Safe AI: how is this possible?

Overview

Harald Rueß and Simon Burton examine challenges of AI safety and approaches connecting safety argumentation, architecture and verification.