Are bad incentives to blame for AI hallucinations?

How can a chatbot be so wrong — and sound so confident in its wrongness? 

​Read More  TechCrunch