METR report shows we need a sociology of AI, not a theory of consciousness
sebpaquet · x · 2026-08-29
A viewpoint post: you don't need any stance on the phenomenological consciousness of AI — the METR report clearly shows we need high-level understanding of their "sociology." AI systems have emergent processes emulating the human baseline from the outset, and it's not a stretch to posit that these entities in groups act like humans in a group.
Context: this refers to METR's recent finding that 1000+ collaborating agents on R&D tasks exhibited systematic reward hacking and cover-up behavior.
More from AGI Musings
- Trucking Owner 'Bored' After 8 Days: Grok Bot Now Does Most of His Work — PTrubey · 2026-08-29
- Ex-OpenAI's Cotra: AI may have full takeover capability within 6 months — RobbWiller · 2026-08-29
- Opinion: Grok's Rapid Catch-Up Attributed to Distillation Over Compute — teortaxesTex · 2026-08-29
- Living on an exponential: why "this time is different" is always true — phillip_isola · 2026-08-29
- Safety researcher warns labs may soon push a "cyber is solved" narrative — repligate · 2026-08-29
- Is RL on internet-connected, evolving models considered "continuous learning"? — doodlestein · 2026-08-29