neglecting artificial intelligence? What happened to mosquito nets? What happened to the welfare of shrimp?
The answer is that recently, with the poise and self-discipline of Shaolin monks, many effective altruists have pivoted away from their long-standing commitments in animal welfare and global health and toward a third theater: “AI safety,” which, according to the estimates of several insiders, is currently consuming about half of the movement’s labor and capital. Holden Karnofsky, the former CEO of heavyweight EA grant maker Coefficient Giving, has left his post to work full-time on AI safety; 80,000 Hours, the influential EA job board, now almost exclusively posts jobs in AI safety. And on the stages at EA Global, four lectures are devoted to artificial intelligence—double what is given to either animal welfare or Global Health and Wellbeing.
AI safety falls into the traditional domain of a branch of effective altruists known as “longtermists,” who first came onto the scene with the 2003 publication of a deeply loopy and kaleidoscopic paper by Oxford philosopher Nick Bostrom, called “Astronomical Waste: The Opportunity Cost of Delayed Technological Development.” The paper proposed colonization of the Virgo Supercluster as the prime directive of effective altruism, and warned against any present-day distractions that might hinder that future.
After several years of refinement, the theory was given more coherent public expression in 2020 by a book called The Precipice, written by a famous EA named Toby Ord. While the whole book is entertaining, much of the thesis is available in its dedication, which reads as follows:
To the hundred billion people before us, who fashioned our civilization;
To the seven billion now alive, whose actions may determine its fate;
To the trillions to come, whose existence lies in the balance.
As Ord goes on to explain, to “foreclose” the lives of those future trillions is morally comparable to murder. Just as our nonaction on the banks of the Shallow Pond makes us responsible for the child’s death, so our nonaction in the present makes us responsible for those in the future: “A person does not matter less, the further away from you they are in space … People matter equally regardless of their temporal location too.”
Longtermism’s preoccupation with the lives of future humans ends up doing wonky things to the utilitarian pleasure-pain calculus because, as Ord’s dedication makes clear, the number of humans alive today will be dwarfed by the number of humans who will live over the next several billion years. The whole EA balancing act—five deworming pills or 250 iron supplements? Sixteen malaria vaccines or three cycles of Group Interpersonal Psychotherapy?—is ultimately steamrolled by that single juggernaut, infinitely valued, weighed against almost nothing.
Events that might threaten the infinite future of our species are referred to as existential risks, or “x-risks.” Ord’s proposed list of x-risks goes like this:
Asteroid or comet impact
Supervolcanic eruption
Stellar explosion
Nuclear war
Climate change
Other environmental damage
“Naturally” arising pandemics
Engineered pandemics
Unaligned artificial intelligence
Among effective altruists, however, artificial intelligence is the foremost concern. Artificial intelligence could electronically order DNA laboratories to print unstoppable bioweapons. It could hack into weapons systems and initiate nuclear war by itself, no matter how elaborate the international deterrence architecture. It might be able to manipulate a whole society by generating misinformation to serve its own aims. It is getting rapidly more intelligent without supervision, and shows both the inclination and ability to deceive its supervisors. But for at least the next few years, it remains under the control of a few logically minded billionaires. Even better, two decades of EA penetration into Silicon Valley means that some of these AI billionaires are already effective altruists themselves.
The so-called “alignment” of AI with human values is the subject of “A High-Level Hope for Managing AI X-Risk,” and is hosted by Ajeya Cotra, then Coefficient Giving’s senior adviser on artificial intelligence. The room is packed, the best attendance I’ve seen for any of the lectures—the air is heavy with suspense and unease, as Cotra unveils a chart with a single line hurtling through the future, up toward “Doom.”
Oh, bother. I’m very glad that someone in the world is keeping their eye on artificial intelligence. But why does it have to be the effective altruists? The virtue of utilitarianism, according to the philosopher Martha Nussbaum, is its ”general radicalism and [its] skepticism about conventional morality, [its] willingness to follow the ethical argument wherever it leads.” Almost to a man, the effective altruists I meet at the conference are smart, optimistic and ambitious. These new themes of tragedy and hopelessness—where did they come from?
“People say to me… ‘What’s your ask?’” Eliezer Yudkowsky, the progenitor of the AI-safety field, read aloud from his phone at a TED Talk in 2023. “I do not have any realistic plan. … My best bad take is that we need an international coalition banning large AI training runs… I say this not expecting that to actually happen. I say this expecting that we all just die.”
As I leave the conference, jobless and annoyed, I can’t help feeling that this turn toward AI safety is barely even utilitarian, since its concern is not with joy and suffering but with lives. It frets over future generations despite not knowing whether they will end up being happy or miserable or just alright. Our descendants might turn on and annihilate each other, after having multiplied and thereby magnified the scale of suffering involved in their destruction. To feed themselves, our descendants might raise quadrillions of animals in filth and misery. Or perhaps they will themselves become enslaved and eaten by an undiscovered alien species, lurking somewhere deep in the Virgo Supercluster. How much crazier would that be than the alternatives? In any case, saving the world seems like a worthier task for politically minded thought leaders than for the marginal effective altruist.
I pose this question to Ronen Bar, a famous undercover investigator of animal farms. He appears momentarily puzzled.
“Ah,” he says carefully. “You are only seeing what is happening now, and not what will be happening. Artificial intelligence is on the rise. People are using it. If it becomes dominant, that’s going to affect everybody, including animals—whether it’s aligned or misaligned. Either way, ignoring AI is not wise, from an animal point of view.”
That is that. A funny sort of unspoken truce. Some effective altruists are analyzing floating-point matrices for signs of the apocalypse, while at the same moment others are installing electrical stunners at shrimp farms in rural Vietnam, and still others are interviewing Nigerian mothers with young children whose intestines have been eaten by roundworms. I am convinced that deep down these groups must loathe each other, but so far as I can tell, they do not.
●
Most people I know are scornful of effective altruism, if they have heard of it at all: over the past decade, the effective altruists have managed to irritate Americans from all walks of life.3 They are too self-righteous for Silicon Valley and not self-righteous enough for the universities. They are too interested in money to be socialists but too interested in social justice to be entrepreneurs. They are too awkward to be activists but too active to be geeks.
