Jerry Kaplan accentuates the potential positive in AI weapons; also see We’re Fighting Killer Robots the Wrong Way (WSJ)
"The world should approach A.I. weapons as an engineering problem — to establish internationally sanctioned weapons standards, mandate proper testing and formulate reasonable post-deployment controls — rather than by forgoing the prospect of potentially safer and more effective weapons.Robot Weapons: What’s the Harm? - The New York Times
Instead of turning the planet into a “Terminator”-like battlefield, machines may be able to pierce the fog of war better than humans can, offering at least the possibility of a more humane and secure world. We deserve a chance to find out."