Why AI safety's extinction framing fails to move regulators
Safety research rarely survives contact with industry lobbying because it leads with extinction risk — a frame that doesn't mobilise public pressure or votes.
Concrete, proximate harms (bias, labour displacement, privacy) move people to act and hold regulators accountable. The missing lever is organised public opinion, not better arguments.
A tech founder argues safety comms should flip: lead with what people care about, anchor regulation in lived harm, and let extinction risk live in the technical footnotes.