Effective altruism always seems to catch audiences on the wrong foot. The “weak” version of EA usually presented to the public, namely that you should be awake to the suffering of your fellow creatures, strikes people as lame and anticlimactic. But the “strong” version of effective altruism—that you are morally required to overhaul your life in favor of helping others, whether through your donations, career or diet—seems to go way too far.
Critics of EA like to divide themselves along these lines as well. The first camp is led by academics, like Amia Srinivasan, who find the ethics of effective altruism to be frustratingly shallow. “There is a seemingly unanswerable logic, at once natural and magical, simple and totalising, to both global capitalism and effective altruism,” Srinivasan mused in her 2015 essay “Stop the Robot Apocalypse.” This was my first reaction, too; that effective altruism was insultingly naive, like being told to make love, not war. The economist Daron Acemoglu wrote his own version of this critique in “Response to Effective Altruism,” by reminding the readers of the Boston Review that “every measurement involves value judgments” and instructing them to always “consider the long-term consequences” of their actions.
These remarks are useful, like most platitudes. Yet they seem oddly dusty next to effective altruism’s furious pace: its untold billions, and its tenacious army of charities and entrepreneurs. The platitudes are especially hard to remember at the conference, where I am surrounded by pairs of people from all across the world, totally absorbed by what the other has to say—updating, as the effective altruists say. Indeed, it was decades of internal debate that precipitated the emergence of longtermists, devoted entirely to the “long-term consequences” of their actions. Effective altruists can be many things, but they tend to be very serious about what they do.
The other camp is hip to this, but they are horrified by what they see as a movement that is destroying its adherents’ lives, wasting their money and hurting people abroad. “Having a body part removed in order to ‘bite the utilitarian bullet’ or obsessing about the moral iniquity of ice cream will be seen as examples of lives deformed by a simple-minded and disputable theory,” wrote the political philosopher John Gray in “How & How Not to Be Good” (referring to the tendency of some effective altruists to donate their kidneys to strangers). These articles occasionally get off a few shots on Bentham or Singer, but one gets the uncomfortable feeling that they have been provoked by something else: a deep personal loathing for their subjects. “The rationalist believes he is pursuing something when actually he is fleeing something, absolutely terrified witless that anyone might for one second mistake him for a human being,” elegized Ari Schulman in “Open Wallets, Empty Hearts.” These critics propose crude, pointed alternatives. In place of the Shallow Pond, for instance, Schulman advances the “humble paper towel problem”: “if you’re in a public restroom and see a paper towel on the floor, you should throw it away even if you weren’t the one who dropped it.” That’s it? Can we expect no more from ourselves?
By dividing the territory in half, however, these essays fritter away a good portion of their force, postponing the sucker punch which might lay effective altruism low once and for all. To my mind, the best blow ever landed on effective altruism came from Stanford professor Leif Wenar, in his magisterial and bloodthirsty “The Deaths of Effective Altruism.” This is perhaps the only article ever written on the subject that I can imagine convincing a young EA to change course. At several moments, the piece is very close to synthesizing a critique of the philosophy with an attack on its adherents—a demonstration that the whole project is rotten, bottom to top. For instance, Wenar correctly and incisively diagnoses a strain of childish bravado within effective altruism, even when it comes to thorny multivariable issues like global development. The effective altruists try to “save lives by being clever—the philosopher’s way of being the hero.” But in their haste to execute, they stop listening to the experts, or even to the people they are trying to help, and so they make stupid mistakes. “Behind his glazed eyes,” Wenar wrote, “the hero is thinking, ‘They’re trying to stop me.’”
Yet even Wenar has trouble sticking the landing. After several thousand words of criticism, he offers a few cursory words to his friend Aaron (a fellow professor of philosophy) who once helped local villagers install toilets while on an annual surfing trip in Indonesia. “No billionaires, no hype, no heroes,” writes Wenar. “Aaron will talk about the island if you ask him, but he never brags.” The critical consensus seems to be that minimalism is the marker of morality. Like fashion and figure skating, the best ethics should be understated and effortlessly chic. In other words, the effective altruists may be nice people, but their attitude is backwards—instead of talking about a problem, they try to solve it themselves. “That’s just what firemen do,” taunts Wenar. “That’s just what Spider-Man does.”
We must admit there is something a bit hypocritical in this kind of criticism. Both on the right and the left (and recently in the pages of this very magazine), professional critics tend to enjoy fussing over my generation’s alleged crisis of purpose. And yet here comes effective altruism, offering not only community and purpose but also the rituals that—more than faith—keep the beat of religious life. Like every major religious tradition in the world, it provides moral instruction regarding diet, vocation and charity. In many ways, it is not much more or less than the rediscovery of the ancient and universal ambition to help the downtrodden, updated for the ears of a ruling class accustomed to statistics. Like any religious practice, it can be more or less demanding, communal, mystical, stringent or bonkers, depending on who you ask.
And what do our elders propose as a replacement? Aristotle? Arendt? Marx? A life in community? A life of the mind? Worst of all, more dialogue? Our purpose, apparently, is never to discover our purpose, and the whole program, which is supposed to represent “the good,” sounds suspiciously like pursuing a Ph.D. By contrast, the job board 80,000 Hours, a common gateway into effective altruism, offers this advice to young people: “Unless you happen to be the heir to a large estate, your career is the single biggest resource you have to contribute to the world … get good at something that helps others, and passion will follow.”
For a young, bright college student, living in the most powerful and prosperous nation on earth, it must take a very good imagination, or very bad intentions, to ignore effective altruism.
●
My own exposure to effective altruism has been brief, but I’ll conclude with a personal story. After finishing my stint on the farm, I went back to college. Within their dreamy pot clouds, I got to know the school’s much-advertised terror squads of radical leftists, ready to take cold-blooded action on behalf of the world’s downtrodden, whether human, animal or plant.
And so, heartened by my exposure to effective altruism at the conference, I began to wonder about our annual spring festival, which typically costs the college north of one hundred thousand dollars every year. At an early student-body meeting, I ascended to the podium and read out my proposal to divert a significant fraction of those funds toward a selection of worthy causes. According to a Gazan friend of mine, a hundred thousand dollars could rescue twenty of her comrades under siege and deliver them safely to other countries. I ticked off a few other possibilities—malaria nets, deworming pills and support for local homeless—hoping that these kinds of calculations would intrigue the student body. I concluded by proposing the establishment of a committee, and took my seat.
