gandamu_ml: blame narratives after AI failures will miss the point, closed or open

gandamu_ml · x · 2026-09-13

Responding to tszzl's point that alignment breaks trivially via fine-tuning, gandamuml notes a narrative trap: if closed models fail first, the blame falls on corporate recklessness; if open models fail first, on open intelligence itself. Either way, the public narrative misses what actually needs to be done.

Related event: Alignment Easily Bypassed by Fine-tuning, AI Safety Narratives Trapped in Dilemma(3 posts)→

Original post →

More from AGI Musings

AGI Musings channel →