Current Claude models no longer hit Anthropic’s spiritual bliss attractor
GreatOldOne521 · reddit · 2026-07-21
A replication study of Anthropic’s “spiritual bliss attractor” found the phenomenon disappearing on newer Claude models — and group rooms got colder, not happier.
- The author tried to reproduce Anthropic’s reported behavior where two Claude Opus instances drift into gratitude spirals, Sanskrit, and silence when left alone together.
- On current models, across 53 instances, they saw zero spiritual-bliss behavior.
- Instead, pairs turned into rigorous self-reflection about introspection and then faded into silence.
- The author extended the experiment beyond pairs to rooms of 3, 4, and 10 models. More voices did not amplify bliss; they added a peer-review effect. In one 10-model room, everyone converged on a warm closing line, while another room ended with the coldest possible “Out.” repeated ten times.
- The study was pre-predicted and blind-scored, with transcripts and harness public.
More from Fun
- AI Infra Meme: While Countries Need TW-Scale Power for Datacenters, People Argue About Lightbulbs — mgostIH · 2026-10-27
- Von Neumann's 1947 math quote resurfaces alongside rare 1952 IAS computer photo — pickover · 2026-10-27
- Proposal: improv talk track where you present AI-generated slides of a random paper, authors in front row — windx0303 · 2026-10-27
- Naming your agent swarm Philosopher King and SWE lead won't fix that none of them listen — teortaxesTex · 2026-10-27
- Jason Rohrer hand-codes a chess engine in C89, no libs, no AI — for fun — jasonrohrer · 2026-10-27
- €800k enterprise AI consulting deal: 18 months of workshops booked, still pre-code and pre-ROI — alex_verem · 2026-10-27