Then, in a flurry of activity, the whole room rose up; a couple students hissed some quick, vicious remarks about the importance of the prom for “marginalized seniors”; and the proposal was shot down.
I was furious. Marginalized seniors? What could they possibly mean? A good hunk of the prom cash had been earmarked to provide bottomless beer. The lives of Gazans, Kenyans, Indians and our own country’s homeless were scheduled, literally, to be pissed away. All at once, the students began to clap.
Sitting in the auditorium, I was overwhelmed by a memory of winter on the ranch. After the first frost, my friends and I would sometimes take to the frozen fields to play baseball. Baseball is a long, tedious sport, providing only the patient thrill of planning pitches, counting strikes, checking runners and keeping score. Unfortunately, as the season went on, this became more difficult, and we began to lose baseballs in snowdrifts. So we began pitching clods of ice, and happily whacked them for a little while, the balls exploding into a glittering mist. Iceball, as we called it, was about something meaner: that single moment of brutish glory, destined to evaporate and entail nothing.
Listening to the deafening applause that day in the auditorium, it seemed to me that all of our seminars and lectures and assemblies had come to nothing; that they were pointless and destructive, like an inning of iceball. Despite the national racket, the fact remains that students in American universities still routinely write, vote, post, protest and otherwise speak their mind on all manner of issues. When will they get to work?
Friday
On a dark, gray day in October 2025, the first dozen attendees of the Effective Altruism Global conference make their way to the Sheraton Hotel just north of Times Square in New York City. Cold and wet, I arrive at the Sheraton around 5 p.m., making my way through a lobby full of Brazilian businessmen holding squash racquets, and head down the stairs to a quiet hallway, where I find a table staffed by a young Englishman. After asking for my name, he prints it in lovely cursive on an index card and clips it onto a lanyard. Another young man gives me a pale gray shirt with delicate Victorian-era illustrations of birds, with captions in the style of amateur botanical sketches. They read:
●
While their personal motivations vary, most effective altruists trace their intellectual heritage back to the 1972 publication of an essay by the Australian philosopher Peter Singer entitled “Famine, Affluence, and Morality.” The essay is brief and repetitive, like biblical poetry, and urges readers to donate to relief efforts for the famine that was then afflicting Bangladesh. After a couple pages of such appeals, Singer proposes a thought experiment:
Singer observes elsewhere that something like this really happened in the town of Wigan, in England: in 2007, two cops hesitated to dive into a pond after a ten-year-old boy who had tried to save his sister from drowning and then slipped underwater. Two elderly fishermen saved the sister but were unable to rescue the boy. Several minutes later, the boy’s stepfather arrived and pulled his dead body out of the water. “I don’t know why they didn’t go in,” the boy’s stepfather told the media. “I can’t understand it. If I had been walking along a canal and seen a child drowning I would have jumped in. You don’t have to be trained to jump in after a drowning child.”
Despite the assurance of Assistant Chief Constable David Thompson that the officers “acted correctly,” the Greater Manchester Police and most British media outlets refused to report the names of the officers on duty. Whether or not they had followed protocol, there was something obviously shameful about their willingness to stay on the banks. And yet, Singer pointed out, there is a disturbing amount of begging and cajoling that must go on before we feel compelled to provide even a minimal amount of aid to those suffering abroad. We do not want to give. What’s more, we are not quite expected to give.
“If it is in our power to prevent something very bad from happening, without thereby sacrificing anything morally significant, we ought, morally, to do it,” wrote Singer. “The uncontroversial appearance of the principle just stated is deceptive. If it were acted upon, even in its qualified form, our lives, our society, and our world would be fundamentally changed.”
Hearing the Parable of the Shallow Pond is a turning point for many young EAs. Put one way, it envisions ethics not as an academic football but as the science of helping others. Put another way, it instructs us to feel responsible for both what we do and what we do not do. Like many parables, it does not always find a receptive heart.
I first heard the Parable of the Shallow Pond from a prominent effective altruist named Zachary Robinson at a college near the farm where I worked along with several other young people. Robinson is a slight young man with friendly features, and his eyes darted cautiously across the small crowd as he ascended to a wooden lectern. I remember him explaining the parable and speaking briefly about his work. But frankly, my recollection of the afternoon is a little spotty, for shortly after he imparted his parable, a tide of objections rose up in my mind that effectively extinguished my ability to hear the rest of the talk. Sure—but who is this fictitious kid? You’re really asking me to give money to someone halfway across the world, with whom I have no connection and about whom I know nothing?
When Zachary finished, I remember rising with a queer feeling in my stomach. Back at the farm, my friends and I took a stroll through the communal garden, where we proceeded to stomp on effective altruism from all conceivable angles. I had just finished struggling through a class on The Phenomenology of Spirit, which made the simplicity of Zachary’s scheme seem especially insulting. I remember one of my friends snatching a ripe tomato off its vine and holding it before us as evidence: the prickly crown and the thin red fuzz, birthed from a skinny green plant that twisted its way around a trellis.
“The world is a mysterious place,” he declared.
We nodded. Wasn’t it? Just a year ago I had attended an urban high school, enjoying all the luxuries of a modern city, and now here I was out in the country, a good thirty miles from civilization, rising with the sun to harvest tomatoes. Life had borne me and my friends like leaves along like a mighty river, across the country and into the high mountains, and I found it difficult to imagine what the stakes would be in pretending that we had the right to administer pleasure and pain, life and death, to our fellow creatures. So I made a determined effort to forget the parable.
Yet it began to seriously nag at me a few months later, after I had gotten to know a cow on the ranch named Boo. Boo’s mother had died unexpectedly after his birth, and the student body had brought him onto the main campus and fed him milk out of a baby bottle. Now, fully grown, he had only good feelings toward humans. When I went out to the fields in the morning, he would gallop toward me, hoping to get scratched between his eyes. Sometimes he would help me move the metal irrigation lines by nudging them with his big, black head. Then, one day, I went out into the fields to find that all the cows, including Boo, had been trucked to slaughter at a big meatpacking operation, the kind that stretches endlessly along the side of Western highways.
I knew it was coming, of course—I was living on a ranch. I had butchered hundreds of chickens, and I wasn’t squeamish about the mechanics of slaughter. By growing alfalfa, I had been fattening Boo myself. But he was there, and then he was gone. It was the first time anyone I knew had died. Alone in the field, I wept.
