Can an algorithm detect sarcasm better than you?

This segment first aired in January of 2016.

David Bamman and his co-author Noah Smith, have created an algorithm that can detect sarcasm online with 85% accuracy.

In their paper, Contextualized Sarcasm Detection on Twitter, they argue that just as with humans, understanding contextual information, like who wrote the text or if it was shared with friends or a wider audience, makes it easier for the machine to recognize sarcasm in Tweets and other text-based communication.