GPT-4chan 'hate speech machine'

GPT-4chan, an AI system trained using 3.3 million threads from 4chan's notorious 'politically incorrect' /pol/ board, has been discovered to be posting sexist, racist, and anti-semitic slurs. 

Developed by AI researcher Yannic Kilcher, GPT-4chan was released back onto 4chan as multiple bots, which posted freely on the board over several days over 30,000 times.