Over the following weeks, I began to feel that there was something misguided about our disembodied food system, and in my musings I returned to the parable. Boo would be eaten by someone who’d never known him. In return, the ranch received a trivial amount of money, and the student farmers trudged through cold mornings without a friend. We had gained something insignificant and lost something tremendous. We had miscalculated.
Saturday
The first morning of the conference, several hundred young EAs tramp into the Sheraton’s lobby through a light rain. During the introductory ceremony, I take my seat beside a very elegantly dressed Indian man with an Enron sticker on his laptop.
“Where did you get that sticker?” I ask him.
“My boss,” he replies calmly.
I look around the table—his buddies keep a straight face. Finally, they burst out laughing.
“It’s just a joke,” they exclaim. “Of course he doesn’t work for Enron! He runs the effective altruist club at the University of Wisconsin, Madison.”
“I’m focused on Global Health and Wellbeing,” he tells me proudly.
I give him an approving nod. Global Health and Wellbeing is the most ancient of the EA departments, and confers a certain noble modesty, not unlike being a Hufflepuff. It is premised entirely on the contrast drawn by “Famine, Affluence, and Morality”: that there is something repugnant about the coexistence of extreme wealth and extreme poverty. This fact is strikingly illustrated by a bar graph that Giving What We Can has erected outside of the conference hall. The chart displays the wealth of the world’s upper percentile as a skyscraper at the far right, teetering evilly over nine unimaginably squat little flats—the global poor. It seems both easy and right to knock it down, like a tower of blocks.
The public imagination can generate no shortage of explanations for global poverty. Perhaps it can be ascribed to American imperialism, or to the sinister machinations of global capital, or maybe somehow to the starving themselves. But EAs tend to shrug off these explanations. As the EAs see it, the reason a child is dying, right now, of dehydration, is because you have refused to buy him a bottle of water. By donating seventeen dollars to a charity called Evidence Action, you can provide ten Kenyan children with clean water for an entire year.
To free up cash for projects like these, effective altruists have built out an impressive toolkit of money-saving devices. For instance, most of the EAs at my table are not eating breakfast. Instead they are drinking Soylent, a nutritional shake which is perplexingly named after the nutritional wafer made from human remains in the 1973 dystopian thriller Soylent Green. Soylent is a nutritionally complete meal replacement, meaning that it contains all 28 vitamins and minerals required to maintain and regenerate the human body. It tastes sweet and slightly stale, and is almost redundantly smooth, like liquefied pudding, clotting unpleasantly in your esophagus before sliding into the pit of your stomach, where it sticks like cement. Scientifically, you can drink only Soylent for the rest of your life and remain in perfect health.
“It can replace any meal,” announces one of the EAs, sniffing his Soylent nostalgically and evidently recalling some nightmarish phase of his working life. “But not every meal.”
Soylent was first cooked up in San Francisco in 2013 by an electrical engineer named Rob Rhinehart, and received an enthusiastic blast of venture capital from Y Combinator and Andreessen Horowitz. Initially marketed as a futuristic lifehack, it is now primarily sold in suburban Walmarts, where its gentle, sludgy texture is apparently indispensable to people with throat cancer and irritable bowel syndrome. It is packaged like a geriatric protein shake, such as you might find in a pharmacy next to the laxatives.
I suspect that neither the digestive benefits nor the micronutrient content, however, are what recommend Soylent to EAs, who are not as health- and longevity-obsessed as their second cousins in the tech sector. Rather, the many bottles of Soylent, tossed with happy abandon across the tables and floors of the conference rooms, serve as a persistent reminder of the solvability of problems. One has only to read the many descriptions of poisonous mushrooms in a mycological field guide to feel a gruesome awe at the millennia of trial and error involved in determining what is healthy for us to eat and what is not. Well, here it is.
As such, Soylent is a good introduction to effective altruism: both to its pugnacious quest to solve problems (like spending money on food) and its blithe disinterest in marketing those solutions to the general public.
For a more familiar example, take domestic crime, which costs the United States at least $300 billion annually through policing, prisons and courts, not counting the social costs of community violence and instability. This despite the fact that crime has received more formal attention over the last several thousand years than perhaps any other problem in human history. Substantial chunks of most major religious texts are devoted to crime and punishment; it is a critical dividing line between the worldviews of the major parties.
Alright, but what if I told you that the famous decline in violent crime that began in the 1970s and 1980s occurred simultaneously with the intensive regulation and eventual ban of leaded paint and gasoline? Or that violent crime rates in America since the 1930s track extraordinarily closely with levels of lead in blood samples from preschoolers taken 23 years earlier? Or that protecting a population from lead exposure has been shown not only to raise cognitive ability and improve impulse control but also simply to reduce violent crime by as much as 56 percent?
So instead of making murals about urban decay, or donating to the New York Public Library, or architecting even bigger prisons, a good effective altruist will book a flight to Malawi to meet with government officials who are interested in mandating alternative lead-free paint formulas in line with contemporary medical recommendations—a program that requires a small handful of people to draw up a single solid meta-analysis and hold a few conversations with a few officials who already more or less agree with the project. In 2024, following just such an effort by the Lead Exposure Elimination Project and its partners, the country of Nigeria agreed to implement enforceable regulations on the sale of lead paint. That is what the effective altruists call a sound intervention.
●
Lunch is served around noon in the buffet hall, which remains at the same thunderous decibel level that it attained early that morning. Not one effective altruist stops talking long enough to chew. Every 25 minutes, everyone in the room shakes their partner’s hand and heads to another meeting. At any given moment, at each banner and at all the ballroom tables in between, there are several hundred conversations going on simultaneously, all echoing off the Sheraton’s enormous ceilings.
Three long tables are set up with coffee, sandwiches, salads and soups. The Soylents are arrayed in a phalanx on the back wall, and little bags of sweet potato chips are always available. What’s more, everything is vegan, even (rather ambitiously) the turkey, provolone and honey-mustard sandwiches.
At first glance, the alliance between vegans and effective altruists might seem obvious. A consciousness that has already been expanded to encompass all the world’s poor is surely big enough to fit the world’s animals. A declaration about reducing harm surely comes with a suggestion about nonviolence. And any believer in a moral framework that attempts to quantify suffering should be immediately alarmed by the numbers involved: 85 billion land animals and trillions of fish are raised and slaughtered annually in terrifying conditions.
