AI isn't about replacing humans. It's about complementing them. That's why it should be called augmented intelligence.
NLP and NLG, the two branches of AI that parse and generate human language, have removed many of the barriers between humans and computers.
A machine learning expert decided to address artificial intelligence's lack of imagination, and overnight, he invented generative adversarial networks (GAN).
Narrow AI technologies are very good at solving specific problems, but they fall short of fulfilling the requirements for human-level AI.