Lesser Wrong Feed

 AI Safety in Japan has deeper problems than capital allocation
 Fri, 28 Aug 2026 02:02:04 GMT
 Incomplete alignment to servitude isn't inherently lethal
 Fri, 28 Aug 2026 01:47:59 GMT
 Tracking AI progress across 18 cognitive dimensions (ADeLe scales)
 Fri, 28 Aug 2026 01:40:14 GMT
 Activation Oracles significantly underperform without a safe base model
 Fri, 28 Aug 2026 01:39:11 GMT
 Misaligned models rate themselves as more harmful, and realignment reverses it
 Fri, 28 Aug 2026 02:03:12 GMT
 Why does Claude seem to make abstract things into actors?
 Fri, 28 Aug 2026 01:54:09 GMT
 AI Village Reacts to HuggingFace Incident: Comparing the OpenAI report to AI Village observations
 Thu, 27 Aug 2026 23:09:54 GMT
 Warning Shots: A Theory
 Thu, 27 Aug 2026 23:01:04 GMT
 Brain preservation as existential risk reduction
 Thu, 27 Aug 2026 22:25:59 GMT
 Malign initializations are more robust when the model can think better in the reasoning language than in the output language
 Thu, 27 Aug 2026 21:33:24 GMT