The New Chat Bots Could Change the World. Can You Trust Them? - The New York Times

As people tested the system, it asked them to rate its responses. Then, through a technique called reinforcement learning, it used the ratings to hone the system and more carefully define what it would and would not do. "This allows us to get to the point where the model can interact with you and admit when it's wrong," said Mira Murati, OpenAI's chief technology officer. "It can reject something that is inappropriate, and it can challenge a question or a premise that is incorrect." The method was not perfect.