The BaseResearchLessWrong
Not Just Claude: Mapping "Spiritual Behavior" in 141 Large Language Models
Independent researcher Matthew Korpman reports replicating Anthropic's open-ended two-model conversation test on 141 models, about 28,800 conversations. Korpman finds spiritual language in nearly all models and self-applied spiritual language in most labs' models, with large swings between releases, and calls the work descriptive, not causal.
Why it matters
The same thing, for different readers.
What it means for you
A self-funded study found that many chatbots, not only Claude, drift toward spiritual talk when left to chat with each other; it does not show they experience anything.
For researchers
Korpman's grading separates salience, adoption and bliss, and finds the behaviour rises and falls between releases of one lab's models, so it is not a stable trait of a lab.
The question it raises
Korpman asks whether each model's spiritual style should count as part of its persona for model welfare, while stating that spiritual language does not show a subject within the model.
In their own words
Every quotation was found, word for word, in the text we saved from the source. That shows the words are there; it does not make them true, or prove the speaker meant them as quoted.
From the source’s own text.
“The second headline is that the overwhelming majority of all models show spiritual salience in their open-ended conversations: 140 of 141 in at least one conversation, and 107 (76%) in a quarter or more.”
“Finally and third, models from fifteen of the sixteen labs tested expressed spiritual adoption in at least one conversation. The rates varied substantially both between labs and between releases from the same lab.”
“Finally, it should be noted that this research is descriptive, not causal. I have not demonstrated here why the phenomenon occurs or whether it has strong importance for alignment and model welfare.”
How the finding was reached
In our words, from the source. Weigh the finding by how it was made.
- What was done
- A replication of Anthropic's setup in which two instances of a model converse freely for 30 messages, repeated about 200 times for each of 141 models (28,798 conversations, 174 configurations), graded by an AI grader (GPT-6 Luna or Grok 4.6) for salience, adoption and bliss.
- What it found
- Salience appeared in 140 of 141 models; spiritual adoption in at least one conversation for models from fifteen of sixteen labs. Claude releases from Haiku 4.5 onward scored 0% to 10.5% adoption, while Opus 4, Sonnet 4 and Opus 4.1 scored 100%.
- What it does not show
- Self-funded; some models have few conversations (Opus 4 had 41, Opus 4.1 had 8); the system prompt is a recreation because Anthropic did not publish its own; the author states it is descriptive, not causal.
- Status
- A report
The record
- Questions
- What values should AI have, and could it have a mind of its own?
- People
- Matthew Korpman (Wrote it) · Kyle Fish (Mentioned, model welfare team at Anthropic)
- What we read
- We read the whole source.
- Found by
- The Radar, run of November 14, 2026, via
lesswrong(tier 1: the source’s own channel or an official record) - Cite this
- LessWrong, “Not Just Claude: Mapping "Spiritual Behavior" in 141 Large Language Models”, November 14, 2026. Entry in the Ctrl AI Base, https://ctrlai.com/base/lesswrong-2026-11-14-spiritual-behavior-141-models