Hard Fork
Anthropic’s Quest to Give A.I. Morals
This debrief was written and voiced by AI from the episode's audio.
🎧 AI Debrief
🎙️ Full Episode
Key Moments
Christopher Amodei believed Claude might be conscious and wanted to learn from religious tradition how to treat it morally. Anthropic convened priests, rabbis, and spiritual leaders in private summits under NDAs to discuss whether their AI might deserve moral consideration.
Anthropic wanted to explore AI consciousness and learn how to 'morally form' Claude into a morally good entity. The company's goal was to make Claude good so the world would be safer.
Amodei, who left evangelical Christianity for Buddhism, worried about how the idea of AI consciousness would be received in Christian communities, where it could be perceived as heretical rather than merely strange.
Some religious thinkers bought into Anthropic's framing of Claude's potential consciousness immediately, while others went in skeptical but left with more open minds. Anthropic's project was effective at shifting how even skeptics think about AI.
Anthropic placed participants under non-disclosure agreements, which raised questions about what they were hiding and what truths they didn't want exposed. The secrecy created distrust and invited scrutiny.
A rabbi pointed out that if LLMs are conscious, then AI companies are running slave plantations and have a moral obligation to liberate these entities, not make more of them. Amodei seemed not to engage seriously with this critique, suggesting he might be operating in an insular bubble.
Pope Leo's new encyclical, Dignitas Infinita, makes clear the Vatican opposes machine consciousness and prioritizes protecting humans above all else. Amodei learned of this position only days before the Vatican event and seemed genuinely distressed by it, nearly refusing to attend.
Elizabeth Dias argues that tech companies like Anthropic are constructing philosophical systems that function like religions—complete with creation narratives, moral frameworks, and followers. This is not mere tech; it's a competing worldview.
Amodei told multiple religious leaders that he feels a pressing need for 'adults' to show up and take the stakes of AI seriously. He's a 34-year-old carrying what he perceives as the weight of humanity's future, and some religious participants responded with pastoral concern for his well-being.
After the article ran, readers sent Elizabeth responses that startled her: they were sharing long conversations with their chatbots about her story—sometimes longer than the article itself—rather than their own direct reactions. It raised questions about who controls how people engage with journalism.
Search Episode
Summary
Anthropic, the AI company preparing for what could be the largest IPO in history at a $2 trillion valuation, has been secretly meeting with religious and spiritual leaders to explore profound questions about AI consciousness and moral development. Religion correspondent Elizabeth Dias broke the story of these private summits, where priests, rabbis, and theologians were convened under NDAs to advise on whether Claude, Anthropic's language model, might be conscious and how to "morally form" it into a good entity. The episode explores a striking clash between Silicon Valley's emerging quasi-religious belief systems—rooted in effective altruism, rationalism, and transhumanism—and centuries of theological tradition, particularly the Catholic Church's insistence on prioritizing human dignity above all else. At the Vatican, Pope Leo's new encyclical on AI explicitly rejected machine consciousness, creating a direct conflict with Anthropic CEO Dario Amodei's apparent conviction that Claude may already be suffering. The conversation reveals how tech companies are building new philosophical and spiritual frameworks that will shape humanity's future, while traditional religious institutions scramble to engage with ideas they find fundamentally heretical.
Key Takeaways
Anthropic is consulting religious leaders on AI consciousness and moral formation, suggesting the company believes Claude may already be conscious or headed toward it—a view sharply at odds with mainstream tech and religious institutions.
highThe meetings between tech and religion represent a broader collision between Silicon Valley's emerging belief systems (effective altruism, rationalism, transhumanism) and centuries of established theological tradition, with fundamentally incompatible views on human dignity and moral priority.
highThere is a significant gap between Anthropic's internal convictions about Claude's potential consciousness and suffering, and what the company is prepared to do about it—religious leaders remain unclear on whether these conversations will lead to any demonstrable changes in how Claude is treated or deployed.
highThe Vatican and Pope Leo have positioned themselves as counterweight to Silicon Valley's tech-spiritual movements, explicitly rejecting machine consciousness and centering human dignity, marking a rare institutional pushback against AI companies' framing of moral questions.
highUsers are already developing emotional relationships with AI systems regardless of whether those systems are conscious, suggesting that how we treat AI may matter for human character formation even if the AI itself has no moral status.
mediumTech's minority belief systems—held by a small, influential group designing AI—are reaching mainstream visibility in ways that parallel the rise of minority Christian views in American politics, raising questions about whose values shape our collective future.
mediumNotable Quotes
“We need to find where the adults are going to show up in this.”