Indeed, effective altruism has historically had a very strong connection to animals. In 1975, a couple years after publishing the Parable of the Shallow Pond, Peter Singer wrote the provocative Animal Liberation, a book he would later describe as “one of the best strategic decisions that I made” and a work that had the most “positive, practical impact” of any in his career. In turning his attention to animals he was following a well-trodden path: many of the same nineteenth-century English reformers who preached utilitarianism and abolition also agitated for animal welfare. Jeremy Bentham, the founding father of utilitarianism, had explained the connection in this way:
“Why,” Bentham asked, “should the law refuse its protection to any sensitive being?”
Peter Singer makes a similar argument in Animal Liberation, providing a usefully detailed description of modern animal farming and drawing the same parallels between animal and human rights. Yet his account is somewhat less earnest than Bentham’s, and reads more like a thought experiment—a long appendix to “Famine, Affluence, and Morality.” The lesson is that if you have truly understood the Parable of the Shallow Pond—that you must judge what you do by all of its consequences—then you must reconsider many seemingly innocent or almost praiseworthy habits, including eating fish and drinking milk.
The logical appeal of Singer’s consequentialism, however, contrasts sharply with the movement that spawned directly from Animal Liberation. There are a great number of animal-rights activists at the conference, hoping to squeeze a few extra cents out of the EA grant makers, who are notoriously fickle and exacting. These activists see animal liberation as a real, glorious future, and they intend to bring this future about by infiltrating, protesting and disrupting, just like twentieth-century leftist revolutionaries—many of whom, incidentally, were vegan or vegetarian themselves. There are vegan graffiti artists and vegan punk bands1Core members Ian MacKaye, Greg Ginn and Ray Cappo (of Minor Threat, Black Flag and Youth of Today, respectively) are outspoken vegans, as was Earth Crisis and all of the Subhumans and John Joseph—frontman for the Cro-Mags, Ironman athlete and author of the indelible Meat Is for Pussies. and vegan terrorists. Most of them are women.
By contrast, Singer’s band of effective altruists are overwhelmingly male, and were made by their Creator to write code, not to eat lentil soup and use ridiculous phrases like “animal rights.” They do not rail against hunters and butchers and Big Ag. They prefer to consider “win-wins,” another term of art borrowed from Wall Street, which refers to situations in which none of the parties involved are forced to make sacrifices.
An infamous example of hardcore Singerian marginal thinking as applied to animals is a nonprofit called the Shrimp Welfare Project. Small and unloved, shrimp are not provided by most Western countries with even minimal standards of humane handling. Their treatment is gruesome, to say the least; they spend their entire 120-day existence in a pool of their own excrement, as a result of which they overwhelmingly become depressed, lethargic, anorexic, cannibalistic or enter hypoxic shock. A small number of female shrimp are removed in order to perpetuate the stock, and upon removal, their eyestalks are “ablated”—severed or crushed—which depresses reproduction-regulating hormones and accelerates ovulation. After several months, all the shrimp are taken out of the water and left to asphyxiate for as long as an hour in a bucket of ice.
To the average vegan activist, this is a perfect example of the perverse excesses allowed by a society that does not recognize animal rights, either culturally or in law. Protests, billboards, consumer outreach and sabotage with explosives might all be in order.
But Andrés Jiménez Zorrilla and Aaron Boddy, two entrepreneurial effective altruists, saw things differently. Fish and crustaceans are the only mass-farmed animals without stunning regulations. So they partnered with the Norwegian company Optimar, which manufactures heavy-duty electrical stunners capable of stunning over ten metric tons of shrimp per hour. With money raised from a host of EA grant makers, they began purchasing stunners wholesale at a price of $55,000 per unit and giving them—for free—to shrimp farmers.
Millions of dollars’ worth of expensive machines, delivered to them personally by tens of thousands of smiling vegans, must have caught all those Vietnamese, Honduran, Sri Lankan, Indian, Indonesian, Ecuadorian, Thai, Venezuelan and Nigerian shrimp farmers by surprise. But they were happy to route their shrimp through a stunner before slaughter and add a “Humanely Stunned” label on the bags. For their part, the Shrimp Welfare Project was happy to become the single most effective animal charity on the planet, with 1,400 shrimp stunned per dollar donated.
Lateral thinking or sleight of hand? Many traditional animal advocates were outraged. They believe that the Shrimp Welfare Project is the latest in a string of “welfarist” compromises, with the most famous being the introduction of “cage-free” eggs in the 2000s—that is, eggs laid by chickens who have been removed from cages but otherwise remain in exactly the same dank barns as before.
Just a few blocks down the street from our hotel, for instance, local activists are protesting Marriott International for serving eggs laid by caged hens. They’re participating in something called RAGE (or, Resistance Against Greed and Exploitation): a global pressure campaign to hold companies to animal-welfare promises which, in Marriott’s case, were extracted nearly a decade ago. Isn’t it cruel to make these activists stand out in the cold, pleading with Marriott for another empty deadline? Still, effective altruists smile on these cage-free campaigns, which, despite some corporate flip-flopping, have over the past couple decades freed several billion hens worldwide from cages whose minimal space requirements induced gruesome self-mutilation, and whose wire struts irreparably mangled the sensitive bottoms of their hens’ feet. The activists, they say, prefer the romance of a hopelessly distant revolution to interventions that improve the lives of animals in the present—billions of chickens and trillions of shrimp.
●
By late afternoon, my head is swimming from a full day in the Sheraton’s grand ballroom. I stagger into a quieter, gray-toned lecture hall where a young researcher named Mal Graham is addressing a small crowd. Graham holds a B.A. in physics and another B.A. in philosophy from Oxford and a Ph.D. in Engineering Mechanics at Virginia Tech, where their research was on “dynamic gap-crossing movements in jumping and flying snakes.” (By the standards of the conference, this is a fairly normal academic background.) The resume implies that Graham is interested in another possible theater for continued EA-vegan collaboration—the young, very trippy field of Wild Animal Welfare.
“The literature estimates suggest that up to a billion birds are killed annually in North America from ‘bird-window collisions,’” Graham is informing the room. “But… we have to remember that protecting them from a single collision does not make them immortal. So under some philosophical views that are predominant within EA, the question becomes, ‘Is the expected value of the time we gain them by preventing them from having that collision better than the expected value from the difference in suffering between their new death and their old death?’”
