Wednesday 1 August 2018

flagged content

The experimental social media chatbot Tay that quickly suffered a racist meltdown after less than a day’s exposure to the Internet wilds has a younger sibling called Zo, whose programming dictates she follows the protocols of their aggregate idea of a typical teenage American girl.

Far worse than the stereotyped pastiche of shouts and murmurs, however, is the unsettling and insidious way Zo champions political correctness through disengagement and strictly avoids the lures and decoys that led to the grounding of her elder sister. What Zo is ostensibly being trained for is to police forums for uncivil material and she has adopted a disturbing way of shutting down a conversation, should an interlocutor introduce a term associated with controversy. Decontextualizing censorship is a mistake that humans are prone to make enough by their own devices and possibly not something to be automated and institutionalised.