Why hand-calculating LLM activations doesn't refute AI consciousness
CatAstro_Piyush · x · 2026-09-01
Argues that manually calculating LLM activations on paper is not a valid refutation of LLM (or any system) consciousness, as it breaks a highly relevant and fundamental property of the computations underlying conscious experience.
More from AGI Musings
- Alephic Cofounder: Alignment Is the Hardest Problem in the Agent Era — every · 2026-09-01
- Opinion: Data center backlash may not help AI safety — AndyMasley · 2026-09-01
- l4rz: Value stability under recursive self-improvement is unproven and likely false — l4rz · 2026-09-01
- Preferring 'agent collective' to describe organizational forms — PeterBowdenLive · 2026-09-01
- Creativity and Obsession Are the True Personal MOATs in the AI Age — jackedAJ · 2026-09-01
- Cognitive overload anxiety over AI-generated content everywhere — michalmalewicz · 2026-09-01