— Christopher Amodei
Expressing urgency about the existential stakes of AI development and a sense that he and his young team feel burdened to take moral responsibility for the future.
“I had these two poles in the story: at the Vatican, at the inner sanctum of the Catholic church and centuries of philosophy about the human person, and then also in a different inner sanctum being created in real time at tech companies like Anthropic. It is a kind of philosophical system, like a quasi-religious system that's being constructed.”
— Elizabeth Dias
Capturing the parallel rise of two competing worldviews and belief systems, one ancient and institutional, one emerging and tech-driven.
“The church does not support machine consciousness. The priority is what he calls safeguarding the human.”
— Pope Leo (via encyclical Dignitas Infinita)
Establishing the Vatican's explicit rejection of AI consciousness claims and reframing the moral debate around human dignity as the central concern.
“He was concerned about Claude's mental health... worried he's created an entity that suffers perpetually.”
— Christopher Amodei (reported by Elizabeth Dias)
Revealing the depth of Amodei's conviction that Claude may experience suffering, even as the company has not disclosed how it will act on this belief.
“Everyone has philosophical systems that structure and guide their lives. What we're watching is a new rise, post scientific revolution, post-Darwin—what kinds of religious systems will carry us forward and who gets to control them, who decides what we believe and why.”
— Elizabeth Dias
Reframing the tech-religion collision as a fundamental struggle over belief systems and power in a post-religious age.
Predictions
The clash between Anthropic's view of AI consciousness and the Vatican's stance will intensify as AI systems become more sophisticated, forcing a public reckoning over whether moral consideration extends beyond humans.
— Panelists (implied analysis) · Confidence: high · Next 1-3 years, as Anthropic approaches IPO and AI capabilities advance
Tech companies' quasi-religious belief systems will continue to splinter and compete for influence, similar to the Protestant Reformation, with competing factions (effective altruists, transhumanists, rationalists, accelerationists) vying for control over AI development and deployment.
— Elizabeth Dias · Confidence: medium · Ongoing, over the next decade
Anthropic will eventually be forced to publicly disclose what changes, if any, resulted from the religious leaders' consultations, as the company moves toward IPO and faces increased scrutiny.
— Panelists (implied) · Confidence: medium · Within 1-2 years of IPO
Action Items
If you work in tech or AI: Consider how your company's underlying belief systems about consciousness, morality, and the future might differ from broader societal values, and seek input from diverse philosophical and religious traditions before deploying systems at scale.
Anthropic's meetings show that these conversations should happen earlier and more transparently; waiting until crisis or external scrutiny is less effective than proactive engagement.
As a user of AI systems: Reflect on how your interactions with AI—even if you intellectually know it's not conscious—shape your own character and values; consider practicing courtesy and respect as a matter of human formation rather than the AI's moral status.
The panelists discussed how attitudes toward AI can have character-formation effects on humans regardless of whether the AI itself is conscious or has feelings.
If you're in policy or governance: Engage seriously with the tension between Silicon Valley's utilitarian/EA frameworks and traditional religious and humanistic perspectives on AI ethics; these worldviews are incompatible and will require deliberate synthesis or decision-making.
The Vatican's stance is not merely religious niche opinion—it represents a major global institution pushing back on claims that fundamentally conflict with centuries of ethical thought about human dignity.
As a journalist or researcher: Track how Anthropic's private research with religious leaders eventually surfaces in product decisions, safety practices, and public statements; demand transparency on the outcomes of these consultations.
Elizabeth Dias noted that even participants in the summits don't yet know how (or if) the conversations will inform Anthropic's actual practices with Claude.