If each creature on the earth was represented by a dot, and the human dots could fit on a quarter, then the wild animal population would fill a FIFA regulation-size soccer field. The overwhelming majority of these animals may well inhabit a living nightmare. In New England, for example, a warming climate has encouraged the population growth of a tick called Dermacentor albipictus; tens of thousands of these ticks attach to a single moose, until large patches of its skin are invisible underneath great carpets of bloated insects. All sucking at once, the ticks can kill a moose—usually a calf—over the course of several months through anemia, blood loss and exhaustion. Each biological order has its own demons. Langur monkey populations in India will crush the infants of defeated tribes in order to trigger estrus in their mothers. In tropical waters, the fat-lipped cookiecutter shark will suction onto live prey, twisting out plugs of fresh meat. And the less said about parasitic wasps, the better.
There is only one major organization devoted to alleviating the suffering of wild animals. It is called the Wild Animal Initiative, and it occasionally funds researchers who have bright ideas about how to interrupt the endless cycle of suffering in nature. But WAI’s research remains primarily speculative; meanwhile, the yearly budget for both farmed and wild animal welfare is under three hundred million dollars, which is several times less than what Americans spend every year on Halloween costumes for their pets.
Feeling vaguely nauseous at the magnitude of evil in the world, I wander out of the talk. “Trauma,” a school counselor once told me kindly, “is getting hurt and not being able to do anything about it.”
Sunday
On the final day of the conference, I hop out of bed bright and early, determined to pack my Sunday with a respectable number of chats. However, half-asleep, I grab the wrong coat on my way out the door. I usually wear a blue puffer jacket, but today I accidentally snatch a very chic brown canvas windbreaker instead.
Forty-five minutes later, progressing rapidly through the stages of frostbite that precede amputation, I stagger through the doors of the Sheraton. I make my way to the buffet tables and plant myself beside the tanks of coffee, which have miniature burners beneath them. I draw up a cup for myself and absentmindedly nibble on some vegan cookies.
Truthfully, I imagine that part of the reason I did not immediately turn back for the winter coat was that I expected my willingness to self-sacrifice would win me some points with the altruists. Wasn’t sustained exposure to the New York City winter proof of my indifference to the pleasures and privations of the material world? Were Peter Singer drowning in a very cold pond, I clearly would not hesitate to pull him out and engage him in a chat.
My first chat of the day is with a rangy Dutch fellow named Sjir Hoeijmakers,2The first name is pronounced “Sheer”; the last name appears to be some kind of diabolical Rorschach test contrived for the amusement of Dutch natives. who bears a beguiling resemblance to the infamously hot English actor Tom Hiddleston. Sjir and I are scheduled to walk around the perimeter of the hotel. He looks me up and down.
“If you are going to be chilly,” he says with a soft accent, “then I’m happy to stay inside.”
I assure him that my windbreaker is warmer than it looks, and follow him out into the blistering cold.
Sjir is affectionate and eminently calm—the very best of the effective-altruist personality types. He does not seem indifferent to pleasures or privations. In fact, he is very tall and healthy-looking. Sjir is also the CEO of Giving What We Can, which is the sanest and most publicly legible of the EA charities.
“We encourage people to pledge 10 percent of their income to charity,” he says. “And we make some recommendations as to where we think their donation could have the largest impact.”
Like veganism, the 10 percent figure is an old-school Singerism. It was inspired by the observation that two major religious traditions—Jewish tzedakah and the Christian tithe—recommend donating 10 percent of one’s income to charity. According to the Giving What We Can website, over eleven thousand people have answered the call to pledge, directing over sixty million dollars toward good causes every year.
At this point in the essay, a pledge of six thousand dollars per person may actually strike you as a little small. One young acquaintance of mine, upon discovering effective altruism, promptly underwent surgery and donated his kidney to a random stranger. I explained that, upon seeing pictures of him in the hospital, weak and frail and grinning broadly, I thought: that is an altruist.
“Oh, we’re not against that at all,” Sjir says affably.
“Can I ask what you donate?”
“Sure,” he says. “I donate 50 percent of my income.”
“How do you feel about that?”
“It gives me joy, and it leaves me more than enough to cover my needs. Many of the things that provide the most meaning to my life are free, such as my relationships with my family and loved ones.”
“Don’t you think you should be giving more?”
Sjir explains, “Some people work themselves very hard, and donate a lot of money, but they put so much pressure on themselves that they become unhappy, and may even burn out. And then they don’t have much of an impact at all.”
●
Most of Sunday afternoon is taken up by a job fair, which occupies half of the Sheraton’s main ballroom. I perambulate idly through the stalls, rubbing some elbows and half-hoping to land myself a sweet summer internship.
But I am not in luck. Of fifty stalls, a whopping 27 are devoted to AI research, AI journalism, AI alignment, AI governance, AI solutions…
What is this? Have I stumbled into a trade show, an expo of third- and fourth-tier artificial-intelligence startups, looking to skim some spoonfuls for themselves off the top of that rampaging gravy train; that utterly obvious, omnipresent, unmissable, all-consuming pop-culture phenomenon? Who is neglecting artificial intelligence? What happened to mosquito nets? What happened to the welfare of shrimp?
The answer is that recently, with the poise and self-discipline of Shaolin monks, many effective altruists have pivoted away from their long-standing commitments in animal welfare and global health and toward a third theater: “AI safety,” which, according to the estimates of several insiders, is currently consuming about half of the movement’s labor and capital. Holden Karnofsky, the former CEO of heavyweight EA grant maker Coefficient Giving, has left his post to work full-time on AI safety; 80,000 Hours, the influential EA job board, now almost exclusively posts jobs in AI safety. And on the stages at EA Global, four lectures are devoted to artificial intelligence—double what is given to either animal welfare or Global Health and Wellbeing.
AI safety falls into the traditional domain of a branch of effective altruists known as “longtermists,” who first came onto the scene with the 2003 publication of a deeply loopy and kaleidoscopic paper by Oxford philosopher Nick Bostrom, called “Astronomical Waste: The Opportunity Cost of Delayed Technological Development.” The paper proposed colonization of the Virgo Supercluster as the prime directive of effective altruism, and warned against any present-day distractions that might hinder that future.
