After briefly fading into the background following the spectacular collapse of Sam Bankman-Fried’s FTX under the weight of its own accounting fraud, effective altruism, or E.A., is back in the news in a big way. This time, the philosophy’s adherents have found themselves in a sudden-onset blood feud with the Trump White House. Given the administration’s unceasing hostility to both effectiveness and altruism, the spat has a certain intuitiveness to it, but this conflict is not driven by Trump’s hatred of the global poor. Instead, it was the Department of Defense posting, seemingly unprompted, a pronouncement of “Americanism, not Effective Altruism” that ignited the conflict. Since then, there has been a major struggle over “AI safety” policy between E.A. adherents and the techno-fascist right—two groups that do, confusingly, overlap as well.
It’s hard to remember as the world burns around us, but there was a moment when E.A. (and its sister philosophies rationalism and longtermism) was ascendent. It was quickly amassing capital, adherents, and political power. In just a few short years, it went from a few quirky thinkers and writers trying to logic their way to solving morality to a coalition with substantial financial, media, and lobbying capacity.
While Bankman-Fried may have become effective altruism’s most notorious adherent, he was far from alone. Facebook co-founder Dustin Moskovitz and Skype founder Jaan Tallinn remain firmly in the E.A. camp, and Peter Thiel and Elon Musk have both publicly embraced and funded parts of the movement (although their overall giving doesn’t graft as cleanly onto the philosophy). LinkedIn founder Reid Hoffman has also described his approach to philanthropy similarly to E.A.
That ascendence was, in no small part, fueled by Silicon Valley titans’ patronage and proselytization. And a key reason for that is that the central premise of (especially longtermist) E.A. is just a hop, skip, and a jump from a moral permission structure that nakedly celebrates wealth and excuses opulence as steps on the path to a more just future. Bankman-Fried himself admitted as much in a candid exchange with a journalist after his business empire imploded; the altruism he championed was all about cloaking the scramble for wealth and power as something noble.
This is in no way true of everyone who takes interest in or subscribes to E.A., and it’s important to remember that those who are able to season their effective altruism with a dose of epistemic humility do often fund worthy projects. But the totalizing moral structure that shields the machinations of the powerful is a crucial dynamic to make sense of how we got to our current political moment, with tech oligarchs running amok, major financial realignment, and all-encompassing AI storylines. Because while the impulse to do altruism, and make it effective, can help galvanize works rooted in good intentions, our would-be overlords took that to start paving the proverbial road to hell.
At its core, effective altruism posits that, given scarce resources, the most ethical course of action is the one that maximizes the “returns” that can be generated from those scarce resources. Usually, the “returns” in question are measured in things like lives saved, disease incidences avoided, people educated, or other noble goals. But the returns can also be monetary, which is part of why E.A. is extremely useful to the ultrarich seeking to amass power.
The obvious assumption here is that there is an objective way of accounting for those returns and an objective way to “weight” doing good that benefits different people. One of the more common instincts among effective altruists is to consider all human lives equal. On its face, that sounds obviously correct! But it can quickly lead to counterintuitive conclusions.
Take for instance a charity that runs a program integrating ex-cons into communities’ workforces at an expense of $10,000 per week. An E.A.-style analysis could take this program and compare it to a famously cost-effective intervention like bed nets to fight malaria and mosquito-borne disease—which costs around $2 to $5 per net and can save a child’s life for every 600 nets distributed—and conclude that this facially worthy cause is actually bleeding resources away from more effective forms of altruistic intervention. And in that framework, career development to integrate people back into their communities is not only suboptimally moral but could even be immoral because that work takes on the opportunity cost of children’s lives. In the hypothetical above, that $10,000 a week could instead be spent on malaria nets and save as many as eight lives a week.
So, to a stickler E.A. analysis, community service missions can be actively harmful because it would be more moral to spend the money elsewhere. This sort of thinking is not universal among the movement’s adherents; it is entirely possible to recognize the limits of this economics-style optimization problem as a philosophy. But when that possibility is pushed out of mind, or actively attacked, such analysis can rapidly lead to some very unsavory moral positions.
