Strange Loop Canon's @krishnanrohit explains why the biggest risk from autonomous AI agents may look less like a takeover and more like a flash crash: "You can look at cyber risk premia falling and note that we have these crazy things happening, but clearly people whose job it is to give you money if you get attacked by cyber don't seem to be worried about it. The question is, what do they know that you don't know?" "The key risk with AI I've always worried about is something similar to flash…