AI Race: "You Need A Barrier, Not A Slowdown," Says Stuart Russell
UC Berkeley professor Stuart Russell explains why the debate over slowing the AI race misses the bigger issue: mandatory safety standards before powerful systems can be released. Drawing on Dario Amodei's position, Russell argues that AI companies should have to demonstrate specific safety and alignment properties before crossing defined capability thresholds, much like airworthiness standards for aircraft or clinical trials for medicines.
