Yes, it's a big problem that chatbots make things up. But there's more than one bright side to those 'hallucinations.' So I spoke up for them. wired.com/story/plaintex…