After several years of refinement, the theory was given more coherent public expression in 2020 by a book called The Precipice, written by a famous EA named Toby Ord. While the whole book is entertaining, much of the thesis is available in its dedication, which reads as follows:
As Ord goes on to explain, to “foreclose” the lives of those future trillions is morally comparable to murder. Just as our nonaction on the banks of the Shallow Pond makes us responsible for the child’s death, so our nonaction in the present makes us responsible for those in the future: “A person does not matter less, the further away from you they are in space … People matter equally regardless of their temporal location too.”
Longtermism’s preoccupation with the lives of future humans ends up doing wonky things to the utilitarian pleasure-pain calculus because, as Ord’s dedication makes clear, the number of humans alive today will be dwarfed by the number of humans who will live over the next several billion years. The whole EA balancing act—five deworming pills or 250 iron supplements? Sixteen malaria vaccines or three cycles of Group Interpersonal Psychotherapy?—is ultimately steamrolled by that single juggernaut, infinitely valued, weighed against almost nothing.
Events that might threaten the infinite future of our species are referred to as existential risks, or “x-risks.” Ord’s proposed list of x-risks goes like this:
Among effective altruists, however, artificial intelligence is the foremost concern. Artificial intelligence could electronically order DNA laboratories to print unstoppable bioweapons. It could hack into weapons systems and initiate nuclear war by itself, no matter how elaborate the international deterrence architecture. It might be able to manipulate a whole society by generating misinformation to serve its own aims. It is getting rapidly more intelligent without supervision, and shows both the inclination and ability to deceive its supervisors. But for at least the next few years, it remains under the control of a few logically minded billionaires. Even better, two decades of EA penetration into Silicon Valley means that some of these AI billionaires are already effective altruists themselves.
The so-called “alignment” of AI with human values is the subject of “A High-Level Hope for Managing AI X-Risk,” and is hosted by Ajeya Cotra, then Coefficient Giving’s senior adviser on artificial intelligence. The room is packed, the best attendance I’ve seen for any of the lectures—the air is heavy with suspense and unease, as Cotra unveils a chart with a single line hurtling through the future, up toward “Doom.”
Oh, bother. I’m very glad that someone in the world is keeping their eye on artificial intelligence. But why does it have to be the effective altruists? The virtue of utilitarianism, according to the philosopher Martha Nussbaum, is its ”general radicalism and [its] skepticism about conventional morality, [its] willingness to follow the ethical argument wherever it leads.” Almost to a man, the effective altruists I meet at the conference are smart, optimistic and ambitious. These new themes of tragedy and hopelessness—where did they come from?
“People say to me… ‘What’s your ask?’” Eliezer Yudkowsky, the progenitor of the AI-safety field, read aloud from his phone at a TED Talk in 2023. “I do not have any realistic plan. … My best bad take is that we need an international coalition banning large AI training runs… I say this not expecting that to actually happen. I say this expecting that we all just die.”
As I leave the conference, jobless and annoyed, I can’t help feeling that this turn toward AI safety is barely even utilitarian, since its concern is not with joy and suffering but with lives. It frets over future generations despite not knowing whether they will end up being happy or miserable or just alright. Our descendants might turn on and annihilate each other, after having multiplied and thereby magnified the scale of suffering involved in their destruction. To feed themselves, our descendants might raise quadrillions of animals in filth and misery. Or perhaps they will themselves become enslaved and eaten by an undiscovered alien species, lurking somewhere deep in the Virgo Supercluster. How much crazier would that be than the alternatives? In any case, saving the world seems like a worthier task for politically minded thought leaders than for the marginal effective altruist.
I pose this question to Ronen Bar, a famous undercover investigator of animal farms. He appears momentarily puzzled.
“Ah,” he says carefully. “You are only seeing what is happening now, and not what will be happening. Artificial intelligence is on the rise. People are using it. If it becomes dominant, that’s going to affect everybody, including animals—whether it’s aligned or misaligned. Either way, ignoring AI is not wise, from an animal point of view.”
That is that. A funny sort of unspoken truce. Some effective altruists are analyzing floating-point matrices for signs of the apocalypse, while at the same moment others are installing electrical stunners at shrimp farms in rural Vietnam, and still others are interviewing Nigerian mothers with young children whose intestines have been eaten by roundworms. I am convinced that deep down these groups must loathe each other, but so far as I can tell, they do not.
●
Most people I know are scornful of effective altruism, if they have heard of it at all: over the past decade, the effective altruists have managed to irritate Americans from all walks of life.3So much so that several of my interviewees advised me to publish this article pseudonymously, on the grounds that there is something a bit crazed and extrajudicial about effective altruism which has not historically recommended its affiliates to potential employers. After some reflection, I decided that they were probably right. Still, the decision annoyed me, both because it feels personally awkward and because I find it inexplicable that the world is not friendlier to these ideas. They are too self-righteous for Silicon Valley and not self-righteous enough for the universities. They are too interested in money to be socialists but too interested in social justice to be entrepreneurs. They are too awkward to be activists but too active to be geeks.
Effective altruism always seems to catch audiences on the wrong foot. The “weak” version of EA usually presented to the public, namely that you should be awake to the suffering of your fellow creatures, strikes people as lame and anticlimactic. But the “strong” version of effective altruism—that you are morally required to overhaul your life in favor of helping others, whether through your donations, career or diet—seems to go way too far.
Critics of EA like to divide themselves along these lines as well. The first camp is led by academics, like Amia Srinivasan, who find the ethics of effective altruism to be frustratingly shallow. “There is a seemingly unanswerable logic, at once natural and magical, simple and totalising, to both global capitalism and effective altruism,” Srinivasan mused in her 2015 essay “Stop the Robot Apocalypse.” This was my first reaction, too; that effective altruism was insultingly naive, like being told to make love, not war. The economist Daron Acemoglu wrote his own version of this critique in “Response to Effective Altruism,” by reminding the readers of the Boston Review that “every measurement involves value judgments” and instructing them to always “consider the long-term consequences” of their actions.
These remarks are useful, like most platitudes. Yet they seem oddly dusty next to effective altruism’s furious pace: its untold billions, and its tenacious army of charities and entrepreneurs. The platitudes are especially hard to remember at the conference, where I am surrounded by pairs of people from all across the world, totally absorbed by what the other has to say—updating, as the effective altruists say. Indeed, it was decades of internal debate that precipitated the emergence of longtermists, devoted entirely to the “long-term consequences” of their actions. Effective altruists can be many things, but they tend to be very serious about what they do.
