With AI on the way, the Leverhulme Centre for the Future of Intelligence is bringing together experts to examine its morality and governance
Picture a self-driving car that sees a pedestrian in the road and has to swerve to avoid them. Now imagine there are cyclists on both sides of the car – and only the one on the right is wearing a helmet. Should the car veer right, to avoid killing the unprotected rider, even if that means punishing safer cycling? Stephen Cave chuckles ruefully. “At least since Socrates we’ve been worrying about moral philosophy and how to describe what’s right and what’s wrong,” he says. “Now suddenly we’ve got to programme this into artificial systems and it’s like, damn, we haven’t got very far.”