I wonder if there has been work on Bayesian spam filtering but for upworthy-style clickbait headlines. Phrase substitution is fun, but I'd love to see some sort of machine learning that could pick up on the clickbaityness of headlines.

For example, I can't remember the last time a headline with "This/These" or "You" wasn't clickbait.

Instead of 'fixing' the hyperbolic headlines, stop following the sources that have shared them, and ignore these sites. Upworthy et. al. do this because it works.