Here's What's Wrong with Effective Altruism

Effective altruism has no room for wolves or mountain climbing

By Emma Marris

Illustrations by Ellen Weinstein

May 16, 2023

Illustration by Ellen Weinstein

First, the good news: The much-maligned tech bros want to be ethical. But, being tech bros, they want a shiny, new ethic—one that’s iconoclastic, counter-intuitive, and algorithmic. Their ideal system is one that allows them to keep making lots of money as long as they give some of it away. The latest philosophical trend, called effective altruism, hits the spot nicely. 

Effective altruism is an ethical approach based on figuring out the best way to reduce suffering in the present and future. Facebook cofounder Dustin Moskovitz is a proponent; Elon Musk has dabbled. Last year, Oxford philosopher William MacAskill published a book-length argument for effective altruism, What We Owe the Future, which immediately hit the New York Times best-seller list and was greeted by (mostly) glowing reviews. Actor Joseph Gordon-Levitt said the book made him weep. Bill McKibben blurbed it. Before he lost his fortune and got slammed with a raft of criminal indictments, former crypto billionaire Sam Bankman-Fried was the poster child for “earning to give” as he amassed a huge fortune that he said he planned to give to charities.

The bad news is that effective altruism is flawed—and not just because its most prominent adherent is an alleged con man. Effective altruism doesn’t play well with most environmental ethics theories, in part because in the universe of effective altruism, only entities that can suffer matter. Trees, rivers, species—none of these are intrinsically valuable. Effective altruism distills all of ethics into an overriding variable: suffering. And that fatally oversimplifies the many ways in which the living world can be valuable. Effective altruism discounts the ethical dimensions of relationships, the rich braid of elements that make up a “good life,” and the moral worth of a species or a wetland.

Effective altruism discounts the ethical dimensions of relationships, the rich braid of elements that make up a "good life," and the moral worth of a species or a wetland.

Effective altruism (EA to devotees) traces its intellectual heritage to philosopher Peter Singer. In his classic 1972 essay “Famine, Affluence, and Morality,” Singer proposes a thought experiment: If you see a child drowning in a pond but saving them would ruin your nice clothes, should you do it? Of course. Then Singer asks another question: If you know that a child is starving halfway around the world and that you could save that child’s life with a modest donation—perhaps the cost of a new outfit—should you? For Singer, the answer is also clearly yes. The two situations, he argues, are morally identical. Helping suffering humans, wherever they may be, is not generosity—it is our moral duty. In 1975, Singer extended his argument to all sentient beings in his landmark book Animal Liberation.

Inspired by Singer, in 2009 MacAskill and another Oxford philosopher, Toby Ord, launched a group called Giving What We Can. The charity has helped donors direct money to “the world’s most effective charities.” Their picks range widely, from the Against Malaria Foundation, which buys bed nets for poor families in Africa and Asia, to TerraPraxis, which works on hydrogen fuel and advanced nuclear power. Along the way, Giving What We Can helped coin the term effective altruism.

But the thing about basing your ethic on a single principle is that you have to follow it to its logical conclusions—and some of those conclusions get a little weirder than a straightforward desire to prevent malaria or tackle climate change. 

What We Owe the Future makes the case that because many, many, many more people are likely to live in the future, working to make their lives better should be among our top priorities. It’s just math. And when MacAskill says “the future,” he doesn’t mean a generation or two out. He means the trillions of humans who he believes will live across many planets millions of years from now. 

MacAskill and his fellow “longtermists” are thus worried about threats to humanity that might play out over millennia, including climate change, nuclear annihilation, and genetically engineered pandemics. Longtermists are especially worried about the prospect of artificial intelligence taking over the planet and destroying humanity. A major criticism of longtermism is that our ability to predict what actions taken in the present will improve well-being in the future is extremely poor. MacAskill says that although the far future is very uncertain, the number of lives at stake is extremely large, so even if the probability that our actions will help is low, we should still try, because the ethical payoff could be incalculable.

But putting any probability on any event more than 1,000 years in the future is absurd. MacAskill claims, for example, that there is a 10 percent chance that human civilization will last for longer than a million years. I’m sorry, but no one can say a thing about what will happen a million years from now—not even an Oxford philosopher. Ronald Sandler, an environmental philosopher at Northeastern University, describes such calculations as “epistemic arrogance.” 

EA is especially problematic when it comes to the issue of wild animal suffering. There are thousands of times more individual animals (especially fish!) than individual humans, and each of them, in some way, suffers. So how do we ensure wild animal welfare? If we take EA seriously—if only the pain and pleasure of sentient creatures matters—then there is nothing wrong, in principle, with completely disassembling ecosystems to save prey from predators. 

Most people will likely find such a conclusion unintuitive, if not completely bonkers. Yet MacAskill suggests that the lives of wild animals are so riddled with suffering that these animals probably shouldn’t even exist. More than 90 percent of fish larvae die within days, he says, “eaten, starved, or suffocated.” As adults, he adds, they are likely to die horribly, to “suffocate as a result of an algal bloom, or be killed by parasites, or die of exhaustion after building their nest or releasing their eggs, or be torn apart or swallowed whole then crushed in a predator’s esophagus.” Since he believes that wild animals’ lives are plausibly “worse than nothing,” he thinks that the serious decline in the abundance of animals is probably a good thing. 

I recently wrote a book about animals, and I do not share MacAskill’s belief that animals’ lives are worse than nothing. Animals experience pleasure and even joy. From lizards basking in the sun to pretty much any hungry creature getting a satisfying mouthful of food, animals can have high highs as well as low lows. Even the fish evoked by MacAskill that struggles upstream to mate probably feels something we would recognize as pleasure in those acts. 

More broadly, sentient individuals are not the only valuable things that exist. There’s something inherently valuable about complex life on Earth, about the way energy flows from body to body, about the diversity of kinds of life, about the relationships between ourselves and other forms of life. This value goes beyond humans’ aesthetic appreciation of the more-than-human world. It is not the value of “naturalness,” because that is a value based on a false nature-human binary. This is a value of a deeply intertwined world that includes us. 

Singer says he welcomes the conversations that effective altruism has opened. In May, he published a thoroughly revised edition of his classic book; the updated Animal Liberation Now includes more material about wild animal suffering. “I do think we need to consider the balance between conservation values and welfare values of nonhuman animals,” he told me. But, “If people want to say that there’s intrinsic value in biodiversity, irrespective of the welfare of animals, there’s no way to compare that to suffering or well-being. That is something that can’t really be brought under the framework. There is no easy answer.”

Effective altruists are right to care about the future, in general. And EA acolytes have to be commended for investing in basic research on which charities help the most people today. An analysis by The Economist found that effective altruists gave a whopping $600 million to charity in 2021. Insofar as these donors would have spent that money on private jets or NFTs if they hadn’t gotten into EA, we can thank the movement for raising serious money for mostly good causes. 

But I would hope that those who are excited by effective altruism and the idea of optimizing doing good allow themselves to be open to the idea that there are many good things in our universe. Focusing on suffering alone ignores relationships between people, between species, between ourselves and place. It ignores the value of autonomy, the value of justice, the unfathomable complexity of an ecosystem. Even the absence of suffering is not always an unqualified good. As environmental philosopher Sandler says, “Unpleasant experiences can be part of meaningful experiences.” Mountain climbing, giving birth, struggling to help others—suffering can make us who we are. “The beauty and wonder of the world is that it isn’t reducible to a single parameter,” Sandler argues. In a world filled with incommensurable goods, we cannot always use an algorithm to determine which action is best.