Mooning the Way We Destroy the Machines

In 2012, Google started developing a form of CAPSHA with snippets of photos from Google Street View. By 2014 they developed a version that trains machine learning models to just recognize objects and understand pictures. This is coming straight up at that previous story of recapture blurry text captures completing a huge archive of our writing. They finished Google Books and more than 150 years of archived New York Times.