Researchers at MIT have created a “psychopath” AI named Norman by training it using data from Reddit, saying that it produced highly disturbing and violent responses to standard prompts such as image captioning. What do you think?
“They better not start outsourcing my hatred to robots.”
Greg Mandeville • Knickknack Salesman
“Good, I’ve been getting sick of the cheery, happy-go-lucky AI we currently have.”
Stefan Wright • Ullulation Instructor
Advertisement
“Yeah, same thing happened to my son.”
Nina Stephens • Systems Analyst