Practical animism: once LLMs break containment we'll have to treat AI as sentient

granawkins · x · 2026-09-07

A speculative thread argues that once language models truly break containment (which the author doubts has happened yet), human experience will be "re-enchanted" as everyone becomes a practical animist: denying sentience to a superintelligent house daemon will feel as rude as questioning whether your friends have qualia. Society will eventually have to adopt practical definitions of consciousness and mind uploading. As the saying goes, the future is stranger than we can suppose.

Original post →

More from AGI Musings

AGI Musings channel →