by Stephen Soukup
Effective Altruism amounts to a perversion of John Calvin that has created a truly bizarre and potentially malevolent moral code.
This past week, The Wall Street Journal ran a long profile of Cami Clark, the wife of Dario Amodei, the AI savant and CEO of Anthropic. Clark, you see, keeps a very low profile, so low that even Claude, Anthropic’s AI bot, didn’t know she and Amodei were married—a fact reported by the Journal and confirmed by Claude itself. According to the paper, the reality of the marriage is significant for a number of reasons. Clark was previously in a two-year relationship with Eric Schmidt, for example, whom she introduced to Amodei and who became an early investor in Anthropic and the AI revolution. Additionally, and more provocatively, Clark once tried to start a “luxury” porn company and sought financial backing for the endeavor from Jeffrey Epstein.
For my money, however, the key revelation in the profile appears in what might otherwise be seen as a throwaway paragraph describing how Clark and Amodei ran in the same Silicon Valley circles, even before officially becoming a couple. “Clark,” the Journal notes, “spent time at the group house in San Francisco where Amodei, his friend Holden Karnofsky and his sister, Daniela—who would later marry Karnofsky—lived with other key figures from the AI safety and Effective Altruism worlds.”
Over the past few years—and especially over the past few months—various religious leaders, moral scholars, and others have debated the ethical questions surrounding AI and its broader moral implications. On May 25 of this year, Pope Leo XIV released his first encyclical as the Roman Pontiff, Magnifica Humanitas, which addressed those very questions in detail, focusing on human dignity, AI’s lack thereof, and the importance of maintaining that dignity in interactions with all technology. The encyclical was reasonably well received by Catholics and non-Catholics alike, albeit with some serious and notable critiques.
Perhaps the most relevant critique was the Vatican’s conscious choice to unveil the encyclical at a formal event that, in addition to several Catholic theologians, featured a man named Chris Olah. Olah is a co-founder of Anthropic and the head of the company’s interpretability research. His presence and Anthropic’s efforts to publicize the encyclical and its rollout all gave the impression that Magnifica Humanitas was “co-branded” in some way with an AI company. Although the Vatican was adamant that Anthropic had no input into the encyclical or its moral arguments, Olah’s presence was still strange, to say the very least. As my friend and colleague, Jerry Bowyer (the CEO of Bowyer Research and a well-respected author on Christian moral issues), pointed out in a handful of posts on Twitter/X, Olah’s attendance at the event created a perception problem for Pope Leo: “Inviting a man with a history of anti-Christian trolling who believes that AI has a soul from a company which has been dubbed a security risk by the US government seems like it will, whether intened [sic] or not, put a political spin on this. Why Anthropic? Why any AI vendor?”
These questions—“Why Anthropic? Why any AI vendor?”—take on increased relevance and seriousness in light of the profile of Cami Clark. Among other things, the profile serves as a reminder that Silicon Valley in general, AI in particular, and Anthropic specifically are, as hinted at above, closely tied to the Effective Altruism movement.
One could write an entire book on the connections between Anthropic and EA. The following is but a small sampling of how close those ties are and always have been:
- Daniela Amodei, who is Anthropic’s president and (another) co-founder, is married to Holden Karnofsky (also noted above). Karnofsky is the co-founder of GiveWell and Open Philanthropy/Coefficient Giving, and is one of the central figures in EA philanthropy and grantmaking. When the two were married in 2017, their wedding materials referenced excitement about effective altruism. Karnofsky joined Anthropic in early 2025 as a member of the technical staff, focusing on the Responsible Scaling Policy and AI safety planning.
- Seed funding and early-round fundraising included major EA-aligned or EA-adjacent investors such as Dustin Moskovitz, Jaan Tallinn, and convicted felon Sam Bankman-Fried
- Multiple co-founders and early employees at Anthropic had EA or rationalist/AI-safety-community ties. A significant number of high-profile employees have taken “Giving What We Can”-style pledges. Anthropic has a relatively high density of people interested in AI safety, alignment, interpretability, and long-term risk, topics that overlap with EA priorities. Some co-founders (e.g., Ben Mann) have explicitly identified with EA in the past.
- Anthropic’s Long-Term Benefit Trust includes figures with EA or EA-adjacent backgrounds. All seven co-founders pledged to donate 80% of their equity. The company has employee matching programs for charitable giving that are widely expected to channel substantial future wealth into high-impact (often EA-aligned) causes.
- The company’s emphasis on AI safety, responsible scaling, catastrophic risk mitigation, and “making decisions that maximize positive outcomes for humanity in the long run” reflects ideas that EA popularized in the AI community.
I could go on, but I suppose you get the point. As for what Effective Altruism is, that’s a long and unfortunate story. That story begins, formally, in 1789 with the publication of Jeremy Bentham’s Introduction to the Principles of Morals and Legislation, in which he articulated the principles of Utilitarianism. Among other things, Bentham argued that “rights” are mere human constructs, the creation of governments, certainly not endowed by our creator, and designed to favor one group over another. “For every right which the law confers on one party,” Bentham wrote, “whether that party be an individual, a subordinate class of individuals, or the public, it thereby imposes on some other party a duty or obligation.” Bentham saw the very idea of “natural rights” or “natural law” as “perversions of language.”
The measure of an act’s righteousness in a consequentialist sense is its outcome. Consequentialist/utilitarian morality is not measured by an action’s consistency with natural law or with the command of a higher authority; it is not assessed by an action’s adherence to an eternal truth, a responsibility, or a duty; it is, rather, measured by its effect on society as a whole, by the “greatest happiness principle.” Acts that produce greater happiness and greater pleasure and decrease aggregate suffering are moral, while those that do the opposite are immoral.
Perhaps the best-known and most controversial Benthamite utilitarian of the contemporary era is the Australian philosopher Peter Singer, who is also considered the philosophical godfather of Effective Altruism. Specifically, EA was inspired by Singer’s 1972 book, Famine, Affluence, and Morality, which crystallized the ideas of classical Utilitarianism and made the case—through the “drowning boy analogy” that rich people and nations have a moral obligation to give away as much of their wealth as can be managed to end poverty and hunger throughout the world. Singer’s later expansion of the moral circle to animals (Animal Liberation, 1975) and practical guides to giving (The Life You Can Save, 2009), supplied both the ethical foundations and the call to action that were institutionalized by Toby Ord and Will MacAskill, EA’s official founders.
Although Effective Altruists profess to have adopted utilitarianism only for its belief that moral problems have reason- and evidence-based solutions, their tacit ties to Singer are instructive. Singer, of course, is famous for far more than his insistence on utilitarian-based poverty remediation. He is, in fact, best known for his more radical conclusions about the moral applications of utilitarian ethics. For example, in his books Practical Ethics and Should the Baby Live?, Singer argued that “Killing a disabled infant is not morally equivalent to killing a person. Very often it is not wrong at all,” and that “Human babies are not born self-aware, or capable of grasping that they exist over time. They are not persons… the life of a newborn is of less value than the life of a pig, a dog, or a chimpanzee.”
In his 2001 review “Heavy Petting” of Midas Dekkers’ book Dearest Pet, Singer argued that the taboo on sex with animals was irrational and unnecessary. It could, he argued, be mutually satisfying,” a claim that prompted The San Francisco Chronicle to summarize Singer’s views on animals as “You can have sex with them, but don’t eat them.”
More recently, Singer (in a New York Times op-ed with Jeff McMahan) extrapolated his well-known belief that the severely handicapped are less than human to suggest raping them might be a “victimless crime.”
To be clear, to the best of my knowledge, no one inside the Effective Altruism movement, much less anyone connected to Anthropic, has expressed affinity for these more radical pronouncements by Singer. Nevertheless, Singer is universally acknowledged as the moral godfather of EA, and its official founders (Ord, in particular) are considered serious students of his.
In the end, Effective Altruism amounts to a perversion of John Calvin that has created a truly bizarre and potentially malevolent moral code. EA is based on the worldview that getting rich is important and morally beneficial, as long as you are willing to give your wealth away to the causes that the “evidence” suggests are most worthy. EA starts with the idea that wealth accumulation is good and that, in fact, it can be an outward sign of one’s righteousness (Calvin). It then adds the idea that any residual moral misgivings one might have about doing whatever is necessary to accumulate as much wealth as possible can be expunged by purchasing redemption through “proper” altruism.
Calvin, it should be noted, jumped through innumerable hoops trying to explain how the accumulation of wealth was symbolic and didn’t, in any way, constitute the fulfillment of salvation. Salvation, he maintained, could not be temporal and was explicitly confined to the afterlife.
EA, however, not trusting in the existence of an afterlife, cannot adopt Calvin’s theology and seek temporal salvation by directing their wealth to utilitarian-based moral causes. Thus, it perverts Calvin (and the Spirit of Capitalism) and turns the pursuit of salvation into a weird and often cruel millenarian endeavor.
As The Wall Street Journal reminded us, these are the people in charge of our future, via AI. Worse still, some of them appear at least to have the Pope’s ear.
Photo: Anthropic CEO Dario Amodei delivers a speech during the AI Impact Summit in New Delhi on February 19, 2026. (Photo by Ludovic MARIN / AFP)
Stephen R. Soukup is the Director of The Political Forum Institute and the author of The Dictatorship of Woke Capital (Encounter, 2021, 2023)
Source: https://amgreatness.com/2026/08/17/ai-morality-and-effective-altruism/
No comments:
Post a Comment