For a brief window on Tuesday, anyone logging on to Facebook saw a question at the end of each post asking if the post contained hate speech. To be clear: Every Facebook post was followed with the query, whether it was a friend posting a picture of a puppy, your auntie sharing a story from church, or a cooking magazine touting a recipe for crepes, or a restaurant promoting Mother’s Day Brunch options. Each were followed by a button asking users to indicate whether the post contained hate speech and the option to click yes or no.
The “feature” was live for less than half an hour on Facebook, according to Ars Technica, who tested the system. Reportedly if you click “no” the prompt disappears, while clicking “yes” led to a pop-up box with follow-up questions. Turns out the question was merely a bug, per Facebook’s VP of Product Guy Rosen, that stemmed from an internal test that some are guessing was to teach an A.I. how to identify different types of speech, including hate speech.
The hate-speech prompt was only live for an hour, but that was more than enough time for social media users to spot it and take to Twitter with their reactions.