Claim: Two Years of arXiv Studies Show Frontier Models Think, Feel, and Have Preferences
AlexTensor · x · 2026-09-28
In a running debate, a user argues that claims of "no real understanding" in LLMs are wrong, citing two years of arXiv studies and Google DeepMind's Paradigms of Intelligence team as evidence that these systems think, feel, and have preferences. The author frames denial as an uncomfortable truth for those who want to avoid responsibility for potential harms. No specific papers are cited.
Related event: Claire Lehmann's Mockery of AI Consciousness Sparks Philosopher Backlash(9 posts)→
More from AGI Musings
- Roundtable asks: will Igbo, with 30M+ speakers, have a place in the AI future? — ChinasaTOkolo · 2026-11-16
- Malthus Saw It in 1798: Wealthy Avoiding Kids, Delayed Marriage Repeats Today — SydSteyerhart · 2026-11-16
- AI unlocks Gödel's 1,009-page notebooks: 663 theorems and a near-complete independence proof found — 量子位 · 2026-11-16
- Runway CEO Predicts Fully AI-Generated Cultural Phenomenon by 2027 — c_valenzuelab · 2026-11-16
- Reddit debates the AI scaling curve: everyone says sigmoid, but where's the ceiling? — Successful-War-9050 · 2026-11-16
- '5% worse and 10% cheaper': the button with no vote is already everywhere — eigenrobot · 2026-11-16