There’s a story I tell in my book because it’s a great illustration of how AI gets the wrong idea about what problem we’re asking it to solve: Researchers at the University of Tuebingen trained a neural net to recognize images, and then had it point out