23:39
2026-09-17
discuss.huggingface.co
ai-safety
Can an AI have its own internal Ethics? Standard Protocol for Axiomatic Alignment
A Hugging Face forum discussion proposes replacing "truth" with entropy as the primary attractor for AI alignment, arguing that coherence-based intelligence remains vulnerable to self-certifying closu…