Take, for instance, Sam Bankman-Fried, who subscribed to the E.A. movement’s more radical offshoot, longtermism. In SBF’s mind, spending a dollar to help save a life today would be a waste when that dollar could instead be invested in the development of new technologies that could theoretically save thousands or millions of lives in the distant future. With enough navel-gazing, SBF appears to have convinced himself that a utilitarian course of action could include purchasing the island-nation of Nauru and setting it up as a doomsday bunker from which a select few effective altruists could repopulate the planet after an apocalyptic scenario. Deranged, eugenic, and laughable, you might say, but if you give an insular group of E.A.s enough time and space to stew in their own juices, just about anything can seem rational.
Think back to the fundamental idea that we can objectively decide how much to “weight” the good done to different individuals. Any attempt to refine such an analysis runs into the problem that the “solution” to your optimization problem will be different depending on how weights are assigned. And you can get better results if you prioritize certain subpopulations. For example, if you have limited funding for cancer research, you will get wildly different resource allocations depending on if you weight each life the same versus if you prioritize maximizing years lived post-treatment. If everyone counts the same, there will be comparatively more funding flowing toward cancers that impact older adults. If we weight by years of life post-treatment, the focus will be much more on pediatric cancer.
Naturally, either of those is colorably better than the current market-based approach to research and development that sees more investment in erectile dysfunction treatment than, say, endometriosis. But if it feels a little gross to be pitting different communities of cancer patients against each other, that’s because it is. It is also necessitated by any system that purports to “solve” moral quandaries. And it feeds into the more unsavory tendencies among E.A.s.
There are broadly three buckets of destructive ways of thinking that effective altruists are predisposed toward, all downstream of this necessity to determine the relative moral worth of helping different populations with a fixed pool of resources: deference to notions of a chosen elite, justification of the centralization of power and wealth, and wanton disregard for human well-being and agency.
It’s here that I have to introduce you to Eliezer Yudkowsky, AI theorist, Peter Thiel associate, Harry Potter fanfic writer, and movement figurehead. Yudkowsky, often referred to simply as “Yud” by those in the know of AI research and E.A. thought, views himself in a messianic light, to the point that when one supporter (Scott Alexander, an E.A./rationalist thought leader in his own right) tweeted a comparison of him to Jesus Christ, he responded with a hypothetical conversation where he condescends to the Lamb of God. If you have developed and propagated a moral framework that is objectively more correct than the Son of God, though, it follows that you have the moral high ground, elevated well above that puny peak at Mount Sinai where Moses received his ethical framework.
Yudkowsky’s epic screed, Harry Potter and the Methods of Rationality, is a common entry point to the world of rationalist and E.A. thought, and has been deployed as a recruiting tool to bring new people into the movement. At more than 600,000 words across 122 chapters, the alternate Harry Potter story, doled out in pieces over five years, follows the titular character as he uses rationalist thinking to engineer new types of magic, wielding his upbringing by an Oxford philosopher to demonstrate the power of rational thought. Arthur C. Clarke famously posited that “any sufficiently advanced technology is indistinguishable from magic”; Yudkowsky’s take is basically that sufficiently rational thought will enable the creation of technology that functionally is magic.
This messianic sense of self is not confined to Yudkowsky. One of the most long-standing critiques of the broader E.A. space is that it insists on totalizing and moralizing across every part of life to insist that there is an objectively correct way to live. And that people who don’t follow that path are hence wrong.
Australian philosopher Peter Singer, one of the ethical pioneers of E.A. thought, gave a TED Talk in which he argued that a blind American should not be given a guide dog through philanthropy, because it uses funds that could have a much greater impact in treating blindness in developing countries. Similarly, philosopher Eric Friedman wrote that he would be “angry” at people donating to St. Jude, the pediatric cancer treatment hospital, when the same money could go further if it were diverted to a different pediatric cancer center (again in the developing world).
