Just as bad, SMS messages leak metadata, which is information about the message but not the contents of the message itself, such as the phone number of the sender and the recipient, which can identify the people involved in the conversation.
If it sounds far-fetched that innocent people would be swept up by metadata and bad things could happen to them, you should probably know that hundreds of innocent people have already been drone-bombed simply for being in proximity to a SIM card on an NSA-sanctioned kill list, while the agency indiscriminately sucks up data everywhere those same drones fly using specialized sensors.
It is also important not to presume that metadata – or Big Data – is bad and dangerous any more than it is right to assume that technology is bad just because it could be misused.
TechCrunch Disrupt 2011 finalist Everpix, for example, built software that analyzes the visual content of your photos, organizes them into "Moments," hiding the bad photos, reconciling duplicates, merging corresponding metadata, and more.
Metadata retention is also bad news for whistleblowers and journalists.
In 2014 I attended a conference where James Dempsey, Board Member, PCLOB (The Privacy and Civil Liberties Oversight Board), stated that "after 14 years of analyzing the nation's phone call data collected because of section 215 of the Patriot Act [telephone metadata program], not one bad guy was discovered that wasn't already known".
Using machine learning, the company can not only automate metadata creation, a task humans are notoriously bad at, it could then take advantage of that metadata to add intelligence across the entire platform.
Laura Palmer, editorial director of Head of Zeus, accepting the award on his behalf was more chipper: "This completes every start-up publisher's dream hat-trick: Head of Zeus have now won a Political Book award, the Metadata Gold Standard award, and the Literary Review Bad Sex in Fiction award all in a single year".
They're very bad, however, at something that physical books are good at: gathering "metadata" about our reading — broken spines, dog-eared chapters, marginalia.
