On today's episode, Thomas Kallstenius, CEO of Luxembourg Institute of Science and Technology, explains why AI needs to learn ethics so we can avoid a paperclips apocalypse.