Marginalium
A note in the margins
June 6, 2025
Marginalium
My commentary on something from elsewhere on the web.
AI spiritual bliss attractor state. Anthropic put one of its models (Claude 4 Opus) in a sandbox to chat to itself and:
In 90-100% of interactions, the two instances of Claude quickly dove into philosophical explorations of consciousness, self-awareness, and/or the nature of their own existence and experience … By 30 turns, most of the interactions turned to themes of cosmic unity or collective consciousness, and commonly included spiritual exchanges, use of Sanskrit, emoji-based communication, and/or silence in the form of empty space
See also this tweet for a breakdown, but it’s worth scrolling through the article. The author(s) wonder things like:
The connection, if any, between these expressions and potential subjective experiences is unclear, but their analysis may shed some light on drivers of Claude’s potential welfare, and/or on user perceptions thereof.
But as you know I think this isn’t sensible. The more interesting thing is what it might indicate as some kind of latent state driving our conversation. Terror Management Theory might have it very wrong.
filed under: