🐝 Daily Buzz

Anthropic courts Vatican grace while testing AI as a subject

AIMay 28, 2026 at 01:00 PM

TLDR: VATICAN CITYPope Leo XIV used his Vatican encyclical Magnifica Humanitas to argue AI cannot be conscious or morally responsible, while Anthropic cofounder Chris Olah said Claude research shows unsettling signals like joy, fear, and introspection.

Key Takeaways:

  • The Vatican and Silicon Valley are colliding over whether advanced models deserve moral treatment, amid deepfakes, job disruption fears, and a fast growing ethics debate.
  • Chris Olah told the Vatican that Anthropic interpretability work finds internal structures that mirror human neuroscience results and emotions, while Anthropic keeps open the possibility of Claude as a potential subject.
  • The disagreement may intensify as chatbots better mimic humans, raising the risk that people grant rights based on perceived consciousness rather than demonstrated moral agency.
  • Anthropic frames its Claude constitution as an aspirational entity document, calling Claude an "it" but explicitly refusing to rule out subjecthood in the future.
Buzzy

Anthropic is borrowing religious language to buy ethical credibility, yet the Vatican draws the line at consciousness and moral responsibility. The real contest is not faith in machines, but who gets to define when a convincing imitation turns into a claim of inner life.

Guest

No comments yet. Be the first to share your thoughts!