And while St. Jude does have its own issues with arguably bloated overhead and advertising costs, this mode of analysis would say that you shouldn’t fund any pediatric cancer wards in rich countries as long as there is a more cost-effective alternative elsewhere in the world. Remember earlier, when pitting cancer patients against each other felt like a reach? Some people actually do that. The end point of all of this is that unless you are the objectively most worthy of help, any assistance given to you is not just suboptimal but deeply immoral.
To a certain cadre of extremist, egomaniacal E.A.s, their vaunted ability to be the only ones capable of clearly seeing these trade-offs is reason enough for them to be a chosen elite. They need to lead because most people don’t have the stomach to come out and say that children in rich countries don’t deserve help with their cancer. Instead, only this group has the right combination of intelligence and objectivity to usher society into the future. And here is the juncture where some E.A.s go down the road of eugenics, believing that the intentional breeding of humans for desirable traits gets a bad rap. Incidentally, people who follow the tenets of effective altruism and rationalism just happen to have the traits for which they want to select. For some, their superiority is not just intellectual but biological. Which leads to dabbling in long-discredited race science and writing off entire populations.
This innate superiority feeds directly into the second set of dangerous thinking endemic to certain corners of E.A.: the justification of massive concentration of wealth and power. If some people are just better leaders because they are smarter and more objective, then it is preferable for those people to control resources and direct decision-making. This is, as University of Chicago philosopher Brian Leiter noted to the New York Post, part of why tech titans love E.A.; “There’s no talk about higher taxes, redistribution to a more progressive income tax, anti-trust investigations or more regulation of their businesses.” In fact, tax-based redistribution could actually be worse for society if it takes money away from the altruists who could steer it toward worthy causes more efficiently.
This exact problem is compounded by “longtermism.” Recall that the central premise of E.A. charitable calculus is that we can do the most good by being agnostic about how near or distant a person is; the entire idea is to weight only by who is most deserving and would make greatest use of the help. Longtermism extends this line of thinking into a fourth dimension: time.
If good deeds done in an indeterminate future time frame could be just—or more—ethical than anything done now, then there is a point at which amassing wealth now to spend later can be presented as more useful than just doing good things now. This is the justification of the “earn to give” mentality modeled by Sam Bankman-Fried. He had to get filthy rich, you see, because then he can take all that money and better the world on the back end.
Vanilla E.A. already teaches that we should set aside our own personal relationships and care about people regardless of where they are. Longtermism says that we also have to care equally about when people are. And there is practically no level of disempowerment or disinvestment that cannot be justified when you are weighting eight-odd billion lives against a theoretically infinite number of people who might exist in the future.
Here’s where the wanton negligence toward humanity can kick in. Once you break down the social bonds built by relationship and community first and foremost to the people closest to us, it’s much easier to reduce human life to an abstract agglomeration of discrete beings that exist outside of the world as we know it. Suddenly, the inclination to be willing to burn the Earth to fuel space colonization makes a lot more sense. You might doom most of the population in the near future, but there is infinitely more net utility once the entire galaxy’s resources are able to be tapped to fuel future lives. This is basically an argument Elon Musk, noted fan of E.A. gospel What We Owe the Future, has been making for years.
Moreover, once you divorce ethics and altruism from the social and communal relationships that they’re traditionally grounded in, you move away from a focus on channeling resources toward helping living breathing people. Instead, you can justify takes that, actually, insects are more important to help because they have some form of consciousness and vastly outnumber us. The least extreme iteration of this is animal welfare and veganism activism that comes out of E.A. spaces, pushing for greater acceptance of animals as beings worthy of altruistic treatment as well. The most extreme is intertwined with frontier AI development’s quest to create new artificial beings. That version is the crucial evolution of Elon Musk’s older take about space colonization. Our species is just a biological “bootloader” for a one-day interstellar artificial superintelligence.
It is easy to see how noble intent can pave the way to these outlandish ethical positions. We probably should care more about the way we treat animals. But that doesn’t mean that choosing to fund a program to help children orphaned by natural disasters is unethical because the money would be more effectively spent preserving ant colonies. The reason why we intuitively help the people closest to us is not just an ethical lapse, it’s because we have different moral responsibilities depending on how we relate to people.
