Posted on July 3, 2017 by New Scientist - NewsPeering inside an AI’s brain will help us trust its decisions A tool that reveals which aspects of a task an artificial intelligence is focusing on will help us understand why machines can be tricked