Could we fight fake news by reconnecting shared images to their origins, using automatic reverse image search to add context to all imagery that summarizes it based on how it has been captioned across the web and offering links to places it has been seen in the past?