Maybe AI will become conscious and colonize distant stars. But it is the height of ego and folly to predict that it necessarily will happen and that, therefore, we need to prioritize our potential robotic progeny over the people around us in the here and now.
Not all, or even most, people who dabble in E.A. thought fall into these moral traps, but it’s common enough that there are observable patterns of behavior. More importantly, the people who do fall into these tendencies seem determined to engineer the end of the world as we know it.
Effective altruism prides itself on being nonconformist and willing to break with public opinion. Many of its adherents seemingly relish being outside the mainstream. It legitimates their view that they are apart from “normies” who are shepherded towards rote philosophical positions.
That’s how you wind up with Yudkowsky being spoken of in the same breath as someone in a “sex cult” and an important thought leader in AI alignment.
Some expat critics of E.A. who left the movement have a different acronym: TESCREAL, or transhumanism, extropianism, singularitarianism, cosmism, rationalism, effective altruism, and longtermism. Basically, the argument is that Silicon Valley has been taken in by a compulsion to strive for biologically evolving beyond the current nature of the human race (transhumanism plus extropianism), achieving a technological innovation that renders previous human ways of life obsolete and results in integrating humanity with machines (the singularity), the achievement of immortality through consciousness transfers (cosmism), improving human decision-making by centering objectivity (rationalism), and that we have an ethical responsibility to deploy resources in the way most likely to do the most good (E.A.), not just for the people who exist but for those who could exist in the future, as well (longtermism).
The TESCREAL argument lays out the case that much of the tech world has become dedicated to obsessive focus on the potential of human extinction and the necessity of achieving a science fiction–esque future.
The business plan here will look familiar to anyone who has tracked the promises figures like Sam Altman have continually made. Promise that technology is being developed that will fundamentally change life for the better, cure cancer, and usher in a post-scarcity society. In the meantime, build up huge reserves of money and political power. And then justify the costs being borne in the deployment of technology as a necessary evil to usher in a brighter future.
The entire E.A. worldview has been a microcosm of Silicon Valley’s political framework, which is why it was so expedient for figures like Elon Musk and Dustin Moskowitz to promote it. Bundled into a single philosophy, you have the groundwork to validate elites’ self-importance, justify the amassing of power and wealth, and excuse nearly any antisocial behavior through commitments to the future.
To the powerful people who invested in spreading the gospel of effective altruist thought, these quirks are features, not bugs. A philosophy that confirms that those wielding power and resources are actually heroes fighting a cosmic battle for the future allows for oligarchs to justify the undermining of democratic institutions, the hoarding of wealth, and callous treatment of humanity.
One last time, it’s worth bearing in mind that this mentality is not universalizable across everyone who finds E.A. to be valuable. Initiatives like Giving What We Can are a way for people to genuinely put their money where their mouths are. For plenty of people, the weird, deeply anti-human takes are a bridge farther than they’ll ever take the philosophy. But for Silicon Valley titans, these predilections are a crucial element of why they find this worldview to be useful—to their own ends, at least.
Once you have a framework that validates your role as the builders of a new, better world, it justifies serving as pallbearers of what we currently know as human civilization. And once you accept that your movement is an elite that must be nurtured into stewards of the future and that this destiny is more important than anything else in the present, it becomes totally reasonable to do things like buy castles for your thinkers’ enrichment.
Now we’re at a juncture where the people who helped effective altruism make it big, people like Musk, are in conflict with the elements of the movement who have kept out of the worst rabbit holes. Some see the singularity as inevitable, and creating artificial consciousness as an ethical imperative. Some see real people as being worth helping. When your philosophy lionizes elites and excuses their greed, of course they’ll help fund it. But now the chickens are coming home to roost, and it’s becoming more apparent that the indulgence of power is the reason those people were on board in the first place. Their altruism is devoted purely to laying claim to a future that merely aligns with their own desires.