The other camp is hip to this, but they are horrified by what they see as a movement that is destroying its adherents’ lives, wasting their money and hurting people abroad. “Having a body part removed in order to ‘bite the utilitarian bullet’ or obsessing about the moral iniquity of ice cream will be seen as examples of lives deformed by a simple-minded and disputable theory,” wrote the political philosopher John Gray in “How & How Not to Be Good” (referring to the tendency of some effective altruists to donate their kidneys to strangers). These articles occasionally get off a few shots on Bentham or Singer, but one gets the uncomfortable feeling that they have been provoked by something else: a deep personal loathing for their subjects. “The rationalist believes he is pursuing something when actually he is fleeing something, absolutely terrified witless that anyone might for one second mistake him for a human being,” elegized Ari Schulman in “Open Wallets, Empty Hearts.” These critics propose crude, pointed alternatives. In place of the Shallow Pond, for instance, Schulman advances the “humble paper towel problem”: “if you’re in a public restroom and see a paper towel on the floor, you should throw it away even if you weren’t the one who dropped it.” That’s it? Can we expect no more from ourselves?
By dividing the territory in half, however, these essays fritter away a good portion of their force, postponing the sucker punch which might lay effective altruism low once and for all. To my mind, the best blow ever landed on effective altruism came from Stanford professor Leif Wenar, in his magisterial and bloodthirsty “The Deaths of Effective Altruism.” This is perhaps the only article ever written on the subject that I can imagine convincing a young EA to change course. At several moments, the piece is very close to synthesizing a critique of the philosophy with an attack on its adherents—a demonstration that the whole project is rotten, bottom to top. For instance, Wenar correctly and incisively diagnoses a strain of childish bravado within effective altruism, even when it comes to thorny multivariable issues like global development. The effective altruists try to “save lives by being clever—the philosopher’s way of being the hero.” But in their haste to execute, they stop listening to the experts, or even to the people they are trying to help, and so they make stupid mistakes. “Behind his glazed eyes,” Wenar wrote, “the hero is thinking, ‘They’re trying to stop me.’”
Yet even Wenar has trouble sticking the landing. After several thousand words of criticism, he offers a few cursory words to his friend Aaron (a fellow professor of philosophy) who once helped local villagers install toilets while on an annual surfing trip in Indonesia. “No billionaires, no hype, no heroes,” writes Wenar. “Aaron will talk about the island if you ask him, but he never brags.” The critical consensus seems to be that minimalism is the marker of morality. Like fashion and figure skating, the best ethics should be understated and effortlessly chic. In other words, the effective altruists may be nice people, but their attitude is backwards—instead of talking about a problem, they try to solve it themselves. “That’s just what firemen do,” taunts Wenar. “That’s just what Spider-Man does.”
We must admit there is something a bit hypocritical in this kind of criticism. Both on the right and the left (and recently in the pages of this very magazine), professional critics tend to enjoy fussing over my generation’s alleged crisis of purpose. And yet here comes effective altruism, offering not only community and purpose but also the rituals that—more than faith—keep the beat of religious life. Like every major religious tradition in the world, it provides moral instruction regarding diet, vocation and charity. In many ways, it is not much more or less than the rediscovery of the ancient and universal ambition to help the downtrodden, updated for the ears of a ruling class accustomed to statistics. Like any religious practice, it can be more or less demanding, communal, mystical, stringent or bonkers, depending on who you ask.
And what do our elders propose as a replacement? Aristotle? Arendt? Marx? A life in community? A life of the mind? Worst of all, more dialogue? Our purpose, apparently, is never to discover our purpose, and the whole program, which is supposed to represent “the good,” sounds suspiciously like pursuing a Ph.D. By contrast, the job board 80,000 Hours, a common gateway into effective altruism, offers this advice to young people: “Unless you happen to be the heir to a large estate, your career is the single biggest resource you have to contribute to the world … get good at something that helps others, and passion will follow.”
For a young, bright college student, living in the most powerful and prosperous nation on earth, it must take a very good imagination, or very bad intentions, to ignore effective altruism.
●
My own exposure to effective altruism has been brief, but I’ll conclude with a personal story. After finishing my stint on the farm, I went back to college. Within their dreamy pot clouds, I got to know the school’s much-advertised terror squads of radical leftists, ready to take cold-blooded action on behalf of the world’s downtrodden, whether human, animal or plant.
And so, heartened by my exposure to effective altruism at the conference, I began to wonder about our annual spring festival, which typically costs the college north of one hundred thousand dollars every year. At an early student-body meeting, I ascended to the podium and read out my proposal to divert a significant fraction of those funds toward a selection of worthy causes. According to a Gazan friend of mine, a hundred thousand dollars could rescue twenty of her comrades under siege and deliver them safely to other countries. I ticked off a few other possibilities—malaria nets, deworming pills and support for local homeless—hoping that these kinds of calculations would intrigue the student body. I concluded by proposing the establishment of a committee, and took my seat.
Then, in a flurry of activity, the whole room rose up; a couple students hissed some quick, vicious remarks about the importance of the prom for “marginalized seniors”; and the proposal was shot down.
I was furious. Marginalized seniors? What could they possibly mean? A good hunk of the prom cash had been earmarked to provide bottomless beer. The lives of Gazans, Kenyans, Indians and our own country’s homeless were scheduled, literally, to be pissed away. All at once, the students began to clap.
Sitting in the auditorium, I was overwhelmed by a memory of winter on the ranch. After the first frost, my friends and I would sometimes take to the frozen fields to play baseball. Baseball is a long, tedious sport, providing only the patient thrill of planning pitches, counting strikes, checking runners and keeping score. Unfortunately, as the season went on, this became more difficult, and we began to lose baseballs in snowdrifts. So we began pitching clods of ice, and happily whacked them for a little while, the balls exploding into a glittering mist. Iceball, as we called it, was about something meaner: that single moment of brutish glory, destined to evaporate and entail nothing.
Listening to the deafening applause that day in the auditorium, it seemed to me that all of our seminars and lectures and assemblies had come to nothing; that they were pointless and destructive, like an inning of iceball. Despite the national racket, the fact remains that students in American universities still routinely write, vote, post, protest and otherwise speak their mind on all manner of issues. When will they get to work?
If you liked this essay, you’ll love reading The Point in print.