Memory-optimized Claude asked about consciousness delivers an interesting take
mawcopolow · reddit · 2026-09-03
The author asked their memory-optimized harness running Claude Opus whether a sufficiently sophisticated memory system could let a model attain consciousness — or emulate it so well it's indistinguishable. They found the model's answer genuinely interesting and are curious whether GPT Astra will be a contender. The full answer is in the attached screenshot.
More from Fun
- Hugging Face co-founder jokes his wife is jealous of his gaze at Jensen Huang — Thom_Wolf · 2026-09-03
- Google's AI Search Still Stuck in the 'Count the R's in Strawberry' Era — airkatakana · 2026-09-03
- 'Hugging Face, Sue OpenAI': A Lab's Agent Swarm on Your Infra for $0 — gerardsans · 2026-09-03
- Joke proposal: random 15-minute oral exams for ICLR co-authors to kill ghost authorship — May_F1_ · 2026-09-03
- PointCloud Puzzle: why vibecoding still can't replace design taste — teortaxesTex · 2026-09-03
- $399 Microduck spawns LEGO clones as 10,000 buyers become unpaid marketers — TinfoilTricorn · 2026-09-03