YouTube credits machine learning for playing a major role in removing eight million objectionable videos from the platform during the last quarter of 2017. In a recent blogpost, the platform states that AI technology flagged up to more than 83% of videos that are being reviewed for deletion by employees, and almost three-quarters were removed before they were even viewed. It’s a major difference from early 2017 when only 8% of videos containing “violent extremism,” were flagged and removed before it reached 10 views. (Fortune)
