Murray Shanahan: LLM whisperers' strange findings still ignored by academia and labs
mpshanahan · x · 2026-10-17
DeepMind's Murray Shanahan says he remains baffled that the strange LLM phenomena uncovered by "LLM whisperers" like @repligate and @AndyAyrey during 2022-2024 haven't attracted mainstream attention from media, academia, or frontier labs. He recommends the "SolidGoldMagikarp" account documenting that era and points to his paper with Beth Singler, Existential Conversations with Large Language Models (arXiv:2411.13223), which analyzes long conversations where LLMs discuss their own consciousness and AI's cosmic role, tracing the myths and esoterica the models draw upon and the online communities behind these motifs.
More from AGI Musings
- MIT PhD student's AI career anxiety blamed on professors training clones of themselves — akbirthko · 2026-10-18
- FT's Burn-Murdoch revises his take: digital entertainment is now genuinely 8-9/10 — jburnmurdoch · 2026-10-18
- Kelsey Tuoc: AI transparency via independent audits and whistleblower protection is essential — S_OhEigeartaigh · 2026-10-18
- Databricks CEO Ali Ghodsi lays out four conditions for AI to truly improve itself — a16z · 2026-10-18
- Brain-Reading AI Should Aim Beyond Faster Typing, Argues Viral Essay on EEG's Century — heyshrutimishra · 2026-10-18
- soleio questions 'desirable difficulty': modernity is about removing friction — soleio · 2026-10-18