Apple has a bit of an egg on its face after pulling an #ArtificialIntelligence-powered news notification feature. In this video, I cover the basic facts of this story before using it as a conversation piece to discuss the importance of quality control (and probably a human-in-the-loop) for doing #AI well in publicly visible, practical settings. I provide a little intuition for the possibility that the hallucination problem (when your AI makes stuff up) is probably not going to get all the way solved by newer and fancier models alone. More advanced AI is even likely to have more subtle and unexpected ways of embarrassing us...