Last month, OpenAI announced a text generating A.I. so scarily accurate they claimed it would be dangerous to release it. Now researchers have developed a tool to help spot text written by bots. Here’s what it means for the future of fake news in an age of smart machines.